T h e Ox f o r d H a n d b o o k o f

ROM A N B R I TA I N
ii
The Oxford Handbook of

ROMAN BRITAIN
Edited by
MARTIN MILLETT
LOUISE REVELL
and
ALISON MOORE

1
iv

3
Great Clarendon Street, Oxford, ox2 6dp,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Oxford University Press 2016
The moral rights of the authors‌have been asserted
First Edition published in 2016
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2016944843
ISBN 978–​0–​19–​969773–​1
Printed in Great Britain by
Clays Ltd, St Ives plc
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
For Richard Reece
vi
 Contents

List of Figures  xi
List of Tables  xxi
List of Contributors  xxiii
Introduction  xxvii

PA RT I   NAT U R E OF T H E E V I DE N C E

1. Early Studies in Roman Britain: 1610 to 1906  3

Richard Hingley
2. Roman Britain since Haverfield  22
Martin Millett
3. Romano-​British Archaeology in the Early Twenty-First Century 43
Pete Wilson
4. The Development of Artefact Studies  63
Ellen Swift
5. The Textual and Archaeological Evidence  95
Henry Hurst
6. The Early Roman Horizon  117
Lacey Wallace
7. Britain at the End of Empire  134
Simon Esmonde Cleary
8. Britain before the Romans  150
Timothy Champion
9. Beyond Hadrian’s Wall  179
Fraser Hunter
viii

viii   Contents

10. Mobility, Migration, and Diasporas in Roman Britain  203

Hella Eckardt and Gundula Müldner
11. Multiculturalism on Hadrian’s Wall  224
Claire Nesbitt
12. Britons on the Move: Mobility of British-​Born Emigrants in the
Roman Empire 245
Tatiana Ivleva
13. Britain, Gaul, and Germany: Cultural Interactions  262
Tom Moore

PA RT I I   S O C I E T Y A N D T H E I N DI V I DUA L

14. Inscriptions and Identity  285

Valerie M.  Hope
15. Ideas of Childhood in Roman Britain: The Bioarchaeological and
Material Evidence  303
Rebecca Gowland
16. The Life Course  321
Alison Moore
17. Status and Burial  341
John Pearce
18. Gender in Roman Britain  363
Melanie Sherratt and Alison Moore
19. Deviancy in Late Romano-British Burial 381
Belinda Crerar
20. Clothing and Identity  406
H. E. M. Cool
21. Cemeteries and Funerary Practice  425
Jake Weekes
22. Identity and the Military Community in Roman Britain  448
Ian Haynes
 Contents   ix

23. Roman Military Culture  464

Lindsay Allason-​Jones

PA RT I I I   F OR M S OF K N OW L E D G E

24. Changing Materialities  481

Andrew Gardner
25. Forms of Knowledge: Changing Technologies
of Romano-​British Pottery  510
Jeremy Evans
26. Metals and Metalworking  532
David Dungworth
27. Medicine  555
Patricia Baker
28. Sociolinguistics  573
Alex Mullen
29. Art in Roman Britain  599
Ben Croxford
30. Names of Gods  619
Amy Zoll
31. Ritual Deposition  641
Alex Smith
32. Christianity in Roman Britain  660
David Petts
33. Memories of the Past in Roman Britain  681
Zena Kamash

PA RT I V   L A N D S C A P E A N D E C ON OM Y

34. ‘By Small Things Revealed’: Rural Settlement and Society  699

Martin Millett
x

x   Contents

35. Rural Transformation in the Urbanized Landscape  720

Martin Pitts
36. The Development of Towns  741
Adam Rogers
37. Urban Monumentality in Roman Britain  767
Louise Revell
38. The Exploitation of Animals in Roman Britain  791
Mark Maltby
39. Arable Farming, Horticulture, and Food: Expansion, Innovation,
and Diversity 807
Marijke van der Veen
40. Coinage and the Economy  834
Philippa Walton and Sam Moorhead
41. Economy and Power in Late Roman Britain  850
James Gerrard

Index  869
 List of Figures

1.1 The Cogidubnus inscription from Chichester, Sussex. From Gale (1723).
Reproduced by permission of Durham University Library. 7
1.2 Plan of a Roman villa excavated by Samuel Lysons at Bignor in Sussex.
Source: Lysons (1815: plate 19). Reproduced by permission of Durham
University Library. 10
1.3 Knook Castle and British villages, Wiltshire. Source: Hoare (1810).
Reproduced by permission of Durham University Library. 12
1.4 The civil and military districts. Source: Haverfield (1912: figure 1).
Reproduced by permission of Durham University Library. 15
1.5 The Roman site at Chesters (Cilurnum) showing the excavated areas
and ‘forum’. Source: Budge (1903: opposite p. 98). Reproduced by
permission of Durham University Library. 16
1.6 Sketch plan of the graves forming a ‘family circle’ at Aylesford, Kent.
Source: Evans (1890: figure 4). Reproduced by permission of Durham
University Library. 17
3.1 Excavation of a major Roman building by Oxford Archaeology South in
advance of housing development just outside the walled town
of Cirencester in 2009. Source: © Oxford Archaeology South, with
thanks to Berkeley Homes for agreement to reproduce. 45
3.2 Volunteers excavating Building 16 at Piddington villa, Northamptonshire,
in 2012. Source: © Pete Wilson. 46
3.3 Volunteers excavating the sequence of barracks in the north-​west
quadrant of the fort at Vindolanda in 2011. Source: © The Vindolanda Trust. 53
3.4 Combined geophysical surveys undertaken by the Landscape Research
Centre with the support of Historic England and the Aggregates Levy
Sustainability Fund on the southern side of the Vale of Pickering,
North Yorkshire; depicts more than 1.5 kilometres of late Iron Age and
Roman trackside ribbon development or ‘ladder settlement’ in addition
to a wealth of other prehistoric and post Roman features. Source:
© Landscape Research Centre. 55
4.1 Gender profiles in artefact studies. (a) Percentage of men and women
with published articles on small finds listed in the Society of Antiquaries
xii

xii   List of Figures

subject catalogue from 1900 to 1988 (end date of catalogue). (b) Percentage
of men and women publishing articles in Britannia 1970–​89 (excluding
reviews). (c) Subject of articles by women in Britannia 1970–​89 (excluding
reviews). Source: © Ellen Swift. 78
5.1 Memorandum on wooden leaf tablet from Vindolanda, referring to
the British use of cavalry (Tab. Vindol. II. 164): dimensions 78 x 186
millimetres. After a punctuation mark in l. 4, the text reads: ‘nec residunt |
Brittunculi ut iaculos | mittant’ (‘nor do the Brits mount in order to throw
javelins’). Source: Full text and translation in Bowman and Thomas (1994:
107–108). © The Trustees of the British Museum. 97
5.2 Sale of a slave girl, Fortunata, recorded on a stylus tablet from London,
drawn by R. Tomlin: dimensions c. 140 x 114 millimetres, probably one
of three tablets. The writing survived as scratches in the wooden tablet,
becoming illegible where a triangular patch of the original wax still
survived; [p]‌uellam Fortunatam are the first two words of l. 3. Source: Full
text and translation in Tomlin (2003, 2011). © Roger Tomlin. 98
5.3 Curse tablet no. 43 from the sanctuary of Mercury at Uley,
Gloucestershire, drawn by R. Tomlin: dimensions c. 95 x 83 millimetres.
Docilinus asks Mercury to drive Varianus, Peregrina, and Sabinianus
to death for injuring his farm animal unless they redeem their action.
Source: Text and translation in Hassall and Tomlin (1989: 329–​331, no. 3)
and Tomlin (2002: 172). © Roger Tomlin. 99
5.4 Tile with personal-​name graffito scratched before firing, Candid[us].
From a fourth-century deposit at Hayton, East Yorkshire. Source: Tomlin
and Hassall (2001: 393). Photograph: M. Millett. 101
5.5 Tombstone of Classicianus from London: pieces as displayed in the
British Museum with reconstruction of original monument drawn by
Richard Grasby (Grasby and Tomlin 2002: figure 21), slightly adjusted.
Reconstructed text (as in RIB 12): ‘Dis | [M]‌anibus | [G(ai) Iul(i) G(ai)
f(ili) F]ab(ia tribu) Alpini Classiciani | ... | ... | proc(uratoris) provinc(iae)
Brita[nniae] | Iulia Indi filia Pacata I[ndiana(?)] | uxor [f(ecit)]’ (‘To the
spirits of the departed (and) of Gaius Julius Alpinus Classicianus, son of
Gaius, of the Fabian voting-​tribe ... procurator of the province of Britain,
Julia Pacata I[ndiana], daughter of Indus, his wife, had this built’). Source:
Photograph © The Trustees of the British Museum. 103
5.6 Europa and the bull with two cupids in front of the presumed
semicircular couch in the apse of the dining room at Lullingstone villa,
Kent. The inscription (RIB 2448.6) reads: ‘Invida si ta[uri] vidisset Iuno
natatus | Iustius Aeolias isset adusque domos’ (‘If jealous Juno had seen the
swimming of the bull, more justly would she have gone to the palace of
Aeolus’). Source: © Alamy. 106
 List of Figures    xiii

9.1 Distribution of Roman military sites in Scotland. (a) Flavian (c. ad

78–86); (b) Antonine (c. ad 139–​165). Source: Reproduced by courtesy of
Professor D. J. Breeze. 183
9.2 The Antonine military complex at Inveresk (East Lothian). Source:
Drawing by Alan Braby. 187
9.3 Distribution of Roman finds from non-​Roman sites north of Hadrian’s
Wall. (a) overall distribution; (b) middle Roman Iron Age c. ad 160–​250
(the sites of Birnie and Traprain Law, mentioned in the text, are marked:
*represents find-spots of denarius hoards); (c) late Roman Iron Age c. ad
250–​400; *marks the major centres of Traprain Law, Edinburgh Castle,
Eildon Hill, and Dumbarton Rock. Source: © Fraser Hunter. 191
10.1 Epigraphically recorded age of travellers to Rome 30–​600 ad as recorded
by Noy (2000) and for the late antique west by Handley (2011). Source:
after Handley (2011: graphs 1 and 2). © Department of Archaeology,
University of Reading. 205
10.2 Strontium and oxygen isotope data of humans from Lankhills/​
Winchester, contrasting burials identified as ‘intrusive’ according to
Clarke’s criteria (1979) with other burials from the site; the two boxes
indicate the local strontium isotope range for Winchester and estimates
for the range of oxygen isotope values consistent with a childhood in
Britain based on human skeletal phosphate (δ18Op) data available in 2009
(thin line) and 2012 (thick line). Source: after Evans et al. (2006) and
Eckardt et al. (2009). © Department of Archaeology, University of Reading. 212
10.3 Reconstruction of the so-​called ‘Ivory Bangle Lady’ from York.
Source: © Aaron Watson. 215
11.1 Map showing areas of the Roman Empire with military units on Hadrian’s
Wall. Source: © Rob Witcher, illustration by Christina Unwin. 229
11.2 Regina’s tombstone. Source: © Claire Nesbitt.  239
12.1 Distribution of British brooches. Source: Brooches distribution partly
after Morris (2010: 86, figure 4.35 and appendix 6); map by author.
© Tatiana Ivleva. 250
12.2 Britons abroad: profession and status. Source: © Tatiana Ivleva.  252
12.3 Distribution of the military diplomas (star), funerary (circle), and votive
(diamond-​shape) inscriptions mentioning British emigrants.
Source: Map by author. © Tatiana Ivleva. 253
13.1 Distribution of cremation burials in late Iron Age Britain. Source: after
Fitzpatrick (1997). © T. Moore. 263
13.2 Comparison of late Iron Age cremation burials from (a) Welwyn Garden
City, Hertfordshire, and (b) Clemency, Luxembourg. Source: © British
Museum and Musée National d’Histoire et d’Art, Luxembourg. 265
xiv

xiv   List of Figures

13.3 Similarities in villa layouts from northern Gaul and Britain.

Source: © T. Moore. 269
13.4 Approximate distribution of Romano-​Celtic temples. Source: after Derks
(1998). © T. Moore. 271
16.1 Percentage of burials with grave goods with each age category at
Colchester, Poundbury, and Lankhills. Source: © Alison Moore. 327
16.2 Percentage of total burials with grave goods at East London, Gloucester,
and York. Source: © Alison Moore. 332
16.3 Percentage of aged burials with grave goods at rural sites in western and
eastern regions of southern Roman Britain. Source: © Alison Moore.  333
17.1 Bronze and ceramic vessels in grave 6260, the cremation burial of an
adult (ad 5–​80), from a rural settlement at Tollgate, Springhead. The
bronze jug, pan, and mixing bowl as well as two ceramic flagons were
placed beneath a table. On this were piled terra nigra and terra rubra
platters, beakers, cups, and local copies. The skull and forelimb of a pig
were found close by, as was the cremated bone. One or more boards with
glass gaming pieces and bone dice lay closer to the grave’s centre. Source:
courtesy of Oxford Archaeology. 350
17.2 A bronze drinking vessel found at Llantilio Pertholey near Abergavenny
with a handle in the form of a snarling leopard with inlaid silver and
amber (PAS NMGW-​9A9D16). Fieldwork subsequent to its discovery by
metal-​detecting revealed the original funerary context. Source: courtesy
of National Museums and Galleries of Wales. 352
19.1 Comparative age of decapitated and non-​decapitated inhumations along
the Fen Edge. Source: © B. Crerar.  391
19.2 Comparative sex of decapitated and non-​decapitated inhumations along
the Fen Edge. Source: © B. Crerar.  392
19.3 Comparative grave-good allocation for decapitated and non-​decapitated
inhumations along the Fen Edge. Source: © B. Crerar. 393
19.4 Comparative coffin allocation for decapitated and non-​decapitated
inhumations along the Fen Edge. Source: © B. Crerar.  394
19.5 Comparative posture of decapitated and non-​decapitated inhumations
along the Fen Edge. Source: © B. Crerar.  395
19.6 Guilden Morden cemetery, showing distribution of decapitated burials,
adapted from Fox and Lethbridge (1926: figure 1) and Lethbridge (1934:
site plan). Source: © B. Crerar. 398
20.1 Chronological and regional distributions of brooches. Note: The area and
type labels are those given in Table 20.1 when not otherwise specified.
Figures 20.1 (a-c) are Correspondence Analysis plots that show which
regions have higher proportions of which types. Source: Data derived
from Mackreth (2011). © Hilary Cool. 414
 List of Figures    xv

25.1 Map showing quantitative evidence for the distribution of Severn Valley
wares, in comparison with the core distribution of the western Iron Age
coinage group. Note: Columns represent the percentage at a site; dashes
represent sites without Severn Valley wares. Known kiln sites producing
Severn Valley wares are shown. Both the contour map and the bar plots
show that there is a very sharp fall-​off at the edge of the distribution of the
fabrics. Source: © J. Evans. 512
25.2 Map showing the core areas of the seven regional coin series of Iron Age
Britain, after Creighton (2000: figure A.1), compared with the distribution
of Dressel 1 amphorae finds, after Fitzpatrick (1985), and the distribution of
pre-​conquest Romanized toilet instruments (Hill 1997). Source: © J. Evans. 516
26.1 Map of Roman Britain showing selected sites with evidence for
goldworking. Source: © David Dungworth. 533
26.2 Principal lead-​silver orefields and sites of primary silver extraction.
Source: © David Dungworth.  536
26.3 Zinc and lead content of Colchester-​type brooches (single piece
construction and made of brass) and Colchester derivative brooches
(multiple component construction and made of leaded bronze).
Source: © David Dungworth.  540
26.4 Silver content of denarii and zinc content of contemporary sestertii
and dupondii (data from Walker 1976–​8; Dungworth 1996).
Source: © David Dungworth. 541
26.5 Principal iron production sites. Source: © David Dungworth. 544
26.6 Dated Roman iron production sites in the Weald, after Cleere (1974).
Source: © David Dungworth. 545
28.1 Kernel density plot of the distributions of ‘broad period Roman’ styli
recorded by the Portable Antiquities Scheme as of 17 January 2014,
plotted against selected constraints. Source: Robbins 2014: figure 4 for
constraints, with kind permission of the author; background map
data from Ordnance Survey/​EDINA supplied service, Crown
Copyright/​database right 2013. Produced by L. Wallace. 580
28.2 Tomlin’s line drawing of Tab. Sulis 18. Source: Reproduced with the kind
permission of the author.  581
28.3 Tomlin’s line drawing of Tab. Sulis 30. Source: Reproduced with the kind
permission of the author.  583
28.4 Tomlin’s line drawing of the lead curse tablet at Red Hill, Ratcliffe-​on-​
Soar. Source: Hassall and Tomlin (1993: 311), reproduced with the kind
permission of the author. 586
28.5 Latin and Ogam-​inscribed stone, St Dogmaels/​Llandudoch,
Pembrokeshire, Wales, (a) photograph, (b) drawing and transcription.
xvi

xvi   List of Figures

Source: © Crown copyright: Royal Commission on the Ancient and

Historical Monuments of Wales. 589
29.1 Finds per site of anthropomorphic sculpture. 601
29.2 Zoning of sculpture quality. 605
30.1 Distribution of inscribed votive altars from Roman Britain. Source: after
Millett (1995a: 110). © Amy Zoll. 623
30.2 Main concentrations of votive altars. Source: after Millett (1995a: 110),
temples after Millett (1995a: 112); TOT rings after Daubney (2010: 112).
© Amy Zoll. 625
30.3 Distribution of votive inscriptions dedicated to the gods Cocidius,
Belatucadrus, and Vitiris/​Veteres, including double-​named variants.
Source: after Zoll (1995a). © Amy Zoll. 631
32.1 Silver ring with early Christian symbol (anchor and fish) from the Roman
fort at Binchester, County Durham. Source: © Durham University. 665
32.2 Potential Roman church from Silchester, Hampshire. Based on Frere
(1975: figure 1). Source: © David Petts. 667
32.3 Painted wall plaster from Lullingstone Roman villa (Kent) showing
chi-​rho symbol flanked by an alpha and omega. Source: © The Trustees of
the British Museum.  669
34.1 Magnetometer survey of part of the Roman landscape in the Vale of
Pickering, North Yorkshire. Across the top of the image a trackway
runs east-​west, flanked on either side by settlement enclosures and
small fields. Beyond these to the south are the fainter traces of a
series of fields bounded by ditches. Superimposed on these, in the
northern zone, the back dots represent grubenhaüser (sunken-​featured
buildings) of Early Medieval date. To the south the land rises towards
the Yorkshire Wolds and the ancient landscape is obscured by deposits
of wind-​blown sand. Source: Courtesy of Dominic Powlesland,
Landscape Research Centre. 702
34.2 Plan of the Roman villa at Fishbourne, Sussex, in the later first century ad
showing the approach from the east and associated structures. Source:
Drawn by Lacey Wallace, after Cunliffe (1971); Cunliffe et al. (1996);
Manley and Rudkin (2003). 705
34.3 Plans showing the development of the Roman villa and associated
structures at Frocester Court, Gloucestershire. Source: Drawn by Lacey
Wallace, based on Price (2000–​10). 707
34.4 Plan of the nucleated rural settlement at Higham Ferrers,
Northamptonshire, in the late third–​fourth centuries ad. A series of
buildings set within enclosures flank the Roman road, which runs along
 List of Figures    xvii

the valley of the river Nene just above the flood plain. Source: Illustration
from Lawrence and Smith (2009), reproduced courtesy of Oxford
Archaeology. 709
34.5 Photograph of a domestic building (Figure 34.4, No. 10810) within the
nucleated rural settlement at Higham Ferrers, Northamptonshire. The
Roman road and the river Nene are visible in the background. Source:
Illustration from Lawrence and Smith (2009), reproduced courtesy
of Oxford Archaeology. 710
34.6 Cut-​away reconstruction drawing of the third-​century aisled hall
excavated at Shiptonthorpe, East Yorkshire. Source: Drawing by Mark
Faulkner, from Millett (2006). 713
34.7 Aerial photograph of part of the Roman landscape at Burnby Lane,
Hayton, East Yorkshire, looking north-​east. The valley of the Burnby
Beck runs down the centre of the photograph, with its relict course
showing as a light crop-​mark against the darker green of the flood plain.
On either side of this damp ground can be seen a series of settlement
enclosures. On the left a length of boundary ditch is also visible. This runs
parallel with the stream, and formed one side of a droveway that linked
the lowlands to the south-​west with the Wolds to the north-​east, allowing
animals to be moved for seasonal grazing. Source: Photograph by Peter
Halkon, courtesy of the Hayton Project. 716
35.1 Correspondence analysis highlighting differences in the composition
of pottery assemblages (top) by vessel form (bottom) in the hinterland
of Roman London and Colchester, c. 50 bc–​ad 250. Note: Details of
abbreviations are outlined in Table 35.1. © Martin Pitts. 727
35.2 Pottery assemblages from Roman London, Essex, and Cambridgeshire
plotted according to the percentage prevalence of lids versus jars, c. 50 bc–​
ad 250 Source: after Perring and Pitts (2013). © Martin Pitts. 729
35.3 Multidimensional scaling analysis of the prevalence of non-​work-related
health conditions in late Roman Britain. Source: after Pitts and Griffin (2012).
© Martin Pitts. 733
35.4 Comparison of the Gini coefficient of inequality with the mean number
of furnishings per grave for selected Romano-​British cemeteries
Source: after Pitts and Griffin (2012), with additions. © Martin Pitts. 736
36.1 Map of southern Britain with the main settlements mentioned in the
chapter. Source: © A. C. Rogers.  744
36.2 Plan of pits, shafts, and timber-fenced enclosures excavated within the
centre of Roman Dorchester (Durnovaria) at the Greyhound Yard site,
1981–​4. Source: adapted from Woodward and Woodward (2004: figure 1).
© A. C. Rogers. 753
xviii

xviii   List of Figures

36.3 Plan of the relationship between the fortress and later colonia at Colchester;
the walls around the town were constructed in the early second century ad;
the theatre and temple were constructed within the fortress annexe. Source:
adapted from Crummy (1993: figure 2.9). © A. C. Rogers. 754
36.4 Plan of the relationship between the town wall circuit and enigmatic
earthworks at Silchester (Calleva Atrebatum) possibly associated with the
earlier oppidum here. Source: adapted from Fulford (1984: figure 85). 757
37.1 Plan of Caerwent forum, with location of probable curia marked.
Source: Adapted from Brewer (2006: 39). 770
37.2 Plan of Silchester forum with known and approximate locations of
inscriptions and statues indicated. Source: P. Copeland; adapted from
Isserlin (1998: figure 9.1) with additions. 771
37.3 Plan of Canterbury. Source: (Millett 2007: figure 5.15). Reproduced by
kind permission of the Kent History Project, Kent County Council, from
plans provided by the Canterbury Archaeological Trust. 775
37.4 Plan of Wroxeter. Source: (White et al. 2013: figure 4.21). 777
39.1 (a) Charred spelt wheat grain, showing infestation by a grain weevil from
a third/​fourth century corn drier at Grateley South, Hampshire; Photo:
Gill Campbell; © English Heritage; reproduced by kind permission of
English Heritage; (b) Charred remains of the granary weevil (Sitophilus
granarius L.) from a deposit filled with charred ‘rubbings’ of malted spelt
in an early/​mid Roman ditch at Northfleet villa, Kent ([D.] Smith 2011);
Photo: David Smith; (c) Charred germinated spelt grains from a late
Roman corn drier at Northfleet villa, Kent ([W.] Smith 2011: Plate 9); ©
High Speed 1 Ltd; image reproduced with the kind permission of High
Speed 1 Ltd; (d) Charred detached spelt grain sprouts from a late Roman
ditch at Northfleet villa, Kent ([W.] Smith 2011: Plate 10); © High Speed
1 Ltd; image reproduced with the kind permission of High Speed 1 Ltd;
(e) Charred coriander; one of a cache of more than 1,000 coriander fruits
found on the floor of a shop in Colchester that was burnt down during
the Boudiccan Revolt of ad 6/61. The material was recovered as part of
the excavations carried out at 45-​6 High Street (Murphy 1977); Photo: Gill
Campbell; © English Heritage; image reproduced by kind permission of
English Heritage; (f) Waterlogged olives from a latrine block at the rear of
a first century tavern at 1 Poultry, London (Davis 2011: figure 39. 275);
Photo: Andy Chopping; image reproduced by kind permission of
Museum of London Archaeology. 811
39.2 Predominant mode of preservation for the three main categories of food
and fibre crops, based on the total number of occurrences of each of these
plants in archaeobotanical assemblages from Roman Britain. Source:
after Van der Veen (2008); Van der Veen et al. (2008). 815
 List of Figures    xix

39.3 Frequency patterns of selected foods, waterlogged records only. a) foods

that initially increase but then decline; b) foods that increase over time;
c) foods that decline over time. N = number of records with waterlogged
remains. Source: after Van der Veen (2008); Van der Veen et al. (2008). 818
39.4 A near complete waterlogged imported pine cone (Pinus pinea) from a
late third/​early fourth century ditch fill containing possible other votive
objects at Clatterford Roman villa (McPhillips 2001). Source: © English
Heritage; image reproduced by kind permission of English Heritage.  824
40.1 Roman coins recorded by the Portable Antiquities Scheme.
Source: © Philippa Walton.  835
40.2 BERK-​65D307—a Roman Republican denarius issued in c. 207 bc found
in Berkshire. Source: © Portable Antiquities Scheme. 837
40.3 The Frome hoard being excavated. Source: © Somerset County Museums
Service. 842
40.4 The Coleraine Hoard (County Antrim) deposited in the first half of the
fifth century ad. Source: © The Trustees of the British Museum. 845
41.1 Plan of the roadside settlement at Higham Ferrers. Source: (after
Lawrence and Smith 2009: figure 2.18). © James Gerrard/​Andrew Agate. 854
41.2 Plan of the Roman villa at Turkdean. Source: (after Holbrook 2004: figure
4). © James Gerrard/​Andrew Agate. 857
41.3 The distribution of villas near Ilchester: 1) Dinnington, 2) Lopen,
3) Seavington St Mary, 4) Ilchester Mead, 5) Batemoor Barn, 6) Lufton,
7) West Coker, 8) East Coker, 9) Westlands. Source: © James Gerrard/​
Andrew Agate.  858
xx
 List of Tables

2.1 The educational networks of the contributors to this volume.  34

4.1 Finds categories in Roman site reports up to the 1960s. © Ellen Swift. 66
11.1 Foreign units serving on Hadrian’s Wall and their provincial origins.
Source: © Claire Nesbitt. 230
16.1 Distribution of grave goods with young/​prime adults and percentage of
items related to personal appearance with female burials at four major
urban cemeteries. Source: © Alison Moore. 328
16.2 Percentage of grave goods with mature/​older male and female burials
at five urban cemeteries. Source: © Alison Moore. 330
19.1 Sites consulted.  390
20.1 Brooches from England and Wales by region and broad date band
Source: Mackreth (2011). © Hilary Cool. 412
21.1 Some significant and published case studies.  426
31.1 Occurrence of selective votive objects in Southern Romano-​British cult
sites. Source: After Smith (2001: 155, figure 5.13). 645
35.1 Codes used in Figures 35.1 and 35.2  728
xxii
 List of Contributors

Lindsay Allason-​Jones is a Visiting Fellow at University of Newcastle upon Tyne, Chair

of the Marc Fitch Fund and the Hadrian Arts Trust
Patricia Baker is a Senior Lecturer in Classical and Archaeological Studies at University
of Kent
Timothy Champion is Emeritus Professor of Archaeology at University of Southampton
H. E. M. Cool is a Director of Barbican Research Associates Ltd
Belinda Crerar is a Curator at the British Museum
Ben Croxford is Historic Environment Officer for Merseyside
David Dungworth is the Head of Archaeological Conservation and Technology at
Historic England
Hella Eckardt is Associate Professor in Roman archaeology at University of Reading
Simon Esmonde Cleary is Professor of Roman Archaeology at University of
Birmingham and has been a Visiting Professor at the University of Toulouse-​le Mirail
Jeremy Evans is a freelance Roman pottery consultant and a Director of Barbican
Research Associates Ltd
Andrew Gardner is a Senior Lecturer in the archaeology of the Roman Empire at
University College London and a Co-​Director of the Caerleon Excavations and of the
West Dean Archaeological Project
James Gerrard is a Lecturer in Roman Archaeology at University of Newcastle upon
Tyne
Rebecca Gowland is a Senior Lecturer in human bioarchaeology at the Department of
Archaeology at University of Durham
Ian Haynes is Professor of Archaeology at University of Newcastle upon Tyne
Richard Hingley is Professor of Roman Archaeology at University of Durham
Valerie M. Hope is a Senior Lecturer in the Department of Classical Studies at Open
University
Fraser Hunter is the Iron Age and Roman Curator at National Museums Scotland
xxiv

xxiv   List of contributors

Henry Hurst is a Reader in Classical Archaeology (Emeritus) at University of

Cambridge
Tatiana Ivleva is the Postdoctoral Marie Curie Fellow at University of Newcastle upon
Tyne
Zena Kamash is Lecturer in Roman Archaeology and Art at Royal Holloway, University
of London
Mark Maltby is a Reader in Archaeology at University of Bournemouth
Martin Millett is the Laurence Professor of Classical Archaeology at University of
Cambridge
Alison Moore is a Sessional Lecturer in Roman Archaeology and an Academic Editor
Tom Moore is a Senior Lecturer in Archaeology at University of Durham
Sam Moorhead is National Finds Adviser for Iron Age and Roman Coins, Department
of Portable Antiquities and Treasure at the British Museum
Gundula Müldner is a bioarchaeologist at University of Reading
Alex Mullen is Assistant Professor in Classical Studies at University of Nottingham
Claire Nesbitt is a Postdoctoral Research Associate in the Department of Archaeology
at University of Durham
John Pearce teaches Roman archaeology at King's College London
David Petts is a Lecturer in Archaeology in the Department of Archaeology, University
of Durham
Martin Pitts is Senior Lecturer in the Department of Classics and Ancient History at
University of Exeter
Louise Revell is Associate Professor in Roman Studies at University of Southampton
Adam Rogers is a Teaching Fellow at University of Leicester
Melanie Sherratt is a Ph.D. candidate at University of Durham
Alex Smith is a Senior Research Fellow at University of Reading
Ellen Swift is a Senior Lecturer in Archaeology at University of Kent
Marijke van der Veen is Professor of Archaeology at University of Leicester
Lacey Wallace is a Postdoctoral Research Associate in Roman Archaeology in the
Faculty of Classics, University of Cambridge
Philippa Walton is a Research Fellow at Wolfson College, University of Oxford
Jake Weekes is a Research Officer at the Canterbury Archaeological Trust
 List of contributors    xxv

Pete Wilson is an Archaeological Consultant and formerly Head of Research Policy

(Roman Archaeology) for English Heritage
Amy Zoll is a Social Sciences Information Technology Specialist and Ph.D. candidate in
Anthropology at the University of Pennsylvania
xxvi
 Introduction

martin millett, louise revell,

and alison moore

The potential reader of this book might be forgiven for an initial response along the
lines of ‘why yet another book on Roman Britain?’ After all, as Sir Mortimer Wheeler
wrote in discussing the subject:

[W]e are slowly approaching a little more nearly to the mind of Roman Britain, for
what that mind be worth-​and all human mentality is presumably worth something.
We must not expect much of great consequence. Hadrian’s Wall, for example, has
long ceased to matter as a major historical problem. … But for the major attainments
of mankind the young scholar, thus trained, must now look to other periods, other
lands. Haverfield’s work was not merely a monument, it was a tombstone. (Antiquity
no. 138, 1961: 157–​8)

We would assert that there are two sound answers to this question. First, whilst there are
many books on Roman Britain, most follow an approach that was first used in the 1930s
with R.G. Collingwood’s contribution to the Oxford History of England, namely using
the archaeological evidence within a chronologically organized narrative of historical
events, complemented by chapters on topics such as art and religion. It is remarkable
how enduring this framework has proved, not only in describing the subject but also in
setting the agenda for how the subject should be approached right down to the radical
reappraisal provided in David Mattingley’s An Imperial Possession (2006). In his review
of Peter Salway’s volume in the Oxford History of England published in 1981 and replacing
Collingwood’s 1937 volume, Richard Reece attacked this approach, memorably saying:

Textbooks on Roman Britain to date make the subject appear like a nice sand-​pit
in which toddlers can safely be left to play. I am thankful that it is a wild overgrown
garden in which anything may happen. And I shall continue to try to prove it.
(Archaeological Journal Vol. 139, 1982: 456).

This book takes its lead from Richard Reece’s critique, and seeks to provide a compre-
hensive review of Roman Britain from an entirely new perspective, that represented by
younger scholars who have embraced new perspectives on the subject since the estab-
lishment of the Theoretical Roman Archaeology Conference (TRAC) in 1991. In this sense
although about Roman Britain, the book offers a new approach that is very different
from the many other volumes already on the shelf.
xxviii

xxviii   Introduction

Second, for all students of the archaeology of the Roman Empire, it must be acknowl-
edged that although Roman Britain itself was arguably marginal to the mainstream, it
remains amongst the most fully explored and best documented of Rome’s provinces.
There are very fully publications of primary data: inscriptions in the Roman Inscriptions
of Britain (1965–​); sculpture in the Corpus Signorum Imperii Romani (1977–​); mosaics
in D. Neal and S. Cosh’s Roman Mosaics of Britain (2002–​10); and place-​names in A.L.F.
Rivet and C. Smith’s Place-​names of Roman Britain (1979). In addition there is a wealth
of synthetic works, finds studies, maps, and site reports. These mean that the province is
worthy of study because with its evidence we can explore ideas and interpretations more
thoroughly than in other less well documented provinces. In that sense, by offering a
Handbook which contains new approaches this book seeks to capitalize on such past
work whilst offering thoughtful signposts for the future.
In summary, given the dramatic increase in new scholarship over the last thirty years in
the field of Roman Britain, we feel the time is ripe for a comprehensive review of the topic.
This Handbook is organized around sections that reflect recent theoretical approaches to
the material. In particular, we hope that it reflects the new questions and interpretations
brought to the subject since the inauguration of TRAC. The sections are structured around
the following themes: (i) the nature of the evidence; (ii) Britain in the Roman Empire; (iii)
society and the individual; (iv) forms of knowledge; (v) landscape, settlement, and the
economy; and (vi) materiality. The individual articles range from surveys to more focused
discussions, and represent the methodological and theoretical breadth of the field.
We are deliberately eschewing a chronological structure that has dominated past
work on Roman Britain but it is of course important to bear in mind how the prov-
ince originated and developed through time. The table below provides a summary of the
textual evidence for the province, thus providing the basic evidence on which histori-
cal accounts are usually based. In many works on Roman Britain there is a tendency to
embellish this primary record with historical inference derived from the study of finds,
principally pottery and coins. This may of course be entirely reasonable in the context
of a particular site narrative or regional study, but we are concerned that generalizing
from such evidence to provide a broader chronological narrative provides a false sense
of secure knowledge rather than acknowledging the fragmentary and mutable nature of
the evidence. For this reason the framework provided below simply uses texts.

Timeline of Roman Britain

55–​54 bc
Julius Caesar, engaged on his conquest of Gaul, made two military expeditions to
Britain. The first in 55 bc was of limited success although it generated a useful propa-
ganda advantage in Rome where Britain was known as a mysterious place on the other
 Introduction   xxix

side of the ‘Ocean’. The larger-​scale campaign in 54 bc was more successful and resulted
in the imposition of treaties on the peoples of the south-​east, which marks the first stage
of a projected Roman annexation. (Caesar, Gallic Wars, V)

54 bc–​ad 43
Events, first in Gaul then at Rome, prevented any capitalization on Caesar’s success.
Diplomatic contacts between British tribes and Rome continued with the arrival in
Rome of tribal leaders deposed in Britain under Augustus (27 bc–​ad 14) and Caligula
(ad 37–​41). It is plausible that Rome considered some of the peoples of the south and
east to have been client states. (Res Gestae, 32; Tacitus, Annals, II). Both Augustus and
Caligula (in ad 40) are said to have considered invading Britain, although neither acted.
(Dio Cassius, LIII; Tacitus, Annals, II)

54 bc–​ad 43
There was also commercial contact, with the export of products from Britain and
imports from the empire, which seems to have been worthwhile for Rome (Strabo,
IV). Rome influenced political evolution in Britain in the period down to AD 43 and
this engendered internal conflicts. Although outside the territory of the empire, some of
the states remained as clients of successive emperors and had close contacts with Rome.
This was a period of rapid development and the growth of one dominant indigenous
people, the Catuvellauni, whose capital lay at Camulodunon (Colchester). Their expan-
sion threatened the adjacent areas, and deposed leaders appealed to Rome to help re-​
establish themselves.

ad 43
The last such leader was Berikos (Verica) who fled to the emperor Claudius leading
the latter to intervene in Britain. Claudius, anxious to obtain prestige through military
success, sent an invasion force of about 40,000 men. The force moved through south-
ern England, and took the principal indigenous centre, Camulodunon (Colchester),
under Claudius’ personal command. Claudius accepted the surrender of eleven British
kings, and was honoured at Rome where he returned immediately. (Dio Cassius, LX;
Suetonius, Claudius, 17; Inscription, CIL V no. 920; Dio Cassius, LX)

ad 43–​60
Following the defeat of the south-​eastern states, campaigning continued in the south-
west under the command of the future emperor Vespasian, who won a series of victories.
xxx

xxx   Introduction

A revolt occurred in East Anglia (in ad 47), but the expansion of the Roman occupa-
tion continued into Wales where it was slowed by the guerrilla tactics of the inhabitants,
and by a revolt in the Pennines. Early in the reign of Nero (ad 54–​68) consideration
was given to withdrawal, presumably because of slow military progress. (Suetonius,
Vespasian, 4; Tacitus, Annals, XII; Suetonius, Nero, 18)

ad 58–​60
Campaigns were undertaken against the Druids on Anglesey. (Tacitus, Annals, XIV)

ad 60
East Anglia rose against the Romans under the leadership of Queen Boudica, with the
destruction of the Roman towns of Colchester, London, and Verulamium. (Tacitus,
Annals, XIV)

ad 61–​71
The crushing of the revolt was followed by a period of consolidation. (Tacitus,
Agricola, XVI)

ad 71–​4
The allied kingdom of the Brigantes in the Pennines became divided internally, so Rome
intervened and conquered the area. (Tacitus, Histories, III, 45 and Agricola, XVII)

ad 74–​7
Expansion of the Roman province continued with the subjugation of south Wales.
(Tacitus, Agricola, XVII)

ad 77–​83
The new governor, Agricola, led a sustained series of campaigns which enlarged the
province substantially. (Tacitus, Agricola, XVIII–​XXXVIII)
 Introduction   xxxi

ad 80
Agricola completed the conquest of Wales and Anglesey, then he turned to the consoli-
dation of northern Britain, before moving into Scotland where he reached the Tay in ad
80. He garrisoned the area up to the Forth-​Clyde isthmus, and dealt with south-​west
Scotland, at the same time suggesting to Rome that Ireland would be easy to conquer.

ad 82
Agricola again advanced northwards up the eastern coastal plain, leading to the estab-
lishment of a garrison as far as the Tay.

ad 83
The great set-​piece battle of Mons Graupius, somewhere in north-​east Scotland, marked
the successful completion of Agricola’s campaigns. At this stage Britain was also circum-
navigated by Agricola’s fleet.

c. ad 87
Following Agricola’s departure, Roman military attention was turned to the Danube,
and as a result troops from Britain were removed, with resultant withdrawal from
Scotland. (Inscription, ILS no. 2719)
The historical sources for the next thirty years are very poor but it is suggested, on the
basis of the distribution of forts and dated pottery, that the army was slowly withdrawn
from most of Scotland and the occupied zone contracted to the area based on a line from
the Tyne to the Solway.

ad 122
Hadrian visited Britain following a military disturbance. The construction of Hadrian’s
Wall, built ‘to separate the Romans from the barbarians’, followed shortly afterwards.
(Scriptores Historiae Augustae (SHA), Hadrian, 5; 11)

ad 139–​42
The new emperor, Antoninus Pius, abandoned Hadrian’s Wall and advanced north-
wards, driving back the barbarians, before constructing a new wall between the Forth
and the Clyde. (SHA, Antoninus Pius, 5)
xxxii

xxxii   Introduction

ad 138–​61
Antoninus Pius may have removed territory from the Brigantes and fought a war in the
Genounian region although this reference may not concern Britain at all. (Pausanius,
Description of Greece, VIII, 43)
The evidence for the following period is ephemeral and ambiguous. The conven-
tional interpretation that the Antonine Wall was abandoned in the mid-150s and
then briefly reoccupied again until 163–​4 has been revised. It is now thought that the
withdrawal back to Hadrian's Wall was piecemeal and took place through the period
between 155 and 163–​4.

ad 181–​4
Tribes invaded across Hadrian’s Wall, but were repulsed. (Dio Cassius, LXXII)

ad 184–​5
The emperor Commodus took the title Britannicus, and issued coins to commemorate
his victory in Britain. (Coins, RIC 437, 440, 451)

ad 185
A mutiny among the troops in Britain led to the downfall of Perennis, the praetorian
prefect in Rome. As a result, Commodus appointed the future emperor Pertinax as
Governor of Britain. (Dio Cassius, LXXII; SHA, Pertinax, 3)

ad 192–​96/​7
Commodus and then his successor Pertinax were assassinated, and in the struggle for
the succession Clodius Albinus, the governor of Britain, was one of three contenders.
For an initial period, Septimius Severus recognized him as his deputy (i.e. Caesar) while
fighting against the third pretender, Pescennius Niger. Once he was defeated, Severus
and Albinus engaged in a civil war, with Albinus taking troops from Britain with him.
Severus won after a major battle outside Lyon. Following his victory Severus divided
Britain into two provinces, probably to ensure that no single governor again had com-
mand of so large an army. (Dio Cassius, LXXIII; Herodian, II, 15; Dio Cassius, LXXV;
Herodian, III, 8)
 Introduction   xxxiii

ad 197–​205
The tribes of the Maeatae and Caledonians in Scotland broke their treaty with Rome
and waged war in the North. Since Roman military concerns lay elsewhere, the Maeatae
were bought off. (Dio Cassius, LXXV)

ad 208–​11
Renewed troubles in Britain provided the opportunity for the emperor Severus to come
in person with his sons to campaign in Scotland. He defeated the Caledonians but this
was followed by a revolt of the Maeatae, and a renewed guerrilla war by the Caledonians.
While preparing for the campaign of 211 Severus died in York. He was succeeded by his
sons, Geta and Caracalla, who terminated the expedition and withdrew from the terri-
tory gained. (Dio Cassius, LXXVI; Herodian, III, 14–​15; Dio Cassius, LXXVII; Herodian,
III, 15)
After 244 the structure of authority within the empire failed, and there was a period
of about forty years during which emperors neither maintained power for long, nor
ensured a peaceful succession. In the period from 259 to 274 Britain formed part of
the Gallic Empire, which had effectively seceded from the remainder of the empire.
Political stability was restored with the accession of Aurelian (27–​5) and consolidated by
Diocletian (from 284), who shared power with Maximian from 286. Within Britain the
events of the mid-​third century up to this period are little known.

ad 286–​9
Carausius was appointed to patrol the English Channel which was infiltrated by ‘barbar-
ians’. His success enabled him to profit from the campaigns, and for exploiting this he
was sentenced to death by the emperor. He responded by declaring himself emperor.
His control of Britain was unchallenged until Maximian launched a naval attack which
seems to have been repulsed. Carausius then occupied parts of northern Gaul. (Aurelius
Victor, 39; Eutropius, IX. 21; Panegyric on Maximian, 11–​12)

ad 293
Constantius, Maximian’s deputy (i.e. Caesar), attacked and took Boulogne.
Carausius was assassinated by one of his ministers, Allectus, who was declared
emperor in Britain and northern Gaul. (Panegyric on Constantius Caesar, 6; Eutropius,
IX, 22)
xxxiv

xxxiv   Introduction

ad 296
Constantius’ army attacked, landing on the south coast, and Allectus was killed.
Constantius was accorded a triumphal welcome in London. (Panegyric on Constantius
Caesar, 13–​20)

ad 305–​6
Constantius, now co-​emperor with Maximian, came to northern Britain to conduct a
military campaign. He was joined by his son Constantine, and reached the far north
of Scotland. While staying at York the emperor died, and Constantine was declared
emperor. (Aurelius Victor, 40; Eutropius, X, 1–​2)

ad 312–​14
Coins show that the emperor Constantine was in Britain. In 315 he took the title
Britannicus indicating that he had won a victory here. (Coins, RIC 133–​41, 142–​3, 144–​5)

ad 342–​3
A winter visit to Britain by the emperor Constans perhaps suggests a crisis here.
(Libianus, Orationes, LIX)

ad 350–​3
A revolt in Gaul toppled Constans, who was replaced by Magnentius. The defeat of the
rebellion was followed by reprisals in Britain. (Ammianus Marcellinus, XIV. 5)

ad 360
Troops were sent to Britain to deal with the Picts and Scots (although the motive may
have been simply to remove the troops from Gaul). (Ammianus Marcellinus, XX. 1)

ad 364
Barbarian raids by Picts, Scots, and Saxons are recorded. (Ammianus Marcellinus,
XXVI, 4)
 Introduction   xxxv

ad 367–​8
A concerted barbarian attack occurred, with the loss of a senior military com-
mander. Count Theodosius was sent to recover the situation and we are told that he
restored the province after a major campaign. (Ammianus Marcellinus, XXVII, 8;
XXVIII.3)

ad 383–​8
Magnus Maximus, a British army commander, led another revolt, killing the emperor
Gratian. Maximus gained control of much of the western empire and ruled it from Trier.
He moved against Italy and was defeated by Theodosius. (Zosimus. IV, 35; 37; Orosius,
VII, 35)

ad 396–​8
The general Stilicho ordered an expedition against the barbarians in Britain. (Claudian,
Stilicho, II, 247–​55)

ad 399
Peace was restored by Stilicho’s expedition. (Claudian, Eutropius, I, 391–​3)

ad 401–​2
Troops were withdrawn by Stilicho to defend Italy. (Claudian, Gothic War, 416–​18)

ad 406–​7
A usurper, Marcus, took power in Britain. Marcus was replaced first by Gratian then by
Constantine III. Constantine was left to defend Gaul against the barbarians. (Zosimus,
VI, 2; Orosius, VII, 40)

ad 408–​9
Britain is attacked by the Saxons. ‘The Britons freed themselves, expelling their Roman
governors and setting up their own administration.’ (Zosimus, VI, 5)
xxxvi

xxxvi   Introduction

ad 410
The emperor Honorius replied to an appeal for help from Britain by telling the cities
to look after their own defence. (There are serious doubts about whether this reference
does refer to Britain: it is more likely to relate to Bruttium in southern Italy.) (Zosimus,
VI, 10)

ad 429
St Germanus visited Britain to counter heresy in the church and defeated a Saxon raid-
ing party. (Constantius, Vita)

ad 435–​7
St Germanus may have visited Britain for a second time. (Constantius, Vita)
The accounts of events after the first decade of the fifth century are difficult to inter-
pret. There were clearly continuing Saxon raids, which resulted in the loss of territory.
By this stage events at Rome prevented any help being sent from outside Britain. It is
likely that from shortly after 400 the government of the province had degenerated, with
effective power moving into the hands of the landed aristocracy who acted as local bar-
ons. From the 440s land was lost to the Saxons piecemeal, as a result of the limited power
of the local lords. Territory in Kent seems to have been lost first, but there remained a
strong sub-​Roman enclave in the west and north which certainly lasted down to the
early sixth century.
 PA RT  I

NAT U R E OF
T H E E V I DE N C E
2
 Chapter 1

Early Studi e s i n
Rom an Bri ta i n
1610 to 1906

Richard Hingley

Introduction: Images of Civilization
and Barbarity

For great part of four hundred years, the Romans occupied this island in
a state of peace and tranquillity: and a colony so fertile, and abounding in
beautiful situations, must have been inhabited by many Roman adventur-
ers, who migrated hither with their families, and built villas or country
seats, where they lived in some degree of opulence and elegance. Even the
Britons of rank might have built houses in the Roman taste. Whenever we
talk of the Romans in Britain, we think of nothing but rapine and hostility.
(Thomas Warton 1783: 59)

This chapter explores the origins of the study of Roman Britain, addressing the period
from the late sixteenth century to the early twentieth. It compares and contrasts the
accounts of Britannia created by William Camden (1610) and Francis Haverfield (1906).
It also reviews some of the variety of other ideas about Roman Britain that developed
between these times, exploring the ways that discoveries of archaeological objects and
sites were used to support and/​or transform a number of semi-​contradictory ideas
about British origins (cf. Hingley 2008a, 2011, 2012a).
This chapter commences with the publication of the first English edition of William
Camden’s seminal work, Britannia (1610). Camden was the first author to provide a
detailed account of the evidence for the Roman province of Britannia, information that
was used to provide the background for his history of Britain up to his own time. It ends
4

4   Richard Hingley

with the publication of Haverfield’s seminal lecture, The Romanization of Roman Britain,
in 1906. This lecture was later published as a small book (Haverfield 1912) that was
republished on a number of occasions (Freeman 2007). In 1907 Haverfield was elected
to the Camden Professorship at Oxford, a post that commemorated William Camden
(Freeman 2007: 164). Haverfield and Camden followed broadly comparable agendas but
in very different historical contexts. Both authors pursued a common theme: addressing
the introduction of ‘civility’ (Camden) or ‘civilization’ (Haverfield) to the native popula-
tion of the British Isles conquered by the Romans.
Despite the considerable difference in the details of the tales told by Camden and by
Haverfield, the common element in the thematic structure of these two highly influential
accounts drew directly upon the Roman writings that had emphasized the introduction
of Roman ways to indigenous Britons under imperial rule. Of particular significance
to both authors was the section of text in Tacitus’ Agricola (21) that described the train-
ing of the sons of British chiefs in Roman ways and their consequential enslavement
(Hingley 2008a: 10). This theme has provided a complex myth that played a significant
role in ideas about the origins of English civilization from the late sixteenth century
to the present day, although the terms in which this debate have been conceived have
by no means remained constant (cf. Hingley 2008b). This origin myth communicated
directly with the classically educated landed elite living to the south of the Antonine
Wall from the late sixteenth century and, as Norman Vance has observed, came to act as
the foundation of Victorian British pride (Vance 1997: 265; cf. Hingley 2010). It helped
to communicate the humble origins of contemporary British greatness and, from the
late sixteenth century to the twentieth, provided ideological support for the conquest
and control of foreign territories incorporated into the British Empire (Hingley 2000,
2008a).
This was not by any means the only myth of origin that was drawn from the Roman
past by the British (for myths of origin, see Samuel and Thompson 1990; Broklehurst and
Phillips 2004; Hingley 2008a: 4). Another powerful image of Roman Britain that domi-
nated in the nineteenth century was soundly dismissed by Haverfield when he observed
that, in ‘Britain, as it has been described by the majority of writers, we have a province in
which Romans and natives were as distinct as modern Englishman and Indian, and the
“departure of the Romans” in the fifth century left the Britons almost as Celtic as their
coming had found them’ (Haverfield 1906: 190; cf. Haverfield 1896: 428–​429). Haverfield
noted that this inaccurate image had arisen as a result of both an over-​reliance by the
Victorians on the writings of Caesar and Tacitus and also ‘the analogies of English rule
in India’. I have defined this image elsewhere as that of the ‘Celtic subaltern and Roman
officer’ (Hingley 2000: 10). In this context, the term ‘subaltern’ refers to representatives
of a perceived inferior race subject to the hegemony of a ruling class (cf. Spivak 1994).
For much of the period that separated Camden from Haverfield, this idea dominated
the perception of the character of Roman Britain, leaving little room for antiquaries to
explore the potential civilizing influence of the Romans on the ancient Britons.
This image of the Celtic subaltern suggested that ancient Britons retained their bar-
barian, or semi-​barbarian, manners throughout the period of Roman rule in Britain;
 Early Studies in Roman Britain: 1610 to 1906    5

it also suggested that Roman officers lived in their Roman stations (forts, towns, and
villas) alongside, but at some remove from, native peoples. Many nineteenth-​century
accounts of Roman Britain had drawn deeply upon the cultural analogy provided by
British rule in India, just as British officers had drawn upon Roman parallels to inform
their actions and policies (Hingley 2008a: 240–​241). Haverfield’s account of Roman
Britain was intended to point out the bias in the British–​India analogy by document-
ing the progressive influence of Roman civilization on the peoples in the south of the
province. Haverfield stressed the common factors that linked the Romanized people of
Britain to populations across the Roman Empire, including urbanism, villas, forts, and
Roman material culture.
Haverfield’s synthetic account of the Roman province mapped ancient British civi-
lization and subservience onto two different geographical areas of the province, the
‘military district’ and the ‘civil district’ (see Hingley 2000; Webster 2001). It elevated the
importance of the civilized Romano-​British populations; stressing the ancestral intro-
duction of civilization to the people of the south and east of the British Isles and empha-
sizing the ancient barbarity of the people of the north and west (cf. Hingley 2008b:
319–​321).

William Camden: Chorography
and British Civility

During the late sixteenth century, a growing appreciation of the value of surviving clas-
sical texts transformed earlier more directly mythical accounts of the early history of
Britain. Central to this new thinking was William Camden’s fundamental contribu-
tion, the first synthetic account of Britain’s Roman past that drew deeply on recently
rediscovered classical texts and material objects. The first edition of his influential vol-
ume, Britannia, was published in Latin in 1586. Subsequent editions published over the
following two and a half decades updated and expanded this review of the surviving
Roman relics across England, Wales, and southern Scotland. Britannia communi-
cated information derived from local informants who recorded and illustrated ancient
sites and objects. The first English edition of Britannia, published in 1610, contained a
detailed account of the Roman province, including the artefacts that had been identi-
fied, Latin inscriptions, and a handful of pre-​Roman coins (Hingley 2008a: 26–​40).
The 1610 edition, subtitled a Chorographicall Description, brought the evidence for
the Roman province into a direct engagement with Camden’s Britain (Hingley 2012a:
8–​9). Chorography is an analytical concept originating in the ancient Mediterranean
world and used by early modern scholars in their accounts of the landscapes of
England. Howard Marchitello observed that chorography delineates ‘topography not
exclusively as it exists in the present moment, but also as it has existed historically’,
since the concept is based on the idea that the character of the land described in
6

6   Richard Hingley

particular places persists through time (Marchitello 1997: 78, 55). By the early seven-
teenth century, chorography had a close relationship with growing notions of landed
property. As a method, it drew upon the history of the past of selected locations to
help to justify the local aristocracy’s lineages and rights to estates (Swann 2001: 101–​
107). By connecting the modern kingdom of England with the civil zone of the Roman
province, Camden constructed a longer ancestry for the civilization and religion of the
Elizabethan and Jacobean English—​an idea that placed the Welsh, Scots, and Irish in a
subservient position.
For Camden and his peers, the Roman history of Britain had a particular relevance as
an ancient context for the introduction of civility and Christianity to Britain. The con-
cept of civility in turn derived from the Latin civilitas, meaning the art of government or
the qualities of citizenship (Bryson 1998: 43–​58). I have discussed the concepts of civil-
ity and civilization/​Romanization in the writings of Camden and Haverfield elsewhere
(Hingley 2008b). It also contributed to a developing Jacobean fixation with exploring
the unity and disunity of the new Great Britain (Hingley 2008a: 53). British policies
in Ireland and North America at this time were informed by ideas derived from these
ancient sources, a developing knowledge that helped to conceptualize and justify colo-
nial exploration by providing models for dominating ‘savages’ and ‘barbarians’ (Hingley
2008a: 60–​66).
This fashion for looking to the Roman past for the origins of contemporary civil-
ity ceased to be popular during the troubled decades of the early seventeenth century,
but the idea was reinvented in the works of the eighteenth-​century antiquaries such
as the Revd William Stukeley (for ‘Augustan’ England, see Ayres 1997: p. xiv). Early
eighteenth-​century society was dominated by a landed aristocracy that drew deeply
on classical Roman models (Ayres 1997: 2–​47). Stukeley’s ideas about the prehistoric
henges of southern Britain, including Stonehenge, are more well known today than his
contribution to Roman studies, which was of equal significance during his own life-
time. Fascinated by the Roman remains of Britain, Stukeley travelled along the Roman
roads of the south, producing a volume entitled Itinerarium Curiosum (Stukeley 1724;
cf. Sweet 2004: 166). This itinerary provided an account of the towns and remains that
lay along the routes, glorifying the surviving remains of Roman civilization across
Britain.
Occasionally Stukeley’s tales about the early origins of Britain communicated the
idea of a continuity of civilization that served to link the classical provincial past with
early eighteenth-​century Augustan England. Very little excavated evidence was avail-
able to Stukeley but, in his account of a place that he calls ‘Mantantonis’ (Chichester), he
describes an inscription that had been found the previous year during the digging of a
cellar in the town (Figure 1.1) (Stukeley 1724: 194). This was the Cogidubnus inscription,
which had already been recognized by another influential antiquary, Robert Gale, to be
of considerable significance. Gale used the inscription to argue that Cogidubnus was a
Roman citizen of British origin, who was ‘Romanized’ and took the name of his benefac-
tor, the emperor Claudius (Gale 1723: 393–​394). Stukeley built on Gale’s account by arguing
that the name ‘Pudens’ in the final line of the inscription referred to a man mentioned in
 Early Studies in Roman Britain: 1610 to 1906    7

Figure 1.1  The Cogidubnus inscription from Chichester, Sussex.

Source: from Gale (1723). Reproduced by permission of Durham University Library.

Martial’s Epigrams (Volume 1: 238-​9). Martial’s Pudens was married to a British woman
called Claudia Peregrina (Martial Epigrams, Volume 2: 277), whom Stukeley assumes to
have been identical to Claudia Rufina. Stukeley drew upon and transformed earlier tradi-
tions to suggest that Claudia was Cogidubnus’ heir and also a Christian (Stukeley 1724:
193; cf. Hingley 2008a: 187–​188). Earlier versions of this legend had claimed that Claudia
Rufina was the daughter of Caratacus and had become a Christian while in Rome with
her father. Stukeley later suggested that Claudia and Pudens invited St Paul to visit them in
Chichester during the mid-​first century ad and that he preached to the local population
(Stukeley 1740: 233).
The reconstruction of the Roman history of Britain inspired Stukeley to reflect on an
ancestral civility that mirrored the growing imperial ambitions of the British elite (Haycock
2002: 119). Stukeley sought to project the Roman remains of Britain into his neo-​classi-
cal present, with the apparent grandeur of these material traces, in turn, reflecting on the
contemporary greatness of Augustan England (Ayres 1997: 96–​97; for the complexities of
the term neo-​classical, see Sachs 2010: 30). Stukeley had access to all the Roman materi-
als studied by Camden, since Britannia was republished in an expanded form in 1722, but
he also recorded a variety of new discoveries. Stukeley was involved in the recording and
illustration of a number of Roman towns, and he also encouraged friends and associates
to uncover and document the remains of several Roman villas, such as Cotterstock and
Weldon in Northamptonshire (Hingley 2008a: 171–17​2). During the 1740s, a local man
called John Stair conducted the first excavations at the Roman city at Silchester and pro-
duced a plan that included the town walls, street system, and central forum (Hingley 2008a:
181–18​4; 2012b).
8

8   Richard Hingley

Roman Settlers and ‘Celtic Subalterns’

These early excavations began to provide evidence for the civil elements of Roman
Britain, including towns and country houses. Important discoveries were reported
at the meetings of the Society of Antiquaries and in regular short papers on Roman
matters in the Philosophical Transactions and the Gentleman’s Magazine. The idea of
the civil province was to be developed further as a result of more extensive excava-
tions of villas and towns conducted at the end of the eighteenth century and during
the early nineteenth. The increasing appreciation of a substantial civilian settlement
across the Roman province was, however, often accompanied by the idea that these
were the homes, not of Romanized Britons, but of Romans who had settled in Britain
from overseas.
How did this idea of two separate populations come about? Until the pre-​Roman
date of ‘Celtic’ metalwork came to be clearly demonstrated during the early nine-
teenth century, the main source for thinking about the ancient Britons was the
classical authors who referred (mostly dismissively) to the semi-​naked and animal
skin-​covered barbarians of ancient Britain (Smiles 1994). In addition, the sixth-​cen-
tury ad writings of the monk Gildas projected the idea of subservient semi-​barbar-
ian and semi-​naked Britons living alongside the Romans, who lost their valour and
fighting spirit before succumbing to the invasion of the fifth-​century Anglo-​Saxons
(Hingley 2012a: 170–​171). Of course, some contrary views idealized the ancient
Britons, including the antiquarian works that tied Stonehenge and the megalithic
monuments in with the Druids (cf. Smiles 1994: 77–​79; Morse 2005: 41–​47). This
image remained highly popular from the sixteenth century to the early twentieth.
Despite Tacitus’ comments about the civilizing (and enslavement) of the Britons
in Agricola (21), it was usually felt that these people did not become particularly
Romanized under Roman tuition.
From the late sixteenth century, antiquaries began to collect and study Latin inscrip-
tions derived from sites within the Roman military frontier zone, including the monu-
ments now known as Hadrian’s Wall and the Antonine Wall. These monuments were
also surveyed and mapped at this time (Hingley 2008a: 110–​114; 2012a). Studies of the
inscriptions derived from Roman sites across northern England and southern Scotland
indicated that individual soldiers had travelled to Britain from different regions of the
Roman Empire. Latin inscriptions from sites in the south of the province, such as the
Cogidubnus example, were far rarer, and the vast majority of named soldiers derived
from overseas (Sweet 2004: 181–​183; Hingley 2008a: 160). The urban sites that devel-
oped in southern Britain in the Roman period often had a military origin, and the Latin
inscriptions referring to these soldiers were often taken to indicate stations occupied by
a military population of incomers. The excavation of Roman towns and villas gradually
led to the interpretation of these buildings as elements of the infrastructure of ‘stations’
occupied by the Roman officers who had settled in Britain as a result of their imperial
and military duties.
 Early Studies in Roman Britain: 1610 to 1906    9

Caesar, Tacitus, and other classical writers had presented accounts of significant
events across the province that fitted with the military emphasis provided by the Roman
inscriptions. As a result, from the seventeenth century to the nineteenth, most antiquar-
ies viewed the Roman occupants of Britain as settlers from overseas who had brought
their Roman identity and culture with them. During the eighteenth century, these ideas
coincided with the increasing militarization of British society that accompanied the
expansion of colonial territories overseas (Hingley 2008a: 161). At this time, a significant
number of antiquaries, including Stukeley, began to develop a fascination with the net-
work of Roman roads and stations across Britain, partly recorded by the Latin itineraries
(Hingley 2008a: 161–​163).
Roman mosaics began to be uncovered and recorded in some numbers during the
early eighteenth century, but were occasionally interpreted as pavements used to floor
the tents of Roman generals (Hunter 1995: 196). More observant antiquaries real-
ized that mosaics were usually associated with substantial buildings, probably villas
(Hingley 2008a: 166–​167, 169–​173). The dominant explanation continued to suggest that
these elaborate buildings represented the homes of Roman generals or Roman gentle-
men from overseas. In 1787, Major Hayman Rooke published an account of the remains
of a substantial Roman building that he had uncovered at Mansfield Woodhouse
(Nottinghamshire). He suggested that the building indicated that ‘the manners of Italy’
had been introduced to Britain by Roman settlers, but did not discuss the idea that these
Romanized individuals could possibly be Britons (Rooke 1787: 375; cf. Hingley 2008a:
235–​236).
Samuel Lysons conducted a remarkable campaign of excavations at the sites of several
Roman villas in southern England between 1789 and 1819. Uncovering the remains of
several substantial buildings, he produced information that led to a reassessment of the
character of Roman culture in Britain. For the first time, remains of the foundations of
buildings comparable to the better-​preserved classical remains of Rome, Pompeii, and
Herculaneum were excavated on a large scale in the British countryside (Hingley 2008a:
247–​253). Lysons inferred that the builders and occupiers of these villas were Roman set-
tlers from overseas, although, in the case of the impressive villa at Bignor, the proximity
of the remains to Chichester caused him to speculate that it might have been the home
of Cogidubnus (Lysons 1815: 219) (Figure 1.2). During the nineteenth century, however,
the concept that the Roman buildings in the cities and villas of Roman Britain were the
homes of Roman officers continued to dominate. Impressive buildings were found at the
Roman towns of Bath, Cirencester, and London during the early and mid-​nineteenth
century, indicating the widespread scale of the Roman investment in Britain, but these
urban centres were usually considered to represent military ‘stations’ with a civilian ele-
ment to their populations (Hingley 2008a: 279–2​83).
Until the late nineteenth century, it was not possible to locate the homes and posses-
sions of the pre-​Roman peoples of Britain. The few coins with inscriptions, including
abbreviated names of several pre-​Roman leaders referred to by classical authors, were
the only pre-​Roman items that antiquaries could identify with any confidence prior
to the nineteenth century (Hingley 2008a: 29–​30). As a result, the image of skin-​clad
10

10   Richard Hingley

Figure 1.2  Plan of a Roman villa excavated by Samuel Lysons at Bignor in Sussex.

Source: Lysons (1815: plate 19). Reproduced by permission of Durham University Library.

semi-​naked barbarian Britons survived well into the twentieth century, when it was
challenged by new information about the settlements and material possessions of ‘Iron
Age’ people (cf. Hingley 2011). In many accounts of the eighteenth and nineteenth cen-
turies, the Britons of the Roman period were described as members of a subservient
population living alongside the Roman settlers. This was an idea that survived into the
twentieth century in some scholarly and popular accounts of pre-​Roman and Roman
Britain (Smiles 1994: 146; Hingley 2012a: 223–22​6). Attested evidence for named ancient
Britons on Roman inscriptions was very rare, and, consequently, most evidence indi-
cated overseas origins for the Romans of Britain.

Contrasting Views of Roman Britain

Although many antiquaries considered that the Roman villas and towns had repre-
sented the homes of Roman settlers from overseas, antiquaries occasionally drew upon
the idea of the civilizing power of Roman rule in Britain. Thomas Warton suggested that
Britons of rank might have built houses in the Roman style (Warton 1783: 59). Warton
emphasized, however, that Roman ‘adventurers’ were mainly responsible for the Roman
 Early Studies in Roman Britain: 1610 to 1906    11

buildings of Britain. Sir Richard Colt Hoare excavated several ancient ‘British villages’
on the chalk downs of Wiltshire, going against the trends of the time by deliberately
exploring these extensive earthwork sites rather than focusing on the excavation of vil-
las and towns (Hingley 2008a: 255). Colt Hoare planned and partly excavated a number
of Roman-​period sites with less impressive remains (Figure 1.3), concluding that it was
‘the wise policy of the Romans to civilize, as well as conquer … after having taking pos-
session of the British settlements, both conquerors and conquered resided together; the
former introducing many arts, comforts and luxuries of life … to which the Britons had
been strangers’ (Hoare 1821: 127).
The potential relevance of the message of the introduction of Christianity to Britain
also caused some Victorian antiquaries and clerics to draw different messages from the
Roman past (Hingley 2008a: 271–​278). Stukeley’s claims for Claudia, Pudens, and St
Paul at Chichester were reinvented and elaborated by John William, the Archdeacon
of Cardigan, while other authors imagined that Claudia was the Christian daughter
of Caratacus or Cogidubnus (Williams 1848; see Vance 1997: 205–​206; Hingley 2008a:
271–27​5). The solicitor and antiquary Henry Coote, who had developed an interest in
Roman Britain, was determined to find a Roman root for British Christianity by draw-
ing on an approach that argued for continuity in the urban centres of the province from
the Roman period to medieval England (Coote 1878). These perspectives tied into a
developing image that supported the concept of the mixed ancestral origins for the con-
temporary population of Britain, including the ancient Britons, Romans, Anglo-​Saxons,
and Normans (Hingley 2000: 91–9​3).
Despite the works of antiquaries such as Colt Hoare and Henry Coote, it remained
difficult to mount a sustained challenge to the idea that the Roman population of Britain
constituted incomers from overseas. Thomas Wright (181–​77) published his popular but
problematic book, The Celt, the Roman and the Saxon, in 1852. Wright (1852: 266–2​71)
portrayed the population of Roman Britain as a collection of distinct races living in their
individual Roman military ‘stations’ among a population of enslaved and downtrodden
Britons. His writing indicates that he considered British peasants and slaves to be geneti-
cally incapable of modifying their ways to accommodate themselves to the civilized
lifestyles of the occupying power. Wright described the cities, villas, and roads, by con-
trast, as being occupied by a series of semi-​independent Roman republics, each derived
from a different part of the empire and with contrasting racial identities (Wright 1852;
cf. Hingley 2008a: 279–2​83). Henry Mengden Scarth published a short book on Roman
Britain in 1883 in which he desperately sought evidence for Christianity in the country’s
early history (Scarth 1883). He struggled to find much archaeological support for the idea
that Roman civilization was transferred to the Britons in any meaningful way, although
he had a greater comprehension of the possibility that Britons might have become
Romanized than many other contemporary accounts (Hingley 2008a: 293). For example,
Bertram Windle (1897: 11) published an account of Roman Britain, in which he observed:

The comparison has justly been made between the Roman occupation of Britain and
our own occupation of India, for in both cases the intention of the conquering race
12

Figure 1.3  Knook Castle and British villages, Wiltshire.

Source: Hoare (1810). Reproduced by permission of Durham University Library.
 Early Studies in Roman Britain: 1610 to 1906    13

has been, whilst firmly holding the dominions of which they have become possessed,
to interfere as little as possible with the natives so long as they were content to submit
quietly to the demands of their conqueror.

One of the most influential works on Roman Britain at the turn of the twentieth
century was Rudyard Kipling’s collection of tales Puck of Pook’s Hill (1906). This pro-
jected a view of the Roman officers in Roman Britain as settlers from overseas or their
descendants. The work had a deep impact on the teaching of Roman Britain in schools
and was recommended as a teaching aid to generations of teachers (Hingley 2012a: 220).
Kipling’s writings indicate that the image of Celtic subalterns and Roman officers was
still current at the beginning of the twentieth century. Nevertheless, academic scholar-
ship was beginning to turn to a different explanation for Roman culture in southern
Britain. Indeed, Haverfield’s contribution to this debate was probably prompted by the
publication of Puck of Pook’s Hill (Rivet 1976: 14).

Romanization: Solving
a Contradiction

By the beginning of the twentieth century, archaeologists were developing a far more
detailed picture of the military works, towns, and villas of Roman Britain, Gaul, and
Germany as a result of a substantial number of new excavations. Knowledge of pre-​
Roman culture was also improving. This accumulating information would gradu-
ally lead to a new view of Roman culture as a transformer of indigenous ways of life.
Francis Haverfield is usually seen as largely responsible for the coherent new idea of
Romanization that arose during the early twentieth century. Haverfield drew deeply
upon the improving knowledge of Roman Britain that had resulted from a number of
recent archaeological projects, including excavations on Hadrian’s Wall, at Aylesford,
Cranborne Chase, and Silchester. He also drew upon the scholarship of the German
ancient historian Theodor Mommsen, who had outlined an approach to Romanization
developed by Haverfield in his own work (Hingley 2008a: 317–​318). In his article of
1906, Haverfield defined the way that Romanization was thought to have operated (cf.
Hingley 2000: 114–​123). He argued that the Roman Empire became fully Romanized and
that ‘the definite and coherent civilization of Italy took hold of uncivilized but intelli-
gent men, while the tolerance of Rome, which coerced no one into conformity, made
its culture the more attractive’ (Haverfield 1906: 188). Discussing the spread of Roman
architecture and culture to the western parts of the empire, he argued that: ‘In mate-
rial culture the Romanization advanced … quickly. One uniform fashion spread from
Italy throughout central and western Europe, driving out native art and substituting a
conventionalized copy of Italian art’ (Haverfield 1906: 188). This new knowledge ena-
bled Haverfield to provide a well-​informed interpretation of the Romanization of the
14

14   Richard Hingley

indigenous inhabitants of the southern and eastern parts of Roman Britain, which he
called the ‘civil district’ (Haverfield 1906: 191–19​4).
Although Haverfield was the first to apply Romanization in a sustained way to
the archaeology of Roman Britain, he was not the first to think about the distinc-
tions between the south and north of the province. The work of the antiquary John
Collingwood Bruce had encouraged generations of northern English antiquaries to
focus on the evidence for Hadrian’s Wall (Hingley 2012a: 200). Bruce drew a distinction
between the Roman ‘camps’ and walls of the north of the province and the ‘cities’ of the
south (Bruce 1860: 343). He also remarked on the ‘comparative security and luxury of
those who were fortunate enough to live in the south’ and noted that no mosaic floors
had been discovered in the three northernmost counties of England (Bruce 1860: 344).
These observations were developed by Haverfield in the definition that he provided of
the military and civil districts (Haverfield 1906: 192) (Figure 1.4).
During the second half of the nineteenth century, excavations in London,
Verulamium, Cirencester, and Silchester indicated that the Roman towns of southern
Britain had complex sequences of deep and sustained occupation (Hingley 2008a: 279–​
293; Hingley 2012b; cf. Hoselitz 2007: 173–17​4). It was realized that the military inscrip-
tions, used by previous generations to provide evidence for the overseas origin of many
‘Romans’ in Britain, dated to the early stages of the military occupation of many sites,
allowing a renewed emphasis upon a potentially indigenous contribution at these loca-
tions. Research on Hadrian’s Wall and the Antonine Wall indicated that the regular
‘stations’ along the lines of these two frontier works represented military forts rather
than civilian or partially militarized cities (Hingley 2012a: 196–​199). Initially, the central
building in the Roman fort at Chesters on Hadrian’s Wall was interpreted as the forum
of a very small classical city (Figure 1.5). During the final years of the nineteenth century,
however, it was argued that buildings of this type actually represented the headquarters
buildings of the Roman forts that occurred at frequent intervals along Hadrian’s Wall.
The excavation of forts in northern England, Scotland, and on the German Limes
during the final years of the nineteenth and the early twentieth centuries led to an
improved understanding of the military organization of the northern parts of Roman
Britain (Hingley 2012a). In the south, the extensive excavations at the Roman town
of Silchester were to prove equally vital to the changing new perspectives outlined by
Haverfield. Extensive excavations were undertaken at the site from 1864 to 1878 by James
Gerald Joyce and from 1890 to 1909 by the Society of Antiquaries. This work led to a
gradually evolving knowledge of the archaeology of this Roman city (Hingley 2008a:
287–​289, 302–​6), providing very little evidence to support the idea of a Roman military
population at any period of its history. Observing the irregular plan of the town and
the variation of the house plans from those of Roman Italy, Haverfield (1894) proposed
that Silchester represented ‘a native copy of a Roman town, such as occurs in countries
ruled by a nation of higher civilization than the subject race’. Using this new approach
to the Romanization of southern Britain, many urban centres were reinterpreted as
civil centres of local self-​rule, the outcome of the transformation of pre-​Roman tribes
into Roman civitates. Early military occupation at some of these sites was explained as
 Isurium
EBURACUM LEG VI and Col

Buxton
DEVA Lindum
LEG XX Col.

Viroconium Leic

ISCA Glos.Col Camulodunum

LEC Verulam Col.
Ciren. Mun
LONDINIUM
Bath Silch.
Winch.
Chich.
Exeter

Bltiens HADRIANS WALL

EBURACUM LEG VI and Col.

Lindum
DEVA
Col
LEC XX E
R
ON SHO

Viroconium

Camulodunum
X

ISCA
A

S
LEG Othona
LONDINIUM
E
OR
H

Anderida S
ON
SAX

Figure 1.4  The civil and military districts.

Source: Haverfield (1912: figure 1). Reproduced by permission of Durham University Library.
16

16   Richard Hingley

Figure 1.5  The Roman site at Chesters (Cilurnum) showing the excavated areas and ‘forum’.
Source: Budge (1903: opposite p. 98). Reproduced by permission of Durham University Library.

conquest period activity, the military units later moving further north and west to estab-
lish and occupy the military zone.
The concept of Romanized Britons was also becoming popular outside the urban
arena. In an important publication of the ‘Late Celtic’ pottery from the Aylesford ceme-
tery in Kent, the archaeologist Arthur John Evans argued that the pre-​Roman ceramics
showed ‘Romanizing influences’—​a phenomenon that was also evident in the con-
temporary coinage (Figure 1.6) (Evans 1890: 351, n.c. 356, 383). Evans and Haverfield
 Early Studies in Roman Britain: 1610 to 1906    17

Figure 1.6  Sketch plan of the graves forming a ‘family circle’ at Aylesford, Kent.

Source: Evans (1890: figure 4). Reproduced by permission of Durham University Library.

were both influenced by the excavations undertaken by Pitt-Rivers when he explored

the homes of people he called ‘Romanised Britons’ on Cranborne Chase (Pitt-Rivers
1888: 65; cf. Hingley 2008a: 298). Pitt-Rivers’s innovative fieldwork drew upon Hoare’s
earlier studies and was, in turn, influential in identifying the adoption of Roman pot-
tery and personal ornaments by Britons who did not appear from the excavated set-
tlement evidence to have had a particularly elevated social status (see also Swift, this
volume, for further discussion on the development of Romano-​British artefact studies).
The concept of Romanization was not used in Britain prior to the final decades of
the nineteenth century, although authors since the early seventeenth century had some-
times adopted the term ‘Romanized’ (Hingley 2008b). I have argued that the concept
of Romanization focused attention onto a much more directly evolutionary interpreta-
tion of Roman identity and cultural change that emphasized progress (Hingley 2000).
Haverfield explored this idea in detail in 1906 to provide an account of how indigenous
peoples in the lowland civil areas of Roman Britain could gradually adopt Roman ways.
The south of the province was then thought to have become fairly fully Romanized, with
even the peasants adopting Roman styles of pottery, artefacts, and building (Haverfield
1906: 198). The elite were seen as administrators of the towns who lived in the villas exca-
vated across the lowlands.
In the military zone, which covered much of Wales and central Britain (northern
England and southern Scotland), evidence for Roman culture was generally found only
18

18   Richard Hingley

on Roman military sites (Haverfield 1906: 191–192). In this region, villas were a very
rare occurrence, and towns were scarce. Where urban centres did occur, evidence was
found for continued military associations. The homes of the indigenous people in the
military districts usually appeared to change relatively little as a result of Roman control.
Excavation work uncovered round-houses, with Roman pottery and material culture
rarely occurring. Consequently, the idea of Celtic subaltern and Roman officer contin-
ues to be popular for the military areas of Roman Britain (Hingley 2004).

Conclusion: Ancestral Tales

The material remains that were available when Camden, Stukeley, and Haverfield wrote
their accounts impacted deeply on their interpretations of Roman Britain, but the mean-
ing of the Roman past had also been transformed by historical circumstances. Although
the remains of Roman sites had been uncovered when Camden was writing, the tech-
niques of excavation and site recording were generally unknown, and there was very
little comprehension of the variety of Roman site types across the Roman Empire. The
only material objects that enabled Camden to tell stories about pre-​Roman and Roman
Britain were the Latin inscriptions and coins that had been found over the centuries.
Haverfield had access to an additional three centuries of information from the survey-
ing and excavation of archaeological sites and the study of inscriptions and classical
texts. The careful excavation that had been undertaken at Roman sites in Britain and
across Europe had identified a variety of different types of site, including towns, forts,
villas, temples, industrial sites, and rural settlements. Also available by this time were
the initial results of work on the chronology of ‘Late Celtic’ (late Iron Age) and Roman
pottery.
My previous work has explored the ways that ideas outlined by Camden, Stukeley,
Haverfield, and others were used to define the national and imperial origin of Britain
and the British (Hingley 2000, 2008a, 2011). This chapter has adopted a different
approach by suggesting that Haverfield helped to establish an intellectually coher-
ent and well-​informed account of the archaeology of Roman Britain, a body of work
that challenged earlier understandings. This is why his work had so much impact upon
later archaeologists (for further discussion of the development of Romano-​British
archaeology after Haverfield, see Millett, this volume). Consequentially, the concept of
Romanization remained popular at least until the 1990s; indeed, ideas about the pro-
gressive character of Roman cultural change remain influential in the second decade
of the twenty-​first century. The establishment of the model for the civil and military
districts also had a major impact on twentieth-​century archaeology (Hingley 2004;
although see James 2001).
Many of the archaeologists working on Roman Britain have now rejected the inher-
ently progressive interpretations propounded by Romanization theory. It is important,
however, to see Haverfield’s work in the context of its time. He was clearly reacting to
 Early Studies in Roman Britain: 1610 to 1906    19

the influential origin myth that drew an analogy between British India and the Roman
rule of Britain. Haverfield’s perspective built upon, contradicted, and transformed
concepts of Roman Britain, developing and giving new attention to an interpretation
that had originated with classical writings and had been worked on during the late six-
teenth century by William Camden. Haverfield set an agenda that has only recently
been challenged by authors who have produced accounts of Roman Britain that adopt
a range of different perspectives (including James and Millett 2001, Webster 2001, and
Mattingly 2006).

Acknowledgements

I am very grateful to the Arts and Humanities Research Council and to Durham
University for the research leave that initially enabled me to develop these arguments
(see Hingley 2008a: p. vii). I am also grateful to the editors of this volume for the invi-
tation to write this chapter and to Christina Unwin for help with editing and with the
illustrations. Hingley (2008a) includes a fuller account of the topics covered in this
chapter. The archaeology of Hadrian’s Wall is not covered in detail here, but is addressed
in Hingley (2012a). For other writings about the antiquarian and archaeological study
of Roman Britain during the period covered in this chapter, see Hingley (2000), Sweet
(2004: 155–188), Freeman (2007), and Hoselitz (2007).

References
Primary Sources
Martial, Epigrams, Volume I, trans. D. R. Shakleton Bailey. Loeb Classical Library. Cambridge,
MA: Harvard University Press, 1993.
Martial, Epigrams, Volume III, trans. D. R. Shakleton Bailey. Loeb Classical Library. Cambridge,
MA: Harvard University Press, 1993.
Tacitus, De Vita Agricolae, ed. R. G. Ogilvie and I. Richmond. Oxford: Clarendon Press, 1976.

Secondary Sources
Ayres, P. (1997). Classical Culture and the Idea of Rome in Eighteenth Century England.
Cambridge: Cambridge University Press.
Broklehurst, H., and Phillips, R. (2004). History, Nationhood and the Question of Britain.
London: Palgrave Macmillan.
Bruce, J. C. (1860). ‘ … observations … associated … with the proceedings of an archaeological
week’, Archaeological Journal, 17: 342–​347.
Bryson, A. (1998). From Courtesy to Civility:  Changing Codes of Conduct in Early Modern
England. Oxford: Oxford University Press.
Budge, E. A.  W. (1903). An Account of the Roman Antiquities Preserved in the Museum at
Chesters, Northumberland. London: Gilbert and Rivington.
20

20   Richard Hingley

Camden, W. (1610). Britannia, or a Chorographicall Description of the Most Flourishing

Kingdomes, England, Scotland, and Ireland., trans. into English by Philémon Holland.
London: Georgii Bishop and Ioannis Norton.
Coote, H. C. (1878). The Romans of Britain. London: Frederic Norgate.
Evans, A. J. (1890). ‘On a Late-​Celtic Urn-​Field at Aylesford, Kent, and on the Gaulish, Illyro-​
Italic and Classical Connexions of the Forms of Pottery and Bronze-​Work there Discovered’,
Archaeologia, 52: 315–​388.
Freeman, P. W. M. (2007). The Best Training-​Ground for Archaeologists: Francis Haverfield and
the Invention of Romano-​British Archaeology. Oxford: Oxbow Books.
Gale, R. (1723). ‘An Account of a Roman Inscription Found at Chicester’, Philosophical
Transactions, 32: 391–​400.
Haverfield, F. (1894). ‘The Foundation of Silchester’, Athenaeum, 3503 (15 December): 836.
Haverfield, F. (1896). ‘Early British Christianity’, English Historical Review, 43: 417–​430.
Haverfield, F. (1906). ‘The Romanization of Roman Britain’, Proceedings of the British Academy,
2: 185–​217.
Haverfield, F. (1912). The Romanization of Roman Britain: New Edition, Revised and Enlarged.
Oxford: Clarendon Press.
Haycock, D. B. (2002). William Stukeley:  Science, Religion and Archaeology in Eighteenth
Century England. Woodbridge: Boydell Press.
Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman
Archaeology. London: Routledge.
Hingley, R. (2004). ‘Rural Settlement in Northern Britain’, in M. Todd (ed.), A Companion to
Roman Britain. Oxford: Blackwell, 327–​348.
Hingley, R. (2005). Globalizing Roman Culture: Unity, Diversity and Empire. London: Routledge.
Hingley, R. (2008a). The Recovery of Roman Britain 1586–​ 1906:  ‘A Colony so Fertile’.
Oxford: Oxford University Press.
Hingley, R. (2008b). ‘Not so Romanized? Tradition, Reinvention or Discovery in the Study of
Roman Britain’, World Archaeology, 40: 428–​444.
Hingley, R. (2010). ‘“The Most Ancient Boundary between England and Scotland”: Genealogies
of the Roman Walls’, Classical Reception Journal, 2/​1: 25–​43.
Hingley, R. (2011). ‘Iron Age Knowledge: Pre-​Roman Peoples and Myths of Origin’, in T. Moore
and X.-​L. Armada (eds), Atlantic Europe in the First Millennium bc: Crossing the Divide.
Oxford: Oxford University Press, 617–​637.
Hingley, R. (2012a). Hadrian’s Wall: A Life. Oxford: Oxford University Press.
Hingley, R. (2012b). ‘“A place that the lover of antiquity will visit with great delight”: From
Caer Segonte to Calleva Atrebatum’, in M. Fulford (ed.), Silchester and the Study of Romano-​
British Urbanism. Journal of Roman Archaeology Supplementary Series 90. Portsmouth,
RI: Journal of Roman Archaeology, 23–​40.
Hoare, R. C. (1810). The History of Ancient Wiltshire: Part I. London: William Miller.
Hoare, R. C. (1821). The Ancient History of Wiltshire: Volume II. London: Lackington, Hughes,
Harding, Mavor and Lepard.
Hoselitz, V. (2007). Imagining Roman Britain:  Victorian Responses to a Roman Past.
Woodbridge: Boydell and Brewer.
Hunter, M. (1995). Science and the Shape of Orthodoxy: Intellectual Change in Late Seventeeth-​
Century Britain. Woodbridge: Boydell Press.
James, S. (2001). ‘Romanization and the Peoples of Britain’, in S. Keay and N. Terrenato (eds),
Italy and the West: Comparative Issues in Romanization. Oxford: Oxbow Books, 187–​208.
 Early Studies in Roman Britain: 1610 to 1906    21

James, S., and Millett, M. (2001) (eds). Britons and Romans:  Advancing an Archaeological
Agenda. York: Council for British Archaeology.
Kipling, R. (1906). Puck of Pook’s Hill. London: Macmillan and Co.
Lysons, S. (1815). ‘Account of the Remains of a Roman Villa, Discovered at Bignor, in Sussex, in
the Years 1811, 1812, 1813, 1814 and 1815’, Archaeologia, 18 (1817), 203–​221.
Marchitello, H. (1997). Narrative and Meaning in Early Modern England: Browne’s Skull and
Other Histories. Cambridge: Cambridge University Press.
Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire. London:
Allen Lane.
Morse, M. (2005). How the Celts Came to Britain:  Druids, Ancient Skulls and the Birth of
Archaeology. Stroud, Tempus.
Pitt-​Rivers, A. L. F. L. F. (1888). Excavations in Cranborne Chase, near Rushmore, on the Borders
of Dorset and Wilts, 1880–​1888: Volume 2. London: Harrison and Sons.
Rivet, A. L. F. (1976). Rudyard Kipling’s Roman Britain: Fact and Fiction. Keele: Keele University
Library.
Rooke, H. (1787). ‘An Account of the Remains of Two Roman villae Discovered near Mansfield
Woodhouse, in May and October, 1786’, Archaeologia, 8: 363–​376.
Sachs, J. (2010). Romantic Antiquity:  Rome in the British Imagination. Oxford:  Oxford
University Press.
Samuel, R., and Thompson, P. (1990). The Myths We Live By. London: Routledge.
Scarth, H. M. (1883). Early Britain: Roman Britain. London: Society for Promoting Christian
Knowledge.
Smiles, S. (1994). The Image of Antiquity:  Ancient Britain and the Romantic Imagination.
London: Yale University Press.
Spivak, G. C. (1994). ‘Can the Subaltern Speak?’, in P. Williams and L. Chrisman (eds), Colonial
Discourse and Post-​Colonial Theory: A Reader. London: Harvester, 66–​111.
Stukeley, W. (1724). Itinerarium Curiosum. Or, an Account of the Antiquities and Remarkable
Curiositys in Nature or Art, Observed in Travels throu’ Great Britain. London: Printed for the
author.
Stukeley, W. (1740). ‘30 Aug., 1740 [diary entry]’, in W. C. Lukis (ed.), The Family Memoirs of
the Rev. William Stukeley MD and the Antiquarian and Other Correspondences of William
Stukeley and Roger and Samuel Gale. London: Surtees Society, 232–​233.
Swann, M. (2001). Curiosities and Texts: The Culture of Collecting in Early Modern England.
Philidelphia: University of Pennsylvania Press.
Sweet, R. (2004). Antiquaries:  The Discovery of the Past in Eighteenth Century Britain.
London: Hambledon and London.
Vance, N. (1997). The Victorians and Ancient Rome. Oxford: Blackwell.
Warton, T. (1783). Specimen of a History of Oxfordshire: The Second Edition. Oxford: J. Nichols,
J. Robson, C. Dilly.
Webster, J. (2001). ‘Creolizing the Roman Provinces’, American Journal of Archaeology,
105: 209–​225.
Williams, J. (1848). Claudia and Pudens: An Attempt to Show that Claudia, Mentioned in St
Paul’s Second Epistle to Timothy, was a British Princess. London: Privately printed.
Windle, B. C. A. (1897). Life in Early Britain. London: David Nutt.
Wright, T. (1852). The Celt, the Roman, and the Saxon: A History of the Early Inhabitants of
Britain, down to the Conversion of the Anglo-​Saxons to Christianity. London: Arthur Hall
Virtue and Sons.
22

Chapter 2

Roman Brita i n si nc e
Haverf i e l d

Martin Millett

Introduction

There has been a considerable amount written in ten years concerning the history of
the archaeology of Roman Britain, with a principal focus on the historical context of the
‘Romanization’ and its relationship with Britain’s own imperial past. Most interest has
focused on the later nineteenth and earlier twentieth centuries, in particular the role of
Francis Haverfield, who can be seen in many senses as the father of the modern disci-
pline (Hingley 2000; Freeman 2007). Rather less attention has been paid to the devel-
opment of the subject since the First World War, or to the trends in thought and the
organization of research that led to the enormous growth of the discipline in the later
twentieth century (see, however, Jones 1987; Freeman 2007: 581–​605; Fulford 2007). In
this chapter I would like to explore two complementary issues in order to provide some
historical context for the rest of the volume. First, who led research and how they were
they networked? Although there is some danger this just provides a catalogue of ‘great
men’ (and this is too often gendered), it would be a mistake to underestimate the impor-
tance of personal influence, teacher–​pupil links, and fieldwork-​based friendships in the
development of the subject. Second, how has the subject changed over the last 100 years
or so and how does this relate to a broader historical context?

Early Twentieth-​C entury Foundations

Although there is a long history of antiquarian interest in Roman Britain, the origins
of the present subject lie very much in the second half of the nineteenth century. Two
key trends can be identified: first, an increase in the amount of evidence uncovered as
 Roman Britain since Haverfield    23

a result of industrialization and the infrastructural expansion of the Victorian period.

Second, the growth of a general interest in archaeology, combined with the accumula-
tion of wealth by some that led to a dynamic period in archaeological exploration, with
sites such as Silchester, Caerwent, Corbridge, and Wroxeter becoming the subject of
substantial excavations, which were aimed at enhancing understanding of the Roman
period. Although these excavations were methodologically crude by modern standards,
the data accumulated have been of long-​term significance in feeding further interest
and subsequent research. Equally, the Victorian and Edwardian impulse towards grand
projects also led to the first systematic attempts at the collation of the archaeological
evidence—​for instance, in the Victoria County Histories. This gathering-​together of evi-
dence forms the foundation for modern understanding of Roman Britain.
In the earlier twentieth century there were comparatively few people active in research
on Roman Britain, and their connections, both personal and educational, are compara-
tively straightforward. Thus, as Camden Professor of Ancient History at Oxford, Francis
Haverfield knew all those engaged in the subject—​some of whom had been students of
his at Oxford—​although, of course, archaeology was not then a subject taught in its own
right, so they were generally classicists or historians. Haverfield also utilized a strong
network of local contacts, as reflected especially through his authorship of a periodic
series of reviews of work on Roman Britain and his later annual reviews of fieldwork
published by the British Academy in 1913 and 1914. The latter were predecessors to the
annual survey published in the Journal of Roman Studies from 1921, which was funded
by the Haverfield Bequest.
The large-​ scale excavations undertaken in this period included those run at
Corbridge with input from Haverfield (Bishop 1994). Elsewhere, the enormous pro-
ject to excavate the whole of Silchester was organized by the Society of Antiquaries of
London, which published the results in its journal Archaeologia. It also initiated publica-
tion of its Research Reports series in 1913, to publish the excavations at Wroxeter. These
volumes provide the first monograph series devoted to archaeology in Britain and were
largely devoted to the Roman period in their earlier years, publishing reports on a series
of key projects. There was also important work done on sites in Scotland at this period,
most notably the large-​scale excavations at Newstead, which were magnificently pub-
lished (Curle 1911). George MacDonald, who went on to be a leading figure in the subject
after the First World War, wrote The Roman Wall in Scotland (1911, revised 1934), provid-
ing the first systematic synthesis of the archaeology of the Roman frontier in Scotland.
Haverfield’s own most enduring written contribution was his Romanization of
Roman Britain, first delivered as a lecture to the (newly founded) British Academy and
published in its Proceedings in 1906, subsequently appearing as a short book in 1912 (see
also Hingley, this volume). This provided a comprehensive and accessible explanation
for the incorporation of Britain into the Roman world, based on ideas about cultural
change in the indigenous peoples rather than the implantation of Roman colonists.
Although very heavily influenced by contemporary views of ‘civilization’, Haverfield’s
ideas influenced the research agenda and academic debate for much of the following
century.
24

24   Martin Millett

The period leading up to the First World War was one of considerable investment in
research on Roman sites and shows a dynamism that is difficult now to appreciate fully
because of the way in which leaders of the upcoming generation, like G. L. Cheesman,
were killed in the conflict of 1914–​18. The tragedy of the loss of the leaders of that gen-
eration was said to have broken Haverfield, who himself died in 1919 (Haverfield and
MacDonald 1924: 31–​32; D. B. Webster 1991: 120; cf. Freeman 2007, 348–​424). We are left
to speculate how the subject might have been different had some of his pupils survived
to build on his intellectual foundations.

The Interwar Years

The period after 1918 saw two key trends in the study of Roman Britain. First, despite
the diminution of the enormous disposable wealth of the Edwardian era, there was a
gradual acceleration in archaeological field research, characterized by a sequence of
key excavations designed to elucidate the history of the province. These research-​based
excavations were important in advancing field methods as well as providing an increas-
ing resource for historical synthesis. Equally, we may note that the subject became more
diversified, with dynamic research taking place in a number of different regions.
The post-​First World War generation was small in number but also closely connected,
with many also Oxford-​linked. The leading intellectual figure was undoubtedly R. G.
Collingwood, an Oxford student of Haverfield’s, who, while professionally an academic
philosopher (becoming Waynfleet Professor of Philosophy at Oxford), also made a
very substantial contribution to the study of Roman Britain, both through his general
writings—​which as elegant literary works have yet to be surpassed—​and through his
work on the fundamental catalogue of inscriptions (The Roman Inscriptions of Britain).
His intellectual influence on archaeology was very substantial too. As explained in his
autobiography, he held that knowledge was based on the logic of question and answer,
with information being only the result of the questions asked (Collingwood 1939). His
work, together with younger scholars, on sorting out the history and chronology of
Hadrian’s Wall (which was the subject of intense research through the 1920s and 1930s)
exemplified this approach, with small-​scale excavations designed to answer particular
questions and the results presented in a tight and disciplined intellectual framework.
Such fieldwork built directly on the traditions of keyhole excavation on Hadrian’s Wall,
which had been developed in Cumberland from 1894 to 1903.
Collingwood’s general writings included a key and enduring practical manual (The
Archaeology of Roman Britain, 1930) and the first serious synthesis on the province
Roman Britain and the English Settlements (volume 1 of the Oxford History of England,
published 1936, with J. N. L. Myers writing the English Settlements section). This set the
stage for many subsequent works, providing a synthesis that drew together information
from classical texts and archaeology to provide a narrative history of the province, com-
plemented by chapters that dealt with themes such as towns and villas. Although later
 Roman Britain since Haverfield    25

scholars have often been critical of his conclusions, it is remarkable how most, down to
the present, have replicated this format. Interestingly, Collingwood—​himself brought
up in a family active in the Arts and Crafts Movement—​was proudest of his chapter on
art in this book (Collingwood 1939: 144). Although a piece of polemical writing that has
attracted a lot of criticism, this has also provided the intellectual template for approaches
to the subject down to today.
The second key figure of the period was Mortimer Wheeler, a man of enormous
energy and organizational ability, which he combined with a flair for publicity (Wheeler
1955; Hawkes 1982). Wheeler’s excavating career began in 1913 and included small-​scale
excavations on the Balkerne Gate at Colchester, underneath a pub floor, while he was a
soldier in 1917. After demobilization, he went on to increasingly bold projects, first in
Wales, where he became director of the National Museum, and later in England, with
notable work at Lydney, Verulamium, and Maiden Castle. Until her premature death
in 1936, he worked jointly with his wife, Tessa Verney Wheeler, who made a fundamen-
tal contribution to the projects (Carr 2012). Mortimer Wheeler’s excavations combined
clear strategic direction with rapid publication, ensuring that the sites had an immedi-
ate impact on thinking. Hence, it is immediately obvious that their Verulamium work
was immensely influential on Collingwood’s account of town in Roman Britain and the
English Settlements (Collingwood and Myres 1936). However, although Wheeler did
write synthetic works of his own—​being responsible for Prehistoric and Roman Wales
(1925) and key works on Roman London during his tenure of the Directorship of the
London Museum—​they were not his forte. Wheeler was driven by grand narrative, and
his excavations were designed to elucidate a story, with the result that they sometimes
cut corners, especially in the presentation of the finds like pottery. Interestingly, his doc-
torate was awarded for work on Roman pottery in the Rhineland, among other things
(Wheeler 1920). Nevertheless, he was a pioneer of large-​scale and systematic fieldwork
in a manner that had not previously featured in the study of Roman Britain. Although
Wheeler worked with a number of people who later made a significant contribution to
the subject, he did not himself develop an active school of research students, even fol-
lowing his foundation of the Institute of Archaeology in London in 1937. Not only did
this enterprise put archaeological education at the centre of things, but, unusually, it
integrated Roman studies with the rest of archaeology (Potter 1987). Wheeler’s work
certainly raised the standard of excavation, and those who had dug with him spread his
methods, so that they dominated in the middle twentieth century.
By contrast, Collingwood was the mentor of the new generation of researchers who
followed his agenda—​the logic of question and answer—​in using small-​scale excavation
and fieldwork to establish the history of Hadrian’s Wall. Most notable was Eric Birley,
who was appointed to a post at Durham University in 1931. He was initially based at its
outpost, Armstrong College in Newcastle (later King’s College then, in 1963, to become
the University of Newcastle-​upon-​Tyne). Through the late 1920s and 1930s work on
Hadrian’s Wall blossomed, and the foundations of what is often known as the Durham
School developed, with a focus on both targeted excavation to establish the history of the
frontier and the meticulous use of epigraphic material (James 2002). Eric Birley bought
26

26   Martin Millett

land around Hadrian’s Wall, including the fort site of Chesterholm (Vindolanda); it is
through this that his son Robin Birley developed the excavations from the 1970s, which
produced the famous writing tablets.
The principal focus of archaeological work in Roman Britain through this period
remained military history and urban sites, with comparatively little interest being taken
in the countryside, in the archaeology of ordinary people, or in art. There were excep-
tions to this, such as the excavations on villas such as Bignor (Winbolt and Herbert 1930)
and Folkestone (Winbolt 1925), or the description of art objects in museum publications
(e.g. British Museum 1922; Wheeler 1930); but, being outside the mainstream, they did
not attract much attention. There was also a series of major excavations that continued
the pre-​First World War traditions. J. P. Bushe-​Fox was the key excavator, a man of con-
siderable skill who first dug with Sir William Finders Petrie in Egypt, then at Corbridge,
taking a key role in the study of the pottery there. He went on to dig at Wroxeter and
Hengistbury Head before the First World War and then undertook the key excavations
at Richborough from 1922 to 1939. Elsewhere, for instance, at Wroxeter and at Caistor-​
by-​Norwich, some rather poor quality work was done in the same tradition.
Another important departure that presaged later development was the growth of res-
cue archaeology in advance of construction work. There had long been archaeological
work undertaken in response to building, and indeed, in London, the London Museum
had employed a field officer to record sites since the 1920s (Fulford 2007: 360–​361). Such
work developed with projects like the construction of the Colchester bypass in 1930–​1,
where Christopher Hawkes—​who had first dug with Wheeler and had been taught by
Collingwood at Oxford—​led a major campaign of excavations, exploring the late Iron
Age settlement at Camulodunum. Although the publication of his major report was
delayed until after the Second World War (and then in a much attenuated form because
of the paper shortage), this proved a model for later work where excavations were
undertaken in response to redevelopment work, as happened with the construction of
airfields during the war.

The 1940s–​1960s

In the period following the Second World War there was a gradual growth in the study
of Roman Britain, which gathered pace through the 1950s and into the 1960s. The single
biggest innovation through this period was the vast growth of archaeology beyond the
purely academic sphere. This involved both excavation in advance of construction work
and its growth as a leisure pursuit.
Research-​based traditions were complemented both by university-​run training exca-
vations and a rash of post-​war ‘rescue excavations’. All continued within a frame of refer-
ence still largely dominated by the desire to write narrative history. A key development
was the excavation of occupied urban sites (unlike the greenfield sites such as Silchester,
previously explored). Bomb damage to historic towns in England led to efforts to
 Roman Britain since Haverfield    27

organize excavations to recover evidence of their early history. This difficult work was
led by Aileen Fox at Exeter, W. F. Grimes in London, Kathleen M. Kenyon in Southwark,
and S. S. Frere at Canterbury. It was revolutionary, not only in changing understanding
of the history of the towns but in beginning to create a new route into archaeology. The
excavations were comparatively poorly funded and organized through local commit-
tees, relying to a greater or lesser extent on volunteer labour. The techniques of deep
urban excavation had to be developed and specialist teams brought together to study
the finds. This pattern of organization was spread to other sites, with Frere subsequently
moving on to major excavations in St Albans in the 1950s–​60s and individuals such as
John Wacher and Alan McWhirr going on from there to undertake the excavations at
Catterick and in Cirencester. The Verulamium excavations led by Frere were the most
influential. As well as the high quality of results from his meticulous excavations, which
provided a new model for understanding the origins and endings of towns in Roman
Britain, many of the people who worked there went on to make major contributions
to the study of Roman Britain. Such summer excavations were not only key training
grounds but created a closely networked generation of younger archaeologists with a
passion for the subject and a hunger for large-​scale excavation, as well as the knowledge
and experience to undertake it.
These excavations were rooted in rescue archaeology but developed research agendas,
as did the field training schools of the period. Important also were big excavations at
Fishbourne and Winchester, led by Barry Cunliffe and Martin Biddle respectively, and
later, too, the pair of contrasting Birmingham University summer schools at Wroxeter
run by Graham Webster (from 1955 to 1985) and Phil Barker (from 1966 to 1990). The
Wroxeter excavations were heavily tied into the long-​standing tradition of liberal adult
education, in which archaeology took a key role in the 1960s and 1970s, bringing in keen
amateurs and training them through participation in major projects.
This blossoming of the archaeology of Roman Britain is reflected in a different way
with developments in the north. Eric Birley had a most distinguished war record, with
service in 1939–​45 at Bletchley Park, where he used the methods developed for the study
of Roman military epigraphy to interpret intelligence material relating to the battle
order of the German army. Under his leadership the period after 1945 saw the study of
Roman Britain flourish at Durham but with an emphasis closely linked to continen-
tal scholarship (James 2002). Significantly, Birley was instrumental in establishing the
Limes Congress (first held in Durham in 1949), which linked him and his students
with others working on the frontiers—​primarily in the German academic tradition—​
creating a broader tradition of Limesforschungen in Britain. The annual training exca-
vation at Corbridge was one of the few places where students could come to learn the
subject. Through these summer seasons from 1947 to 1973 generations of students
formed enduring networks.
Excavations generated significant volumes of finds that demanded study. Pottery, in
particular, was identified as key in dating sites, a matter fundamental to the questions
of historical reconstruction with which excavation was primarily concerned. Ceramic
chronology, through the study of samian ware, had long been promoted by Eric Birley
28

28   Martin Millett

(through whose initiative Stanfield and Simpson’s Central Gaulish Potters came to
fruition in 1958). By contrast, coarse pottery was comparatively poorly understood.
Important work was promoted on Bushe-​Fox’s excavations, but the first major typo-​
chronological studies were undertaken in the late 1930s and 1940s. The publication of
Kathleen Kenyon’s study in the Jewry Wall volume (1948) and Christopher Hawkes and
M. R. Hull’s work on the pottery from the pre-​war excavations at Camulodunum (1947)
had set the agenda for such studies in southern England—​the latter in particular link-
ing to continental literature. In Newcastle, John Gillam, a key figure in the Corbridge
excavations and a student of Birley and Richmond, developed this field of study with
the publication of his Types of Roman Coarse Pottery Vessels in Northern Britain, first
published in Archaeologia Aeliana in 1957, and subsequently revised as a book in 1968.
Gillam’s typological approach, like that reflected in Stanfield and Simpson’s work on
samian ware, built on the historical chronology of the northern military sites, which
had been established by previous excavations and created a monolithic framework that
proved enduring, notwithstanding later changes in the dating of sites upon which it
was based.
The towering figure of the period was undoubtedly Ian Richmond, an Oxford
graduate who had been sent by Collingwood to dig with Wheeler at Segontium
in 1922. He was briefly Director of the British School at Rome (BSR) (1930–​2),
and taught at the University of Durham’s King’s College, Newcastle, from 1935 to
1956—​b ecoming Professor in 1950—​w here he did key work on the frontier. He
went on to be appointed as the first holder of the chair in the Archaeology of the
Roman Empire at Oxford in 1956, a position that he held until his early death
in 1965. His work was geographically and intellectually wide-​ranging, but he put
great energy into his fieldwork in Britain, working on a range of military sites,
including his notable excavations at Hod Hill (1951–​8 ) and Inchtuthill (1952–​65).
At Oxford he inspired a new generation of students who went on to make key con-
tributions to the subject. In contrast to some contemporaries, Richmond was also
much wider ranging in his portfolio of publications. This included the first major
post-​w ar synthesis—​Roman Britain, first published in 1955—​w hich remained
in print in the Penguin History of Britain series for many years, very effectively
popularizing the subject. Interestingly, the format and content of the book were
very much influenced by Collingwood’s template, but, given its enduring charac-
ter, Richmond’s little book had the effect of ‘fixing’ the orthodoxy for many years.
Another legacy was indirect: as at Oxford, he had set about preparing a synthesis
of the temple complex at Bath. With his death this project passed to the ener-
getic young Barry Cunliffe, who ensured its rapid completion and publication in
1969, in the same way he had done with the fifth volume of Bushe-​Fox’s work at
Richborough a year earlier, thereby establishing himself not only as the excava-
tor of the spectacular site at Fishbourne, but as a key figure in Romano-​British
studies. Uniquely for this period, Cunliffe’s training had been in Archaeology and
Anthropology (rather than Classics) at Cambridge. His excavations at Fishbourne
 Roman Britain since Haverfield    29

and Portchester were also an important training ground for the next generation of
archaeologists.
Another interesting figure closely associated with Richmond at the British School
at Rome (BSR) in the 1920s and 1930s was Jocelyn Toynbee, who came from a distin-
guished intellectual family. She was influenced at the BSR by Eugénie Sellers Strong and
prepared a magnificent book based on her Oxford doctoral thesis on Hadrianic sculp-
ture. She had been an undergraduate in Cambridge, then taught at Oxford and Reading,
before returning to Cambridge, where she was a Fellow of Newnham College and rose to
become the Laurence Professor of Classical Archaeology in 1951–​62. She was not only a
significant art historian but made a series of major contributions to Roman archaeology
more broadly. In the context of Roman Britain, she wrote the first influential contri-
bution to the study of Christianity in Britain (1953) as well as organizing a key exhibi-
tion in London on art in Roman Britain in 1962. The catalogue of this show (Toynbee
1962), and the major book that followed two years later (Toynbee 1964), provided the
first full-​length treatment of the subject and had an enduring impact, not only by draw-
ing together the evidence in one place where it could be more fully appreciated but also
by establishing a genre of study.
In many ways the Guildhall Museum exhibition organized by Jocelyn Toynbee is but
one symbol of the dynamism and diversification of the subject in this period. Not only
did other great projects come to fruition—​including the publication of the first volume
of The Roman Inscriptions of Britain (1965), I. D.  Margary’s Roman Roads in Britain
(1955–​7), the influential third edition of the Ordnance Survey Map of Roman Britain
(1956), as well as the Ordnance Survey Map of Hadrian’s Wall (1964) —​but there was a
series of other meetings and publications of key work that set an agenda for the future.
An important conference held in Leicester in 1963 drew together the results of work on
the towns of the province under the title The Civitas Capitals of Roman Britain (edited
by J. S. Wacher, 1966). Religion, already noted among Toynbee’s interests, but that had
hitherto been little explored, benefited from the synthesis of temples by M. J. T. Lewis
(1966), a student of hers. Similarly, a contrasting approach, which laid the ground for
later work and sought to emphasize indigenous traditions, appeared in the form of Ann
Ross’s Pagan Celtic Britain (1967).
The countryside, which had not figured very largely in previous work, was brought
to the fore in contrasting ways. An epoch-​making volume Town and Country in Roman
Britain was published in 1958 by A. L. F. Rivet, who was the Assistant Archaeology
Officer of the Ordnance Survey and, as such, had been responsible for the production
of the third edition of the Map of Roman Britain. He was a pupil of Collingwood and
had dug with Wheeler. He used this work to present a coherent and powerful narra-
tive on the Romano-​British countryside, providing an account that has never been sur-
passed for its originality. Later, the villas of the province were also the subject of his
The Roman Villa in Britain (edited 1969), drawing together a variety of new evidence.
Perhaps of most enduring value within this volume was David Smith’s synthesis of the
evidence of mosaics, in which he sought to identify a series of regional artistic schools
30

30   Martin Millett

in the province. Indigenous aspects of the countryside were also brought into focus for
the first time through the Council for British Archaeology conference held in Oxford
in 1965, which resulted in the publication, in 1966, of Rural Settlement in Roman Britain
(edited by Charles Thomas).
These initiatives were generally associated with the civilian areas of Roman Britain,
and there was remarkably little synthesis of the extensive work done on the Roman
frontiers in the north, although, in Wales, V. E. Nash Williams produced his key study
The Roman Frontier in Wales (1954, revised M. G. Jarrett 1969). Work on Hadrian’s
Wall and the Antonine Wall continued apace, but was mostly published in a stream
of articles in the Proceedings of the Society of Antiquaries of Scotland, Archaeologia
Aeliana, and the Transactions of the Cumberland and Westmoreland Archaeological
Society. A  much-​read general review of the military was provided in Graham
Webster’s book The Roman Imperial Army (1969), which was, in effect, largely based
on Roman Britain. In contrast, new attitudes to the relationship between the Roman
army and the indigenous peoples were first seen in Peter Salway’s volume The Frontier
People of Roman Britain (1965) and, more generally, in the fieldwork by George Jobey
in Northumbria. A whole range of different aspects of the archaeology of the prov-
ince were also increasingly revealed through campaigns of aerial photography, most
notably those initiated by J. K. St Joseph based in Cambridge. As well as revealing a
whole swathe of new sites, especially forts and villas, he worked systematically to use
aerial photography to explore the Roman military campaigns in the north, contrib-
uting much to the explicit research agenda of reconstructing the narrative history of
the province.
The culmination of this period of energetic research came in 1967 with the publica-
tion of Sheppard Frere’s Britannia: A History of Roman Britain. He had been a student
of Classics at Cambridge, embarking on his major excavations in the post-​war period
while teaching at Lancing (where incidentally Haverfield had also taught), before hold-
ing a Readership at the Institute of Archaeology from 1955 (in succession to Wheeler),
being promoted to Professor in 1963, then moving to the Chair of the Archaeology of
the Roman Empire in Oxford in 1966. He had a unique command of the subject and
Britannia was a masterful volume, written in a style that summarized existing knowl-
edge and understanding firmly and authoritatively. It not only told us what was known,
but demonstrated how it should be understood, very much within the tradition of
Britain’s imperial power. Following the format set by Collingwood, it remained as the
key work for a generation.

The 1970s and 1980s

Even as the undoubted successes of the post-​war generation were coming together,
there was a series of changes taking place that was to have fundamental implications
for the study of Roman Britain. These changes affected not only the evidence base of
 Roman Britain since Haverfield    31

the subject but also how the province was to be studied. The trends of the 1960s contin-
ued for some time, with the publication of major syntheses like John Wacher’s Towns
of Roman Britain (1974) and substantial works of reference including Rivet and Smith’s
fundamental volume The Place-​Names of Roman Britain (1979) and A. R. Birley’s Fasti
of Roman Britain (1981). The British volumes of the Corpus Signorum Imperii Romani
also began to appear in 1977. The year 1970 also saw the publication of the first volume
of the journal Britannia devoted to the study of Roman Britain, which was very much
Frere’s creation. Finally, in a completely unpredicted way, excavations at Vindolanda in
1973 produced a mass of new textual evidence to enrich our knowledge of the Roman
military (Bowman 1994). Nevertheless, the landscape had already begun to change, and
orthodoxy was being questioned.
With the economic growth of the 1960s, many new archaeological discoveries were
made as a result of redevelopment work, whether building projects or motorway con-
structions. The scale of this activity put an enormous strain on the established response
in these circumstances, which had seen central government, through the Ministry of
Public Buildings and Works (later the Department of the Environment), fund indi-
vidual archaeologists to dig threatened sites. It was becoming evident that this ad hoc
system, which had involved both freelance diggers and academics being brought in
to run the excavations, was now no longer sufficient, and increasing political pressure
mounted for better provision. It was in these circumstances that the ‘rescue’ movement
was born, with the objective of creating a professional state-​funded archaeological ser-
vice to deal with the problem (Jones 1984). In a very British way the solution was not to
create a state service, but a plethora of local solutions emerged with local authorities,
museums, universities, and independent organizations establishing ‘excavation units’
piecemeal across the country to dig sites funded by the state. This boom did result in
the wholesale recording of sites threatened by destruction and the establishment of a
network of professional excavation teams, but there was massive variation in its extent
and quality, and whether it was ever published. At its best, work of this period produced
key new insights into urban development and hitherto poorly understood aspects of
Roman Britain. For instance, work at Colchester gave a clear insight into the origins of
the colonia (Crummy 1988).
The move towards large-​scale rescue excavation went hand-​in-​hand with a philo-
sophical shift among excavators. Whereas previous generations, whether explicitly or
implicitly, had mostly excavated to answer research questions and contribute to under-
standing the past, there was a strong belief among many of the rescue era that excavating
could be a ‘neutral’ activity that ‘saved’ evidence through ‘objective’ recording so that
it could be thought about and interpreted later, accordingly promoting the concept of
‘preservation by record’. This form of excavation, promoted intellectually by Phil Barker
(1977), especially in his excavations at Wroxeter, had a disastrous effect on publication,
with many practitioners only belatedly discovering that, if you do not understand what
you are digging while the evidence is in front of you, you will find it almost impossible
to make sense of it later in publication. As a result, and with an overwhelming number
of excavations, few sites made any real impact on the understanding of Roman Britain.
32

32   Martin Millett

However, the rescue boom did demonstrate that Roman Britain had been very densely
occupied by rural sites, so that few building projects did not find Roman material. This,
in turn, finally disposed of the idea that the Romano-​British landscape had been domi-
nated by vast tracts of uninhabited woodland with isolated clearings for towns and
villas.
The major samples of the landscape revealed by the construction of motor-
ways also led to an increasing interest in field systems and the organization of the
countryside—​an interest that moved the focus of rural archaeology away from elite
buildings and towards a desire to understand whole landscapes (Fowler 1975). The
intellectual impact of this was to encourage new forms of landscape research that
attempted to see settlement more holistically and diachronically. This had a nota-
ble impact in some areas: for instance, through the pioneering work of Cunliffe at
Chalton in Hampshire (Cunliffe 1973) and David Miles in the Upper Thames valley
(Miles 1988).
The downside of the excavation philosophy of this period was, to some extent,
counterbalanced by the development of a view of the equality of the evidence.
Previously, excavators had often been interested principally in finds that were of obvi-
ous value in establishing the chronology of their sites (namely, pottery and coins),
or material that had some intrinsic ‘artistic’ interest. Now excavators who were less
focused on reconstructing a narrative history sought to analyse a fuller range of finds,
leading to a growth of studies not only of pottery and small finds but, increasingly, of
the animal bones and botanical remains that held the key to understanding the agrar-
ian economy and the environment (King 1978; Dimbleby and Jones 1981). This led to
a greater appreciation of the potential of excavated finds in order to understand past
society and stimulated approaches that increasingly drew on the natural sciences for
the development of new modes of study. Notable in this respect were changes in the
study of pottery pioneered by David Peacock at Southampton University. Building
on his geological training, he developed methods that not only promoted the char-
acterization and sourcing of ceramics, but also insisted that they should be used for
understanding the past economy and not simply for dating (Peacock 1982). Similarly,
traditional approaches to the coin evidence were complemented by a range of new
studies pioneered by Richard Reece and John Casey, which sought to use the evi-
dence to look at broader issues of archaeological interpretation (Casey and Reece
1974; Reece 1987).
In all these areas the influence of alternative intellectual trends represented by the
‘New Archaeology’ began to be felt in the study of Roman Britain (Johnson 1999: 12–​33).
Thoughts increasingly turned to thinking about Roman Britain as a social and economic
system, while studies like those by Ian Hodder showed how the evidence could contrib-
ute to broader debates in archaeology (e.g. Hodder 1972; Hodder and Orton 1976). The
sharp response to such work in Sheppard Frere’s valedictory lecture at Oxford in 1987
well illustrates how the study of Roman Britain had become a heavily contested subject.
The changing landscape of higher education in the UK played an important role
in the shift in emphasis away from historical narrative, as archaeology degrees in
 Roman Britain since Haverfield    33

universities become more widespread. This was linked to the growth of a much
broader-​based higher education system during the 1960s with the proliferation
and growth of universities and a diversification of subjects taught. This widening of
undergraduate education involved the founding of new universities and of archae-
ology departments. Existing departments grew and began to offer undergraduate
degrees in archaeology (where most had previously taught it only as a subsidiary
subject). In addition, newly created universities such as Southampton, Newcastle,
and Leicester saw it as a key subject worthy of support. Thus, through the late 1960s
and into the early 1970s, new degree courses came into being, and academic jobs
were created and filled, many by those active in the study of Roman Britain. Some
existing centres had also expanded or reinforced the discipline. Oxford had already
created a chair in the Archaeology of the Roman Empire in 1956, bringing the study
of Roman Britain to classics and history students. Durham moved to build a fully-​
fledged Department of Archaeology, while also retaining and expanding its work on
Roman Britain. From the 1970s onwards, there was a series of growing centres of
Roman teaching and research in British universities (most notably London’s Institute
of Archaeology, Cardiff, Southampton, Reading, Leicester, Durham, Birmingham,
Manchester, Oxford, and Newcastle-​upon-​Tyne, plus others that emerged later).
Each of these came to produce groups of graduates who made distinctive contri-
butions to particular areas of the subject, while appointments to lectureships and
movements between universities created a complex web of interactions, so that
today many of those in posts teaching Roman archaeology in Britain share linkages
through who have taught them and which excavations they took part in (Table 2.1).
Two particularly influential figures in the teaching of the subject were the contrast-
ing figures of Richard Reece and Barri Jones.
At the Institute of Archaeology, undergraduate archaeology was taught from 1968
with sub-​departments offering specialist training in their own areas. From 1974, the
Roman Department comprised Professor John Wilkes (a Durham Ph.D.  graduate
who had taught at Birmingham), Mark Hassall, a product of the Verulamium excava-
tion who had studied in Oxford then London, and Richard Reece, who had studied
biochemistry at University College London and taught in schools before doing his
doctoral research on Roman coinage at Oxford. He had a unique, Socratic teaching
style, stimulating critical thought and discussion among his students. In this period,
the Institute attracted students who were already strongly committed to the subject,
and, among the Romanists, Richard Reece developed a very strong following, with the
result that significant numbers of them went on to take leading roles in the subject.
Although Richard Reece was not himself an excavator on a large scale, his encour-
agement of generations of students to engage critically with primary material meant
that many went on to do important work in the field and on finds, especially pottery
and coins.
Barri Jones, an Oxford pupil of Sir Ian Richmond’s, was also highly influential in the
subject. He was a key player in the UK rescue movement in the early 1970s, having done
important work in Italy and North Africa. He held a lecturing post at the University of
34

Table 2.1 The educational networks of the contributors to this volume

1960s 1970s 1980s 1990s 2000s 2010s

Oxford • * [C. F. C.Hawkes] * [S. S. Frere] • •

* [S. S. Frere] * [S. S. Frere/​ * [A. Wilson]
J. Lloyd]

Cambridge • - ••• • # #

* [J. Clackson]
* [M. Millett]
* [M. Millett]
* [M. Millett]

Birmingham • •

Cardiff • * [W. Manning] #

Sheffield •# * [G. Jones/​ •#
D. Gilberstson]

Institute of ••• • • # # # * [R. Reece]

Archaeology * [R. Reece] * [R. Reece]
* [C. Lockyear/​
S. Moorhead

Newcastle • •# • #
* [J. Crow]
* [P. van der
Eijk]

Durham • • • • # # # # # • • # # #

* [M. Millett/​C. * [C. Haselgrove]
Haselgrove/ * [R. Hingley/​
​C. Caple/​ A. Leone]
J. Bayley] * [S. Lucy/​A.
* [M. Millett/​ Millard]
J. Price] * [R. Hingley/​R.
Witcher]

Bradford • * [R. Jones]

* [R. Jones]
* [R. Jones]
Southampton * [T. Champion] #
* [M. Millett]
* [L. Revell]

Reading #
* [M. Fulford/
​A. Wallace-​Hadrill]
* [M. Fulford]
* [M. Fulford/​H.
Härke]
 Roman Britain since Haverfield    35

Table 2.1 Continued
1960s 1970s 1980s 1990s 2000s 2010s

Warwick •

Kent • • #
* [E. Swift/​A.
Ward]

York • #
* [S. Roskams]
* [D. Perring/​S.
Roskams]

Wales–​ * [Aldhouse-​
Newport Green]

Bournemouth * [Darvill/​
O’Connor]

Leicester

Other •# • #
European * [J. Bintliff/
​B. van der
Meer]

North America • •• * [Wailes]

Note: • = Undergraduate; # = Masters; * = Ph.D. [Supervisor(s)].

Manchester from 1964 and became Professor in 1971, after playing a key role in estab-
lishing the degree programme there. He was a dynamic fieldworker and an inspirational
teacher, many of whose students from the 1970s went on to make key contributions to the
study of Roman Britain, both as academics and in the newly growing excavation units.
One of the key long-​term results of the growth of archaeology degrees and the boom
in excavation in Britain was a change in attitude to the Roman past. Hitherto, most
engaged in its study had been trained and immersed in Classics, thus giving priority
to ancient texts in their studies and tending towards approaches that laid emphasis
on reconstructing the narrative history from a Roman perspective. The new gener-
ation, which had often discovered the subject through digging, was brought up on
ideas about using material evidence to approach the past in a variety of different ways,
often putting emphasis on social and economic history, seeking the ordinary people
rather than the elites and willing to expend efforts on new types of evidence. Equally,
the structure of higher education meant that more people gained the opportunity to
undertake doctoral research, so a range of major new studies could be undertaken,
stretching the subject in new directions and often developing approaches that owed
as much to the research traditions of prehistory as to the Classical world. Important
in this regard was a series of Ph.D. projects that produced syntheses of aspects of the
36

36   Martin Millett

archaeology of Roman Britain. For example, Mike Fulford produced a study of New
Forest Roman pottery (1975), Tom Blagg wrote his thesis on architectural ornament
(1982, published 2002), and Robert Philpott drew together the evidence for burial
rites (1991). Thus narrative historical approaches no longer dominated the subject in
the way that they had up to the 1970s.
As these new approaches grew, there was a series of conferences and seminars
that broke away from the consensus of the 1960s, questioning both conclusions and
assumptions while offering examples of the value of new approaches. These included
new studies of cemeteries (Reece 1977); different approaches to imperialism (Burnham
and Johnson 1979); studies of the agrarian landscape (Miles 1982); and a questioning of
the role of the military in the development of the province (Blagg and King 1984). Many
of these ideas were also taken up in some of the professional excavation units, which
saw a role not simply for gathering information but for forwarding understanding.
Important studies of the military archaeology in the north of England and Scotland
not only swept away the previous consensus about the chronology and function of
the frontiers (Breeze and Dobson 1976; Hanson and Maxwell 1983), but also sought to
explore the complex relationship between the army and local communities. The leading
proponent of this approach was David Breeze, who had been a student of Eric Birley’s
at Durham and went on to a career in heritage management in Scotland, but never lost
sight of research priorities (Breeze 1982).
The culmination of these processes was a group of alternative publications that came
out as the members of the new generation established themselves in their careers (Henig
1984; Allason-​Jones 1989; Esmonde Cleary 1989; Hingley 1989; Mattingly and Jones
1990). Among these was my own Romanization of Britain (Millett 1990), designed as
an attempt to provide a coherent alternative narrative to that with which I had been
brought up (Millett 2003–​4).

The 1990s and Beyond

In the academic sphere, the period since 1990 has been dominated by much spirited
debate, which has found voice principally through the Theoretical Roman Archaeology
Conference (TRAC), which began in Newcastle in 1991. Discussion has focused on cri-
tiques of the concept of Romanization, but, increasingly, TRAC has become a forum
for a range of ideas that can loosely be characterized as post-​imperial approaches. The
other strength of these meetings has been their increasing emphasis on more sophis-
ticated interpretations of artefacts, moving them from the margins to the centre stage
for discussion. Equally, there has been a strong trend towards evaluating the history of
ideas in the subject and promoting alternative approaches. This has meant that research
on Roman Britain now has a rare dynamism and excitement that is recognized by other
branches of archaeology.
 Roman Britain since Haverfield    37

In commercial archaeology a substantial change also took place in 1990 with the
implementation of a system whereby, if required by the local planning authority,
the developer of a site, rather than the state, was expected to pay for its archaeologi-
cal investigation prior to destruction (see also Wilson, this volume). This completely
changed the nature of commercial archaeology, as the pressure of competition on
price came to bear. Excavation was no longer determined solely by academic strate-
gies, and considerable autonomy was given to local authority advisers in determin-
ing what was required at particular sites. A positive benefit of this has been that, in
some areas, extremely skilled and experienced archaeological teams have been able to
carry through big projects on good budgets with excellent results that are changing our
understanding. So, for instance, London is now probably among the most extensively
explored and fully published of the cities in the Roman Empire. Equally, because of the
need to justify the work undertaken, there has been a renewed emphasis, at a national
and a regional level, on debating what we know about Roman Britain and encourag-
ing commercial excavations to contribute to knowledge by addressing specific research
agendas (James and Millett 2001). This has also required the beginnings of some sys-
tematic re-​evaluation of what is known about the province, as much information is
not now published but archived. This all takes place against a background in which
many full-​time field archaeologists are of the highest calibre and are contributing to
understanding Roman Britain on a massive scale. Against this there are also areas,
often in regions where property values are not high or where local authorities are weak,
where the quality of archaeological work is poorer and little is being added to collective
understanding. Equally, the lack of systematic publication makes it all but impossible
to maintain a clear overview of the subject (Fulford 2011; Fulford and Holbrook 2011).
In parallel with this there has also been a resurgence of excellent primary research,
combining both the synthesis of existing evidence with first-​class new fieldwork and
excavation. The publication of inscriptions has progressed with the completion of vol-
umes 2 and 3 of The Roman Inscriptions of Britain, as well as completion of the definitive
publication of the Vindolanda tablets (Bowman and Thomas 1983, 1994, 2003; Bowman,
Thomas and Tomlin 2010, 2011). Anthony Birley has produced an updated version of
his Fasti, albeit under a new title (2005). The four volumes of David Neal and Stephen
Cosh’s Roman Mosaics in Britain are now in print, and the British volumes of the Corpus
Signorum Imperii Romani are all now published. Further important and pioneer-
ing works have appeared, like Paul Tyers’s Roman Pottery in Britain (1996) and Jeremy
Taylor’s Atlas of Roman Rural Settlement in Britain (2007), making this among the best-​
published provinces in the empire.
Archaeology has also seen the development of new techniques of research—​some
of which are now commonplace—​like large-​scale geophysics, which has been used
extensively to map sites in Roman Britain, providing a wealth of information about a
range of site types. Other new techniques like isotope studies are altering the questions
we can answer, focusing attention on aspects of demography that had hitherto been a
closed book.
38

38   Martin Millett

Similarly there are now a series of alternative yet authoritative syntheses of the sub-
ject that provide contrasting and stimulating approaches to the province (Cool 2006;
Creighton 2006; Mattingly 2006). These, and a rich range of other works, show how
Roman Britain is now a contested subject and one in which much stimulating work is
taking place.
Given the scope and content of the present volume, which is designed to show the
range and diversity of contemporary approaches to Roman Britain, it is unnecessary
to provide any summary of current research here. It is worth noting how there is now
a wide recognition that knowledge is the product of addressing questions, not simply
data-​gathering, and that the questions we ask and the research we produce are them-
selves a product of contemporary perspectives. This leads to a postmodern diversity of
approaches. Against this background we should also note that recognition of our infor-
mation base is now much larger than ever before, not only as a result of commercial
developer-​funded excavations but also because of the evidence from metal-​detecting
now brought together by the ‘Portable Antiquities Scheme’ (see also Wilson, this vol-
ume). The challenge is how to make sense of this information in ways that help elucidate
varied aspects of the history of Roman Britain.

References
Allason-​Jones, L. (1989). Women in Roman Britain. London: British Museum Press.
Barker, P. (1977). Techniques of Archaeological Excavation. London: B. T. Batsford.
Birley, A. R. (1981). Fasti of Roman Britain. Oxford: Clarendon Press.
Birley, A. R. (2005). The Roman Government of Britain. Oxford: Oxford University Press.
Bishop, M. C. (1994). Corstopitum: An Edwardian Excavation. London: English Heritage.
Blagg, T. F.  C. (2002). Roman Architectural Ornament in Britain. BAR British Series 329.
Oxford: Archaeopress.
Blagg, T. F. C., and King, A. C. (1984) (eds). Military and Civilian in Roman Britain. BAR British
Series 136. Oxford: Archaeopress.
Bowman, A. K. (1994). Life and Letters on the Roman Frontier:  Vindolanda and its People.
London: British Museum Press.
Bowman, A. K., and Thomas, J. D. (1983). Vindolanda: The Latin Writing-​Tablets. London:
Britannia Monograph Series 4.
Bowman, A. K., and Thomas, J. D. (1994). The Vindolanda Writing Tablets (Tabulae
Vindolandenses II). London: British Museum Press.
Bowman, A. K., and Thomas, J. D. (2003). The Vindolanda Writing Tablets (Tabulae
Vindolandenses III). London: British Museum Press.
Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2010). ‘The Vindolanda writing-​tablets
(Tabulalae Vindolandenses IV, part 1)’. Britannia, 41: 187–​235.
Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2011). ‘The Vindolanda writing-​tablets
(Tabulalae Vindolandenses IV, part 2)’. Britannia, 42: 113–​144.
Breeze, D. (1982). The Northern Frontiers of Roman Britain. London: B.T. Batsford.
Breeze, D., and Dobson, B. (1976). Hadrian’s Wall. London: Allen Lane.
 Roman Britain since Haverfield    39

British Academy (1977–​ to date). Corpus Signorum Imperii Romani. Oxford:  Oxford
University Press.
British Museum (1922). Guide to the Antiquities of Roman Britain. London: Trustees of the
British Museum.
Burnham, B. C., and Johnson, H. B. (1979) (eds). Invasion and Response: The Case of Roman
Britain. BAR British Series 73. Oxford: Archaeopress.
Carr, L. C. (2012). Tessa Verney Wheeler:  Women in Archaeology before World War II.
Oxford: Oxford University Press.
Casey, P. J., and Reece, R. M. (1974) (eds). Coins and the Archaeologist. BAR British Series
4. Oxford: Archaeopress.
Collingwood, R. G. (1930). The Archaeology of Roman Britain. London:  Methuen and
Company.
Collingwood, R. G. (1939). An Autobiography. Oxford: Oxford University Press.
Collingwood, R. G., and Myres, J. N. L. (1936). Roman Britain and the English Settlements.
Oxford: Clarendon Press.
Collingwood, R. G., Wright, R. P., Frere, S. S., Roxan, M. M., Tomlin, R., and Hassall, M. W. C.
(1965–​to date). The Roman Inscriptions of Britain. Oxford: Clarendon Press.
Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge
University Press.
Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge.
Crummy, P. (1988). ‘Colchester’, in G. Webster (ed.), Fortress into City. London: Batsford, 24–​47.
Cunliffe, B. W. (1968). Fifth Report on the Excavations of the Roman Fort at Richborough, Kent.
Oxford: Oxford University Press.
Cunliffe, B. W. (1969). Roman Bath. London: Society of Antiquaries of London.
Cunliffe, B. W. (1973). ‘Chalton, Hants.: The Evolution of a Landscape’, Antiquaries Journal,
53: 173–​190.
Curle, J. (1911). A Roman Frontier Post and its People: The Fort of Newstead in the Parish of
Melrose. Glasgow: James Maclehose and Sons.
Dimbleby, G., and Jones, M. K. (1981) (eds). The Environment of Man: The Iron Age to the Anglo-​
Saxon Period. BAR British Series 87. Oxford: Archaeopress.
Esmonde Cleary, A. S. (1989). The Ending of Roman Britain. London: B. T. Batsford.
Fowler, P. J. (1975) (ed.). Recent Work in Rural Archaeology. London: Adams and Dart.
Freeman, P. (2007). The Best Training-​Ground for Archaeologists: Francis Haverfield and the
Invention of Romano-​British Archaeology. Oxford: Oxbow Books.
Frere, S. S. (1967). Britannia: A History of Roman Britain. London: Routledge and Kegan Paul.
Frere, S. S. (1987). Roman Britain since Haverfield and Richmond: A Lecture Delivered in
All Souls College, 23 October 1987. Oxford:  Reprinted from History and Archaeology
Review.
Fulford, M. G. (1975). New Forest Roman Pottery. BAR British Series 17. Oxford:
Archaeopress.
Fulford, M. G. (2007). ‘The Grand Excavation Projects of the Twentieth Century’, in S. Pearce
(ed.), Visions of Antiquity: The Society of Antiquaries of London 1707–​2007. London: Society
of Antiquaries (=Archaeologia, 111), 33–​81.
Fulford, M. G. (2011). ‘The Impact of Commercial Archaeology on the UK Heritage’, in B. W.
Cunliffe (ed.), History for the Taking? Perspectives on Material Heritage. London:  British
Academy, 33–​53.
40

40   Martin Millett

Fulford, M. G., and Holbrook, N. (2011). ‘Assessing the Contribution of Commercial

Archaeology to the Study of the Roman Period in England, 1990–​2004’, Antiquaries Journal,
91: 323–​345.
Gillam, J. P. (1957). ‘Types of Roman Coarse Pottery Vessels in Northern Britain’, Archaeologia
Aeliana, 4th ser., 35: 180–​251.
Gillam, J. P. (1968). Types of Roman Coarse Pottery Vessels in Northern Britain. Newcastle-​
upon-​Tyne: Oriel Press.
Hanson, W., and Maxwell, G. (1983). Rome’s North-​ West Frontier:  The Antonine Wall.
Edinburgh: Edinburgh University Press.
Haverfield, F. (1912). The Romanization of Roman Britain. Oxford: Clarendon Press.
Haverfield, F., and MacDonald, G. (1924). The Roman Occupation of Britain. Oxford: Oxford
University Press.
Hawkes, C. F.  C., and Hull, M. R. (1947). Camulodunum:  First Report on the Excavations
at Colchester, 1930–​1939. Oxford:  Report of the Research Committee of the Society of
Antiquaries of London, 14.
Hawkes, J. (1982). Mortimer Wheeler:  Adventurer in Archaeology. London:  Weidenfeld and
Nicolson.
Henig, M. (1984). Religion in Roman Britain. London: B. T. Batsford.
Hingley, R. (1989). Rural Settlement in Roman Britain. London: Seaby.
Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman
Archaeology. London: Routledge.
Hodder, I. R. (1972). ‘Locational Models and the Study of Romano-​British Settlement’, in D. L.
Clarke (ed.), Models in Archaeology. London: Methuen, 887–​909.
Hodder, I. R., and Orton, C. R. (1976). Spatial Analysis in Archaeology. Cambridge: Cambridge
University Press.
James, S. T. (2002). ‘Writing the Legions: The Development and Future of Roman Military
Studies in Britain’, Archaeological Journal, 159: 1–​58.
James, S. T., and Millett, M. (2001) (eds). Britons and Romans: Advancing an Archaeological
Agenda. CBA Research Report, 125. York: Council for British Archaeology.
Johnson, M. H. (1999). Archaeological Theory: An Introduction. Oxford: Blackwell.
Jones, G. D. B. (1984). Past Imperfect: The Story of Rescue Archaeology. London: Heinemann.
Jones, R. F. J. (1987). ‘The Archaeologists of Roman Britain’, Bulletin of the University of London
Institute of Archaeology, 24: 85–​97.
Kenyon, K. M. (1948). Excavations at the Jewry Wall site, Leicester. Oxford:  Report of the
Research Committee of the Society of Antiquaries of London. 15.
King, A. C. (1978). ‘A Comparative Survey of Bone Assemblages from Roman Sites in Britain’,
Bulletin of the University of London Institute of Archaeology, 15: 207–​232.
Lewis, M. J. T. (1966). Temples in Roman Britain. Cambridge: Cambridge University Press.
MacDonald, G. (1911). The Roman Wall in Scotland. Glasgow: J. Maclehose and Sons.
Margary, I. D. (1955–​7). Roman Roads in Britain. 2 vols. London: Phoenix House.
Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Allen Lane.
Mattingly, D., and Jones, G. D. B. (1990). An Atlas of Roman Britain. Oxford: Blackwell.
Miles, D. (1982) (ed.). The Romano-​ British Countryside. BAR British Series 103.
Oxford: Archaeopress.
Miles, D. (1988). ‘Villas and Variety: Aspects of Economy and Society in the Upper Thames
Landscape’, in K. Branigan and D. Miles (eds), The Economies of Romano-​British Villas.
Sheffield: Department of Archaeology and Prehistory, 60–​72.
 Roman Britain since Haverfield    41

Millett, M. (1990). The Romanization of Britain: An Essay in Archaeological Interpretation.

Cambridge: Cambridge University Press.
Millett, M. (2003–​4). ‘The Romanization of Britain:  Changing Perspectives’, Kodai, 13/​14:
169–​173.
Nash Williams, V. E. (1954). The Roman Frontier in Wales. Cardiff: University of Wales Press.
Neal, D. S., and Cosh, S. R. (2002–​2010). Roman Mosaics of Britain. Vols. 1–​4. London: Society
of Antiquaries.
Ordnance Survey (1956). Map of Roman Britain. Chessington: Ordnance Survey.
Ordnance Survey (1964). Map of Hadrian’s Wall. Chessington: Ordnance Survey.
Peacock, D. P. S. (1982). Pottery in the Roman World. London: Longman.
Philpott, R. (1991). Burial Practices in Roman Britain. BAR British Series 219.
Oxford: Archaeopress.
Potter, T. W. (1987). ‘The Institute and Roman Archaeology: Past, Present and Future—​
an Outsider’s View’, Bulletin of the University of London Institute of Archaeology, 24:
71–​84.
Reece, R. M. (1977) (ed.). Burial in the Roman World. CBA Research Report 22. London: Council
for British Archaeology.
Reece, R. M. (1987). Coinage in Roman Britain. London: Seaby.
Richmond, I. (1955). Roman Britain. London: Penguin Books.
Rivet, A. L. F. (1958). Town and Country in Roman Britain. London: Hutchinson University
Library.
Rivet, A. L. F. (1969) (ed.). The Roman Villa in Britain. London: Routledge and Kegan Paul.
Rivet, A. L.  F., and Smith, C. (1979). The Place-​Names of Roman Britain. London:  B.T.
Batsford.
Ross, A. (1967). Pagan Celtic Britain: Studies in Iconography and Tradition. London: Routledge
and Kegan Paul.
Salway, P. (1965). The Frontier People of Roman Britain. Cambridge:  Cambridge
University Press.
Stanfield, J. A., and Simpson, G. (1958). Central Gaulish Potters. Oxford:  Oxford
University Press.
Taylor, J. (2007). Atlas of Roman Rural Settlement in Britain. York:  Council for British
Archaeology.
Thomas, C. (1966) (ed.). Rural Settlement in Roman Britain. London:  Council for British
Archaeology.
Toynbee, J. M. C. (1953). ‘Christianity in Roman Britain’, Journal of the British Archaeological
Association. 3rd ser., 16: 1–​24.
Toynbee, J. M. C. (1962). Art in Roman Britain. London: Phaidon.
Toynbee, J. M. C. (1964). Art in Britain under the Romans. Oxford: Oxford University Press.
Tyers, P. (1996). Roman Pottery in Britain. London: B. T. Batsford.
Wacher, J. (1966) (ed.). The Civitas Capitals of Roman Britain. Leicester:  Leicester
University Press.
Wacher, J. (1974). Towns of Roman Britain. London: B.T. Batsford.
Webster, D. B. (1991). Hawkeseye: The Early Life of Christopher Hawkes. Stroud: Alan Sutton.
Webster, G. (1969). The Roman Imperial Army. London: Black.
Wheeler, R. E. M. (1920). Comparative Notes on the Rhenish Pottery of the Roman Period (with
an Appendix on the Balkerne Gate, Colchester). D.Litt. thesis, University College London.
Wheeler, R. E. M. (1925). Prehistoric and Roman Wales. Oxford: Clarendon Press.
42

42   Martin Millett

Wheeler, R. E. M. (1930). London in Roman Times. London: London Museum.

Wheeler, R. E.  M. (1955). Still Digging:  Interleaves from an Antiquary’s Notebook.
London: Michael Joseph.
Winbolt, S. E. (1925). Roman Folkestone: A Record of Excavation of Roman Villas at East Wear
Bay. London: Methuen and Co.
Winbolt, S. E., and Herbert, G. (1930). The Roman Villa at Bignor, Sussex [with a Plan of the
Villa]. Oxford: Clarendon Press.
 Chapter 3

Romano-​B ri t i sh
Archaeol o g y i n t h e
E arly T wen t y-​Fi rst
Century

Pete Wilson

Introduction

The early twenty-​first century is a period of rapid change even for such a venerable a
subject as Romano-​British archaeology. Britannia has been studied since at least the
time of Camden but, as a subject, can be argued to have been resistant to new thinking,
dominated by a succession of ‘greats’ whose influence, in some cases unwittingly, often
became an obstacle to the development of new interpretations and approaches. The ten-
dency for the great and the good to dominate the subject extended well into the twenti-
eth century: one has only to think of Eric Birley’s influence in relation to Hadrian’s Wall
or the colossus-​like contributions of Sheppard Frere across a wide variety of subjects.
Frere’s 1974 edition of Britannia is still this archaeologist’s first choice when checking
‘basic facts’ on a wide range of topics.

Archaeology in the United Kingdom

Prior to 1990, responsibility for archaeology in the United Kingdom (UK) lay largely
with the Department of the Environment, which funded most work on threatened
sites. Some archaeological research, normally on unthreatened sites, was undertaken
by universities and local archaeological societies. In November 1990, the government
introduced Planning Policy Guidance 16: Archaeology and Planning (PPG 16) (HMSO
44

44   Pete Wilson

1990), which established archaeology as a ‘material consideration’ in the planning

process and offered two options for important sites and remains: ‘preservation in situ’
(the preferred option) or excavation and ‘preservation by record’. It also established
the principle of ‘polluter pays’, transferring the responsibility for funding excavation in
advance of destruction from the state to the developer. In 2010, PPG 16 was superseded
by Planning Policy Statement 5: Planning and the Historic Environment (PPS 5) (DCLG
2010), which was, in turn, replaced in 2012 by the National Planning Policy Framework
(DCLG 2012).
At a national level, following the creation of the Scottish Parliament and National
Assembly for Wales in the late 1990s, responsibility for the historic environment (includ-
ing archaeology) lies: in England, with Historic England (a non-​departmental public
body sponsored by the Department for Culture, Media, and Sport); in Wales, with Cadw
(the Welsh government’s historic environment service); and, in Scotland, with Historic
Scotland (an executive agency of the Scottish government). These bodies are responsible
for recommending sites for statutory protection through designation as scheduled monu-
ments and for acting as their respective government’s national advisors. However, most
archaeological sites—​recorded in Historic Environment Records (HERs), which are main-
tained by local authorities—​are undesignated and, where threatened by development, are
provided with some level of protection through the local planning process, which can
demand changes to development proposals or the funding of excavations in advance of
partial or complete destruction (‘preservation by record’). Development-​led archaeology
in the UK is largely undertaken by archaeological units, some of which may be established
as charities or trusts, while others are fully commercial operations. Other bodies active
in archaeological research include some university archaeology departments, long-​estab-
lished local archaeological societies, and, increasingly, local community-​based groups,
which are often funded—​at least in part—​by the Heritage Lottery Fund, and which may
work with professional archaeologists who can act as mentors and advisors.

How We Got to Where We are Now

While this chapter is focused on Romano-​British archaeology in 2016, it is perhaps

worth spending a little time identifying and summarizing those factors that have been
the key drivers of change. From the perspective of this writer, a major turning point is
the increasing professionalization of field archaeology; first, through the ‘Ministry of
Works Excavators’ such as John Wacher, Ernest Greenfield, and Philip Rahtz, and then
through the rise of rescue archaeology that widened the pool of practitioners as regional
units developed (see also Millett, this volume). Without decrying the contributions of
universities, museum-​based colleagues, and amateur societies all of whom had to cede
space to these new players; new thinking, more funding, and a greater intensity of work
all contributed to a period of rapid change.
 Romano-British Archaeology in the Early Twenty-First Century    45

The next step change came with the advent of developer funding. With the best will in
the world, the resources that the Department of Environment was able to deploy, even
when supplemented in the 1970s and 1980s by Manpower Service Commission money,
could not cope with the scale of destruction resulting from development. The advent of
PPG 16 in 1990 can be seen to have set the scene in development-​related archaeology
up to the present day, and the practice and issues relating to developer-​funded work
have been reviewed many times (see, e.g. Everill 2007). Parallel developments included
a decline in amateur fieldwork and in particular excavations—​a decline often blamed
on the rise of professionalism and developer funding—​and the apparent flight of many
university-​based archaeologists to projects outside the UK, driven by a perceived
greater value in research terms of foreign projects and, in some cases, the lower costs of
fieldwork in some countries in comparison with the UK.
Therefore, by the millennium, the UK had a substantial commercial sector under-
taking work that was predicated on developers’ ambitions with respect to choice of
location, albeit mitigated by the work of curatorial archaeologists and the planning
and designation systems (Figure 3.1). University-​led research was, by that time, more
restricted than it had been when Newcastle regularly excavated on Hadrian’s Wall and

Figure 3.1  Excavation of a major Roman building by Oxford Archaeology South in advance of

housing development just outside the walled town of Cirencester in 2009.
Source: © Oxford Archaeology South, with thanks to Berkeley Homes for agreement to reproduce.
46

46   Pete Wilson

Figure 3.2  Volunteers excavating Building 16 at Piddington villa, Northamptonshire, in 2012.

Source: © Pete Wilson.

Birmingham worked at Wroxeter over a thirty-​year period; although Mike Fulford and
Reading’s ongoing association with Silchester (Fulford 1989; Fulford and Clarke 2011)
and Martin Millett and Peter Halkon’s work in the East Riding (Halkon and Millett
1999; Millett 2006) serve to demonstrate that important university-​led work continues.
Bleaker still was the situation with respect to amateur fieldwork; this reflected in part the
generally ageing profile of amateur societies, but also the difficulties or perceived dif-
ficulties of undertaking project work. However, again there were exceptions: for exam-
ple, the Monmouth Archaeological Society at Caerleon (Burnham 2001: 215)—​where
University of Wales College (Newport) was also active (Burnham 2001: 315–​316)—​or the
long-​running Piddington excavations (Friendship-​Taylor and Friendship-​Taylor 2011)
(Figure 3.2).

Research Frameworks

In the decade from 2000 to ​2010 considerable effort and resource were expended on the
development of regional research frameworks and strategies, which will be considered
later in this chapter. However, the recognition of a need for lacunae to be identified and
priorities defined goes back to at least the latter part of the twentieth century. For the
purposes of this chapter, the invitation from the Directorate of Ancient Monuments and
 Romano-British Archaeology in the Early Twenty-First Century    47

Historic Buildings to the Roman Society in 1975 to form a committee to ‘advise on aca-
demic priorities with respect to the rescue archaeology of the Roman period’—​which
led to the publication of Priorities for the Preservation and Excavation of Romano-​British
Sites (Roman Society 1985)—​can perhaps be seen as pivotal. The priorities defined in
1985 might, by today’s standards, be seen as generalized and limited, with a cynical
reader perhaps feeling that he or she could identify the research interests of each of
the committee members writ large in the document. Other researchers also promoted
research issues that reflected their interests; for example, Vivien Swan in her The Pottery
Kilns of Roman Britain (1984) devotes a chapter to ‘Kiln Studies: Current Techniques
and Future Research’, providing cogent arguments that influence her subject to this day.
In 1996, English Heritage launched Frameworks for Our Past (Olivier 1996), which
envisaged the creation of regional and specialist research frameworks consisting of
three elements: resource assessment, research agenda, and research strategy. Intellectual
ownership was intended to lie with the research community that developed the frame-
work, and, as a consequence, the documents reflect a level of variation, with several
covering the (former) government regions, while colleagues in the south-​east have
opted for a series of subregional documents. At the time of writing some areas lacked
full coverage: for example, the research-​strategy element of the Yorkshire framework
is still to be produced. Outside the basic regional structure, Stonehenge, Avebury, and
Hadrian’s Wall have been the subject of targeted volumes (AAHRG 2001; Darvill 2005;
Symonds and Mason 2009)—​the Hadrian’s Wall volume being cross-​referenced with
the frameworks for the north-​west and north-​east (Brennand 2006, 2007; Petts and
Gerrard 2006). Similarly some county-​level frameworks have been produced, such
as those for Bedfordshire (Oake et al. 2007) and Surrey (Bird 2006), as well as others
for smaller areas—​notably, for the Roman period, that for Fishbourne and its environs
(Manley 2008). If we look beyond England, a research framework for Wales is in place
(IfA Wales/​Cymru 2008) and one for Scotland is under development <http://​www.
socantscot.org/​scarf.asp?Menu> (accessed 25 May 2014).
Pivotal in the development of thinking about research priorities for Roman Britain
across the wider profession was a session at the Roman Archaeology Conference in
1999, which led to the publication of Britons and Romans: Advancing a Research Agenda
(James and Millett 2001). More recently, Historic England has developed its own
Research Strategy for the Roman Period Historic Environment http://​www.english-​her-
itage.org.uk/​professional/​research/​strategies/​research-​strategies/​roman (accessed 25
May 2014), drawing on the available regional research frameworks and in consultation
with the sector.
While it is not in doubt that the development of research frameworks has sharpened
thinking with respect to academic objectives in projects of all types and at all levels, pro-
vided support for bids for funding, and assisted local and national curators in sustain-
ing arguments in favour of appropriate and academically sound mitigation strategies,
it is possible to argue those benefits are perhaps not their greatest value. The process of
developing research strategies, or at least those that might be regarded potential models
of good practice, have generally been inclusive, bringing together archaeologists from
48

48   Pete Wilson

across the sector. This has led to colleagues from universities, curatorial organizations,
museums, and commercial units—​as well as independent archaeologists and those from
the voluntary sector (contra Moore 2006: 3)—​cooperating and debating in a way that
few, if any, other processes or mechanisms have accomplished. In that sense, the research
framework process can be seen to be crucial to the rekindling of some of the links and
interactions between different elements in the sector that, as suggested earlier in this
chapter, have been lost or weakened during the professionalization of archaeology.

The Specialist

Beyond the regional and other location-​based research frameworks, specialist groups
have developed research frameworks for their area of interest; notably, for the Roman
period, the Study Group for Roman Pottery (SGRP) (Perrin 2011). The SGRP document
acts as a powerful voice for archaeological specialists, not only setting out priorities for
Roman pottery studies but also identifying the challenges that we face as a sector in
ensuring that we retain and enhance the specialist knowledge that is crucial to our sub-
ject, as the issues identified for pottery studies are equally applicable to most other areas
of work. Detailed consideration of the difficulties faced by specialists in general is pro-
vided by the Survey of Archaeological Specialists 2010–​11 (Aitchison 2011), but, if we look
at the Roman period, many of the issues are all too clear: an ageing cadre of specialists
who are often singleton researchers; limited opportunities for mentoring or developing
a new generation of specialists; a decreasing number of posts (in common with the rest
of the sector); and, while we remain in the grip of a recession, less work from developer-​
funded projects.

Developer-​funded Archaeology

Prior to the publication of PPG 16 in 1990, it was generally recognized that resources for
fieldwork in response to development were overstretched, and, depending on the inter-
ests and/​or location of commentators, it could be argued that significant archaeology
was lost with limited or even no record. The rise of commercial archaeology increased
the amount of resources going into fieldwork, particularly excavation (see, e.g. Darvill
and Russell 2002), but also led to some asking questions of the process, given that sites
were selected by developers. What is clear is that developer funding has led to even
greater emphasis on the site (normally defined by the limits of the development) and,
in extreme cases, to an atomization of the resource, with sites either side of a modern
boundary feature being excavated by competing contracting units with little or no coop-
eration or exchange of ideas. Equally, the practice of different phases of the same project
(evaluation, excavation, and post excavation) sometimes being undertaken by different
 Romano-British Archaeology in the Early Twenty-First Century    49

organizations as a result of being separate tendered can further exacerbate the difficul-
ties and, potentially, lead to an incomplete understanding of what has been excavated
or destroyed. A further perception is that difficulties can arise from the willingness of
many commercial units to range over a wide area in pursuit of work, fuelling an oft-​
repeated concern with respect to a potential lack of local knowledge. The argument is
that operating in different parts of the country could lead to a lack of understanding of
the local context, which could hamper the development of a project design appropri-
ate to the circumstances of the site, impede understanding of the emerging results, and
result in flawed analyses and interpretations in post-​excavation. These issues have been
debated widely, and, in addition to Everill’s paper (2007), attention can be drawn to Tom
Moore’s consideration of PPG 16 and cultural landscapes, wherein he suggests: ‘we are in
danger of having archaeological landscapes made up of “dots” representing archaeologi-
cal evidence, often of a fragmentary nature, which are not understood as part of a whole’
(Moore 2006: 5).
The potential of developer-​funded work has been demonstrated on numerous occa-
sions and in many circumstances, with site monographs from key urban centres mak-
ing significant contributions to our understanding of Romano-​British sites. Examples
include Causeway Lane, Leicester (Connor and Buckley 1999), or the more recent series
of monographs on sites in London from Museum of London Archaeology Service/​
Museum of London Archaeology, Pre Construct Archaeology, Wessex Archaeology,
and others. Similarly, in the context of the rural areas of Roman Britain, while individual
site reports may enlighten us about the specifics of individual sites (for example, the
recent volume on Faverdale, Darlington (Proctor 2012)), perhaps our greatest advances
have come from work on linear schemes such as roads and pipelines. Taking Yorkshire
as an example, the combined data for the Iron Age and Romano-​British periods from the
M1–​A1 Link Road (Roberts et al. 2001), Holmfield/​Ferrybridge Interchange (Roberts
2005), and the A1(M) Darrington to Dishforth scheme (Brown et al. 2007), represent an
unprecedented and highly informative sample of the rural landscape in an area where
other major projects are working towards publication. As samples of landscape exam-
ined in detail and analysed and published at length (with substantial discussion sections
occupying between 10 and 20 per cent of the volumes), they complement and inform
extensive survey projects, such as those undertaken on the crop marks of the Magnesian
Limestone (Roberts et al. 2010) or the Swale–​Ure Washlands (Bridgland et al. 2011) and
funded through the Aggregates Levy Sustainability Fund.
The wide range of projects undertaken and the number of contractors undertak-
ing work with developer funding make it rather invidious to pick out ‘highlights’, but
it is perhaps necessary to cite some examples. Looking to the south of England, Oxford
Archaeology’s work at Lankhills (Booth et al. 2010) has challenged our understanding of
a site that has figured large in the literature of late Roman Britain. Similarly, the work of
Cotswold Archaeology (formerly Cotswold Archaeological Trust) over many years and
increasingly through developer-​funded work has, with others, built on the work of the
Cirencester Excavation Committee (see, e.g. Holbrook 1998, 2008). It would be possible
to go on, but in the context of this chapter it is clear that commercially funded projects
50

50   Pete Wilson

of all sizes, where the results are appropriately analysed and disseminated, are adding
substantially to our understanding of most aspects of Roman Britain.
On a less positive note, Moore (2006) addresses the issue of grey literature and the lack of
integration of the results of commercially funded work into academic and other syntheses.
Since Moore’s paper was published, we have had Richard Bradley’s ground-​breaking study
(2007), which sought to integrate the evidence for prehistoric periods from the grey lit-
erature with pre-​existing data. For the Roman period, we have already seen the results
of the first two phases of the Roman Grey Literature Project (Holbrook 2010a, b; Fulford
and Holbrook 2011; Hodgson 2012). Reading University and Cotswold Archaeology, with
funding from Leverhulme Trust and English Heritage, are now engaged in the third phase
of the project: an England-​wide study of data from grey literature and fully published
reports intended to produce a new account of the rural settlement of Roman Britain.

Non-​publication

Recognition of the potential of development archaeology is nothing new. For many

years the Ministry of Works and its successors (the Department of the Environment
(Ancient Monuments Branch) and, most recently, Historic England) have—​through
targeted support for ‘unexpected discoveries’—​ put considerable resources into
responding to threats posed by development, including the work of the ‘Ministry of
Works Excavators’. The 1970s are viewed by many as the ‘heroic age’ of rescue archaeol-
ogy that led to the development and perceived stability of regional archaeology units
and, subsequently, to the present-​day commercially led situation, as already discussed.
However, all too often the promise of the fieldwork was not fully realized, projects were
not adequately analysed or published, and the phenomenon known as ‘backlog’ was on
us. The problem also extended to much of the work done during the Manpower Service
Commission-​funded fieldwork boom in the late 1970s. Many of the publications from
that time (of both Roman and other sites) were a product of the individual determina-
tion of those responsible; but all too often important work remains largely unknown
and substantially inaccessible. The issue is particularly difficult with respect to some
of our most important towns and cities, although work on key rural sites also remains
unpublished. Set against that situation, the issues around grey literature already dis-
cussed seem almost esoteric. The information in grey literature, for all the difficulties
of using it, could be accessed through HERs or the Archaeology Data Service (ADS)
if the researchers whose work might benefit from it were to take the trouble to seek it
out. That said, after the replacement of PPG 16 with Planning Policy Statement 5 (PPS
5), it would now be difficult to justify non-​publication of projects while the associated
Practice Guide remains in force, as it states (in the specification for the ‘content of a writ-
ten scheme of investigation’) that it should contain: ‘proposals for the preparation and
publication of a suitable report on the investigation, its results and the advancement in
understanding that those results bring’ (DCLG 2010: 38). It is hoped that assurances
 Romano-British Archaeology in the Early Twenty-First Century    51

that the force of PPS 5 will survive the implementation of the National Planning Policy
Framework (DCLG 2012) are borne out.

Research for its Own Sake

While it has to be accepted that all archaeology, at least if well done and appropriately dis-
seminated, contributes to our collective knowledge and/​or academic discourse, in times
of resource pressure those bodies responsible for managing the archaeological resource
(be they national agencies, curatorial archaeologists, or organizations that own the
resource such as the National Trust) are increasingly forced to focus on what is needed
for management purposes or what can be regarded as ‘reasonable’ in terms of a brief for
a development-​related site. Although academic knowledge will be gained, it may also be
constrained in terms of the resources available for analysis (if a commercial project) or
limited by the project design (if driven by management questions). One example with
a positive benefit, the recent Miner–​Farmer Landscapes Project in the North Pennines
(an Area of Outstanding Natural Beauty), sought to examine environmental threats from
erosion and climate change and, as part of that study, has produced a re-​examination of
Whitley Castle Roman fort and the areas surrounding it (Ainsworth and Went 2009).
While a level of ongoing engagement of university-​based colleagues has already been
acknowledged, it is possible to argue that, in recent years, there has been re-​engage-
ment on the part of universities with the archaeology of Roman Britain. This sugges-
tion is prompted, in part, by the development of a number of high-​profile projects,
including those undertaken by the University of Newcastle within the Hadrian’s Wall
World Heritage Site (initially with Historic England), and work by Durham University
at Binchester, the University of Cambridge at Aldborough, and the University of
Nottingham at Caistor St Edmund. It is notable that all those projects have included
opportunities for non-​student volunteers, delivering that better engagement referred to
earlier in the chapter. Given the success of the South Shields and Vindolanda volunteer
programmes—​which demonstrate an international enthusiasm for Roman Britain—​it
is perhaps surprising that there are not more joint projects in the UK along the lines
of the Binchester partnership involving Stanford University (see <https://​sites.google.
com/​site/​binchesterromanfort/​> (accessed 13 July 2012)).

Community Archaeology, the Media,

and Key Stage 2

Public engagement with archaeology in all its forms has never been higher, and, in
part, this reflects the profile of archaeology on television. In the UK, the success of
52

52   Pete Wilson

programmes such as Meet the Ancestors and Time Team made archaeology mainstream
entertainment, and Roman-​period sites seem to have a particular fascination. In part
this may simply reflect the fact that Roman-​period structures can be quite ‘tele-​visual’,
often being constructed of masonry and frequently associated with substantial num-
bers of objects, many of which are readily recognizable and visually attractive. The fact
that the Roman period figures in the national curriculum in Key Stage 2 may well also
influence the thinking of programme-​makers. Roman-​period stories play well with
news media, often for the same reasons of good visual appeal and recognizable material
culture, although the latter is too often reduced (at least in the headlines) to a focus on
‘value’ expressed in purely financial terms, and the media may not always treat our sub-
ject in ways with which we are comfortable! The Daily Mail (22 May 2002) reported our
publication of Catterick (Wilson 2002 a, b) with the headline ‘Meet Crossus Dressus—​
our first transvestite’, in reference to Hilary Cool’s suggestion that a burial might rep-
resent a gallus (a priest of Cybele). Since the 1950s, when Sir Mortimer Wheeler and
Glyn Daniel were unlikely media stars, archaeologists have grappled with the conflict
between the necessity of working with the media to popularize and ‘make the case for’
our subject and the charges of ‘dumbing down’ from colleagues suspicious of possibly
simplistic ‘popular’ treatments of often complex issues. These arguments continue to the
present day (see, e.g. Clack and Brittain 2007; Whittington 2010).
The attributes of Roman-​period archaeology that appeal to the media also make it
attractive to amateur, community, and public archaeology projects. The value of the con-
tribution of the voluntary sector is difficult to gauge because much of the work is almost
invisible at a national level (and also locally in some cases), even if the groups themselves
are not unknown: for example, the Council for British Archaeology (CBA) has identi-
fied the existence of in excess of 2,000 community archaeology groups (Thomas 2010).
Good relationships exist between many professionals and non-​professional groups; in
many cases Heritage Lottery funding or other sources facilitate direct cooperation and,
where necessary, mentoring. In terms of the present-​day archaeological landscape, pro-
fessional archaeologists and academics need to engage with the enthusiasm that the vol-
untary sector represents. This is not simply as a potential ‘project resource’ but, more
pressingly, to ensure that the results that derive from voluntary-​sector projects are fed
into the literature and other resources available to researchers and also to offer, where
needed, support to ensure that those results are of a standard that represents a return of
lasting value for the work and enthusiasm expended. PPS 5 has pushed the profession as
a whole towards better and higher-​quality engagement with the public; our challenge
is to deliver that without compromising the standards and academic integrity that we
value so highly.
The nature of many Roman sites and associated collections provides us with an almost
unrivalled ‘marketing tool’ for our subject. The significance of Hadrian’s Wall as an icon
of the past cannot be overemphasized; along with Stonehenge, it may be seen to repre-
sent ‘archaeology’ in the UK. The ability of Hadrian’s Wall to engage the wider public is
demonstrated by its popularity as a tourist draw, but, more significantly in the context
of this chapter, the wish of people to engage further is difficult to overstate. For example,
 Romano-British Archaeology in the Early Twenty-First Century    53

Figure 3.3  Volunteers excavating the sequence of barracks in the north-​west quadrant of the
fort at Vindolanda in 2011.
Source: © The Vindolanda Trust.

the long-​term association of South Shields with Earthwatch brings paying international
volunteers into industrial Tyneside <http://​www.earthwatch.org/​exped/​bidwell.html>
(accessed 13 July 2012), while all 650 places for paying voluntary diggers during a recent
season at Vindolanda (Figure 3.3) were filled in less than three hours <http://​www.
guardian.co.uk/​uk/​the-​northerner/​2011/​nov/​10/​vindolanda-​archaeology-​hadrian-​s-​
wall-​housteads-​andrew-​birley-​haltwhistle> (accessed 13 July 2012).

Portable Antiquities Scheme (PAS)

One area where amateurs dominate is amongst metal detectorists, and it is not neces-
sary here to re-​examine the history of the often troubled relationships between archae-
ologists and detectorists. The looting of Wanborough Roman temple (Surrey) and the
continued depredations at Icklingham Roman town (Suffolk) are well known (Thomas
2009; Wilson 2015) and serve to illustrate the vulnerability of Roman-​period sites
to illicit detecting. On a more positive note, responsibly detected and appropriately
recorded metal-​detector finds are adding much to our knowledge, particularly with
respect to numismatic studies (see, e.g. Walton 2012). Detectorists are also revealing
the existence of previously unknown sites and, where these are responsibly detected,
54

54   Pete Wilson

the finds recorded, and the data passed to the appropriate Historic Environment
Record, they are contributing significantly to research (see, e.g. Brindle 2009). For
example, the Portable Antiquities Scheme (PAS) records 90 Masters dissertations and
63 Ph.D. theses either in progress or completed using Scheme data <http://​finds.org.
uk/​research> (accessed 7 July 2012). However, perhaps only 40 per cent of detectorists
record their finds with the PAS. Monitoring of eBay by PAS has revealed significant
levels of non-​reporting of potential treasure finds, which there is a clear duty in law
to report; so one has to ask: ‘what is being lost unrecorded?’ In addition, detector-
ists often argue that they are ‘saving the heritage’ by removing metalwork from the
soil where it is under attack from fertilizers and other chemicals, citing the poor state
of recent finds compared to those made twenty or so years earlier in the same loca-
tion. While it is probable that agricultural and other chemicals are having an adverse
effect on buried material culture, it may also be the case that sites are getting ‘detected
out’, with only the poorly preserved material remaining after decades of detecting.
Metal detectors are not going to go away, but they are contributing quite legally to the
erosion of the archaeological resource, particularly when finds are not recorded. This
presents an ongoing challenge to archaeologists. Should we encourage and support
‘responsible metal detecting’ while also demanding greater levels of recording, or risk
seeing our common past—​of all metal-​using periods—​disappear into the antiquities
trade or private collections?
Even given the success of the PAS, legal metal detecting provides other challenges to
archaeologists and museum professionals, not least the pressure to acquire key finds
for public display. While not necessarily taking resources directly away from curato-
rial work or archaeological projects, the staff time and effort required to acquire finds
such as the Frome hoard of Roman coins (Moorhead et al. 2010), which under the
Treasure Act process was valued at £320,250 with a further £100,000 needed for con-
servation, represents a major challenge. The state of museum funding (particularly
acquisition budgets) means that many museums may struggle to find the resources
to acquire even single locally found objects that would enhance their collections and
displays.

Approaches and Techniques

The advent of the metal detector is only one of many innovations of which archaeologists
of all types have had to take account in recent decades. The introduction of new technol-
ogies and developments has made immense and far less controversial contributions. The
potential of aerial photography has long been known and, through the work of Professor
St Joseph, David Wilson, and many others, Romano-​British sites have provided some
of its most productive targets for many years, with many new sites identified and much
new detail about known sites added. Since 1988, large-​scale intensive mapping of air
photographs by the English Heritage National Mapping Programme and others has
 Romano-British Archaeology in the Early Twenty-First Century    55

added immensely to our stock of knowledge. For example, taking the National Mapping
Programme alone, in excess of 40 per cent of England had been mapped by 2012
<http://​w ww.english-​heritage.org.uk/​professional/​research/​landscapes-​and-​areas/
​national-​mapping-​programme/​> (accessed 25 May 2014), which allows present-​day
researchers widened opportunities to think at a landscape level. Furthermore, the
potential to combine the aerial photographic evidence with satellite and multi-​spectral
imagery, LIDAR data <http://​www.english-​heritage.org.uk/​professional/​research/​
landscapes-​and-​areas/​aerial-​survey/​archaeology/​> (accessed 25 May 2014) and large-​
scale geophysical surveys (see Powlesland 2009; <http://​www.landscaperesearchcentre.
org/​atlas/​LRC_​VOP.html> (accessed 25 May 2014), provides the potential for comple-
mentary datasets, which can give previously unimaginable detail (Figure 3.4).
The contributions of individual remote-​sensing techniques are well known in the
literature: for example, the extensive geophysical surveys on the forts and vici of the
Hadrian’s Wall World Heritage Site, at Birdoswald (Biggins and Taylor 1999; 2004a),
Maryport (Biggins and Taylor 2004b), and elsewhere, or those undertaken as part of
the Cadw-​sponsored Roman Forts Environs Project (see Hopewell 2005). As already
suggested, the use of aerial photography in combination with geophysical survey
has revolutionized our knowledge of many sites and their context (see, e.g. Wilson
2002a:  34–​45) for work designed to contextualize excavation data). Furthermore,
projects designed from the start to utilize a range of remote-​sensing techniques in
combination with surface collection, excavation, and geographical and topographic
modelling, as at Wroxeter (Gaffney and White 2007) or in the Vale of Pickering
(Powlesland et al. 2006), allow a further step change in understanding and provide
models for future work.

Figure 3.4  Combined geophysical surveys undertaken by the Landscape Research Centre with
the support of Historic England and the Aggregates Levy Sustainability Fund on the southern
side of the Vale of Pickering, North Yorkshire; depicts more than 1.5 kilometres of late Iron Age
and Roman trackside ribbon development or ‘ladder settlement’ in addition to a wealth of other
prehistoric and post Roman features.
Source: © Landscape Research Centre.
56

56   Pete Wilson

Developing Understanding

Having moved on from being a subject dominated and, some would say, constrained
by the ‘great and the good’, Romano-​British archaeology in the second decade of the
twenty-​first century is in an interesting position. We undoubtedly have a new genera-
tion of ‘greats’, but one perhaps less dominating and more open to debate. The appear-
ance of books and papers that challenge or question old orthodoxies, in whole or in
part, is now common, and those new works and their authors are likely to be met with
questions and searching reviews from colleagues (see, e.g. Freeman (1993) on Martin
Millett’s The Romanization of Britain (1990)), rather than facing offhand dismissal as
‘upstarts’ by those whose opinions they challenge. That, viewed with the fact that there
is space in the literature for a volume running to over 600 pages that seeks to present ‘an
experimental, speculative and heretical’ view of Roman Britain (Mattingly 2006: p. xv),
suggests strongly a subject that has matured nicely. This openness to new ideas may,
in part, reflect the success of the Theoretical Roman Archaeology Conference (TRAC)
from its inception in 1991 in providing exposure for new ideas, different approaches, and
(mostly) younger researchers.
Writers such as Mattingly (2006: 491–​528) and Taylor (2007) beat the drum for the
recognition of diversity within Roman Britain. On one level this is not new—​as Mike
Fulford (2007) states in his review of Mattingly (2006), where he points to the Peoples of
Roman Britain series published in the 1970s and 1980s by Duckworth—​and, in addition,
regional variation in aspects of material culture has long been recognized. However,
what we are seeing now is the presentation of a more nuanced and complex picture
that is beginning to show variables operating at a local level. Such understanding can
be enhanced only by the full incorporation by researchers of all types of data, including
that held by the PAS and data derived from voluntary-​sector work and commercially led
interventions.
Roman Britain continues to spring surprises, such as the discovery or recognition in
the decade to 2010 of a further three Roman military sites at Calstock, Lostwithiel, and
near St Austell, in Cornwall (Ordnance Survey 2011); overturning long-​held academic
belief but seemingly having little impact on popularly accepted ‘facts’—​hence ‘The
Romans never settled here’ <http://​www.guardian.co.uk/​travel/​2012/​jun/​24/​free-​corn-
wall-​independence-​campaign> (accessed 25 May 2014). Furthermore, Roman Britain
is, or at least those studying Roman Britain are, gradually reconnecting with the wider
empire. The Limes Congress has, since 1949, brought together archaeologists working
on Roman military sites, and related topics and scholars of Roman Britain have figured
large in the various Congress programmes. This may be a result of the obvious research
potential of similarities and differences between military sites and broad issues of mili-
tary strategy, deployment, and supply across the empire. In contrast, for many years
British archaeologists working on the civilian archaeology of Roman Britain were
apparently less willing to make those linkages and engage with colleagues working on
 Romano-British Archaeology in the Early Twenty-First Century    57

similar subjects elsewhere in the Roman Empire. That said, the fact that there were
exceptions among excavators and others is accepted, as is the strength of international
cooperation on particular topics such as the Corpus Inscriptionum Latinarum or terra
sigillata. The reasons behind this more limited engagement are debatable, but for dec-
ades it appeared that most archaeologists of Roman Britain were talking to themselves,
or rather to like-​minded colleagues whose thinking was insular by inclination. By
1966, Sir Mortimer Wheeler saw ‘Romano-​British archaeology [as having] reached an
uncommonly high level of technical excellence, but [having] diverted into a relatively
insignificant channel talent which might have been more amply employed’ (Freeman
2007: 625). While we might debate vigorously the insignificance of our study, I for one
would accept that in seeking to better understand a minor province on the outer edge
of the empire, archaeologists of Roman Britain need to engage more closely with those
working on other provinces: twenty-​first-​century Roman Britain cannot be an insular
study.
However, engagement is happening increasingly through contacts with foreign col-
leagues and also through greater direct cooperation and the internationalization of key
conferences such as the Roman Archaeology Conference (RAC) and Theoretical Roman
Archaeology Conference (TRAC). While there have often been non-​UK speakers at
both conferences, the holding of the conferences outside the UK (for example, TRAC
2008 in Amsterdam, RAC/​TRAC in Ann Arbor in 2009 and Frankfurt in 2012) pro-
motes greater exposure of UK archaeologists to a much wider archaeological commu-
nity and to a broad range of work and approaches, both academic and in terms of field
procedures.

Challenges for the Future

We live in uncertain times, and across all parts of the historic environment sector we
are seeing unprecedented change. We cannot yet know what the long-​term impact of
changes to university fees will mean for individual archaeology departments and for the
expectations of students still drawn to the subject.
Similarly, the implications of the Government’s 2010 comprehensive spending review
on national bodies are still being worked through and at the time of writing the full
impacts are as yet unclear, and, in a climate where further cuts are possible, we face con-
tinuing uncertainty. There is also uncertainty as to the impact of the proposed Welsh
Heritage Bill or the possible merger of the Royal Commission on the Ancient and
Historical Monuments of Scotland with Historic Scotland <http://​www.historic-​scot-
land.gov.uk/​news_​article.htm?articleid=36834> (accessed 13 July 2012). In England, as
elsewhere, local authorities are under immense pressure from austerity measures, and
it is Historic Environment Services, museums, and other non-​statutory services that are
bearing much of the pressure. Romano-​British archaeology is not affected by this any
more than the archaeology of any other period, but we are possibly looking at a future
58

58   Pete Wilson

with a very different curatorial regime that will potentially have severe consequences for
all periods.
In conclusion, it is difficult to try and predict where Romano-​British archaeology is
going—​at least in the short term—​given the issues that have been outlined. However,
there is a sense that our subject (while suffering along with the rest of the heritage sector)
has a resilience that will ensure it retains its popularity with the public—​as providing
potentially good television, evocative places to visit, and a strong focus for community
and public archaeology. Those in academic posts, unit-​based archaeologists, and profes-
sionals in curatorial roles can perhaps only ‘wait and see’ as to how university depart-
ments fare, what impact the National Planning Policy Framework and the Localism Act
in England have on development-​led archaeology, and what the future prognosis for
funding is. The lights may be flickering, but, hopefully, whatever reinventions the sector
goes through, new and exciting work will emerge in the commercial sector as devel-
opment returns and resources will be found to continue and develop both university-
and community-​based field projects focused on our Roman past. Academically, we are
stronger than ever, but it is possible that, across the sector, we need to demonstrate that
Britannia is far from being a ‘grand old lady’ that has ‘ceased to matter as a major histori-
cal problem’ (Wheeler 1961: 159).

References
AAHRG (Avebury Archaeological and Historical Research Group) (2001). Archaeological Research
Agenda for the Avebury World Heritage Site. Salisbury: Trust for Wessex Archaeology <http://​www.
wiltshire.gov.uk/​archaeological-​research-​agenda-​avebury.pdf> (accessed 25 May 2014).
Ainsworth, S., and Went, D. (2009). ‘Whitley Castle, Tynedale, Northumberland:  An
Archaeological Investigation of the Roman Fort’, English Heritage: Research Department
Report 89/​2009.
Aitchison, K. (2011). Survey of Archaeological Specialists 2010–​11. Sheffield: Landward Research
<http://​www.heacademy.ac.uk/​assets/​hca/​documents/​archaeology/​survey_​of_​archaeolog-
ical_​specialists_​heh06.pdf> (accessed 25 May 2014).
Biggins, J. A., and Taylor, D. J.  A. (1999). ‘A Survey of the Roman Fort and Settlement at
Birdoswald, Cumbria’, Britannia, 30: 91–​110.
Biggins, J. A., and Taylor, D. J.  A. (2004a). ‘Geophysical Survey of the vicus at Birdoswald
Roman Fort, Cumbria’, Britannia, 35: 159–​178.
Biggins, J. A., and Taylor, D. J. A. (2004b). ‘The Roman Fort and vicus at Maryport: Geophysical
Survey, 2000–​2004’, in R. J. A. Wilson and I. D. Caruna (eds), Romans on the Solway. Essays in
Honour of Richard Bellhouse. Cumberland and Westmorland Antiquarian and Archaeological
Society Extra Series 31. Kendal: Trustees of the Senhouse Roman Museum, Maryport, 102–​133.
Bird, D. (2006). Surrey Archaeological Research Framework 2006. Kingston on Thames and
Guildford:  Surrey County Council and Surrey Archaeological Society <http://​www.sur-
reycc.gov.uk/​_​_​data/​assets/​pdf_​file/​0008/​167813/​SARF2006.pdf> (accessed 25 May 2014).
Booth, P., Simmonds, A., Boyle, A., Clough, S., Cool, H. E. M., and Poore, D. (2010). The Late
Roman Cemetery at Lankhills, Winchester:  Excavations 2000–​2005. Oxford Archaeology
Monograph 10. Oxford: Oxford Archaeology.
 Romano-British Archaeology in the Early Twenty-First Century    59

Bradley, R. (2007). The Prehistory of Britain and Ireland. Cambridge:  Cambridge

University Press.
Brennand, M. (2006) (ed.). ‘The Archaeology of North-​West England:  An Archaeological
Research Framework of North-​West England. Volume 1: Resource Assessment’, Archaeology
North-​West, 8/​18 <http://​www.liverpoolmuseums.org.uk/​mol/​collections/​archaeology/​arf/​
index.aspx> (accessed 25 May 2014).
Brennand, M. (2007) (ed.). ‘Research and Archaeology in North-​ West England:  An
Archaeological Research Framework of North-​West England. Volume 2: Research Agenda
and Strategy. Archaeology North-​West, 9/​19  <http://​www.liverpoolmuseums.org.uk/​mol/​
collections/​archaeology/​arf/​accessingresearchagenda.aspx> (accessed 25 May 2014)  and
<http://​www.liverpoolmuseums.org.uk/​mol/​collections/​archaeology/​arf/​researchstrategy.
aspx> (accessed 25 May 2014).
Bridgland, D., Innes, J., Long, A., and Mitchell, W. (2011). Late Quaternary Landscape Evolution
of the Swale-​Ure Washlands, North Yorkshire. Oxford: Oxbow Books.
Brindle, T. (2009). ‘Amateur Metal Detector Finds and Romano-​ British Settlement:
A Methodological Case Study from Wiltshire’, in M. Driessen, S. Heeren, J. Hendriks, F.
Kemmers, and R. Visser (eds), TRAC 2008: Proceedings of the Eighteenth Annual Theoretical
Roman Archaeology Conference, Amsterdam 2008. Oxford: Oxbow Books, 53–​72.
Brown, F., Howard-​Davis, C., Brennand, M., Boyle, A., Evans, T., O’Connor, S., Spence,
A., Heawood, R., and Lupton, A. (2007). The Archaeology of the A1 (M)  Darrington
to Dishforth DBFO Road Scheme. Lancaster Imprints 12. Lancaster:  Oxford
Archaeology North.
Burnham, B. C. (2001). ‘Roman Britain in 2000. 1.  Sites Explored. 1.  Wales’, Britannia,
32: 312–​318.
Clack, T., and Brittain, M. (2007). Archaeology and the Media. Walnut Creek, CA: University
College London Institute of Archaeology Publications.
Connor, A., and Buckley, R. (1999). Roman and Medieval Occupation in Causeway Lane,
Leicester. Leicester Archaeology Monograph Series 5. Leicester: Leicester Archaeology.
Darvill, T. (2005). Stonehenge World Heritage Site:  An Archaeological Research Framework
<http://​www.english-​heritage.org.uk/​publications/​stonehenge-​whs-​research-​framework/> ​
(accessed 25 May 2014).
Darvill, T., and Russell, B. (2002). Archaeology after PPG 16: Archaeological Investigations in
England 1990–​1999. Bournemouth University School of Conservation Sciences Research
Report 10. Bournemouth and London: Bournemouth University/​English Heritage.
DCLG (Department of Communities and Local Government) (2010). PPS5: Planning for the
Historic Environment Practice Guide. London: HMSO.
DCLG (Department of Communities and Local Government) (2012). National Planning Policy
Framework. London: HMSO.
Everill, P. (2007). ‘British Professional Field Archaeology:  Antiquarians and Labourers;
Developers and Diggers’, in Y. Hamilakis and P. Duke (eds), Archaeology and
Capitalism: From Ethics to Politics. Walnut Creek, CA: Left Coast Press, 158–​187.
Freeman, P. W. M. (1993). ‘“Romanisation” and Roman Material Culture’, Journal of Roman
Archaeology, 6: 438–​445.
Freeman, P. W. M. (2007). The Best Training-​Ground for Archaeologists. Oxford: Oxbow Books.
Frere, S. S. (1974). Britannia: A History of Roman Britain. London: Cardinal.
Friendship-​ Taylor, R. M., and Friendship-​ Taylor, D. E. (2011). Iron Age and Roman
Piddington: Ninth Interim Report. Piddington: Upper Nene Archaeological Society.
60

60   Pete Wilson

Fulford, M. (1989). The Silchester Amphitheatre. Excavations of 1979–​85. Britannia Monograph

Series 10. London: Roman Society.
Fulford, M. (2007). ‘An Insular Obsession’, Britannia, 38: 367–​369.
Fulford, M., and Clarke, A. (2011). Silchester: City in Transition. The Mid-​Roman Occupation
of Insula IX c.AD 125–​250/​300: A Report on Excavations undertaken since 1997. Britannia
Monograph Series 25. London: Roman Society.
Fulford, M. G., and Holbrook, N. (2011). ‘Assessing the Contribution of Commercial
Archaeology to the Study of the Roman Period in England, 1990–​2004’, Antiquaries Journal,
91: 1–​23.
Gaffney, V. L., and White, R. H. (with Goodchild, H.) (2007). Wroxeter, the Cornovii, and
the Urban Process:  Final Reports on the Wroxeter Hinterland Project 1994–​1997. Volume
1:  Researching the Hinterland. JRA Supplementary Series 68. Portsmouth, RI:  Journal of
Roman Archaeology.
Halkon, P., and Millett, M. (1999). Rural Settlement and Industry:  Studies in the Iron Age
and Roman Archaeology of Lowland East Yorkshire. Yorkshire Archaeological Report
4. Leeds: Yorkshire Archaeological Society.
HMSO (Her Majesty’s Stationery Office) (1990). Planning Policy Guidance 16: Archaeology and
Planning. London: HMSO.
Hodgson, N. (2012). ‘Assessing the Contribution of Commercial Archaeology to the Study
of Roman South and West Yorkshire, 1990–​2004’, Yorkshire Archaeological Journal,
84: 38–​58.
Holbrook, N. (1998) (ed.). Cirencester: The Roman Town Defences, Public Buildings and Shops.
Cirencester Excavations V. Cirencester: Cotswold Archaeological Trust.
Holbrook, N. (2008) (ed.). Excavations and Observations in Roman Cirencester 1998–​2007.
Cirencester Excavations VI. Cirencester: Cotswold Archaeological Trust.
Holbrook, N. (2010a). ‘The Contribution of Commercial Archaeology to the Study of Roman
Somerset, 1990–​2004’, Proceedings of the Somerset Archaeological and. Natural History
Society, 154: 33–​50.
Holbrook, N. (2010b). ‘The Contribution of Commercial Archaeology to the Study of Roman
Essex, 1990–​2004’, Essex Archaeology and History (Fourth series), 1: 1–​15.
Hopewell, D. (2005). ‘Roman Fort Environs in North-​West Wales’, Britannia, 34: 225–​269.
IfA Wales/​Cymru (2008). A Research Framework for the Archaeology of Wales <http://​www.
archaeoleg.org.uk/​index.html> (accessed 25 May 2014).
James, S., and Millett, M. (2001) (eds). Britons and Romans:  Advancing an Archaeological
Agenda. CBA Research Report 125. York: Council for British Archaeology.
Manley, J. (2008) (ed.). The Archaeology of Fishbourne and Chichester:  A  Framework for
its Future. Lewes:  Sussex Archaeological Society <http://​sussexpast.co.uk/​ research/​
fishbourne-​conservation-​framework> (accessed 25 May 2014).
Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 BC–​AD 409.
London: Allen Lane.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Millett, M. (2006) (ed.). Shiptonthorpe, East Yorkshire: Archaeological Studies of a Romano-​
British Roadside Settlement. Yorkshire Archaeological Report 5.  Leeds:  Yorkshire
Archaeological Society.
Moore, T. (2006). Following the Digger:  The Impact of Developer-​Funded Archaeology on
Academic and Public Perceptions of Cultural Landscapes. Newcastle: Cultural Landscapes in
 Romano-British Archaeology in the Early Twenty-First Century    61

the 21st Century. Forum, UNESCO, University and Heritage 10th International Seminar,
International Centre for Cultural Heritage Studies. Newcastle:  International Centre for
Cultural Heritage Studies.
Moorhead, S., Booth, A., and Bland, R. (2010). The Frome Hoard. London:  British
Museum Press.
Oake, M., Luke, M., Dawson, M., Edgeworth, M., and Murphy, P. (2007). Bedfordshire Archaeology.
Research and Archaeology: Resource Assessment, Research Agenda and Strategy. Bedfordshire
Archaeology Monograph 9. Bedford: Bedfordshire Archaeology <http://​www.persona.uk.com/​
A5dunstable/​deposit-​docs/​DD026-​DD050/​DD-​047.pdf> (accessed 25 May 2014).
Olivier, A. C. H. (1996). Frameworks for our Past: A Review of Research Frameworks, Strategies
and Perceptions. London: English Heritage.
Ordnance Survey (2011). Roman Britain Historical Map. 6th edn. Southampton:  Ordnance
Survey.
Perrin, R. (2011). A Research Strategy and Updated Agenda for the Study of Roman Pottery in
Britain. Study Group for Roman Pottery Occasional Paper 1 <http://​www.romanpotterys-
tudy.org/​SGRPPublications/​strategy/​Strategy.pdf> (accessed 25 May 2014).
Petts, D., and Gerrard, D. (2006). Shared Visions: The North-​East Regional Research Framework
for the Historic Environment. Durham: Durham County Council <http://​www.dur.ac.uk/​
archaeology/​research/​projects/​?mode=project&id=485> (accessed 25 May 2014).
Powlesland D. (2009). ‘Why Bother? Large Scale Geomagnetic Survey and the Quest for “Real
Archaeology”’, in S. Campana and S. Piro (eds), Seeing the Unseen: Geophysics and Landscape
Archaeology. Abingdon: Taylor and Francis, 167–​182.
Powlesland, D., Lyall, J., Hopkinson, G., Donoghue, D., Beck, M., Harte, A., and Stott, D. (2006).
‘Beneath the Sand: Remote Sensing, Archaeology, Aggregates and Sustainability: A Case
Study from Heslerton, the Vale of Pickering, North Yorkshire, England’, Archaeological
Prospection, 13/​4: 291–​299.
Proctor, J. (2012). Faverdale, Darlington: Excavations at a Major Settlement in the Northern
Frontier Zone of Roman Britain. London: Pre-​Construct Archaeology.
Roberts, I. (2005) (ed.). Ferrybridge Henge: The Ritual Landscape. Archaeological Investigations
on the Site of the Holmfield Interchange of the A1 Motorway. Yorkshire Archaeology 10.
Leeds: Archaeological Services WYAS.
Roberts, I., Burgess, A. and Berg, D. (2001) (eds). A New Link to the Past: The Archaeological
Landscape of the M1–​A1 Link Road. Yorkshire Archaeology 7.  Leeds:  West Yorkshire
Archaeology Service.
Roberts, I., Deegan, A., and Berg, D. (2010). Understanding the Cropmark Landscapes of the
Magnesian Limestone. Leeds: Archaeological Services WYAS.
Roman Society (1985). Priorities for the Preservation and Excavation of Romano-​British Sites.
London: Roman Society.
Selkirk, A. (2010). ‘Who Champions the Amateur?’, Current Archaeology, 247: 48–​49.
Swan, V. G. (1984). The Pottery Kilns of Roman Britain. RCHME Supplementary Series
5. London: HMSO.
Symonds, M. F.  A., and Mason, D. J.  P. (2009) (eds). Frontiers of Knowledge:  A  Research
Framework for Hadrian’s Wall. Durham: Durham County Council and Durham University
<https:// ​ w ww.dur.ac.uk/​ a rchaeology/​ r esearch/​ p rojects/​ ? mode=project&id=485>
(accessed 25 May 2014).
Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151.
York: Council for British Archaeology.
62

62   Pete Wilson

Thomas, S. (2009). ‘Wanborough Revisited:  The Rights and Wrongs of Treasure Trove
Law in England and Wales’, in S. Thomas and P. G. Stone (eds), Metal Detecting and
Archaeology: ‘Heritage Matters’ [Volume 2]. Woodbridge: Boydell, 153–​165.
Thomas, S. (2010). Community Archaeology in the UK: Recent Findings. Council for British
Archaeology <http://​www.britarch.ac.uk/​research/​community> (accessed 25 May 2014).
Walton, P. J. (2012). Rethinking Roman Britain:  Coinage and Archaeology. Wetteren,
Belgium: Moneta 137.
Wheeler, R. E. M. (1961). Review of: The Journal of Roman Studies, 50 (1960) Jubilee Volume,
Antiquity, 35 (138): 157–​159.
Whittington, K. (2010). ‘Through a Glass Lens Darkly’, Archaeologist, 77 (Autumn): 41.
Wilson, P. (2002a). Cataractonium:  Roman Catterick and its Hinterland. Excavations and
Research 1958–​1997: Part 1. CBA Research Report 128. York: Council for British Archaeology.
Wilson, P. (2002b). Cataractonium:  Roman Catterick and its Hinterland. Excavations and
Research 1958–​1997: Part 2. CBA Research Report 129. York: Council for British Archaeology.
Wilson, P. (2015). ‘Metal Detectors: Friends or Foes?’, in P. Everill and P. Irving (eds), Rescue
Archaeology 40 Years On. Hertford: Rescue, 159–172.
 Chapter 4

The Devel opme nt

of Artefact St u di e s

Ellen Swift

Introduction

This chapter surveys the evolution of artefact studies from the antiquarian period to the
early twenty-first century. Following a brief overview of antiquarian interest in objects
and collecting, I investigate the place of artefact studies within the initial development
of archaeological methodology during the later nineteenth century. The classification
of objects and the development of find reports within published site excavation reports
are then considered, followed by an analysis of the main trends in scholarly interpre-
tation. Each section is informed by a consideration of the brief biographical details of
scholars engaged in artefact research in order to place trends in artefact studies within a
wider social perspective. Particular themes relating to data presentation and interpreta-
tion and the status of artefact studies are then discussed at more length. As will become
apparent, the position of artefact research in archaeology results from its complex his-
torical development alongside other aspects of archaeological method and interpreta-
tion, and is influenced not only by trends in academic research but by a range of other
factors, including the lingering influence of antiquarian and early twentieth-​century
studies, the status and profile of artefact researchers, and the wider conditions of aca-
demia and museums in the later twentieth century.

The Antiquarian Interest in Artefacts

Antiquarian interest in artefacts was very much the interest of the collector. Following
the founding of the Society of Antiquaries, early members sometimes met to discuss
items in their personal collections (Herendeen 2007: 90), although initially there was
little in the nature of systematic description and recording. In general, collections
64

64   Ellen Swift

were private ones, and antiquarians were hampered by their lack of access to artefacts
beyond their own objects. Although information could be exchanged with other schol-
ars, established connections with other collectors, access to their finds, and an ability
to travel were essential in developing artefact research. In an early attempt to dissemi-
nate knowledge (also followed by others (C. Evans 2007: 274)), the late eighteenth–​early
nineteenth-​century French numismatist Mionnet made and sold cast copies of Roman
coins (Babelon 1901: 200). The illustration of artefacts, however, was already increas-
ing during the seventeenth century (Hingley 2008: 40). Watercolours and engravings
were the principal methods used until the introduction of photographs in the mid-​
nineteenth century (Smiles 2007: 124).
Eighteenth-​century collectors and antiquarians, such as Richard Warner, who pub-
lished a collection of antiquities from Roman Bath (Warner 1797), focused on inscrip-
tions and coins, as well as the more large-​scale and spectacular finds such as mosaic
floors and pieces of sculpture. Above all, they were interested in items that could
be linked to Roman history (Kendrick 1950:  166), yet the tendency to see artefacts
through the prism of ancient history led to some spectacular misconceptions. Faussett
(1856 [1773]), for instance, thought Anglo-​Saxon artefacts were Roman. Thanks to the
Western admiration for Classical culture, however, some Roman artefacts, such as
coins, were well documented from comparatively early on: for example, the substantive
publication of coin collections during the eighteenth century across Europe (Babelon
1901). The discovery of sites such as Pompeii and Herculaneaum was an added stimu-
lus to eighteenth-​century interest in Roman remains (Hingley 2008: 233–​4). The initial
focus on historical artefacts and their existence as tangible links to the dramas of past
times was similar to the value attached more widely by antiquarians to ‘curiosities’ as a
prop for storytelling (Crane 1999: 187).
As a consequence of the history of collecting, artefact studies played a comparatively
large role in research: for instance, forming about half of all the papers published in the
Society of Antiquaries journal Archaeologia between 1820 and 1850 (Hingley 2007: fig-
ure 59). Of nearly 600 entries on small finds collected for this chapter from the Society of
Antiquaries subject index (maintained until 1988), around 40 per cent were nineteenth
century or earlier. As Tyers (1996: 2) notes for early studies of Roman pottery, in general
there were two principal approaches in antiquarian publications, each of which was to
persist in later artefact research: first, the interpretation of objects through reference to
ancient texts in which they were mentioned and, secondly, a focus on the documenta-
tion and description of objects as a material record of the historical past.

The Beginning of Archaeological

Approaches to Artefact Study

During the nineteenth century, and particularly its second half, more systematic
approaches to artefact studies developed, thanks especially to two major figures: Charles
 The Development of Artefact Studies    65

Roach-Smith and Augustus Lane Fox Pitt-Rivers. Both were private collectors who
became interested in the site context of objects and their activities are recognizable as
archaeology, rather than antiquarianism. A dependence on ancient texts is still evident
in this period, but this was influenced by nineteenth-​century scholarly endeavours
across a range of parallel subjects, from natural history to ethnology. For instance, in
his catalogue of artefacts from Richborough, Roach-​Smith draws upon Pliny’s descrip-
tion of spoon handles being used to pierce eggshells, but goes on to compare this to folk
customs in East Anglia, in which eggshells were apparently pierced to prevent witches
sailing in them (Roach-​Smith 1850: 206). This illustrates the contemporary interest in
ethnographic analogy and the implicit identification of ancient peoples with those who
were perceived to be their modern equivalents. Pitt-​Rivers collected both antiquities
and ethnographic material and participated in the learned societies of both disciplines.
The comparison of ancient and modern artefacts was fundamental to his methodology
(Chapman 1985). Pitt-​Rivers was also an early exponent of experimental reconstruction,
but his most important contribution to artefact studies, influenced by figures such as
Darwin, was the development of typology as a means of understanding the evolution
and dating of objects (Bowden 1991: 54–​55). Both Pitt-​Rivers and Roach-​Smith also had
a good understanding of issues related to dating: Pitt-​Rivers pointed out the significance
of artefacts in dating archaeological sequences and Roach-​Smith knew that only the
latest coins in a deposit were useful for dating (Bowden 1991: 120; Roach-​Smith 1868a:
18–​19). Interest in geographical distribution is also seen in this period; for instance,
Akerman’s listing of coin finds in south-​east Britain (Akerman 1844; J. Evans 1956: 275).
Although classification of natural history was, by this date, well established, it was
only with the recovery of greater quantities of artefacts as a result of the building of
nineteenth-​century railways and canals that it became possible for more systematic
archaeological classification to be made (Piggott 1989:  153). Reports on categories of
artefacts as part of a site report start appearing towards the end of the nineteenth cen-
tury (C. Evans 2007: 287). Lucas (2001: 68) suggests that, deriving from the previous tra-
dition of artefact collection, finds took ‘pride of place’ in early site reports; yet when we
examine reports for the Roman period they do not always include the finds (e.g. Ashby
et al. 1904). Where finds are listed, it is evident from references in passing to ‘other finds’
that only a small selection is actually being published (Pevensey Excavation Committee
1908: 14; Bushe-​Fox 1913: 31). The prevailing attitude is summed up nicely in Fox’s 1893
report on excavations at Silchester, in which he asserts that ‘the objects in bone and glass
call for no special remark’ and omits them (Fox 1893: 471). Bone artefacts start to be sys-
tematically included in site reports only from the 1920s (see Table 4.1).
In addition to deciding what to include, those publishing catalogues had to choose
how to arrange their material. This was not just a presentational matter. There was an
evidently dialectical relationship between classification and interpretation (discussed in
more detail later in this chapter). Roach-​Smith (1850, 1859) arranged artefacts mostly in
categories according to their perceived function, which suited his use of ancient texts
to discuss how the artefacts had been used in everyday Roman living. Pitt-​Rivers, the
other pioneer of artefact studies in the nineteenth century, used illustrated captions in
66

Table 4.1 Finds categories in Roman site reports up to the 1960s

Contents and order
of occurrence of
finds categories
(excluding animal
bone, human bone, Sections with
botanical remains, authorship
Reference (in etc., unless these Main organizing credited to
chronological include worked principle of small Stratigraphic others apart from
order) Site objects) finds context of finds excavator/​s

Fox (1893) Silchester Gold finger-​ring/​ None No Appendices on

intaglios/​coins/​ coin hoards
other metal small and silver
finds/​pottery/​coin hoard
hoard/​silver hoard

Pevensey Pevensey Pottery/​stamped None Partly (location No

Excavation tiles/​ coins/​bronze of selected
Committee small finds/​other finds noted in
(1908) small finds, mostly description of
metal excavations)

Curle (1911) Newstead Stone objects Functional Yes Appendices

including categories on coins and
inscriptions/​ vegetable
dress and armour/​ remains
weapons/​
Samian/​other
pottery/​tools
and implements/​
transport
and harness/​
miscellaneous/​
ornaments/​
appendix on
vegetable remains
including baskets/​
appendix on coins

Haverfield Corstopitum Inscribed Material Only for the All, including

(1911) objects/​other pottery pottery and
stone objects/​ coins
bronze small
finds (starting
with brooches)/​
iron small finds/​
lead small finds/​
miscellaneous
objects of wood,
ceramic or
enamel/​samian/​
decorated grey
ware pottery
sherd/​coins
Table 4.1 Continued
Contents and order
of occurrence of
finds categories
(excluding animal
bone, human bone, Sections with
botanical remains, authorship
Reference (in etc., unless these Main organizing credited to
chronological include worked principle of small Stratigraphic others apart from
order) Site objects) finds context of finds excavator/​s

Bushe-​Fox Wroxeter Small finds Material Yes Pottery and

(1913) (starting with coins
metal finds,
brooches listed
first)/​glass/​
samian/​other
pottery/​coins

Baddeley (1922) Cirencester Bronze small Material No No

finds/​coins/​
pottery

Winbolt (1925) Folkestone Coins/​samian/​ Material Yes No

villa other pottery/​
bronze small
finds/​bone
(including bone
artefacts among
animal bone)/​
silver/​iron/​glass/​
tile stamps/​
architectural
Bushe-​Fox Richborough Sculptured & Material Yes Pottery and
(1926) architectural (though not very coins
stonework/​lead pig systematic)
with inscription/​
bronze brooches/​
other metal small
finds/​other small
finds/​lamps/​glass
vessels/​samian/​
other pottery/​coins

Wheeler and Caerleon Inscriptions on Material Yes Inscriptions

Wheeler (1928) amphitheatre stone and lead/​ and stamped
stamped tiles/​ tile
statuettes/​
brooches/​other
metal and bone
small finds/​
intaglios/​glass/​
lamps/samian/​
other pottery/​coins
(continued)
68

Table 4.1 Continued

Contents and order

of occurrence of
finds categories
(excluding animal
bone, human bone, Sections with
botanical remains, authorship
Reference (in etc., unless these Main organizing credited to
chronological include worked principle of small Stratigraphic others apart from
order) Site objects) finds context of finds excavator/​s

Fieldhouse, Tiddington Bronze small finds Material No No

May, and (starting with
Wellstood brooches)/​bronze
(1931) and iron small
finds/​iron small
finds/​other metal
small finds/​glass,
stone and bone
objects/​pottery/​tile

Whiting, Ospringe Samian/​other Material Yes Pottery

Hawley, and cemetery pottery/​bronze
May (1931) small finds
(starting with
brooches)/​lead
and pewter
objects (coins
were published
separately in an
earlier article)

Bushe-​Fox Richborough Bronze brooches/​ None Yes Small finds,

(1932) miscellaneous samian,
other small finds/​ potter’s
glass/​lamps/​ stamps, coin
pottery summary, coin
hoard

Corder and Kirk Langton villa Bronze small Material No Coins

(1932) finds (starting
with brooches)/​
bone small finds/​
iron small finds/​
stone small finds/​
pottery/​coin
Table 4.1 Continued
Contents and order
of occurrence of
finds categories
(excluding animal
bone, human bone, Sections with
botanical remains, authorship
Reference (in etc., unless these Main organizing credited to
chronological include worked principle of small Stratigraphic others apart from
order) Site objects) finds context of finds excavator/​s

Wheeler and Lydney Architectural and Material Yes Coins found

Wheeler (1932) worked stone/​ (though not very before 1928-​
bronze small systematic) 9, coin hoard 1
finds (starting
with brooches)/​
miscellaneous small
finds/​iron objects/​
inscribed objects/​
pottery/​coins

Kenyon (1948) Leicester Samian/​other Material Yes Samian,

Jewry wall pottery/​bronze potter’s
small finds stamps and
(starting with coins
brooches)/​bone
small finds/​jet and
shale small finds/​
gold finger-​ring/​
bronze figurine/​
spindle-​whorls/​
miscellaneous
small finds
including those
of iron and stone/​
tiles/​coins

Bushe-​Fox Richborough Bronze terret/​ Material Yes Small finds,

(1949) bronze small pottery and
finds (starting coins
with brooches,
intaglios also
in this section)/​
miscellaneous
small finds/​glass
vessels/​lamps/​
iron/​pottery/​coins
(continued)
70

Table 4.1 Continued
Contents and order
of occurrence of
finds categories
(excluding animal
bone, human bone, Sections with
botanical remains, authorship
Reference (in etc., unless these Main organizing credited to
chronological include worked principle of small Stratigraphic others apart from
order) Site objects) finds context of finds excavator/​s

Oswald (1952) Margidunum Via Quintana/​ Context Yes No

area extending to
south rampart/​
four outer ditches/​
defensive area of
outer ditches

Wedlake (1958) Camerton Samian/​coarse Mainly by type of Samian, coins

pottery/​figurines/​ object
brooches/​coins/​
architectural
fragments/​querns/​
whetstones/​
spindlewhorls/​
counters/​beads/​
bracelets/​rings/​
spatula/​styli/​
miscellaneous
bronze objects/​
studs and nails/​
toilet accessories/​
glass/​locks
and keys/​pins/​
needles/​spoons/​
shale objects/​
miscellaneous
stone objects/​
iron knives/​
miscellaneous
iron objects/​Belgic
brick or daub/​
Roman brick

Gould (1966–​7) Wall Coins/​glass/​ Mainly material Yes Coins, glass,

brooches/​samian/​ brooches,
coarse pottery/​ samian,
medieval pottery/​ military
military bronzes/​ bronze
other metal
objects/​stone and
shale objects
 The Development of Artefact Studies    71

Table 4.1 Continued
Contents and order
of occurrence of
finds categories
(excluding animal
bone, human bone, Sections with
botanical remains, authorship
Reference (in etc., unless these Main organizing credited to
chronological include worked principle of small Stratigraphic others apart from
order) Site objects) finds context of finds excavator/​s

Cunliffe (1968) Richborough Masonry and Material Yes All except

(director of sculptural Roman coarse
excavation B. W. fragments/​ ware
Pearse) brooches/​other
bronze small
finds/​bone small
finds/​iron objects/​
stone mortars/​
Saxon sword/​
iron age pottery/​
Roman coarse
pottery/​samian/​
other pottery/​
coins

© Ellen Swift.

his series Excavations in Cranborne Chase (C. Evans 2007: 286), arranging the objects
shown in the plates in material categories (Pitt-​Rivers 1887). Hierarchies of material that
were already beginning to appear in the first antiquarian studies were consolidated in
the publication of reports that listed inscriptions and figurative art objects first; followed
by pottery and then the other material or functional categories (e.g. Roach-​Smith 1859).
These hierarchies were still evident over half a century later:  for instance, inscribed
objects came first in the 1911 Newstead and Corstopitum reports; inscriptions and
stamped tile were listed first in the Caerleon amphitheatre report of 1928; while sculp-
tural/​architectural stonework and inscribed objects were first in the Richborough report
of 1926 (see Table 4.1). Lucas (2001: 79) observes that objects that were susceptible to
typological analysis also tended to receive special attention, creating a self-​reinforcing
cycle in which they were then perceived as worthy of further study. This is certainly true
of pottery and brooches in the Roman period, which quickly become prominent in the
categories of object recorded, a convention still evident in Roman site reports in the late
twentieth century (see Table 4.1).
Publication of site reports and catalogues had the inevitable effect of distancing
people from the actual artefacts (C. Evans 2007: 288–​289), although it was also very
72

72   Ellen Swift

enabling to research (especially in cases where the illustrations were of high quality),
reducing the necessity of travelling to visit collections and disseminating knowledge of
artefacts to a much wider audience. Artefact expertise remained, however, in the hands
of those actually in contact with the material—​invariably private collectors of inde-
pendent means. Those most active in publishing notes on Roman artefacts during the
nineteenth century (drawn from a survey of entries on categories of small finds in the
Society of Antiquaries of London subject catalogue) include Albert Way, Revd Charles
King, Revd John Cox, Augustus Wollaston Franks, and Robert Soden Smith. All were
university-​educated, from relatively privileged backgrounds, and probably developed
their interest in Roman antiquities through education in the Classics. Only King held
a university position, as a research fellow at Cambridge (Wilson 1997: 3; Burton 2004;
Nurse 2004a, b; Wroth 2004). The presence of King and Cox is a good indication of the
preponderance of clergymen among nineteenth-​century antiquarians interested in the
Roman period. Only two definite instances of women reporting on Roman antiquities
in this period can be found: the Dowager Duchess of Cleveland, who seems to be report-
ing items on behalf of ‘His Grace’ (Cleveland 1850) and Hannah Jackson Gwilt, who was
more closely involved with the Royal Astronomical Society (Beech 2006).
As artefacts were gradually transferred from private to public collections, museum
curators—​who were likely to have previously been private collectors (as with Soden
Smith, Franks, and A.  J. Evans)—​became the first ‘professional’ artefact specialists.
Franks, for instance, in spite of a wide range of other chronological interests, published
a large number of notes on Roman artefacts in antiquarian journals, which are indexed
in the Society of Antiquaries subject catalogue. These are typical of the scholarship of
the day: most merely note the find-​spot and record the presentation of the artefact(s)
at a society meeting. Where Franks contributes a more substantial article, the artefacts
are evaluated through reference to Classical texts and depiction on Roman monuments,
and parallels from other sites are noted (see, e.g. Franks 1858, 1867–​70). Typological
studies are also evident: for example, a consideration of brooch development by Arthur
John Evans, Keeper of the Ashmolean Museum (A. J. Evans 1896). It would be a mistake,
however, to characterize later-​nineteenth-century ​studies as obsessed with typology.
Pitt-​Rivers, for example, proposed new fields of enquiry, such as the study of pottery
fabric and distribution and site assemblages as a means to evaluate the status of the
occupants (Tyers 1996: 7–​8). These were not taken up as serious endeavours, however,
until well into the twentieth century. Roach-​Smith meanwhile (who, unusually, was not
a man of private means but a chemist) made a ground-​breaking study of samian pottery
(Tyers 1996: 4–​6). This study—​using evidence from kiln sites, samian stamps, and the
form and decoration of the vessels—​established that samian was an import into Britain
from Roman Gaul (Roach-​Smith 1868b).
It is interesting to contrast the background of later-​nineteenth-​century archaeology
in Britain—​dependent on knowledgeable amateurs—​to that in Germany, where archae-
ology enjoyed more institutional support (Freeman 2007: 131–​132). Dragendorff, for
instance, who published a very important paper on samian classification towards the
 The Development of Artefact Studies    73

end of the nineteenth century (Tyers 1996: 8–​9), was an academic who had completed a
Ph.D. on the subject and been taught by professors of archaeology in Germany.

Approaches to the Classification

and Interpretation of Artefacts in 
the Earlier Twentieth Century

In the early twentieth century Romano-​British archaeology was dominated by Francis

Haverfield, the first university-​based specialist on Roman Britain (Freeman (2007) is
a detailed treatment of his life and work). Although he is more well known for his syn-
thetic work and work on inscriptions, he had a good knowledge of categories of artefacts
such as brooches and samian pottery. He tried to encourage the publication of museum
catalogues and corpuses of material (Freeman 2007: 233–​235). With regard to artefacts,
he set an example through his publication detailing the Romano-​British holdings of
provincial museums in Britain (Haverfield 1891), and towards the end of his career he
published a catalogue of the brooches in the Tullie House Museum (Haverfield 1919).
Museum catalogues incorporating Romano-​British finds within wider artefact catego-
ries also started to be produced by the British Museum (e.g. Walters 1908; Marshall 1911;
Dalton 1912); although Dalton himself noted that the authors were yet to establish exper-
tise in their subjects (Miller 1973: 316; Freeman 2007: 90), and some catalogues were
later criticized for their inaccuracies (see, e.g. Jenkins 1985). Independent, often wealthy,
figures such as the lawyers James and Alexander Curle (Graham Ritche 2002: 19–​20)
and Bushe-​Fox also contributed significantly to the publication of excavated finds. As
Lucas (2001: 65–​66) notes, in the nineteenth and early twentieth centuries finds assem-
blages were usually written up by the excavators themselves, while as the twentieth cen-
tury progressed an increasing proportion was allocated to specialists. We do commonly
find separately authored sections on coins and/​or pottery from the very early twentieth
century, however: e.g. Newstead and Richborough (Curle 1911; Bushe-​Fox 1926) (see
also Table 4.1, a larger sample of site reports than considered by Lucas, whose focus is
not specifically on the Roman period). Site finds were sometimes published separately
to the excavation report (see, e.g. J. Evans 1894; Haverfield 1911).
In addition to contributions discussing particular types of artefacts, Haverfield incor-
porated discussion of brooches and pottery into his influential synthetic account of the
‘Romanization’ of Britain (Haverfield 1912). He focused particularly on the ‘Celtic’ influ-
ence in these categories of artefacts, evident in their artistic style (Haverfield 1912: 39–​
40); this had also been remarked on previously by others (e.g. A. J. Evans 1896; Romilly
Allen 1901; Curle 1911: 317) and is something of a recurring theme as discussed later in
this chapter. Artefacts were also included in a popular and influential synthesis The
Roman Era in Britain (1911) by John Ward, Keeper of the National Museum of Wales,
74

74   Ellen Swift

who included a much wider range of objects than Haverfield. Ward’s arrangement of the
chapters on artefacts is an inconsistent mixture of material and functional categories.
Other than the pottery section, which discusses kilns and methods of manufacture as
well as descriptions of form, the information is limited to a description of the artefacts
and their varied forms, rather than discussing how the finds inform our wider picture of
Roman Britain.
This generation of scholars was from a very similar background to those of the nine-
teenth century: mostly educated at public schools and Oxford or Cambridge but without
the opportunity to study at university, let alone doctoral level in archaeology, and gain-
ing knowledge through a mixture of practical experience and informal training from
others in the field. Bushe-​Fox, for instance, gained his experience through excavation.
On the basis of his knowledge, he was later appointed Inspector of Ancient Monuments
(Wheeler 1955: 154–​155).
Although such accounts fade out as the twentieth century progresses, we encounter
some interesting imaginative reconstructions in early twentieth-​century site reports,
based, in part, on the information from the finds assemblage. These show the contin-
ued existence of a nineteenth-​century strand of interpretation from the perspective of
ancient history. Curle (1911), for instance, tends to begin each finds section of his report
from Newstead with a preamble that uses the artefacts to inform a suggested picture
of daily life at the fort. On the evidence of the finds, and against the received wisdom
that was to prevail for many succeeding decades, he suggests the presence of women
at the fort—​an idea that was not revisited until the late twentieth century (Curle 1911:
317; Allason-​Jones 1995). An integrated approach to finds is also seen in Winbolt’s dis-
cussion (1925: 167–​180) of everyday living at the Folkestone Roman villa—​an account
very reliant on sources like Pliny but also drawing on the small finds from the excava-
tions. The incorporation of finds is likely to stem, partly, from the functional arrange-
ment of material facilitating the reconstruction of site-​level activities and correlation
with ancient texts and, partly, from the fact that the authors also wrote the finds sections.
Despite these early examples of functional categorization, it became much more com-
mon in the twentieth century to arrange artefact catalogues instead by material (see
Table 4.1). The origins of this, according to Lucas (2001: 69), lie in nineteenth-​century
and even earlier antiquarian approaches to classification of all kinds of objects. Division
into material categories fostered the development of separate branches of artefact spe-
cialism, which became increasingly inward-​looking (focused, for example, on typo-
logical studies), with little reference across material categories or consideration of more
holistic questions; for instance, artefacts as site assemblages (Webster 1977: 318; Lucas
2001: 79). Material categories did not map easily onto possible function and thus activ-
ity at a site. Classification by material inevitably, then, constrained interpretation. It is
also important to note, though, that the traditional site-​report format, while discourag-
ing contextual approaches, did not actively prevent them. Many relatively early finds
reports list the material by context, in addition to the finds sections and/​or provide basic
contextual details after each find entry. Oswald’s report (1952) on Margidunum presents
the finds only by context, though this is unusual (see Table 4.1). The information was
 The Development of Artefact Studies    75

there for contextual studies, but, despite this, they were not a focus of research: context
was valued only for dating purposes.
It may be useful briefly to consider how the term ‘small finds’ became established. As
noted by Evans (C. Evans 2007: 287–​288), the ‘site-​report formula’ came into existence
only from the early twentieth century. Examining these early reports, particularly the
series of research reports of the Society of Antiquaries, we can trace the development of
the term. Initially, the word finds refers not only to the artefacts within a site report but
also to structural evidence and other material: that is, to anything that has been found.
We see this, for instance, in the title of some early publications (e.g. Donovan 1935). At
around the same time, it is evident that the term is apparently becoming a colloquial
way of referring to the artefacts from an excavation. Some site reports from the 1920s
and 1930s use it to refer to the collection of artefacts found, but place it in quotation
marks, as if it is not quite formal written English or established as a term (Winbolt 1925,
1935; Wheeler and Wheeler 1932). Others use it as a section or volume title for special-
ist reports (Corder and Kirk 1932; Nash-​Williams 1932). However, there is a good deal
of variety in practice, and many excavation reports use other headings: for instance,
‘Objects of Metal and Bone’ (Wheeler and Wheeler 1928: 161); ‘Objects Found during
the Excavations’ (Baddeley 1922); ‘Various Metal and Bone Objects’ (Whiting et al. 1931:
99). ‘Small’ seems to be used first as an adjective. Haverfield (1911), for instance, subti-
tles his report on artefacts from Corstopitum ‘Smaller Objects’ and Bushe-​Fox’s series
(1913, 1926, 1932, 1949) of Society of Antiquaries research reports use the heading ‘Small
Objects in Metal, Bone, etc.’ or ‘Small Objects in Metal, Bone, Glass, etc.’. We first see
‘small finds’ as part of a longer phrase ‘Brooches and Other Small Finds’ in the excava-
tion report on Lydney by Mortimer and Tessa Wheeler (Wheeler and Wheeler 1932: 68).
From the 1940s, ‘small finds’ is used fairly consistently by Kathleen Kenyon in her site
reports (Kenyon 1948, 1950, 1959) and now seems to have become an accepted term for
the artefacts section of a report other than pottery. Kenyon (1952: 132, 149) defines the
category in her methodological guide to excavation as objects that are not bulk finds
‘such as coins, brooches, ornaments’ or as referring to ‘special objects’. Its use in a guide
such as this presumably helped to cement the term in common usage, although we can
note that its usage to include coins as well does not match with the current established
terminology, which tends to use it for small artefacts other than coins. There is still
plenty of variation in practice for most of the twentieth century: for instance, the Society
of Antiquaries research reports using a main heading ‘The Finds’ for the section of the
report dealing with artefacts and subheadings ‘Objects of . . . ’ using material categories
(e.g. Hawkes and Hull 1947; Wacher 1969). Following the use of the title ‘Small Finds’
for the Archaeology of York series volumes on this material (e.g. Macgregor 1976, 1978)
and for Crummy’s seminal 1983 report (Crummy 1983), it then becomes established in
use and is especially seen in many excavation report finds volumes from the 1990s (e.g.
Manning et al. 1995; Bishop 1996; Cool and Philo 1998; Rees et al. 2008).
In coin studies, a shift to a different method of classification relatively early on was par-
ticularly enabling to research. The most widely used coin reference book in the initial years
of excavation reports was Cohen’s complete alphabetically arranged list of the western
76

76   Ellen Swift

Roman coinage, achieved in the mid-​nineteenth century (Babelon 1901: 219). In the 1920s,
the authors of the Roman Imperial Coinage (soon to become the standard reference work)
introduced a new system: listing coins by emperor and date or mint (Mattingly et al. 1923–​
84). This facilitated research, since this method of classification corresponded to a much
greater extent with the factors involved in the original production of the coinage.

Scholars and Their Studies from 

the 1930s to 1970s

The continuation of previous approaches driven by ancient history, in which daily life
is the subject of study, can be seen in synthesis books produced through most of the
twentieth century that do incorporate finds in their discussion of everyday activities
(Quennell and Quennell 1937; Birley 1964; Liversidge 1968). Liversidge, in particular,
draws heavily on Romano-​British artefacts in addition to ancient texts, material from
Pompeii, and so on, including discussion of artefacts found in context such as shop
and workshop contents, grave assemblages, and votive deposits. However, in contrast
to the work of her contemporary Jocelyn Toynbee (e.g. Toynbee 1962), this work was
later undervalued and not incorporated to any significant degree in later synthesis vol-
umes (Collingwood and Richmond 1969; Todd 1989). Liversidge was also omitted in a
later survey of the major figures of Romano-​British archaeology (Jones 1987; Wallace
2002). The wider failure to value her work may be related to the lesser status of Romano-​
British archaeologists among Cambridge academics of the period (Richard Reece and
Catherine Hills, pers. comm.) and/​or the focus of her research on social history.
If we turn to scholarly articles on artefacts from the 1930s up to the end of the 1960s,
the main focus is still on the compilation and descriptive publication of material,
sometimes also discussing distribution and dating (e.g. Hawkes 1940; Hildyard 1950;
Charlesworth 1959; Hawkes and Dunning 1961; Boon 1966). The article by Hawkes and
Dunning, though it also includes a typology, is rare in its consideration of the archae-
ological context of the material. Continuing from the nineteenth-​century practice of
publishing notes on small finds displayed at learned societies, most published papers
are short descriptive ‘notes’ on individual or small groups of finds (data compiled from
Society of Antiquaries subject catalogue). In pottery research, there is a focus on pottery
kilns and distribution from the 1930s onwards (Tyers 1996: 16), applying the earlier work
done on samian to other pottery types. Articles on small finds also sometimes show an
interest in trade routes and/​or the origin of artefacts or of characteristic forms or motifs
(e.g. Dalton 1922; Tonnochy and Hawkes 1931; Curle 1933; Hildyard 1945). An interest in
culture-​historical style is also seen, evident through continued discussion of the ‘Celtic’
nature of Romano-​British brooches (e.g. Collingwood 1930a; Savory 1956). O’Neil’s
paper (1935) on coins is an early attempt to evaluate and compare the coin site-​find pro-
file of four different sites (not taken further until Reece’s work of the 1970s onwards),
 The Development of Artefact Studies    77

although he himself points out the difficulties: namely, the old-​fashioned publication
convention of publishing only selective data and the scarcity of people available to carry
out the research needed (O’Neil 1935: 76–​79). We will return to these topics later, but this
comment also raises the question of who is engaged in finds research during this period.
If we examine the profile of those contributing articles on artefacts to scholarly jour-
nals, it is evident that it is dominated by museum personnel, who were also instrumental
in passing on knowledge and providing opportunities for new generations of artefacts
scholars. Dalton and Tonnochy held positions at the British Museum, as did Hawkes
until he transferred to a chair at Oxford in 1946 (Champion 2004); Stevenson was at the
National Museum of Antiquities Scotland; Liversidge was an Honorary Keeper of the
University Museum of Archaeology and Anthropology at Cambridge before she was
appointed as a Research Fellow of Newnham College and later a Cambridge faculty lec-
turer (Archives Hub 2012; Catherine Hills, pers. comm.); Hull was Curator at Colchester
Museum (Hawkes 1982); Charlesworth worked in a number of museums and at the
Ancient Monuments Inspectorate (Serpell 1982); Boon followed Nash-​Williams at the
National Museum of Wales (Manning 1995); Harden employed Kirk at the Ashmolean
(Young 2007) and later moved to the London Museum (Painter and Thompson 1994).
Collingwood is rare as someone with expertise on artefacts who had always been based
in a university. We can also note the appearance of women; to comply with changes in
the law, women were allowed to become Fellows of the Society of Antiquaries from 1920
(J. Evans 1956: 388–​9). Although there is a noticeable dip in the 1930s and 1940s, prob-
ably linked in some way to the disruptions of the Second World War, women increase
noticeably as a proportion of those working on small finds from the 1950s onwards (see
Figure 4.1(a)).
As before, most of these scholars came from privileged backgrounds and were pri-
vately educated, mostly at public schools, before studying at Oxford or Cambridge;
although few had the opportunity to study for a doctorate. Boon is typical in that he had
a degree in Latin but no postgraduate qualifications, building up his expertise through
museum posts (Manning 1995:  p.  xiii). Webster continued the tradition of the self-​
taught expert. Initially an engineer, he later gained a post at the Grosvenor Museum,
Chester, based on his experience gained through archaeological excavation (Henig and
Soffe 2002).
It is instructive to compare the two editions of the core textbook The Archaeology of
Roman Britain (Collingwood 1930b; Collingwood and Richmond 1969) to gauge archae-
ological approaches to artefact studies across the same period. Both editions include
chapters on coins, samian, coarse pottery, brooches and a chapter on weapons, tools, and
utensils. In neither is there any attempt to integrate the artefacts into a wider picture, and
their value is mainly seen in terms of dating. However, the potential of pottery and coins
as economic evidence is acknowledged. Collingwood’s chapter (1930b: 185–​199) on coins
begins with a discussion of the archaeological study of coinage, as opposed to numis-
matic history or art-​historical approaches. It is evident in the revised chapter and its bib-
liography that, while hoarding has been studied further, more work has been done in the
intervening forty-​odd years on the structure of the coinage and the monetary system:
78

(a)
100
90
80
70
60
Percentage

Women
50
Men
40
30
20
10
0
9

8
–0

–1

–2

–3

–4

–5

–6

–7

–8
00

10

20

30

40

50

60

70

80
19

19

19

19

19

19

19

19

19
(b)
100
90
80
70
60
Percentage

Women
50
Men
40
30
20
10
0
1970s 1980s

(c)
Coins, pottery or small finds
Animal bone, botanical
remains or scientific analysis
Mix of finds and other
Other

Figure 4.1  Gender profiles in artefact studies.

(a) Percentage of men and women with published articles on small finds listed in the Society of
Antiquaries subject catalogue from 1900 to 1988 (end date of catalogue). (b) Percentage of men
and women publishing articles in Britannia 1970–​89 (excluding reviews). (c) Subject of articles by
women in Britannia 1970–​89 (excluding reviews).
Source: © Ellen Swift.
 The Development of Artefact Studies    79

that is, not those areas Collingwood prioritized as ‘archaeological’. There is still no men-
tion of studies of coins as site finds, despite O’Neil’s work. We can see here the influence
of long traditions of numismatic study overriding a nascent awareness of the potential
of other approaches. In the chapter on samian in the later edition, continuity of tradi-
tions of study is also evident. Samian was first valued for dating; it is noted that deco-
rated sherds can now be dated more precisely and stratified decorated samian is seen as
the publication priority (Collingwood and Richmond 1969: 236, 249–​250). The revised
chapter on coarse pottery includes a new section on production centres and a greater
range of fabric-​type descriptions, reflecting the work done on these areas by scholars
such as Gillam and Hartley in the intervening period. The other finds sections show very
few changes, which is surprising, since Richmond himself encouraged his students to
take up artefact research (Richard Reece, pers. comm.). In 1969 (as in 1930), the impor-
tance of brooches lies in typology, dating, and distribution, but, whereas Collingwood,
in 1930, is also interested in ‘artistic style’ and its importance for understanding cultural
history, Richmond dismisses it as not the proper focus of archaeologists (Collingwood
1930b: 243; Collingwood and Richmond 1969: 286). The final chapter, on weapons, tools,
and utensils, is descriptive in both cases—​a listing of the various types of items found in
the Roman period—​but the later edition incorporates more information on manufac-
ture and use of the artefacts (Collingwood 1930b: 261–​274; Collingwood and Richmond
1969: 304–​325). Frere’s synthesis volume of the same period incorporates finds research
to a much greater extent, integrated within the narrative rather than occurring in sepa-
rate chapters. Artefact research contributes, for instance, to questions of production and
trade: for example, the market competition of the various pottery industries and consid-
eration of imported and exported objects (Frere 1967: 29–​294). Social history of the kind
done by Liversidge was not easily incorporated within the framework of political and
economic history that structured these books.

The Development of Artefact Studies

from the 1970s Onwards

Alongside the continuing production of essential corpora of small finds (e.g. Henig
1974; Manning 1976; Green 1978), the 1970s saw a number of developments in studies
of both pottery and coins broadly in line with wider developments in archaeological
interpretation that focused on evidence as scientific data (notably D. Clarke 1968).
The Study Group for Roman Pottery was founded in 1971, and new techniques in pot-
tery, such as thin-​section analysis, combined with massive increases in available data
(Tyers 1996: 2–​22), contributed to established research areas (e.g. origin and distribu-
tion). This combination of valued research questions, new data, and developments in
methodology—​including an emphasis on scientific methods—​ensured that pottery
studies flourished, focusing on the economy and trade. Webster (1977) saw the priorities
80

80   Ellen Swift

as the use of more objective descriptive terms (replacing the undesirable subjectivity
of extant fabric descriptions, for instance), and more work on manufacture and distri-
bution. The increase in data from the 1970s onwards also necessitated a consideration
of quantification methods (Tyers 1996: 22–2​3). In coin studies, new methodologies for
the study of site finds were developed by Reece and Casey, including the allocation of
coins to specific period divisions, which enabled statistical comparison (e.g. Reece 1973;
Casey 1974), and culminating in the edited volume Coins and the Archaeologist (Casey
and Reece 1974), setting out specifically archaeological—​as opposed to numismatic—​
approaches to coin finds. The edited book Roman Crafts (Strong and Brown 1976) illu-
minates the contribution of scientific and technical studies to artefacts research in this
period. Reference to contemporary craft and manufacture enabled a detailed consider-
ation of production techniques. The title reflects the wider popularity of handcrafts, folk
art, and so on in the 1970s.
Of the contributors on artefacts in Roman Crafts and elsewhere, Manning, Price, and
Wild have a university affiliation, showing a gradual spread of artefact researchers into
academia—​a trend also demonstrated by coins and pottery specialists such as Reece,
Casey and Peacock. As Jones (1987: 90) notes, most of the 1970s generation of academ-
ics in Roman archaeology were unable to study archaeology as undergraduates, but,
in career trajectory, they benefited from the expansion of universities in the 1960s and
1970s. Yet they are still outnumbered by contributors based in museums or otherwise
outside the universities. Contributors to the edited volume Coins and the Archaeologist
(Casey and Reece 1974) also include only three university researchers, among many
museum-​based researchers.
In the 1980s, approaches to pottery include comparisons to ethnographic data
in addition to a strong interest in the details of production methods and technolo-
gies (Peacock 1982). More systematization was achieved through new guidelines for
recording and the increasing use of computer programs (Tyers 1996: 23). There was
also renewed interest in functional pottery studies in this decade (Tyers 1996: 43).
We can see pottery studies diverging further from Roman archaeology as a particu-
lar subdiscipline in the foundation of the Journal of Roman Pottery Studies in 1986. In
small-​finds studies, Crummy’s Colchester report (1983) was the first major finds cata-
logue since Curle’s Newstead report (1911) to arrange artefacts in functional rather than
material categories—​potentially encouraging the study of artefacts in relation to each
other as functional assemblages rather than in the abstract as typological sequences.
An interest in artefact function is evident in some journal articles of the 1980s (e.g.
Jackson 1985); however, change was very slow, with a continued emphasis on tradi-
tional approaches in small-​finds studies (typology, distribution) and the publication of
previously unpublished artefacts (Swift 2007: 22; data from the Society of Antiquaries
subject catalogue). Notwithstanding new cataloguing trends, the dominance of mate-
rial categories in already published site reports (see Table 4.1) also continued to affect
the way that material was interpreted more widely. For example, in Philpott’s analy-
sis of burial rites, burials with vessels made from different materials are considered
in separate chapters, even though Philpott himself points out that the material from
 The Development of Artefact Studies    81

which grave goods are made might not be ritually significant (Philpott 1991: 127, 117).
The separation inevitably fosters a discussion of status as expressed through more or
less prestigious materials, whereas a consideration of graves with similar vessel forms,
irrespective of material, might have resulted in a greater emphasis on ritual practice. In
journal articles, books, and theses there was a continuing tendency to study only one
type of artefact or material category (e.g. Lloyd-​Morgan 1981; Allason-​Jones 1989a),
rather than integrated studies of assemblages of diverse finds, although, as we have
seen in earlier periods, a separate strand of research concerned with everyday life does
sustain a consideration of artefacts as functional groups (e.g. Allason-​Jones 1989b).
Museum expertise continues in the 1980s in both the documentation and interpreta-
tion of finds (e.g. Johns et al. 1980; Boon 1983), and independent scholars also continue
to make significant contributions (e.g. Hattatt 1982, 1985). We also see publications based
on theses on Roman finds completed in the later 1970s and 1980s (e.g. Lloyd-​Morgan
1977; Green 1978; White 1988; Allason-​Jones 1989a). Indeed, artefact study comprises
around half of all doctoral research in Roman archaeology in the 1980s (Swift 2007: 21–2​
2); this increase in research perhaps contributed to the founding of the Roman Finds
Group in 1988. Yet there are still comparatively few artefact specialists in university
posts, and these are mostly coin and pottery researchers (e.g. Casey, Reece, Fulford,
Greene, Hartley, Peacock, Gillam) who were appointed in the 1970s and earlier. It is
important to note the wider dearth of new posts in archaeology at this time (Mattingly
2008). Although there was a good deal of finds research being carried out by the 1980s,
the wider context was different; unlike pottery researchers in the period of university
expansion in the 1970s, researchers on finds in the 1980s were unable to gain university
posts, thanks to contraction and retrenchment during this time.
To consider gender profiles:  by the 1980s, women are apparently overrepresented
in finds studies, making up around a quarter of all finds researchers (Figure 4.1(a)).
Although this is only slightly greater than the proportion of women publishing in
the journal Britannia on any subject in the 1970s and 1980s (Figure 4.1(b)), it is evi-
dent that these published Britannia contributions by women are themselves heavily
biased towards artefact and other specialist reports (Figure 4.1(c)). By contrast, in the
same period only about a quarter of the articles by men in Britannia are on finds top-
ics. The connection between women and artefact studies was noted at the time (Jones
1987: 95) and is further discussed later in this chapter.
A synthesis volume summarizing the major research in Romano-​British archaeology
between 1960 and 1989 (Todd 1989) provides an opportunity to examine the integra-
tion of finds research into the wider picture of Roman Britain in this period. Although
there are no chapters specifically concerned with artefacts, or with daily life, a number
of contributors incorporate research on pottery, coins, and/​or precious metal objects
into overviews of various topics, including art, the economy, and religion in both the
early and late Roman periods (Blagg 1989: 209–​210; Esmonde Cleary 1989; Fulford
1989; Henig 1989). We can see here the continued existence of hierarchies in material.
Research on other objects is mentioned by only a couple of authors (Breeze 1989: 38;
Keppie 1989: 68–​69). In the bibliography, publications on artefacts are concentrated in
82

82   Ellen Swift

a section on industries and crafts (Todd 1989: 261–26​2), probably because this had been
the focus of new research in the 1960s and 1970s (most notably Strong and Brown 1976).
The 1990s and early 2000s saw a number of developments both in methodologi-
cal issues and in interpretation. Roman finds work continued to be characterized by a
strongly empirical approach (Hingley and Willis 2007: 7–​8). Systematization of data
recording continued, with the establishment, for instance, of the National Roman
Pottery Fabric Reference Collection (Tomber and Dore 1998) and a research framework
for Roman pottery (Willis 2004: 1). The application of comparative statistical methods
to artefacts datasets is seen, and in this we also see computer analysis moving beyond
coins and pottery to other classes of artefact (e.g. Cool and Baxter 2002). From the late
1990s onwards, site reports more widely adopt the functional artefact categories used by
Crummy (1983).
More widely in Roman archaeology, the 1990s is the decade that saw the increas-
ing incorporation of theoretical approaches that had already become established in
other branches of archaeology (James 2003: 180). Explicit theoretical approaches are,
unsurprisingly, most evident in finds-​based articles in the proceedings volumes of the
Theoretical Roman Archaeology Conference (TRAC) from 1991 onwards. Earlier arte-
fact-​based contributions to the TRAC series focus on questioning established assump-
tions (e.g. Allason-​Jones 1995; S. Clarke 1997), while in later volumes there is more of
a tendency to bring in theoretical frameworks more explicitly, particularly in rela-
tion to social, cultural, and ethnic identity. This is clearly the dominant focus of inter-
est, incorporating, for instance, postcolonial or consumption theory (see Fincham et
al. 2000; Fenwick et al. 2008). A particularly influential book is Hodder’s Symbols in
Action (1982). Most of the articles consider one artefact type or category such as pot-
tery vessels, despite the call for a more integrated approach to finds assemblages in the
first TRAC proceedings (J. Evans 1995). Some old foci of interest are revisited, notably
the ‘Celtic’ nature of some Romano-​British brooches, reframed as a possible example of
‘resistance’ to Roman culture (Jundi and Hill 1998). In addition, many articles focus on
contextual information to reinterpret what might be termed ‘pragmatic’ or functionalist
explanations of evidence in more post-​processual terms, particularly focusing on rit-
ual behaviour (e.g. S. Clarke 1997; Hutcheson 1997; Dungworth 1998; van Driel Murray
1999) and clearly heavily influenced by Hill’s work (1995) on structured deposition and/​
or Merrifield’s Archaeology of Ritual and Magic (1987).
Ironically, given the developing focus here and elsewhere on contextual site studies of
Roman material culture (e.g. Allison 2004; Hingley and Willis 2007), a major resource
for new small finds and coins data from the later 1990s has been the Portable Antiquities
Scheme (PAS), founded in 1997, which records stray finds—​mostly metal-​detected
material—​and publishes a web database that has seen rapid expansion (Clark 2008; PAS
2012). Distribution maps, which used to be the product of laborious data collection,
can now be mapped instantly and thus are likely to continue to dominate PAS research,
as they have wider artefact research for much of the twentieth century, though within
new interpretative frameworks such as landscape studies. PAS data will be particularly
important in transforming our picture of rural sites (Hingley and Willis 2007: 11–​12;
 The Development of Artefact Studies    83

Clark 2008:  19; PAS 2012) and listings on the website of finds research undertaken
using PAS data confirm that it is flourishing as a research area. However, PAS data are
only slowly being incorporated into wider artefact research (e.g. Eckardt and Crummy
2008: app. 6; Swift 2010: app. 4), rather than being the subject of stand-​alone projects.
In an article in 2007, I surveyed, through a questionnaire, prevailing views of Roman-​
finds research and attempted to evaluate the profile of artefact researchers. About a
quarter of researchers had a university affiliation at either staff or student level; slightly
more were engaged as freelance or unit-​based specialists, and 30 per cent were based
in museums. Strong gender bias was evident, with women over-​represented in finds
research compared to their numbers in archaeology overall and with a sharp divide
between men specializing in coins or pottery and women focusing on small finds or
small finds and pottery (interestingly, the gender bias in coins study is not evident on
the continent: Richard Reece, pers. comm.). A widespread negative perception of finds
research was documented, and it was suggested that this might be related to: first, the
persistence of old-​fashioned views of what finds research was or could do; secondly,
the position of finds research on the margins of academic study (including its position
as ‘women’s work’); and, thirdly, wider trends in theoretical approaches that attempted
to undermine empiricism. More broadly, it was also suggested that the negative per-
ception of finds research was connected with hierarchies of value within archaeologi-
cal research, in which fieldwork and archaeological theory have been dominant, and
in historical research, which has privileged political and economic history over social
history. The prevalence of women in small-​finds research in particular, documented
in this chapter to exist in the 1970s and 1980s as well as more recently, is clearly con-
nected to this. Women, historically, have tended to focus on social history and may also
have taken a different attitude to established research hierarchies or been constrained by
those hierarchies and wider prejudices into working in what have been perceived to be
‘less important’ areas (Swift 2007). My consideration of a longer time span in the current
chapter also provides additional evidence that social history has been a focus of research
by women scholars that has generally been sidelined in Romano-​British archaeology.
In addition, a negative view of small-​finds research becomes unsurprising when one
considers that archaeologists may have been basing their opinions, for instance, on the
extremely limited picture, barely updated from the 1930s, given in one of the major text-
books of the later twentieth century (Collingwood and Richmond 1969).
Both in my article and elsewhere multiple causes are suggested for well-​documented
declining numbers of artefact specialists and artefact research (see, e.g. Allason-​Jones
2001; Willis 2004); the general undervaluing of finds studies, which is a contributory
factor (among other things such as retrenchment in the 1980s) to reduced numbers of
researchers in universities; researchers in museums not being accorded the same sta-
tus as university academics; structural problems, such as the lack of a defined career
path for artefact researchers outside the universities; work commitments in unit and
museum archaeology necessitating a focus on essential catalogues of material rather
than synthetic and/​or interpretative articles; and an increasing drift in museum exper-
tise and priorities away from artefact knowledge and research towards education and
84

84   Ellen Swift

outreach (Cooper 2007; Johns 2007; Swift 2007). Since there have been fewer research-
ers altogether, often only one person has studied a particular artefact type in depth
(Allason-​Jones 2001: 24). Consequently, methodologies and interpretations have not
been debated, and the development of research—​or lack of it in a particular area—​has
been very dependent on a single individual.
There have, however, been some attempts to address essential skills shortages in arte-
fact identification; for instance, through an Arts and Humanities Research Council grant
scheme for doctoral funding specifically limited to finds research (Fulford 2007) and
through training as part of the PAS. The latter has contributed enormously in changing
the picture in very recent years, not only providing training but reconnecting museums
and artefact research through museum-​based posts, encouraging the reporting of new
material, and providing new datasets for M.A. and Ph.D. research projects (Clark 2008).

Discussion

A number of recurring, interlinked issues that have affected artefact studies can be
drawn out for further discussion, evident from the earliest days of antiquarian studies
to the early twenty-first century: the availability and presentation of data; the status and
profile of researchers; traditions of scholarship, and the influence of wider interpretative
trends in academia.
Artefact research is unlike text-​based scholarship in that it depends on material objects.
Given that few people have been able to travel extensively to view collections, publication
and description of the objects have always been priorities: from the coins dictionaries of
the eighteenth and nineteenth centuries, through the publication of catalogues of major
museum collections of artefacts in the twentieth, to today’s web databases. Researchers
have adopted new technologies accordingly: from cast copies of coins to 3-​D scanning
of artefacts. Yet there have always been problems with selectivity of publication and in
the categorization and quantification of data, which have particularly affected inter-​site
comparison of material. Hierarchies in the way that data are published and arranged have
been very persistent, affecting what has been studied and how, with some finds catego-
ries suffering persistent neglect. Effectively, one generation has unwittingly decided what
could be of interest to the future scholars of another. In the earliest site reports, functional
grouping of finds assemblages related initially to a wider interest in the interpretation of
objects through reference to ancient texts, yet enabled the integration of the artefacts into
the interpretation of a site. Later, the almost universal decision to shift to material catego-
ries—​together with an increasing quantity of data—​encouraged fragmentation into nar-
row research specialisms and publications on one category of artefact or type of material,
discouraged the integration of finds data into site interpretations, and fostered particular
interpretations based on material differences. Conventions established in the 1970s have
also not had solely positive effects. Specialist new methodologies have meant that arte-
fact researchers have effectively become further separated from generalist interpreters
 The Development of Artefact Studies    85

and excavators (Johns 2007: 30). As Tyers (1996: 22) notes, the use of data from quanti-
fied pottery reports has been hampered by the frequent decision not to quantify samian
and mortaria and by a plethora of different methods of quantification. In coins studies,
the use of different period divisions by Casey and Reece (followed by those they taught)
potentially affects site comparisons; though, since full coins list have usually been pub-
lished, in practice this is not so much of a problem. Lockyear (2007: 212–2​14) discusses
other issues with regard to the cataloguing of coins: for instance, condensation of data
and the provision of summary information only.
To reiterate: artefact research is unlike text-​based scholarship in that it depends on
material objects. This is, undoubtedly, one reason why artefact specialists have more
often been private collectors, museum curators, and independent finds experts work-
ing in unit archaeology rather than university-​based academics. Museum scholarship
has had different priorities from university research, with, for instance, an emphasis on
connoisseurship (Johns 2007: 29–​31) and the publishing of catalogues of essential data.
Arguably, this is another factor that contributes to the differing trajectory of artefact
research (especially small-finds research) compared to other scholarly work on Roman
Britain, in which there has been more of a drive towards synthesis and broad interpreta-
tion. Notwithstanding, there are a number of core topics to which finds evidence can
make a strong contribution, and interpretation has consisted of different iterations of
these topics: studies of style; the study of production and trade; the reconstruction of
daily life; studies of ancient technology; and consideration of artefact function. We
can see a slightly differing emphasis on one topic or another through time, related to
wider trends in archaeological interpretation such as culture history, processualism, and
post-​processualism, which have also affected the way the topic has been approached;
but each topic in itself has been present from the nineteenth to the twenty-​first centu-
ries. The documentation and description of material, with further interpretation limited
to typology and distribution, have also been persistent traditions. This is not to deni-
grate the value of this work; such studies are a fundamental stage in any finds research.
Documentation through images and description is obviously essential, and, as regards
typology and distribution, we cannot hope to understand the objects in their social,
cultural, and economic context unless this is defined as closely as possible in time and
space. However, we can see that the focus on description and typology as an end in itself
retarded the take-​up of new approaches. In all areas of finds research, there was innova-
tion in both methodology and interpretation that was overlooked for decades because it
could not be accommodated within prevailing habits of scholarship. It is also clear that,
while work done on aspects such as production, technology, and trade was acknowl-
edged more widely in archaeological scholarship, artefact research on everyday life, in
which the archaeological context of material was considered, has been less valued. This
has particularly affected small-​finds research.
If we take the long view of artefact research across the twentieth century and notwith-
standing the many problems that remain, it actually seems quite encouraging that so much
has been achieved in the face of so many obstacles. The first decade of the twenty-first
century has seen a developing, more positive approach to material culture studies—​at
86

86   Ellen Swift

least in the domain of academic archaeology (see, e.g. Burnham et al. 2001: 74–​75; Hingley
and Willis 2007). New edited compilations focus on artefact studies in particular and not
just on pottery or coins (Aldhouse-​Green and Webster 2002; Hingley and Willis 2007;
Allason-​Jones 2011). Multiple categories of finds data are increasingly featuring in pub-
lished artefact research on Roman Britain (including the study of finds assemblages), and
artefacts are more often integrated with other types of evidence (e.g. Cool 2006; Gardner
2007; Pitts 2010). There is more effort to provide in-​depth interpretation of a finds assem-
blage in relation to the site overall, particularly with regard to contextual analysis of nota-
ble deposits (e.g. Cool 1998, 2004; Eckardt 2006; Rees et al. 2008), although the failure to
do this more widely or to point out this shortcoming in reviews of excavation reports is
still seen as a significant problem (Cool 2006: 244–​245). The reconciliation of empirical
methodologies with post-​processual interpretative approaches has been essential to the
integration of theory within finds studies that is needed for innovation in artefact research
(Hingley and Willis 2007: 8; Swift 2007). The first decades of the twenty-​first century may
be the beginning of a new phase in small-​finds research—​similar to the step change in
coins and pottery studies in the 1970s—​where a new value for material culture studies and
social history coincides with a massive expansion in available and high-​quality data, new
analytical methods for its interrogation, and theoretically informed approaches to inter-
pretation—​that is, provided the problems that we inherit can be overcome.

Acknowledgements

Many thanks to Peter Guest for providing me with an overview of coins literature in
relation to Roman Britain from the 1970s onwards, Anne Alwis and Richard Reece for
useful comments on the text, and Catherine Hills for providing additional detail on Joan
Liversidge.

References
Akerman, J. (1844). Coins of the Romans Relating to Britain, Described and Illustrated. 2nd edn.
London: J. R. Smith.
Aldhouse-​Green, M., and Webster, P. (2002). Artefacts and Archaeology: Aspects of the Celtic
and Roman Worlds. Cardiff: University of Wales Press.
Allason-​Jones, L. (1989a). Earrings in Roman Britain. BAR British Series 201.
Oxford: Archaeopress.
Allason-​Jones, L. (1989b). Women in Roman Britain. London: British Museum Press.
Allason-​Jones, L. (1995). ‘“Sexing” Small Finds’, in P. Rush (ed.), Theoretical Roman Archaeology:
Second Conference Proceedings. Aldershot: Avebury, 22–​32.
Allason-​Jones, L. (2001). ‘Material Culture and Identity’, in S. James and M. Millett (eds),
Britons and Romans:  Advancing an Archaeological Agenda. CBA Research Report 125.
York: Council for British Archaeology, 19–​25.
 The Development of Artefact Studies    87

Allason-​Jones, L. (2011). Artefacts in Roman Britain:  Their Purpose and Use.

Cambridge: Cambridge University Press.
Allison, P. M. (2004). Pompeian Households:  An Analysis of Material Culture. Los
Angeles: Cotsen Institute of Archaeology, University of California, Los Angeles.
Archives Hub (2012). Joan Liversidge: Administrative/​Biographical History <http://​archiveshub.
ac.uk/​data/​gb1932lcca-​lp-​5> (accessed 27 January 2012).
Ashby, T., Hudd, A. E., and Martin, A. T. (1904). ‘VII:  Excavations at Caerwent,
Monmouthshire, on the Site of the Romano-​British city of Venta Silurum, in the years 1901-​
1903’, Archaeologia, 59: 87–​124.
Babelon, E. (1901). Traité des Monnaies Grecques et Romaines [volume  1]. Bologna:  Forni
Editore.
Baddeley, W. S. (1922). ‘The Excavation at Cirencester, February–​March 1922’, Transactions of
the Bristol and Gloucestershire Archaeological Society, 44: 101–​115.
Beech, M. (2006). ‘The cometarium by John Taylor’, Bulletin of the Scientific Instrument Society,
88: 1–​6.
Birley, A. (1964). Life in Roman Britain. London: Batsford.
Bishop, M. C. (1996). Finds from Roman Aldborough:  A  Catalogue of Small Finds from the
Romano-​British Town of Isurium Brigantium. Oxford: Oxbow Books.
Blagg, T. (1989). ‘Art and Architecture’, in M. Todd (ed.), Research on Roman Britain, 1960–​
1989. Britannia Monograph Series 11. London: Society for the Promotion of Roman Studies,
203–​218.
Boon, G. C. (1966). ‘Gilt-​Glass Beads from Caerleon and Elsewhere’, Bulletin of the Board of
Celtic Studies, 22: 104–​109.
Boon, G. C. (1983). ‘Potters, Oculists and Eye-​Troubles’, Britannia, 14: 1–​12.
Bowden, M. (1991). Pitt Rivers:  The Life and Archaeological Work of Lieutenant-​General
Augustus Henry Lane Fox Pitt Rivers, DCL, FRS, FSA. Cambridge:  Cambridge
University Press.
Breeze, D. J. (1989). ‘The Northern Frontiers’, in M. Todd (ed.), Research on Roman Britain,
1960–​89. Britannia Monograph Series 11. London:  Society for the Promotion of Roman
Studies, 37–​60.
Burnham, B., Collis, J., Dobinson, C., Haselgrove, C., and Jones, M. (2001). ‘Themes for Urban
Research’, in S. James and M. Millett (eds), Britons and Romans: Advancing an Archaeological
Agenda. CBA Research Report 125. York: Council for British Archaeology, 67–​76.
Burton, A. (2004). ‘Smith, Robert Henry Soden (1822–​1890)’, Oxford Dictionary of National
Biography. Oxford: Oxford University Press (online edn).
Bushe-​Fox, J. P. (1913). Excavations on the Site of the Roman Town at Wroxeter, Shropshire in
1912. Research Report of the Society of Antiquaries of London, volume 1. Oxford: Oxford
University Press for the Society of Antiquaries of London.
Bushe-​Fox, J. P. (1926). First Report on the Excavation of the Roman Fort at Richborough,
Kent. Research Report of the Society of Antiquaries of London, volume 6. Oxford: Oxford
University Press for the Society of Antiquaries of London.
Bushe-​Fox, J. P. (1932). Third Report on the Excavation of the Roman Fort at Richborough,
Kent. Research Report of the Society of Antiquaries of London, volume 10. Oxford: Oxford
University Press for the Society of Antiquaries of London.
Bushe-​Fox, J. P. (1949). Fourth Report on the Excavation of the Roman Fort at Richborough,
Kent. Research Report of the Society of Antiquaries of London, volume 16. Oxford: Oxford
University Press for the Society of Antiquaries of London.
88

88   Ellen Swift

Casey, P. J. (1974). ‘The Interpretation of Romano-​British Site Finds’, in P. J. Casey and R. Reece
(eds), Coins and the Archaeologist. BAR British Series 4. Oxford: Archaeopress, 37–​51.
Casey, P. J., and Reece, R. (1974). Coins and the Archaeologist. BAR British Series
4. Oxford: Archaeopress.
Champion, T. (2004). ‘Hawkes, Charles Francis Christopher (1905–​1992)’, Oxford Dictionary of
National Biography. Oxford: Oxford University Press (online edn).
Chapman, W. (1985). ‘Arranging Ethnology: A. H. L. F. Pitt Rivers and the Typological Tradition’,
in G. W. Stocking (ed.), Objects and Others: Essays on Museums and Material Culture. History
of Anthropology Series, volume 3. Wisconsin: University of Wisconsin Press, 15–​48.
Charlesworth, D. (1959). ‘Roman Glass in Northern Britain’, Archaeologia Aeliana, 4th edn.
37: 33–​58.
Clark, K. (2008). ‘A Review of the Portable Antiquities Scheme’, Portable Antiquities Scheme
Website. Museums, Libraries and Archives Council <http://​finds.org.uk/​blogs/​centralunit/​
files/​2008/​11/​pas-​final.pdf> (accessed 23 May ​2012).
Clarke, D. (1968). Analytical Archaeology. London: Methuen.
Clarke, S. (1997). ‘Abandonment, Rubbish Disposal and “Special” Deposits at Newstead’,
in K. Meadows, C. Lemke, and J. Heron (eds), TRAC 96:  Proceedings of the Sixth
Annual Theoretical Roman Archaeology Conference, Sheffield 1996. Oxford:  Oxbow
Books, 73–​81.
Cleveland, Dowager Duchess of (1850). ‘Note on Four Gold Rings (One Roman) Found
at Pierse Bridge, Durham and at Wanborough, Prudhoe Castle, and Colchester near
Cambridge, Northumberland’, Archaeological Journal, 7: 191–​192.
Collingwood, R. G. (1930a). ‘Romano-​Celtic Art in Northumbria’, Archaeologia, 80: 37–​58.
Collingwood, R. G. (1930b). The Archaeology of Roman Britain. London: Methuen.
Collingwood, R. G., and Richmond, I. A. (1969). The Archaeology of Roman Britain. 2nd edn.
London: Methuen.
Cool, H. E. M. (1998). ‘Life in Roman Castleford’, in H. E. M. Cool and C. Philo (eds), Roman
Castleford. Volume 1: The Small Finds. Yorkshire Archaeology Series 4. Wakefield: Yorkshire
Archaeology Service, 355–​373.
Cool, H. E. M. (2004). The Roman Cemetery at Brougham. Britannia Monograph Series 21.
London: Society for the Promotion of Roman Studies.
Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge
University Press.
Cool, H. E.  M., and Baxter, M. J. (2002). ‘Exploring Romano-​British Finds Assemblages’,
Oxford Journal of Archaeology, 21/​4: 365–​380.
Cool, H. E.  M., and Philo, C. (1998) (eds). Roman Castleford. Volume 1:  The Small Finds.
Yorkshire Archaeology Series 4. Wakefield: Yorkshire Archaeology Service.
Cooper, N. J. (2007). ‘Promoting the Study of Finds in Roman Britain: Democracy, Integration
and Dissemination: Practice and Methodologies for the Future’, in R. Hingley and S. Willis
(eds), Roman Finds: Context and Theory. Oxford: Oxbow Books, 35–​52.
Corder, P., and Kirk, J. (1932). A Roman Villa at Langton, near Malton, East Yorkshire.
Leeds: Roman Antiquities Committee of the Yorkshire Archaeological Society.
Crane, S. (1999). ‘Story, History and the Passionate Collector’, in M. Myrone and L. Pelz
(eds), Producing the Past:  Aspects of Antiquarian Culture and Practice 1700–​ 1850.
Woodbridge: Ashgate, 187–​203.
Crummy, N. (1983). The Roman Small Finds from Excavations in Colchester. Colchester
Archaeological Reports 2. Colchester: Colchester Archaeological Trust.
 The Development of Artefact Studies    89

Cunliffe, B. (1968). Fifth Report on the Excavations of the Roman Fort at Richborough, Kent.
Research Report of the Society of Antiquaries of London, volume 23. Oxford:  Oxford
University Press for the Society of Antiquaries.
Curle, J. (1911). A Roman Frontier Post and its People: The Fort of Newstead in the Parish of
Melrose. Glasgow: James Maclehose and Sons for the Society of Antiquaries of Scotland.
Curle, J. (1933). ‘An Inventory of Objects of Roman and Provincial Roman Objects Found on
Sites in Scotland not Definitely Associated with Roman Constructions’, Proceedings of the
Society of Antiquaries of Scotland, 66: 277–​400.
Dalton, O. M. (1912). Catalogue of the Finger-​Rings, Early Christian, Byzantine, Teutonic,
Medieval and Later, in the British Museum. London: Longmans.
Dalton, O. M. (1922). ‘Roman Spoons from Dorchester’, Antiquaries Journal, 2: 89–​92.
Donovan, H. (1935). ‘Roman Finds in Bourton-​on-​the-​Water’, Transactions of the Bristol and
Gloucestershire Archaeological Society, 57: 234–​259.
Dungworth, D. (1998). ‘Mystifying Roman Nails: Clavus annalis, defixiones and Minkisi’, in
C. Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 97: Proceedings of the Seventh Annual
Theoretical Roman Archaeology Conference, Nottingham 1997. Oxford:  Oxbow Books,
149–​159.
Eckardt, H. (2006). ‘The Character, Chronology and Use of the Late Roman Pits: The Silchester
Finds Assemblage’, in M. Fulford, A. Clarke, and H. Eckardt (eds), Life and Labour in Late
Roman Silchester:  Excavations in Insula IX since 1997. Britannia Monograph Series 22.
London: Society for the Promotion of Roman Studies, 221–​244.
Eckardt, H., and Crummy, N. (2008). Styling the Body in Late Iron Age and Roman
Britain: A Contextual Approach to Toilet Instruments. Instrumentum Monograph Series 36.
Montagnac: Instrumentum.
Esmonde Cleary, S. (1989). ‘Constantine I  to Constantine III’, in M. Todd (ed.), Research
on Roman Britain, 1960–​1989. Britannia Monograph Series 11. London:  Society for the
Promotion of Roman Studies, 235–​244.
Evans, A. J. (1896). ‘On Two Fibulae of Celtic Fabric from Aescia’, Archaeologia, 55: 179–​198.
Evans, C. (2007). ‘“Delineating Objects”: Nineteenth Century Antiquarian Culture and the
Project of Archaeology’, in S. Pearce (ed.), Visions of Antiquity: The Society of Antiquaries of
London 1707–​2007. London: Society of Antiquaries, 267–​305.
Evans, J. (1894). ‘IX: On Some Iron Tools and Other Articles Formed of Iron Found at Silchester
in the Year 1890’, Archaeologia, 54: 139–​156.
Evans, J. (1956). A History of the Society of Antiquaries. Oxford: Oxford University Press.
Evans, J. (1995). Roman Finds Assemblages: Towards an Integrated Approach?’, in P. Rush (ed.),
Theoretical Roman Archaeology: Second Conference Proceedings. Aldershot: Avebury, 33–​58.
Faussett, B. (1856 [1773]). Inventorium Sepulchrale: An Account of some Antiquities Dug up at
Gilton, Kingston, Sibertswold, Barfriston, Beakesbourne, Chartham, and Crundale, in the
County of Kent, ed. C. Roach-​Smith. London: T. Richards.
Fenwick, C. Wiggins, M., and Wythe, D. (2008) (eds). TRAC 2007: Proceedings of the Seventeenth
Annual Theoretical Roman Archaeology Conference, London 2007. Oxford: Oxbow Books.
Fieldhouse, W. J., May, T., and Wellstood, F. C. (1931). A Romano-​British Industrial Settlement
near Tiddington, Stratford-​upon Avon. Birmingham: Corporation of Stratford-​upon-​Avon
and the Trustees and Guardians of Shakespeare’s Birthplace.
Fincham, G., Harrison, G., Holland, R., and Revell, L. (2000) (eds). TRAC 99:  Proceedings
of the Ninth Annual Theoretical Roman Archaeology Conference, Durham 1999.
Oxford: Oxbow Books.
90

90   Ellen Swift

Fox, G. E. (1893). ‘Excavations on the Site of the Roman City at Silchester, Hants, in 1894’,
Archaeologia, 54: 439–​494.
Franks, A. (1858). ‘On Bosses of Roman Shields Found in Northumberland and Lancashire’,
Archaeological Journal, 15: 55–​58.
Franks, A. (1867–​70). ‘Note on a Roman Enamel Bowl Found at Broughing, Herts’, Proceedings
of the Society of Antiquaries of London. 2nd ser. 4: 514–​516.
Freeman, P. W. M. (2007). The Best Training Ground for Archaeologists: Francis Haverfield and
the Invention of Romano-​British Archaeology. Oxford: Oxbow Books.
Frere, S. S. (1967). Britannia: A History of Roman Britain. London: Routledge and Kegan Paul.
Fulford, M. (1989). ‘The Economy of Roman Britain’, in M. Todd (ed.), Research on Roman
Britain, 1960–​1989. Britannia Monograph Series 11. London: Society for the Promotion of
Roman Studies, 175–​202.
Fulford, M. (2007). ‘Foreword’, in R. Hingley and S. Willis (eds), Roman Finds: Context and
Theory. Oxford: Oxbow Books, pp. ix–​x.
Gardner, A. (2007). An Archaeology of Identity: Soldiers and Society in Late Roman Britain.
Walnut Creek, CA: Left Coast Press.
Gould, J. (1966–​7). ‘Excavations at Wall (Staffordshire) 1961–​63 on the Site of the Early Roman
Forts and of the late Roman Defences’, Transactions of the Lichfield and South Staffordshire
Archaeological and Historical Society, 8: 1–​38.
Graham Ritche, J. N. (2002). ‘James Curle (1862–​1944) and Alexander Ormiston Curle (1866–​
1955):  Pillars of the Establishment’, Proceedings of the Society of Antiquaries of Scotland,
132: 19–​41.
Green, M. J. (1978). A Corpus of Small Cult Objects from the Military Areas of Roman Britain.
BAR British Series 52. Oxford: Archaeopress.
Hattatt, R. (1982). Ancient and Romano-​British Brooches. Oxford: Dorset Publishing Company.
Hattatt, R. (1985). Iron Age and Roman Brooches:  A  Second Selection of Brooches from the
Author’s Collection. Oxford: Oxbow Books.
Haverfield, F. (1891). ‘Roman Remains in Local Museums’, Antiquary, 24: 168–​172.
Haverfield, F. (1911). ‘Report of the 1910 Excavations at Corstopitum:  Smaller Objects’,
Archaeologia Aeliana, 3rd ser. 7: 176–​202.
Haverfield, F. (1912). The Romanization of Roman Britain. 2nd edn. Oxford: Clarendon Press.
Haverfield, F. (1919). ‘The Tullie House fibulae’, Transactions of the Cumberland and
Westmorland Antiquarian and Archaeological Society, ns 19: 1–​16.
Hawkes, C. F. C. (1940). ‘An “Aylesford” La Tene III Brooch from Arundel Park and the Dating
of the Type’, Antiquaries Journal, 20: 492–49​5.
Hawkes, C. F. C. (1982). ‘Mark Reginald Hull, 1897–​1976’, Essex Archaeology and History, 3/​
14: 1–​2.
Hawkes, C. F. C., and Hull, M. R. (1947). Camulodunum: First Report on the Excavations at
Colchester 1930–​39. Research Report of the Society of Antiquaries of London, volume 14.
Oxford: Oxford University Press for the Society of Antiquaries of London.
Hawkes, S., and Dunning, G. (1961). ‘Soldiers and Settlers in Britain’, Medieval Archaeology,
5: 1–​70.
Henig, M. (1974). A Corpus of Roman Engraved Gemstones from British Sites. BAR British Series
8. Oxford: Archaeopress.
Henig, M. (1989). ‘Religion in Roman Britain’, in M. Todd (ed.), Research on Roman Britain,
1960–​1989. Britannia Monograph Series 11. London: Society for the Promotion of Roman
Studies, 219–​234.
 The Development of Artefact Studies    91

Henig, M., and Soffe, G. (2002). ‘Graham Webster—​Archaeologist’, Bulletin of the Association
for Roman Archaeology, 12: 3–​5, 8.
Herendeen, W. (2007). William Camden: A Life in Context. Woodbridge: Boydell.
Hildyard, E. J. W. (1945). ‘The Ancestry of the Trumpet-​Fibula’, Antiquaries Journal, 25: 154–​158.
Hildyard, E. J. W. (1950). ‘The Date of the Malton Brooch’, Antiquaries Journal, 30: 74.
Hill, J. D. (1995). Ritual and Rubbish in the Iron Age of Wessex. BAR British Series 242.
Oxford: Archaeopress.
Hingley, R. (2007). ‘The Society, its Council, the Membership and Publications, 1820–​50’,
in S. Pearce (ed.), Visions of Antiquity:  The Society of Antiquaries of London 1707–​2007.
London: Society of Antiquaries, 173–​198.
Hingley, R. (2008). The Recovery of Roman Britain. Oxford: Oxford University Press.
Hingley, R., and Willis, S. (2007). Roman Finds: Context and Theory. Oxford: Oxbow Books.
Hodder, I. (1982). Symbols in Action:  Ethnoarchaeological Studies of Material Culture.
Cambridge: Cambridge University Press.
Hutcheson, A. J.  R. (1997). ‘Native or Roman? Ironwork Hoards in Northern Britain’,
in K. Meadows, C. Lemke, and J. Heron (eds), TRAC 96:  Proceedings of the Sixth
Annual Theoretical Roman Archaeology Conference, Sheffield 1996. Oxford:  Oxbow
Books, 65–​72.
Jackson, R. (1985). ‘Cosmetic Sets from Late Iron Age and Roman Britain’, Britannia, 16: 165–​192.
James, S. (2003). ‘Roman Archaeology: Crisis and Revolution’, Antiquity, 77: 178–​184.
Jenkins, F. (1985). ‘A Group of Silvered Bronze Horse-​Trappings from Xanten’, Britannia,
16: 141–​164.
Johns, C. (2007). ‘The Last Chance’, in R. Hingley and S. Willis (eds), Roman Finds in Context
and Theory. Oxford: Oxbow Books, 29–​34.
Johns, C., Thompson, H., and Wagstaff, P. (1980). ‘The Wincle, Cheshire, Hoard of Roman
Gold Jewellery’, Antiquaries Journal, 60: 48–​58.
Jones, R. F.  J. (1987). ‘The Archaeologists of Roman Britain’, Bulletin of the Institute of
Archaeologists of London, 24: 85–​97.
Jundi, S., and Hill, J. D. (1998). ‘Brooches and Identities in First Century AD Britain: More
than Meets the Eye?’, in C. Forcey, J. Hawthorne, and R.Witcher (eds), TRAC 97: Proceedings
of the Seventh Annual Theoretical Roman Archaeology Conference, Nottingham 1997.
Oxford: Oxbow Books, 125–​137.
Kendrick, T. D. (1950). British Antiquity. London: Methuen.
Kenyon, K. (1948). Excavations at the Jewry Wall Site, Leicester. Research Report for the Society
of Antiquaries of London, volume 15. Oxford: Oxford University Press for the Society of
Antiquaries of London.
Kenyon, K. (1950). ‘Excavations at Breedon Hill, Leicestershire 1946’, Transactions of the
Leciestershire Archaeological Society, 26: 17–​82.
Kenyon, K. (1952). Beginning in Archaeology. London: Phoenix House.
Kenyon, K. (1959). Excavations in Southwark 1945–​ 47. Research Papers of the Surrey
Archaeological Society 5. Guildford: Surrey Archaeological Society.
Keppie, L. (1989). ‘Beyond the Northern Frontier: Roman and Native in Scotland’, in M. Todd
(ed.), Research on Roman Britain 1960–​1989. Britannia Monograph Series 11. London: Society
for the Promotion of Roman Studies, 61–​74.
Liversidge, J. (1968). Britain in the Roman Empire. London: Routledge and Kegan Paul.
Lloyd-​ Morgan, G. (1977). Mirrors in Roman Britain. BAR British Series 41.
Oxford: Archaeopress.
92

92   Ellen Swift

Lloyd-​Morgan, G. (1981). ‘Jet and Shale in the Archaeological Collections of the Grosvenor
Museum, Chester’, Journal of the Chester Archaeological and Historical Society, 64: 41–​46.
Lockyear, K. (2007). ‘Where do we Go from here? Recording and Analysing Roman Coins
from Archaeological Excavations’, Britannia, 38: 211–​224.
Lucas, G. (2001). Critical Approaches to Fieldwork: Contemporary and Historical Archaeological
Practice. London and New York: Routledge.
Macgregor, A. (1976). Finds from a Roman Sewer System and an Adjacent Building in Church
Street. The Archaeology of York, volume 17, fasc. 1. York: Council for British Archaeology for
the York Archaeological Trust.
Macgregor, A. (1978). Roman Finds from Skeldergate and Bishopshill. The Archaeology of York,
volume 17, fasc. 2. York: Council for British Archaeology for the York Archaeological Trust.
Manning, W. H. (1976). Catalogue of Romano-​British Ironwork in the Museum of Antiquities,
Newcastle upon Tyne. Newcastle upon Tyne:  Department of Archaeology, University of
Newcastle upon Tyne.
Manning, W. H. (1995). ‘George C. Boon (1927–​1994)’, Britannia, 26: pp. xiii–​xiv.
Manning, W., Price, J., and Webster, J. (1995). The Roman Small Finds: Report on the Excavations
at Usk 1965–​1976, Volume 7. Cardiff: University of Wales Press.
Marshall, F. H. (1911). Catalogue of the Jewellery, Greek, Etruscan and Roman, in the Departments
of Antiquities, British Museum. London: British Museum Department of Greek and Roman
Antiquities.
Mattingly, D. (2008). ‘John Dore:  Archaeologist of Northern Britain and Libya who was a
Specialist in Mediterranean Ceramics’, Independent, 25 June 2008.
Mattingly, H., Sydenham, E., Sutherland, C., and Carson, R. (1923–​84) (eds). The Roman
Imperial Coinage. Volume 1: From 31 BC to AD 69. London: Spink and Son.
Merrifield, R. (1987). The Archaeology of Ritual and Magic. London: Batsford.
Miller, E. (1973). That Noble Cabinet: A History of the British Museum. London: André Deutsch.
Nash-​Williams, V. E. (1932). ‘The Roman Legionary Fortress at Caerleon:  Report on the
Excavations Carried out in the Prysg Field 1927–​29. Part II:  The Finds’, Archaeologia
Cambrensis, 87: 48–​104.
Nurse, B. (2004a). ‘Way, Albert (1805–​ 1874)’, Oxford Dictionary of National Biography.
Oxford: Oxford University Press (online edn).
Nurse, B. (2004b). ‘Cox, John Charles (1843–​1919)’, Oxford Dictionary of National Biography.
Oxford: Oxford University Press (online edn).
O’Neil, J. (1935). ‘Coins and Archaeology’, Archaeological Journal, 92: 64–​80.
Oswald, F. (1952). Excavation of a Traverse of Margidunum. Nottingham:  Nottingham
University Press.
Painter, K., and Thompson, H. (1994). ‘Obituary: Donald Harde’, Independent, 29 April 1994.
PAS (Portable Antiquities Scheme) (2012). <http://​finds.org.uk/​> (accessed 22 May 2012).
Peacock, D. P.  S. (1982). Pottery in the Roman World:  An Ethnoarchaeological Approach.
London: Longman.
Pevensey Excavation Committee (1908). Excavations on the Site of the Roman Fortress at
Pevensey: First Report. Lewes/​Eastbourne/​East Grinstead: Privately printed for subscribers.
Philpott, R. (1991). Burial Practices in Roman Britain:  A  Survey of Grave Treatment and
Furnishing. BAR British Series 219. Oxford: Archaeopress.
Piggott, S. (1989). Ancient Britons and the Antiquarian Imagination: Ideas from the Renaissance
to the Regency. London: Thames and Hudson.
 The Development of Artefact Studies    93

Pitt-​Rivers, A. H. L. F. (1887). Excavations in Cranborne Chase near Rushmore on the borders
of Dorset and Wiltshire. Volume 1: Excavations in the Romano-​British Village of Woodcuts
Common and Romano-​British Antiquities in Rushmore Park. Privately Printed.
Pitts, M. (2010). ‘Artefact Suites and Social Practice: An Integrated Approach to Roman Finds
Assemblages’, Facta: A Journal of Roman Material Culture Studies, 4: 125–​152.
Quennell, M., and Quennell, C. H. B. (1937). Everyday Life in Roman Britain. 2nd edn. The
Everyday Life Series. London: Batsford.
Reece, R. (1973). ‘A Short Survey of the Roman Coins Found on Fourteen Sites in Britain’,
Britannia, 3: 69–​76.
Rees, H., Crummy, N., Ottaway, P. J., and Dunn, G. (2008). Artefacts and Society in Roman
and Medieval Winchester:  Small Finds from the Suburbs and Defences, 1971–​ 1986.
Winchester: Winchester Museums Service.
Roach-​Smith, C. (1850). Antiquities of Richborough, Reculver and Lympne. London:  John
Russell Smith.
Roach-​Smith, C. (1859). Illustrations of Roman London. Unpublished: Private Subscription.
Roach-​Smith, C. (1868a). Collectanea Antiqua:  Etchings and Notices of Ancient Remains,
Illustrative of the Habits, Customs, and History of Past Ages. London: John Russell Smith.
Roach-​Smith, C. (1868b). ‘On the Red-​Glazed Pottery of the Romans’, Journal of the British
Archaeological Association, 4: 1–​19.
Romilly Allen, J. (1901). ‘Romano-​British Fibulae Showing Late-​Celtic Influence’, Reliquary
and Illustrated Archaeology, ns 7: 195–​199.
Savory, H. N. (1956). ‘Some Sub-​Romano-​British Brooches from South Wales’, in D. B. Harden
(ed.), Dark Age Britain. London: Methuen and Company, 40–​58.
Serpell, M. (1982). ‘Obituary: Dorothy Charlesworth’, Antiquaries Journal, 62: 487–​504.
Smiles, S. (2007). ‘Art and Antiquity in the Long Nineteenth Century’, in S. Pearce (ed.), Visions
of Antiquity: The Society of Antiquaries of London 1707–​2007. London: Society of Antiquaries,
123–​145.
Strong, D. and Brown, D. (1976) (eds.). Roman Crafts. London: Duckworth.
Sutherland, C., and Carson, R. (1923) (eds). The Roman Imperial Coinage. Volume 1: From 31 BC
to AD 69. London: Spink and Son.
Swift, E. (2007). ‘Small Objects, Small Questions? Perceptions of Finds Research in the
Academic Community’, in R. Hingley and S. Willis (eds), Roman Finds in Context and
Theory. Oxford: Oxbow Books, 18–​28.
Swift, E. (2010). ‘Identifying Migrant Communities:  A  Contextual Analysis of Grave
Assemblages from Continental Late Roman Cemeteries’, Britannia, 41: 237–​282.
Todd, M. (1989). Research on Roman Britain, 1960–​89. Britannia Monograph Series 11.
London: Society for the Promotion of Roman Studies.
Tomber, R., and Dore, J. (1998). The National Roman Fabric Reference Collection: A Handbook.
London: Museum of London Archaeology Service <http://​www.molas.org.uk/​projects/​fab-
rics_​tei.asp?report=nrfc> (accessed 25 October 2011).
Tonnochy, A. B., and Hawkes, C. F. C. (1931). ‘The Sacred Tree Motif on a Roman Bronze from
Essex’, Antiquaries Journal, 11: 123–​128.
Toynbee, J. (1962). Art in Roman Britain. London: Phaidon.
Tyers, P. (1996). Roman Pottery in Britain. London: Batsford.
van Driel Murray, C. (1999). ‘ “And did those feet in ancient time . . . ”: Feet and Shoes as a
Material Projection of the Self ’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC
94

94   Ellen Swift

98: Proceedings of the Eighth Annual Theoretical Roman Archaeology Conference: Leicester

1998. Oxford: Oxbow Books, 131–​140.
Wacher, J. (1969). Excavations at Brough on Humber 1958–​1961. Research Report of the Society
of Antiquaries of London, volume 25. Oxford: Oxford University Press for the Society of
Antiquaries of London.
Wallace, C. (2002). ‘Writing Disciplinary History:  Or why Romano-​British Archaeology
Needs a Biographical Dictionary of its Own’, Oxford Journal of Archaeology, 21/​4: 381–​392.
Walters, H. B. (1908). Catalogue of the Roman Pottery in the Department of Antiquities, British
Museum. London: British Museum Department of Greek and Roman Antiquities.
Ward, J. (1911). The Roman Era in Britain. London: Methuen and Co.
Warner, R. (1797). An Illustration of the Roman Antiquities discovered in Bath. Bath: The Mayor
and Corporation.
Webster, G. (1977). ‘Reflections on Romano-​British Pottery Studies, Past, Present and Future’,
in J. Dore and K. Greene (eds), Roman Pottery Studies in Britain and Beyond. BAR British
Series 30. Oxford: Archaeopress, 317–​333.
Wedlake, W. J. (1958). Excavations at Camerton, Somerset: A Record of Thirty Years’ Excavation
Covering the Period from Neolithic to Saxon Times, 1926–​56. Bath: Privately published.
Wheeler, R. E. M. (1955). ‘Anniversary Address’, Antiquaries Journal, 35: 153–​161.
Wheeler, R. E.  M., and Wheeler, T. V. (1928). ‘The Roman Amphitheatre at Caerleon,
Monmouthshire’, Archaeologia, 78: 111–​218.
Wheeler, R. E. M., and Wheeler, T. V. (1932). Report on the Excavation of the Prehistoric, Roman and
Post-​Roman Site in Lydney Park, Gloucestershire. Research Report of the Society of Antiquaries
of London, volume 9. Oxford: Oxford University Press for the Society of Antiquaries.
White, R. H. (1988). Roman and Celtic Objects from Anglo-​Saxon Graves: A Catalogue and an
Interpretation of their Use. BAR British Series 191. Oxford: Archaeopress.
Whiting, W., Hawley, W., and May, T. (1931). Report on the Excavation of the Roman Cemetery
at Ospringe, Kent. Research Report of the Society of Antiquaries of London, volume
8. Oxford: Oxford University Press for the Society of Antiquaries.
Willis, S. (2004). ‘The Study Group for Roman Pottery: Research Framework Document for
the Study of Roman Pottery in Britain, 2003’, Journal of Roman Pottery Studies, 11: 1–​20.
Wilson, D. M. (1997). ‘Augustus Wollaston Franks: Towards a Portrait’, in M. Caygill and J.
Cherry (eds), Augustus Wollaston Franks:  Nineteenth-​Century Collecting and the British
Museum. London: British Museum Press, 1–​5.
Winbolt, S. E. (1925). Roman Folkestone: A Record of Excavation of Roman Villas at East Wear
Bay. London: Methuen.
Winbolt, S. E. (1935). ‘Report on Recent Finds at Chichester’, Sussex Archaeological Collections,
76: 136–​171.
Wroth, W. W. (rev. Henig, M.) (2004). ‘King, Charles William (1818–​1888)’, Oxford Dictionary
of National Biography. Oxford: Oxford University Press (online edn).
Young, C. (2007). ‘Obituary: Joan Kirk’, Society of Antiquaries Online Newsletter, 176 <http://​
www.sal.org.uk/​salon/​index_​html?id=742> (accessed 20 January ​2012).
 Chapter 5

The Textua l a nd
Archaeol o gica l Ev i de nc e

Henry Hurst

Introduction

Because we like stories, in common with all human beings, and also because of our
logocentric culture, in which we communicate most information in the form of writing,
texts are usually given a privileged status when they survive from a period of the past in
which we are interested.
For Roman Britain the longest-​known and, until recently, most accessible texts were
those of classical authors who wrote about the province, in Latin or Greek, always from
a distance and in passages that were secondary to the wider purpose of their works.
The collective name for these passages, ‘sources’, reflects the way they were tradition-
ally regarded as the fount of knowledge. They are not particularly abundant for Roman
Britain. What survives most fully is coverage of the decades between the invasion of
ad 43 and the governorship of Agricola (ad 78–​85), with a particular focus on military
history. For later events shorter passages and spasmodic references survive, and there is
a body of geographical information in ancient writers and road itineraries (Mann and
Penman 1996; Ireland 2008; summarized in Mattingly 2006: 26–​38). These have been
supplemented by ingenious scholarship, partly about texts found within Britain and
partly about the range of other material evidence found by archaeology. Until recently
most of the known texts from Britain were inscriptions on stone, of which 2,400 were
published collectively as Volume I of the Roman Inscriptions of Britain (hereafter RIB:
Collingwood and Wright 1965; indexes by Goodburn and Waugh 1983). These tended to
skew the overall view even more towards military history, since the army had a stronger
‘epigraphic habit’ than civilians (neatly shown in the case of religion by Millett (2005:
figures 88–​89)) and more of what it inscribed survives, because it spent most time in
those parts of Britain where stone suitable for inscribing is freely available. The way in
which archaeology was done until the 1980s, with its emphasis on military, urban, and
96

96   Henry Hurst

villa sites and an excavation-​dominated approach to their study, also tended to reinforce
this. A traditional format for histories of Roman Britain was, therefore, a military and
political narrative with annexed chapters on ‘industry’, commerce, and cultural life (see
Salway 1981; Frere 1987; Wacher 1998). As discussed more fully elsewhere in this vol-
ume (see chapters by Wilson, Millett, and Swift, this volume), a reaction set in from the
1980s, leading to a series of different emphases, mostly approaching the history more
from a ‘native’ standpoint; these works sometimes identified themselves as ‘postcolo-
nial’, in contrast with earlier scholarship, seen as linked with British imperialism, which
had itself been partly influenced by Roman texts (see Millett 1990; Mattingly 1997, 2006;
Hingley 2000, 2005; Gardner 2007). These approaches tended to emphasize the links
between prehistoric and Roman Britain, through a whole range of archaeological stud-
ies. Such links were strengthened by the observation that the vast bulk of the Romano-​
British population was illiterate.

Types of Text from within Britain

Meanwhile a quiet revolution was taking place, with the publication of large quantities
of new textual evidence. As a result of the extensive recovery in excavations of anaer-
obically preserved material, we have learnt that a simple and cheap writing technol-
ogy, using carbon ink on thin sheets of wood, was widespread in Roman Britain (as
probably it was throughout the Roman world). The earliest publication of a text writ-
ten in ink from Britain may be that by E. G. Turner (1956), of a deed of sale on a sty-
lus tablet of larch wood from the well of a villa at Chew Stoke in Somerset; a letter of
ink on wood, sent from Rochester to London, was published four years later (Turner
and Skutsch 1960), but it was only with the discovery of the texts from the military
site at Vindolanda just south of Hadrian’s Wall, from 1973 onwards, that the abundance
of writing in these materials, and its significance, came to be appreciated (Figure 5.1;
Bowman and Thomas (1983: 33–​37) give a list of the evidence known at that time). Our
previous assumption, that most Roman letter-​and other writing for routine purposes
was incised with a metal stylus on wax in a specially made tablet, was more restrictive
to the practice of writing, since these materials required several skills and resources
in their production. Stylus tablets and, more so, metal styli are nevertheless also fairly
widespread finds in Roman Britain, but the tablets offer us written content only where
the stylus has penetrated into the wood below the wax (Figure 5.2). The first collective
volume giving the texts of the ink inscriptions on wood from Vindolanda was Bowman
and Thomas (1983), two more volumes followed in 1994 (Bowman and Thomas 1994),
and two further primary publications of texts in 2010 and 2011 (Bowman et al. 2010,
2011). These have been accompanied by many ancillary publications, of which Bowman
(2003) is the most comprehensive. This site has produced about 400 substantial, intel-
ligible texts (with fragments of hundreds of others), which form the largest body of
written documentation of life in a military community from anywhere in the Roman
 The Textual and Archaeological Evidence    97

Figure 5.1  Memorandum on wooden leaf tablet from Vindolanda, referring to the British use
of cavalry (Tab. Vindol. II. 164): dimensions 78 × 186 millimetres. After a punctuation mark in l. 4,
the text reads: ‘nec residunt | Brittunculi ut iaculos | mittant’ (‘nor do the Brits mount in order to
throw javelins’).
Source: full text and translation in Bowman and Thomas (1994: 107–​108). © The Trustees of the British Museum.

world. They were mostly discarded material from the commander’s house at the end
of the first century ad, though others have also been found in a barrack block and the
workshop of the fort (Pearce (2004: 48–​49) describes the complexities of their deposi-
tion) and are not so much formal documents about military organization, though such
98

98   Henry Hurst

Figure 5.2  Sale of a slave girl, Fortunata, recorded on a stylus tablet from London, drawn by R.
Tomlin: dimensions c. 140 × 114 millimetres, probably one of three tablets. The writing survived
as scratches in the wooden tablet, becoming illegible where a triangular patch of the original wax
still survived; [p]‌uellam Fortunatam are the first two words of l. 3.
Source: full text and translation in Tomlin (2003, 2011). © Roger Tomlin.

material is present, as a written reflection of life in this military community from the
viewpoint of the commander’s household. They therefore contain domestic material
ranging from accounts and arrangements for supplies to personal correspondence. To
the extent that they offer a view of life in Britain from the standpoint of a military com-
munity in a frontier region and in one well-​known case refer to the native fighters in the
diminutive as Brittunculi (‘Brits’: Figure 5.1), they might seem to bear out a stereotype
of the Romans as colonialists. However, as will be seen, they are more remarkable for
what they reveal of the complexity of communications and contact, and indeed of the
colonial process itself.
There are also major collections of writing tablets from London and Carlisle (esti-
mated for stylus and ink tablets combined as 200+ and 150+ respectively by Pearce
(2004:  48), although smaller numbers have been published). The London finds are
mostly stylus tablets and from two main areas, the river bank and areas adjacent to
the Walbrook stream; the small number with readable texts have included legal and
 The Textual and Archaeological Evidence    99

business documents, including the sales of a wood in Kent (Tomlin 1996) and a slave
girl (Figure 5.2). The military site at Carlisle has yielded tablets inscribed in ink, from
Annetwell Street, which are mainly fragments of accounts and letters, published by
Tomlin (1998); stylus tablets from two other sites have revealed only addresses. The
date range of these finds is later first to mid-​second century (Pearce 2004: 49).
A contrasting set of texts, also published from the 1980s on, are the curses incised
on lead sheet or defixiones (Figure 5.3). The contents of only about ten of these were

Figure 5.3  Curse tablet no. 43 from the sanctuary of Mercury at Uley, Gloucestershire, drawn
by R. Tomlin: dimensions c. 95 × 83 millimetres. Docilinus asks Mercury to drive Varianus,
Peregrina, and Sabinianus to death for injuring his farm animal unless they redeem their action.
Source: text and translation in Hassall and Tomlin (1989: 329–​331, no. 3) and Tomlin (2002: 172). © Roger Tomlin.
100

100   Henry Hurst

known until the finds made at the shrine of Mercury at Uley in Gloucestershire, the
first three of which were published in Britannia 10 for 1979, and the major collection
of 130 from the sacred spring at Bath, whose texts were fully published in 1988 (Tomlin
1988). Uley produced over 100, but because of the difficulty of unrolling the corroded
lead these texts have been published more slowly: five, including the three in Britannia
10, were published in the main excavation report (Tomlin 1993), and seven more since,
in Britannia 20, 23, 26, 27, and 29. Since then others have come to light, and in all over
300 are now known (Mattingly 2006: 310; the contents of sixty-​nine are summarized
in his Table 11, including thirty-​three from Bath and twelve from Uley). In response
to an appeal by the Oxford Centre for the Study of Ancient Documents, thirty-​five
sites are now known to have produced one or more defixiones. Many remain unread—​
where it is either difficult or impossible to unroll the corroded lead sheets—​but this
is nevertheless a growing source of textual information. As will be discussed, their
contents have a broadly uniform framework, where most frequently the help of a deity
is invoked to punish one or more thieves of personal property, but, even if filtered
through conventions of form or expression and the medium probably of a scribe, this
is as near as we can ever come to hearing the voices of the non-​elite population of
Roman Britain.
A third major publication of writing on materials other than stone in the 1990s were
the eight fascicules (and a ninth of indexes) belonging to RIB Volume II (Collingwood
and Wright 1990–​5). This ranged from official tile stamps and indications of weights
and contents of goods on containers, through commercial makers’ stamps and other
marks on products from silver to leather to scratched graffiti (see Figure 5.4), many
simply recording ownership, but also containing dedications and other information.
RIB II performed a most valuable service by bringing this disparate material up to
date in conveniently collected form. These fascicules have an eccentric title page, in
that their collective heading of ‘Instrumentum domesticum’ (‘private documents’ in
a free translation) under-​represents the range of material and the naming of R. G.
Collingwood (died 1943)  and R.  P. Wright (retired 1976, died 1992)  as principal
authors piously records their pioneering contribution, but sends a misleading signal
about scholarly works of the 1990s. The fascicules can be seen as a long-​term follow-​
up to the design of the broader Corpus Inscriptionum Latinarum (CIL), which from
an early stage published other types of written evidence as well as public inscriptions
on stone: so, for example, Volume IV (1871) contains graffiti and painted inscriptions
on the walls of Pompeii and Volume XV (1891) is Heinrich Dressel’s study of stamped
bricks from Rome published under the heading of Instrumentum domesticum. The
first volume of RIB (Collingwood and Wright 1965) had covered (mainly) inscriptions
on stone found up to the start of 1955; and a further 550 inscriptions on stone found up
to the end of 2006, RIB III, were published in 2009 (Tomlin et al. 2009), bringing their
total to just short of 3,000. At the present time, then, epigraphic evidence of all sorts
from Roman Britain is, to a greater extent than ever before, fully published in a con-
venient form; and, as has been the case for many years, new discoveries are added each
 The Textual and Archaeological Evidence    101

Figure  5.4 Tile with personal-​name graffito scratched before firing, Candid[us]. From a
fourth-​century deposit at Hayton, East Yorkshire.
Source: Tomlin and Hassall (2001: 393). Photograph: M. Millett.

year in the journal Britannia under the heading ‘Roman Britain in [the year preceding
the issue date]’.
If we add to these classes of writing inscriptions on coins and potters’ stamps, espe-
cially on imported samian (Hartley 1969; Hartley and Dickinson 2008–​12)—​it becomes
apparent that, illiterate as the great majority of people in Roman Britain undoubtedly
were, they encountered writing at virtually every turn. Even in the countryside coins
and stamped and graffito-​marked pots and other objects were common; main roads had
inscribed milestones, cult centres contained inscribed dedications and defixiones and
(see Beard 1991 for a general discussion of writing and Roman religion; for further dis-
cussion on inscriptions and language see Hope and Mullen, this volume). There is some
evidence generally for an association between writing and votive settings in settlement
sites (Pearce 2004: 51, with references). In urban and military settlements, the volume
and range of writing became immense, from inscribed funerary monuments on the
approaches and monumental inscriptions over entry gates or arches to formal and infor-
mal writing on buildings and then functional writing for legal, commercial, and admin-
istrative purposes.
102

102   Henry Hurst

Ancient Uses of, and Responses

to, Writing

A first challenge therefore lies in the abundance of written evidence. In a very straight-
forward way the mass of writing we now have challenges the old view that there are, on
the whole, only some frustratingly short texts of classical authors, with little written evi-
dence of note from within Roman Britain. ‘Our written sources are primarily external
ones… comparatively little documentation survives that was produced in Britain itself,’
says Mattingly (2006: 21), even if in subsequent pages of the same volume he goes on to
discuss the Vindolanda evidence and the defixiones in some detail. That statement is only
true if ‘documentation’ is taken to mean historical narration in some form. Why, then,
should writing be so common? Viewed from our culture of mass literacy, a natural inclin-
ation might be to assume that writing reflects some degree of functional literacy, using
this term (and subsequently in this chapter) with a meaning as defined by the 1978 Unesco
General Conference: ‘A person is functionally literate who can engage in all those activ-
ities in which literacy is required for effective functioning of his group and community
and also for enabling him to continue to use reading, writing and calculation for his own
and the community’s development’ (see also Harris 1989: ch. 1). But the reality in Roman
Britain, as in the modern world, was clearly more nuanced than that. In the latter case, the
‘New Literacy Studies’ of the 1990s emphasized literacy and numeracy as social practices,
embedded in social settings. ‘Literacy practices’ might then be exercised at ‘literacy events’,
defined as ‘any occasion in which a piece of writing is integral to the nature of the partici-
pants’ interactions and their interpretative processes’ (Street 1984). The merit of this is that
it provides a theoretical framework for all sorts of limited literacy. ‘The literacy as applied,
practised and situated approach questions the validity of designations of individuals as “lit-
erate” or “illiterate”, as many who are labelled illiterate are found to make use of significant
literacy practices for specific purposes in their everyday lives’ (Unesco 2006).
We need not doubt the view that full functional literacy was confined to a small
minority in the ancient world. In his wide-​ranging study, Harris (1989) argued that in
very broad terms less than 10 per cent of the population of the Roman Empire is likely
to have been literate. This broad order of magnitude has not been challenged since, and
is thought to apply even where writing survives in such abundance as from Pompeii or
Roman Egypt. For Roman Britain we need only note momentarily that, even if the entire
resident army of some 40,00–​50,000 soldiers were fully literate (which would certainly
not have been the case) and there was a comparable number of literate local elites, trad-
ers, and officials (which there would certainly not have been), these 80,00–​100,000
would be less than 5 per cent of a population estimated at between two million and six
million (Millett 1990: 181–​186); the real figure was probably smaller. This is not altered by
the discovery of short texts written in cursive script on tiles before they were fired (RIB
2491), including ‘a,b,c’ texts as of learners (2491.135–​145), metrical fragments, among
them the phrase ‘conticuere omnes’ (‘all were silent’) which begins Aeneid book II, and
evidence that the civitas we used to know as the Coritani may, more correctly, have been
 The Textual and Archaeological Evidence    103

called the Corieltauvi (2491.150; Tomlin 1983), interesting as all of these are as anecdotal
instances. The reality was that the great bulk of the population was not only untaught but
had no general need to use reading or writing actively in their daily lives: a situation that
in Britain was brought to an end only with the mass education of modern times, and in
the world as a whole has been a major concern of Unesco up to the present time.
The question of why writing should be so common in a largely illiterate society has
been addressed for many ancient and early medieval societies, and answers have usu-
ally included the theme that it was an expression of power by mostly elite members of
society: they shared written content, which was attainable only indirectly if at all by
the majority; public displays of writing had symbolic meaning accentuated by various
visual devices, from settings to the aesthetics of lettering (Goody 1977; McKitterick
1990; Bowman and Woolf 1994). All of this manifestly applied to Roman Britain (as, for
example, Figure 5.5; an analysis of the lettering and layout of this inscription is given in

Figure 5.5  Tombstone of Classicianus from London: pieces as displayed in the British Museum
with reconstruction of original monument drawn by Richard Grasby (Grasby and Tomlin 2002:
figure 21), slightly adjusted. Reconstructed text (as in RIB 12): ‘Dis | [M]‌anibus | [G(ai) Iul(i) G(ai)
f(ili) F]ab(ia tribu) Alpini Classiciani |… |… | proc(uratoris) provinc(iae) Brita[nniae] | Iulia
Indi filia Pacata I[ndiana(?)] | uxor [f(ecit)]’ (‘To the spirits of the departed (and) of Gaius Julius
Alpinus Classicianus, son of Gaius, of the Fabian voting-​tribe… procurator of the province of
Britain, Julia Pacata I[ndiana], daughter of Indus, his wife, had this built’).
Source: Photograph © The Trustees of the British Museum.
104

104   Henry Hurst

Grasby 2012), but, before it is considered further, a methodological issue also arises—​
that of regarding texts as a type of artefact. This has been well argued by Moreland (2001)
using examples of the early medieval period, but it applies equally to Roman Britain. We
can see how writing was not always being used as we would use it. It is helpful, therefore,
to think of it like, say, pottery or coins, and to try to reconstruct its ancient uses. Our
understanding of Roman Britain would be limited by treating it as a ‘prehistoric’ (in the
sense of unlettered) society, but it is almost equally limiting to focus just on the content
of the surviving writing, without considering how its occurrence might have worked
among the people who created and experienced it.
The prime set of documents from Roman Britain indicating functional literacy are
the Vindolanda texts, studied by Bowman and others: the following observations are
dependent on their work (especially Bowman 2003). As Bowman has commented, a
striking feature of this assemblage is the insight it gives into the existence of a literate ‘sub-​
elite’. Flavius Cerialis, the commander of a Batavian cohort by Hadrian’s Wall, and prob-
ably Lepidina his wife and their friends were not from the top level of Roman provincial
society; they may only have been first-​or second-​generation ‘Romans’ originating from
parts of modern Netherlands and north Germany. Yet he and Lepidina’s friend Claudia
Severa seem to have been using written Latin for business and social purposes, as Cicero
might have. As is seen from comparison of the hands in their personal correspondence,
their circle often made use of scribes, adding only greetings at the beginning and end, but
part of Cerialis’ correspondence in particular also seems to have been written in his own
hand, including preliminary drafts. The variety of expression in the greetings of his and
his wife’s group shows that individuals were able to express themselves directly in writing
(Bowman 2003: 85). The geographical range of their contacts, from other military sites
on Hadrian’s Wall to the garrison city of Carlisle to London and probably beyond, and
the number of different hands—​‘several hundred’ (Bowman 2003: 85)—​suggest that the
Vindolanda collection is the tip of a considerable iceberg of literate communication. The
advantage of a locally available writing technology, of carbon ink on wood, has already
been emphasized (stylus tablets were also present but in smaller numbers (Pearce 2004:
figure 37)). There is, of course, a general military angle, that the Roman army looks to
have been as paper-​bound as its modern successors, as shown by micromanagerial docu-
ments in the Vindolanda collection, complementing the military documents on papy-
rus from Dura Europus on the Eastern frontier (Welles et al. 1959), so that a career as an
officer would develop an individual’s ability to express himself on paper.
Illiterate ordinary soldiers also must have been steered in the direction of writing by
belonging to an organization that used written communication so much. Retired aux-
iliary troops received written discharge certificates (RIB 2401), and veterans from all
parts of the army probably also possessed other written documents. The more enterpris-
ing seem likely to have become functionally literate, so that when they joined civilian
communities, not always close to where they had served (an extreme example is CIL
XVI, 69, a discharge diploma of a British auxiliary found at Brigetio in Hungary), they
may have contributed to the diffusion of literacy. The impact would have been strongest
in the period of high military mobility in the first half-​century after the invasion of 43.
 The Textual and Archaeological Evidence    105

Functional literacy would have also been spread among certain civilian professional
people. The Vindolanda documents suggest this for traders doing significant busi-
ness with the army (but see also Mairs 2012 on ‘interpretation’). Two deeds of sale from
London (Figure 5.2; Tomlin 1996) and that from Somerset (Turner 1956) imply a class of
functionally literate lawyers and their scribes, as might be expected. Literate individu-
als would also have been involved at all levels of the administration, from the governor
and provincial procurator’s office, through civic units of coloniae, civitates, and so on,
down to the organizers of craft and commercial guilds. But numbers here were almost
certainly small: they need only have been hundreds overall.
From the repertoire of figurative imagery on mosaics, especially of the fourth century,
it seems likely that the Romano-​British elite shared the intellectual and spiritual tastes of
other parts of the empire (cf. Scott 2000), even if the interpretation of individual cases
is often arguable. The very limited direct evidence so far for their literary culture takes
the form of inscriptions on mosaics and wall paintings. Writing at villa sites is shown by
numerous finds of metal styli (Hanson and Conolly 2002; Mattingly 2006: 461), but we
cannot say by whom or for what purpose. In a survey of the inscriptional evidence, Ling
(2007) divided the content into signatures of artists and owners, good luck messages,
labels for images portrayed, and ‘literary or pseudo-​literary quotations’. The last group
included part of a quotation from the Aeneid on a wall painting in a villa at Otford,
Kent (Aeneid 1. 313, or 12. 165) and the mosaics from the villas at Lullingstone, Kent,
and Frampton, Somerset. In an earlier review of literary allusions from Roman Britain,
Barrett (1978: 309–​313) argued that the Lullingstone floor, with the image of Europa and
the bull and the inscriptional allusion to Juno’s jealousy in Aeneid 1. 50 (Figure 5.6), writ-
ten in the style of Ovid, should be seen as the original expression of a literate man. Ling
suggests that such a polished piece of writing may not have been original, while two
possible ‘botched hexameters’ recorded by Lysons in the early nineteenth century on
the Bellerophon mosaic at Frampton ‘are questionable in syntax and metre, and could
have been made up by the patron or one of his peers’ (Ling 2007: 87). We cannot know,
but this line of argument echoes a former view about Romano-​British art, that, if it was
refined, it was probably imported, and, if it was blundering, as R. G. Collingwood once
put it, it was British, and it seems unduly negative when placed in the wider context of
fourth-​century classical culture in Roman Britain. One aspect of our appreciation of this
has been the degree of content we try to read into the evidence, as illustrated by Henig’s
suggestion (1997) that the Lullingstone inscription contained Christian cryptograms
(cf. Neal and Cosh 2009: 385, on this) and the arguments for Gnostic interpretations
of Frampton (Perring 2003; Tite 2010). In detail, these ideas often seem far-​fetched, but
the principle, that complex thought-​processes are present in the subject matter that we
struggle to understand, is surely right.
RIB II illustrates many instances where the use and comprehension of written sym-
bols were operating in a functional, but narrowly targeted, way, as, for example, the
indications of contents, quantities, and sources of goods, in dipinti, stamps, and graffiti.
Makers’ stamps on pottery and other goods, even leather (RIB 2445) and owners’ names
scratched as graffiti, can be seen as combining this sort of limited functionality with
106

106   Henry Hurst

Figure 5.6  Europa and the bull with two cupids in front of the presumed semicircular couch in
the apse of the dining room at Lullingstone villa, Kent. The inscription (RIB 2448.6) reads: ‘Invida
si ta[uri] vidisset Iuno natatus | Iustius Aeolias isset adusque domos’ (‘If jealous Juno had seen the
swimming of the bull, more justly would she have gone to the palace of Aeolus’).
Source: © Alamy.

other important roles of writing: symbolic in both cases; and probably (to some degree)
talismanic in the second case.
The symbolic use of writing in Roman Britain can be seen to cover a spectrum from
statements of power by the imperial authority to local applications. Lettering for Latin
inscriptions can be regarded as one of the major original creations of Roman culture,
with high levels of aesthetic subtlety (Gordon 1983): borrowing from a term applied
to Roman narrative art (as by Brilliant 1984), it could be described as ‘imagistic’ in the
sense that its appearance was intended to convey a message before its content (Figure
5.5). A striking example would have been the four-​way arch overlooking the coast at
Richborough, Kent, at the entry to Britannia. The survival of a single large bronze letter
A suggests that this monument to imperial power included a visually impressive inscrip-
tion, with bronze letters glinting against their marble setting (Strong 1968: 68). Most of
the 3,000 surviving inscriptions carved on stone of RIB I and III are also, to some extent,
concerned with the visual impact of writing. Writing accompanying an image usually
amplified the symbolism of the image but was not essential to its basic meaning. Thus,
 The Textual and Archaeological Evidence    107

for example, the message of a triumphant Roman cavalryman is conveyed by the tomb-
stone of Rufus Sita at Gloucester (RIB 121), irrespective of the accompanying inscrip-
tion. The context of the writing could also send a message, as on building inscriptions,
milestones, and many epitaphs: it was unnecessary to read the gateway inscriptions, dis-
tance slabs, and building stones of Hadrian’s Wall to grasp its meaning. However, for
those who could read, formal inscriptions contained an extra layer of meaning, espe-
cially important for their authors, as Woolf (1996) argues. We can indeed see monumen-
tal writing being used to provide controlled access to meaning in a manner not unlike
the stages of entry into a Roman house: a first stage is provided by its existence and the
material used for it; a second stage by the context or any associated image; a third stage,
of reading the content, was accessible only to a privileged minority. Their privilege
would be enhanced by their ability to pass this information on to others, just as one
might describe dining in a rich man’s house. This was a power game to be played by all
who ‘had the epigraphic habit’. Their authorship is shown in RIB I and III: overwhelm-
ingly the army or those in a military setting, with named individuals (predominantly
with non-​Celtic names) and civic, commercial, and social entities following after them.
Officially issued Roman coins give a miniaturized version of this, first as emblems of
authority expressed in different materials (gold, silver, brass, and copper), then through
imagery of the emperor’s head (easy to recognize) and reverse type (gradations of dif-
ficulty), inscriptions, and mintmark. A responsive use is also seen in local copies of the
official coinage, as, for example, in the Claudian-​copy coins issued locally when there was
a shortage of small change among coin-​using communities attached to military sites, in
ad 48–​64 (Kenyon 1993). The emperor’s head remains recognizable if not in detail, while
the reverse image and inscription are represented with varying degrees of abstractness.
Similar tendencies can be seen with pre-​Roman Iron Age coins imitating Roman issues.
Williams (2002) showed that in their style some inscriptions were influenced by, or ran
parallel to, potters’ stamps on amphorae, as well as Arretine and Gallo-​Belgic wares.
Whether a coin offers an illegible inscription or a cartouche in the manner of a potter’s
stamp, the emphasis is towards the ‘imagistic’ or symbolical aspect of writing. Terra sigil-
lata pottery stamps themselves develop from a significant minority presence of illiterate
stamps or symbols in Arretine ware to more uniformly literate stamps of the Gallic pot-
ters of the mid-​first to late second centuries ad. This can be seen as a refinement of the top
stage, of ‘content’, above the symbolism established previously. With the Gallic samian
there is indeed a classic Romanizing process, where ‘recent Romans’, the potters, with
Gallic rather than Roman names in many cases, carry the message of empire to those a
step or two less Romanized than them (on the impact of this pottery, see Woolf 1998: ch.
7, especially pp. 193–​203, and, on linguistic aspects, Mullen 2013). Non-​literate stamps
in British imitations of samian (Biddulph 2013) and in British-​produced mortaria (Hull
1963: 112–​114; Tyers 1996: 116–​135) also exemplify the symbolic role of potters’ stamps.
Generally there is a huge volume of naming by makers and users on artefacts, from
precious-​metal items down to coarse pottery: pre-​casting or pre-​firing graffiti as well as
stamps can be distinguished from graffiti scratched on a finished product (Figure 5.4).
Most of these names are of single individuals, roughly equally divided between Latin and
108

108   Henry Hurst

Celtic names. There are also collective users as military units and some evidently com-
missioned objects, not necessarily of high value (for example, RIB 2496.2 and 3, mor-
taria, with graffiti cut before firing, the first for a century at Colchester, the second for a
contubernium at Usk). A small minority of objects are marked with dedications to deities.
An analysis of graffiti on pottery by Evans (1987) showed that about three-​quarters were
personal names, followed by numbers and a collective ‘other’ category, and that names
were more frequently scratched on fine tablewares. Naming on artefacts by users, as a
mark of ownership or identity, develops the ‘responsive’ use of writing one stage further,
for many of these individuals must have been illiterate, either capable just of scratching
their name on the item or of getting somebody else to do that. That they chose to ‘borrow’
this medium is testimony to its power. It possibly also reflects a sense, exploited in a nega-
tive way in curse tablets, that a name embodied the person (cf. Gager 1992: 14), and thus a
stronger bond would be set up between the object and person than would be the case for
us writing our name in a book or sewing a nametape onto an article of clothing.

Curse Tablets

The British examples of curse tablets or defixiones represent about a fifth of the total
of 1,600 known throughout the Greco-​Roman world from the fifth century onwards
(Gager 1992; Martin 2010). Their material is a first point of interest. Most surviving
examples, including those from Britain, are on lead sheet or, in the case of those analysed
from Bath, a pewter-​like alloy (Tomlin 1988: 82–​84). This might be thought to fit with
their dark messages, invoking illness or death, and deposition in the ground, and occa-
sionally the texts make such a connection—​‘cold lead and cold name’, cold and useless or
heavy as the lead on which this is written (examples cited by Gager 1992: 3–​4). However,
these texts also survive on other materials, from papyrus to potsherds, and lead had
from early times a general use as a writing medium, so, while evidently favoured for this
role, it was not used exclusively. As regards subject matter, various cursing situations are
known: competitive, as in the theatre or circus; sexual; legal and political; commercial;
and pleas for justice and revenge, especially relating to the theft or loss of possessions. As
Tomlin (1988, 2002) has pointed out, the British examples almost exclusively belong to
this last category (the contents of sixty-​nine, from twenty-​one possible sites, were sum-
marized by Mattingly (2006: table 11); only one, from Old Harlow, possibly refers to los-
ing in love). A common formula is to name the alleged culprit, offer the stolen article
to the deity (as Sulis, Sulis Minerva, Mars or Mercury at Bath, or Mercury at Uley) and
invite vengeance, paid in blood, through the loss of sleep and health or through death.
As an example of their contents, Tomlin highlights an interesting pair, possibly of the
same individual from Bath and from Uley:

Docilianus, son of Brucetus to the most holy goddess Sulis. I devote him who has
stolen my hooded cloak, whether man or woman, whether slave or free. May the
 The Textual and Archaeological Evidence    109

goddess Sulis inflict him with the greatest death, and not allow him children now nor
in the future, until he has brought my hooded cloak to the temple of her divinity.
(Tabellae Sulis, no. 10: text and translation in Tomlin (2002: 167))

Uley no. 43 is from Docilinus to Mercury: ‘Varianus and Peregrina and Sabinianus who
have brought evil on my farm animal… I ask that you drive them to the greatest death,
and do not allow them health or sleep unless they redeem from you what they have
administered to me’ (Figure 5.3).
These are formulaic texts, following a preordained pattern, and there is the use often
of judicial phraseology in them. It might be supposed that they were produced to order
by scribes, and in certain non-​British examples this has been indicated by groups of
texts with similar handwriting. In the British material, notably the large group from
Bath, the variety not only of different hands, but of capacities to write, with elegant capi-
tals in some cases but also blundering ones as if clumsily imitating a stone cutter, and of
cursive from flowing to ‘pseudo-​texts’ whose writers may have been illiterates, has led
to the suggestion of more direct involvement by the authors of the texts. There is also
enough variation of expression and in some cases of language to suggest that they reflect
to some degree the feelings of their authors (two seem to be rare examples of written
Celtic (Mullen 2007)). The writing of some tablets could reflect something similar to a
‘literacy event’ in the modern world, but many were probably produced by scribes and
literate ‘friends’. The Bath assemblage probably covered two centuries, so a variety of
hands over this time is not so surprising (see Tomlin 1988: 98–​100).
Behind their conventional expressions, these tablets record private conversations
with the gods, since they were rolled up after being written and deposited in a shrine
for the deity’s eyes only. With their belief in divine powers of supra-​human sight and
capacity to inflict punishment, their theology fits within perceptions of divine power
as expressed in classical literature. The unexceptional nature of many of the objects
stolen—​an iron pan, a cloak, gloves, a small amount of money—​might be taken to
reflect poverty. The brutality of the punishments proposed—​the shedding of blood,
loss of health or death—​even if in itself conventional, combined with the use of pseudo-​
judicial language as if proposing a sentence, evokes the rough justice meted out to the
poor of early modern times, where the theft of a sheep could result in a hanging. So it
is tempting to relate these voices to the image given by the physical remains of Roman-​
Britons, as having lived mostly short, undernourished lives, and suggest that they are
the thoughts of unprivileged members of a poorly policed society. Mattingly’s discus-
sion (2006: 315) introduces an ‘us and them’ distinction by pointing out that only one
of over 300 curse tablets known was from a military site: ‘soldiers enjoyed better pro-
tection in and above the law.’ But this seems likely to be an oversimplification. As Gager
(1992: 177) points out, the depositional context of most of the surviving British tablets—​
in the sacred spring at Bath or buried at a sanctuary—​may give us a skewed view: we
may simply lack those with different contents deposited in theatres or in the circus or
in other contexts. As regards justice, as Martin (2010: 68) has said, these documents are
unsubstantiated allegations. They show aggravation, then, but not necessarily injustice.
110

110   Henry Hurst

With regard to imperialism, it is striking that these people used one of its prime tools—​
writing in Latin—​as a medium to express their thoughts. The content of these defixiones
can perhaps be viewed in a similar way to letters to agony aunt columns in newspa-
pers: they show people’s concerns, but give no balanced view of society; the modern
concerns are, however, expressed publicly.

Texts Mainly Viewing Britain from 

the Outside: Issues of History
and Archaeology

Writing about Britain includes a range of material from poetry to road itineraries, all of
it fragmentary. At its core are the accounts of three historians: Tacitus, a Roman senator,
writing around the turn of the first and second centuries, Dio Cassius, another sena-
tor, who came originally from Nicaea (modern Iznik in Turkey) and wrote in Greek a
century after Tacitus, and Ammianus Marcellinus, from Antioch (Antakya in south-​
east Turkey), who lived c. 33–​395 and wrote (in Latin) a history of Rome from ad 96
in thirty-​one books. These and shorter passages in other texts, sometimes as short as
a phrase, but derived from nearly 100 different authors, together with a substantial
amount of geographical and place-​name information, have been studied for genera-
tions past. As noted, this material offers military and political history between ad 43
and the governorship of Agricola (ad 77–​85), slanted in his favour because of the sur-
viving writing by his son-​in-​law Tacitus, including parts of his biography, the Agricola.
Relatively little survives covering later events and nothing of any substance about, or
resulting from, cultural life in Britain: no Latin works of British authors survive until the
fifth-​century writings of Pelagius and St Patrick and the probably sixth-​century work
of the monk Gildas (Dumville 1984). In the classical authors there are a handful of cli-
chés (topoi) about Britain being across the ocean with backward peoples who lived off
meat and milk and painted themselves blue in battle (as Caesar, Gallic War, V. 14), and,
in Tacitus, some rhetorical swipes about Roman power, of which enslavement through
comfort (Agricola 11) is perhaps one of the best known; there is also some information
from Ptolemy’s Geography of the second century ad. As already commented, these
meagre texts were supplemented by epigraphy and archaeology to make up ‘histories’
of Roman Britain, and a political history for the pre-​conquest period was also recon-
structed from inscribed British coins. Military and political history will always be a part
of the study of Roman Britain, because this is one way—​even if, emphatically, not the
only one—​to place this province within a grand narrative of the Roman Empire. Textual
and archaeological evidence will be combined within this, but this is often not straight-
forward. The directness of the tie-​up between the mention by Tacitus, in Annals 14, 38,
of Classicianus, procurator of Roman Britain at the time of the Boudiccan uprising, and
the tombstone found in London (Figure 5.5: cf. Grasby and Tomlin 2002), is exceptional.
 The Textual and Archaeological Evidence    111

Combining textual and archaeological evidence requires a high command of both

historical and archaeological thinking as well as data: its challenge tends to be under-
rated, because most of us lean towards one or other of these, and there is a tendency to
oversimplify the methodology of the less familiar discipline. The problem is celebrated
in obiter dicta, such as ‘historians who shop in the supermarket of archaeology’ (Pucci
1983: 106). In Roman Britain, because of the relative survival of evidence, it has been
mostly archaeologists shopping in the supermarket of surviving texts. The flaw seems
to be a certain wished-​for historical neatness, of tying archaeological evidence into ran-
domly surviving documentation, with which it usually has no more than an indirect
relationship. It is all too easy to find examples of misguided attempts at this, and it would
be dispiriting to dwell too much on them, but we should nevertheless draw the lesson
from one episode of military archaeology in early Roman Britain, of how it led to distor-
tion of both the archaeological and documentary information and in the medium term
harmed the subject. In the 1960s and 1970s there was an attempt to reconstruct the mili-
tary activities of individual governors in the early years of Roman Britain by assigning to
their term of office particular forts or fortresses (as Webster 1970). The dating and loca-
tion of such sites were compatible with the brief documentation in Tacitus. However,
archaeology was uncovering mostly semi-​permanent winter quarters, as opposed to
marching camps, which did not reflect every twist and turn in campaigning; their dating
was from finds of coins and pottery rather than from inscriptions, so, at best, fluid to a
decade or so; finds from periods of maximum use were usually not abundant, and fewer
than those from the moment of disuse; and overviews of dispositions had to assume a
more-​or-​less textbook composition of units, whereas the reality was soon discovered
to be more untidy. To be brutal, this choice of research target was unattainable, it led to
inappropriate interpretations of evidence, and it diverted attention away from questions
on which archaeological evidence could have thrown light—​as, for example, the interre-
lationships between military sites, the degree to which they were integrated locally, and
so on. It also encouraged the discussion of non-​existent meanings in the text of Tacitus,
as, for example, whether his rhetorical flourish about the Silures being ‘castrisque legio-
num premenda’ (‘suppressed [or oppressed] by legionary camps’) (Annals 12. 32), meant
one or more forts or fortresses (cf. Hurst 1985: 121, for the details). For a while after the
1970s, for reasons partly of contemporary culture, but also because of disappointment
engendered by the approach described, there was a reaction against research on the
Roman army in Britain (cf. James 2002). Perhaps ironically, the subject was rescued by
the discovery of new texts, for the Vindolanda documents forced us to take a more com-
plex, and more rewarding, view of the Roman army in Britain and a more realistic view
of the relationship between textual and archaeological evidence.
A well-​known comment, made against the general background described, in a sym-
posium on the forts of the Saxon Shore, was Richard Reece’s: ‘We must ask historical
questions of historical sources and archaeological questions of archaeological sources,
and never get the two mixed up’ (quoted by Johnston 1977: p. v).1 That is perfect advice
at the basic methodological level when studying texts and/​or material remains, but, as
I hope has been demonstrated, it is insufficient as a wider statement of the relationship
112

112   Henry Hurst

between history and archaeology. Moreland’s argument that texts are also artefacts
should be remembered, and it should be obvious that, for example, the Vindolanda
tablets, the Bath and Uley defixiones, and the Classicianus tombstone belong to both
fields of study. In any case, the aims of ‘archaeology’ and ‘history’ for Roman Britain are
not fundamentally different. Archaeology might be said to explore human agency as
well as wider socio-​economic or technological processes in order to understand what
happened. So also does document-​based history, except that more names tend to be
attached. That does not mean that there are not a host of special studies, where the two
disciplines can and should be kept apart, but the results are ultimately to be synthesized.
That will be fruitful as long as the conceptual frameworks and disciplinary requirements
of both fields of study are respected.

Conclusion

The whole field of textual and archaeological evidence for Roman Britain has seen a
transformation, both of available material and of approaches over the last three dec-
ades. The texts most studied used to be those written at a distance about Britain, with—​
for the most part—​only formal inscriptions on stone from within the province, and
they led to a Rome-​based, top-​down view, against which there were sharp, and vary-
ingly focused reactions. Now there is a greater body of writing from within Roman
Britain than about it, nearly all of it well published, and this writing is representative
of many sectors of society from the imperial administrators and army to persons made
desperate by the loss of a cloak. It is a rich assemblage, and it challenges us to think
actively about how writing was used and to face the complexities of life in the province
with, perhaps, fewer preconceptions than have been apparent in some earlier studies.
As regards the synthesis of archaeology and history, the archaeology of Roman Britain
is, inescapably, the study of people in whose lives writing was a major presence, even if
most could not read; they are also a people and period often written about (by modern
historians) within two broad frameworks, of the history of Britain and of the Roman
Empire. It will be possible to advance on the archaeological front by developing better
thinking, with the written evidence being given its due weight; and, on the historical
front, by adjusting the type of narrative written about Roman Britain to the range of
evidence that is now available.

Acknowledgements

Professor Martin Millett and Dr Roger Tomlin kindly read earlier drafts and made sug-
gestions for improvement; mistakes, however, remain my own. Thanks are also offered
to those who helped in providing images: Chris Sutherns of the British Museum images
 The Textual and Archaeological Evidence    113

for Figure 5.1 and the photograph in Figure 5.5; Dr Tomlin for Figures 5.2 and 5.3;
Professor Millett for Figure 5.4; the staff of Alamy for Figure 5.6.

Abbreviations
CIL  Corpus Inscriptionum Latinarum
RIB  The Roman Inscriptions of Britain

Note
1. Professor Martin Millett kindly located the concealed setting of this well-​known remark.

References
Barrett, A. A. (1978). ‘Knowledge of the Literary Classics in Roman Britain’, Britannia,
9: 307–​313.
Beard, M. (1991). ‘Ancient Literacy and the Function of the Written Word in Roman Religion’,
in J. H. Humphrey (ed.), Literacy in the Roman World. JRA Supplementary Series 3. Ann
Arbor: Journal of Roman Archaeology, 35–​58.
Biddulph, E. (2013). ‘The Blind Potter: The Evolution of Samian Ware and its Imitations’, in M.
Fulford and E. Durham (eds), Seeing Red: New Economic and Social Perspectives on Terra
Sigillata. London: Institute of Classical Studies, 368–​380.
Bowman, A. K. (2003). Life and Letters on the Roman Frontier: Vindolanda and its People. 2nd
edn. London: British Museum Press.
Bowman, A. K., and Thomas, J. D. (1983). Vindolanda: The Latin Writing-​Tablets. Britannia
Monograph Series 4. London: Society for the Promotion of Roman Studies.
Bowman, A. K., and Thomas, J. D. (1994). The Vindolanda Writing-​ Tablets (Tabulae
Vindolandenses). Volumes 2 and 3. London: British Museum Press.
Bowman, A. K., and Woolf, G. (1994) (eds). Literacy and Power in the Ancient World.
Cambridge: Cambridge University Press.
Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2010). ‘The Vindolanda Writing-​Tablets
(Tabulae Vindolandenses IV, Part 1)’, Britannia, 41: 187–​224.
Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2011). ‘The Vindolanda Writing-​Tablets
(Tabulae Vindolandenses IV, Part 2)’, Britannia, 42: 113–​144.
Brilliant, R. (1984). Visual Narratives:  Story-​Telling in Etruscan and Roman Art. Ithaca,
NY: Cornell University Press.
Collingwood, R. G., and Wright, R. P. (1965). The Roman Inscriptions of Britain. Volume
I: Inscriptions on Stone. Oxford: Clarendon Press.
Collingwood, R. G., and Wright, R. P. (1991–​5). The Roman Inscriptions of Britain, Volume
II: Instrumentum Domesticum. Fascicule 1: Gloucester: Alan Sutton. Fascicules 2 (1991), 3
(1991), 4 (1992), 5 (1993), 6 (1994), 7 (1995), 8 (1995), indexes (1995): Stroud: Alan Sutton for
the Administrators of the Haverfield Bequest.
Dumville, D. (1984). ‘The Chronology of De Excidio Britanniae, Book I’, in M. Lapidge and D.
Dumville (eds), Gildas: New Approaches. Woodbridge: Boydell Press, 61–​84.
114

114   Henry Hurst

Evans, J. (1987). ‘Graffiti and the Evidence of Literacy and Pottery Use in Roman Britain’,
Archaeological Journal, 144: 191–​204.
Frere, S. S. (1987). Britannia: A History of Roman Britain. 3rd edn. London: Routledge and
Kegan Paul.
Gager, J. G. (1992). Curse Tablets and Binding Spells from the Ancient World. New York and
Oxford: Oxford University Press.
Gardner, A. (2007). An Archaeology of Identity: Soldiers and Society in Late Roman Britain.
Walnut Creek, CA: Left Coast Press.
Goodburn, R., and Waugh, H. (1983). The Roman Inscriptions of Britain [edited] by R. G.
Collingwood and R. P. Wright. I: Inscriptions on Stone: Epigraphic Indexes. Gloucester: Alan
Sutton.
Goody, J. (1977). The Domestication of the Savage Mind. Cambridge:  Cambridge
University Press.
Gordon, A. E. (1983). Illustrated Introduction to Latin Epigraphy. Berkeley and Los Angeles, and
London: University of California Press.
Grasby, R. D. (2012). Processes in the Making of Roman Inscriptions. Study 7: RIB 12. Oxford: The
Centre for the Study of Ancient Documents.
Grasby, R. D., and Tomlin, R. S.  O. (2002). ‘The Sepulchral Monument of the Procurator
C. Julius Classicianus’, Britannia, 33: 43–​75.
Hanson, W. S., and Conolly, R. (2002). ‘Language and Literacy in Roman Britain:  Some
Archaeological Considerations’, in A. E. Cooley (ed.), Becoming Roman, Writing Latin?
Literacy and Epigraphy in the Roman West. JRA Supplementary Series 48. Portsmouth,
RI: Journal of Roman Archaeology, 151–​164.
Harris, W. V. (1989). Ancient Literacy. Cambridge, MA: Harvard University Press.
Hartley, B. R. (1969). ‘Samian Ware or Terra Sigillata’, in R. G. Collingwood and I. A. Richmond,
The Archaeology of Roman Britain. London: Methuen, 235-​251.
Hartley, B. R., and Dickinson, B. M. (2008–​12). Names on Terra Sigillata:  An Index of
Makers’ Stamps and Signatures on Gallo-​Roman Terra Sigillata (Samian ware). 9  vols.
London: Institute of Classical Studies.
Hassall, M. W.  C., and Tomlin, R. S.  O. (1989). ‘Roman Britain in 1988. II. Inscriptions’,
Britannia, 20: 327–​345.
Henig, M. (1997). ‘The Lullingstone Mosaic: Art, Religion and Letters in a Fourth-​Century
Villa’, Mosaic, 24: 4–​7.
Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman
Archaeology. London: Routledge.
Hingley, R. (2005). Globalizing Roman Culture:  Unity, Diversity and Empire. London:
Routledge.
Hull, M. R. (1963). The Roman Potters’ Kilns of Colchester. Reports of the Research Committee
of the Society of Antiquaries of London, XXI. Oxford: Oxford University Press.
Hurst, H. (1985). Kingsholm. Cambridge: Gloucester Archaeological Publications.
Ireland, S. (2008). Roman Britain: A Sourcebook. 3rd edn. London: Routledge.
James, S. (2002). ‘Writing the Legions: The Past, Present and Future of Roman Military Studies
in Britain’, Archaeological Journal, 159: 1–​58.
Johnston, D. E. (1977) (ed.). The Saxon Shore. CBA Research Report 18. London: Council for
British Archaeology.
Kenyon, R. F. E. (1993). ‘The Copying of Bronze Coins of Claudius I in Roman Britain’.
Unpublished Ph.D. thesis, Institute of Archaeology, University College London.
 The Textual and Archaeological Evidence    115

Ling, R. (2007). ‘Inscriptions on Romano-​British Mosaics and Wall-​Paintings’, Britannia,

38: 63–​91.
Mairs, R. (2012). ‘“Interpreting” at Vindolanda: Commercial and Linguistic Mediation in the
Roman Army’, Britannia, 43: 17–​28.
Mann, J. C., and Penman, R. G. (1996) (eds). Literary Sources for Roman Britain. 3rd edn.
London: London Association of Classical Teachers.
Martin, M. (2010). Sois maudit! Malédictions et envoütements dans l’antiquité. Paris: Errance.
Mattingly, D. J. (1997) (ed.). Dialogues in Roman Imperialism: Power, Discourse and Discrepant
Experience in the Roman Empire. Journal of Roman Archaeology Supplementary Series 23.
Portsmouth, RI: Journal of Roman Archaeology.
Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409.
London: Allen Lane (Penguin Books).
McKitterick, R. (1990) (ed.). The Uses of Literacy in Early Medieval Europe.
Cambridge: Cambridge University Press.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Millett, M. (2005). Roman Britain. London: Batsford.
Moreland, J. (2001). Archaeology and Text. London: Duckworth.
Mullen, A. (2007). ‘Evidence for Written Celtic from Roman Britain: A Linguistic Analysis of
Tabellae Sulis 14 and 18’, Studia Celtica, 41: 31–​45.
Mullen, A. (2013). ‘The Language of the Potteries:  Communication in the Production and
Trade of Gallo-​Roman terra sigillata’, in M. Fulford and E. Durham (eds), Seeing Red: New
Economic and Social Perspectives on Terra sigillata. London: Institute of Classical Studies,
97–​110.
Neal, D. S., and Cosh, S. R. (2009). Roman Mosaics of Britain. Volume III: South-​East Britain,
Part 2. London: Society of Antiquaries.
Pearce, J. (2004). ‘Archaeology, Writing Tablets and Literacy in Roman Britain’, Gallia, 61: 43–​51.
Perring, D. (2003). ‘Gnosticism in Fourth-​ Century Britain:  The Frampton Mosaics
Reconsidered’, Britannia, 34: 97–​127.
Pucci, G. (1983). ‘Pottery and Trade in the Roman Period’, in P. Garnsey, K. Hopkins, and C. R.
Whittaker (eds), Trade in the Ancient Economy. London: Chatto and Windus, 105–​117.
Salway, P. (1981). Roman Britain:  The Oxford History of England, Volume 1A.
Oxford: Clarendon Press.
Scott, S. (2000). Art and Society in Fourth-​ Century Britain:  Villa Mosaics in Context.
Oxford: Oxford University School of Archaeology.
Street, B. V. (1984). Literacy in Theory and Practice. New York: Cambridge University Press.
Strong, D. (1968). ‘The Monument’, in B. W. Cunliffe (ed.), Fifth Report on the Excavations of
the Roman Fort at Richborough, Kent. Reports of the Research Committee of the Society of
Antiquaries 23. Oxford: University Press for the Society of Antiquaries, 4–​73.
Tite, P. L. (2010). ‘“Reading” and “Re-​Reading” the Frampton Mosaics: Religious Innovation
and the Construction of Cultural Identity in Roman Britain’, Bolletino di Archeologia Online.
Volume Speciale E/​E10/​4: 41–​55.
Tomlin, R. S. O. (1983). ‘Non Coritani sed Corieltauvi’, Antiquarians Journal, 63: 353–​355.
Tomlin, R. (1988). ‘The Curse Tablets’, in B. Cunliffe (ed.), The Temple of Sulis Minerva at Bath.
Volume 2: The Finds from the Sacred Spring. Oxford Committee for Archaeology Monograph
16. Oxford:  Oxford Committee for Archaeology, 59–​277. [Also published separately as
Tabellae Sulis: Roman Inscribed Tablets of Tin and Lead from the Sacred Spring at Bath.]
116

116   Henry Hurst

Tomlin, R. S.  O. (1993). ‘The Inscribed Lead Tablets:  An Interim Report’, in A. Woodward
and P. Leach (eds), The Uley Shrines: Excavation of a Ritual Complex on West Hill, Uley,
Gloucestershire, 1977–​9. London: English Heritage, 113–​130.
Tomlin, R. S.  O. (1996). ‘A Five-​Acre Wood in Roman Kent’, in J. Bird, M. Hassall, and
H. Sheldon (eds), Interpreting Roman London:  Papers in Memory of Hugh Chapman.
Oxford: Oxbow Books, 209–​216.
Tomlin, R. S. O. (1998). ‘Roman Manuscripts from Carlisle: The Ink-​Written Tablets’, Britannia,
29: 31–​84.
Tomlin, R. S. O. (2002). ‘Writing to the Gods in Britain’, in A. E. Cooley (ed.), Becoming Roman,
Writing Latin? Literacy and Epigraphy in the Roman West. JRA Supplementary Series 48.
Portsmouth, RI: Journal of Roman Archaeology, 165–​179.
Tomlin, R. S. O. (2003). ‘“The Girl in Question”: A New Text from Roman London’, Britannia,
34: 41–​51.
Tomlin, R. S. O. (2011). ‘Stylus Writing Tablets’, in J. Hill and P. Rowsome (ed.), Roman London
and the Walbrook Stream Crossing: Excavations at 1, Poultry and Vicinity. MOLA Monograph
37. London: Museum of London Archaeology, 514–​517.
Tomlin, R. S.  O., and Hassall, M. W.  C. (2001). ‘Roman Britain in 2000. II. Inscriptions’,
Britannia, 32: 387–​400.
Tomlin, R. S. O., Wright, R. P., and Hassall, M. W. C. (2009). The Roman Inscriptions of Britain.
Volume 3: Inscriptions on Stone Found or Notified between 1 January 1955 and 31 December
2006. Oxford: Oxbow Books.
Turner, E. G. (1956). ‘A Roman Writing-​Tablet from Somerset’, Journal of Roman Studies,
46: 115–​118.
Turner, E. G., and Skutsch, O. (1960). ‘A Roman Writing-​Tablet from London’, Journal of
Roman Studies, 50: 108–​111.
Tyers, P. (1996). Roman Pottery in Britain. London: Batsford.
Unesco (2006). ‘ Understandings of Literacy’, in Education for All. Unesco Global Monitoring
Report 2006, ch. 6.
Wacher, J. S. (1998). Roman Britain. 2nd edn. Stroud: Sutton.
Webster, G. (1970). ‘The Military Situations in Britain between ad 43 and 71’, Britannia, 1:
179–​199.
Welles, C. B., Fink, R. O., and Gilliam, J. F. (1959). The Parchments and Papyri: The Excavations
at Dura-​Europos. Final Report 5, Part 1. New Haven: Yale University Press.
Williams, J. H. C. (2002). ‘Pottery Stamps, Coin Designs and Writing in Late Iron Age Britain’,
in A. E. Cooley (ed.), Becoming Roman, Writing Latin? Literacy and Epigraphy in the Roman
West. JRA Supplementary Series 48. Portsmouth, RI:  Journal of Roman Archaeology,
135–​150.
Woolf, G. (1996). ‘Monumental Writing and the Expansion of Roman Society in the Early
Empire’, Journal of Roman Studies, 86: 22–​39.
Woolf, G. (1998). Becoming Roman: The Origins of Provincial Civilization in Gaul. Cambridge:
Cambridge University Press.
 Chapter 6

The Early Rom a n H ori z on

Lacey Wallace

Introduction

The early Roman horizon in Britain refers both to archaeological deposits from a par-
ticular time period (in the south-​east of Britain, it corresponds roughly to the genera-
tion who lived during and after the Claudian invasion in ad 43) as well as to the material
remains of an emerging Romano-​British culture. To question what happened to the
lives of the indigenous population in Britain when their land was incorporated into the
Roman Empire requires not only the scrutiny of this evidence to investigate how people
survived, adapted, and resisted in a period of rapid change, but also an elusive chron-
ological specificity. Isolating the ‘early Roman’ strata for examination, exploring their
relationship to the late pre-​Roman Iron Age (LPRIA), and drawing meaning and signifi-
cance from the study of artefacts, features, and deposits is a complex feat, and one that
can be touched upon only briefly here (see also Creighton 2006). Bearing this caveat in
mind, this chapter will explore a few of the problems and biases encountered by archae-
ologists studying the ‘early Roman horizon’ and examine the nature of the surviving
archaeological evidence for change and continuity, particularly within the early urban
centres destroyed, and thereby preserved, in the Boudiccan Revolt of ad 60/61.

Problems with the Evidence

and Overcoming our Biases

One of the main obstacles one encounters when exploring the ‘early Roman horizon’ is
the artificial distinction of the LPRIA and the Roman period, divided by the Claudian
invasion of ad 43. The temptation to put weighty emphasis on ad 43 ought to be avoided,
as few sites bear evidence with which an archaeologist could identify changes resulting
from the invasion.
118

118   Lacey Wallace

The academic separation of pre-​Roman and Roman archaeological remains stems,

simply put, from the classical education prevalent in the modern period, which led
to British archaeology of the Roman period being interpreted within the literary and
historical framework of classical studies (See also chapters by Millett and Wilson, this
volume). Traditional syntheses of Roman Britain often use the histories (e.g. Tacitus’
Annals and Agricola, Caesar’s Gallic Wars, and Cassius Dio’s Histories) to create a nar-
rative for the archaeological evidence, which does a disservice to both disciplines by
oversimplifying the relationship between them (see Creighton 2001). Archaeological
investigations can include analysis of ancient texts as part of the research arsenal, but
ought not to use this extremely limited and incomplete body of evidence to create an
analytical framework. The LPRIA archaeological record cannot be ‘matched up’ to mili-
tary incursions and administrative changes, such as the invasions of Caesar and pos-
sible installation of loyal vassal kings (Creighton 2000). Similarly, the movements of
troops described by ancient historians ought not to be linked directly to the building of
roads and founding of towns; a strong bond ties those scholars who have emphasized
the historical accounts to those who would see the Roman military as the mechanism
by which social and economic change occurred in Britain. An academic tradition of
linking archaeological remains to events described by ancient authors—​most of whom
had never travelled to Britain and who wrote with their own agendas and intended
audiences—​is entrenched in our field to this day. This connection has led to assump-
tions and outmoded theories, which can take decades to unseat and must be challenged
and confronted with rigorous studies of evidence.
We can add to these pitfalls the regional diversity of the effects of contact with the
Roman Empire: it is a certainty that the ‘early Roman period’ was experienced differ-
ently in many parts of Britain as ideas, objects, and people moved through the new
province. The experience of people living in east Kent, for example, who had been in
direct contact with the Roman Empire for at least a century before the Claudian inva-
sion, would have differed greatly from people living in locations distant from the trade
routes bringing continental goods to Britain. No two sites exhibit the same ‘early Roman
horizon’, and generalizing would result, at best, in a vague account. If we can agree to call
the ‘early Roman period’ in the south-​east of Britain that which it is possible to date to
the decades immediately following the Claudian invasion, we will immediately see that
it was experienced drastically differently in the countryside from the way it was in the
oppida, emerging urban centres, and military installations.
Finally, one must also consider the advantages and issues of studying a particular
time period as brief as the ‘early Roman horizon’, as well as those types of evidence that
can offer the possibility of assigning dates. Although this chapter avoids the vagaries
of archaeological dating, including the sometimes-​circular logic for the assignation
of artefact dates—​for example, historical events used to identity strata, which contain
object types that are then used to link strata elsewhere with historical dates, eventually
creating a corpus of artefacts assigned a particular date range—​a healthy scepticism of
the precision of such techniques ought to be maintained.
 The Early Roman Horizon    119

Overcoming these Issues in Londinium

Few sites offer robust evidence to isolate archaeological strata from the early Roman
period; Londinium, however, is one such site. Because there was no significant LPRIA
occupation in the area that would become the Roman town, the issue of identifying the
early Roman transition is avoidable. Overcoming the historical narrative remains a sig-
nificant obstacle to interpreting early Londinium, but it is one that can be explored with
a huge quantity and variety of evidence. Being placed in the south-​east at the centre of
transport, trade, and communication routes both over land and by sea, London can be
seen as an ideal site for exploring the urban experiences of the generation who lived
through the Claudian invasion. Moreover, the ad 60/​61 Boudiccan Revolt burnt depos-
its provide a clear archaeological cut-​off point that can be used to isolate the first decade
(or so) of occupation there.
In London, as elsewhere, scholars have made hypotheses (and assumptions) about
the early Roman levels based on existing historical, toponym, and anecdotal evidence
before many sites had been excavated. By the 1950s, the notion of a pre-​Roman London
was out of favour, and a major compilation of the evidence (Merrifield 1965) put archae-
ology at the forefront of future investigations, although there were few sites then known
representing the earliest levels of Roman London. After this time, the academic context
was characterized by the popularity of administration-​driven urbanization following
military installations (the fort-​to-​town hypothesis—​see Webster 1966 and Frere 1967).
So it was suggested by many that a camp, fort, or supply base provided the location,
impetus, and form for the earliest urban settlement, linked to the military movements
of Plautius as described by Cassius Dio (a third-​century historian who wrote in Greek
and had never been to Britain). Hanging the interpretations on the historical evidence,
many thought that, because Plautius crossed the Thames somewhere, it was most likely
that he crossed at London and set up a camp for his troops while they waited for the
arrival of Claudius. The ancient literature was not often subject to critical evaluation,
but rather taken to be a relatively unbiased record of events, as understood by their
authors.
The ubiquitous use of the word ‘conquest’, until relatively recently, employed to
describe the result of Claudius’ efforts in ad 43, is a reflection of how scholars have
allowed their interpretations to be skewed by the propaganda of the time. The effort to
‘match up’ the events in the surviving accounts—​for example, to link the construction
of roads to the movements of troops described—​led to a situation where archaeologists
decided what they were going to find before they dug it up. In London, historians and
archaeologists expected to find the military installation of regular streets crossing at a
principia (underlying the later forum) with rows of barracks and store buildings, sur-
rounded by defensive ditches, all thought to have given form to the later town. The effect
of London and its bridge was seen to have been caused by the presence of a Claudian
base of some kind.
120

120   Lacey Wallace

One example demonstrates how otherwise-​ambiguous evidence could be forced

into the narrative derived from the scraps of ancient histories. In 1968–​9, an excavation
north of Fenchurch Street revealed a small amount of occupation debris, burnt sedi-
ment, a copy of a Claudian coin, and some coarseware pottery (Philp et al. 1977: 7). No
contemporaneous structures were found. The interpretation of this meagre evidence
was described thus:

All of the material recovered from these primary deposits is of mid-​first-​century

date and thus in accord with the long-​established view that permanent settlement
began in London in 43, at the time of the Roman conquest. The probability is that
this primary occupation was of a purely military character, perhaps reflecting either
a temporary camp on this site or activity by troops living nearby. (Philp et al. 1977: 7)

There is no way that this confident interpretation, slotted uncomfortably into a histori-
cal paradigm, would have been made without the ancient texts; it is entirely dependent
on them and acts as an illustration of what is already ‘known’, not as a search for new
evidence. Certainly there are examples where towns founded in the 40s and 50s occupy
earlier military settlements (for example, the veteran settlement at Colchester), but
the assumption that urbanization could not be a force driven by civilians robs the vast
majority of the population of their agency in forming the landscape of the new province.
At the time, only a handful of sites with possible early Roman levels had been found,
and most of these were merely observed as they were destroyed by construction, so
hypotheses could not be formed by placing evidence within the context of wider pat-
terns. Over the next four decades, another ninety sites with features pre-​dating the
Boudiccan fire would come to light. These 118 sites have produced more than 14,000
recorded archaeological contexts (although many more have been destroyed, lost, or
never recorded) (see Wallace 2014). We now have sufficient evidence to begin to dis-
cern patterns in the early Roman levels and to produce interpretations that are separate
from the historical narrative. Where a site has not been investigated, allowing assump-
tions and the written histories to compensate for lack of evidence may be valid. In the
case of Londinium, however, some (e.g. Perring 2011) have allowed their passion for the
historical narrative to overcome the robust data, which are among the finest and most
complete of any Roman town in the empire.

Nature of Change

One of the best sources of evidence for the ‘early Roman horizon’ comes from urban con-
texts, like Londinium. Two other towns in the south-​east that were occupied in the dec-
ade following the Claudian invasion can also be compared: Colchester (Camulodunum/​
Colonia Claudia Victricensis) and St Alban’s (Verulamium). Each of these three also
offers a unique opportunity to investigate a distinct early Roman time period because
 The Early Roman Horizon    121

of the fate these settlements suffered during the Boudiccan Revolt of ad 60/​61. The
destruction of these towns during the Boudiccan Revolt and subsequent rebuilding has
sealed the archaeological deposits within and immediately beneath the fire debris asso-
ciated with the sacking of these towns, allowing us to discuss a cohesive and comparable
period relating to the experiences of the generation who lived through the invasion.
Although their foundations have been linked to the pacification of the south-​east
and the movement of the army north and west during the 40s ad (e.g. Frere 1972) and
although affected by the revolt at approximately the same (archaeologically indistin-
guishable) time, the reasons for the foundation of these early urban sites up to ad 60/​61
were driven by different groups in dramatically different ways. Colchester, founded by
the Roman provincial administration as a veteran colony (Colonia Claudia Victricensis)
in ad 49 atop the earlier legionary fortress, was sited in an area that had not previously
been used for settlement but that lay within a larger oppidum (Camulodunon) area
delineated by major dykes and occupied by scattered ‘hamlets’ as well as high-​status
enclosures and agricultural/​livestock areas (Gascoyne and Radford 2013). In contrast,
a strong ruling elite and a group of people (tribe?) who had occupied a relatively large
and well-​populated oppidum area (Verulamion) in the Ver river valley for some time
then chose to begin to construct an urban settlement (Verulamium) in a symbolic and
significant location where some people would live and work in much closer proximity
to one another than they ever had before (Niblett 2001; Niblett and Thompson 2005).
The earliest phases of development at Londinium also suggest a different agenda. Here,
an implanted population of indigenous Britons and continental merchants and crafts-
men not local to that place came to build a ‘town’ on an unoccupied site, and organized
labour and communal development, interpreting what it meant to live in a town in dif-
ferent ways (Wallace 2014).

Verulamium
That Verulamium was already established and linked into the provincial road networks
prior to the Boudiccan Revolt is attested to, for example, by the burnt remains of timber
workshops in Insula XIV (Frere 1972: 19), which overlay the road onto which they faced
and also overlay evidence of Claudian-​period occupation. Frere, therefore, made the
link between the ‘foundation’ of Verulamium as an urban centre (which he erroneously
believed to overlay an earlier fort) and the creation of the colonia at Colchester as both
being part of an imperial policy of urbanization.
In the pre-​Boudiccan period, the nascent town of Verulamium lay at the centre of
the later urban core, but was a small area (less than 10 hectares) enclosed by ditches
and banks. A quarter of its area was taken up by the pre-​Claudian ‘Central Enclosure’,
which pre-​dated the town and was a location where blanks for British coins were made
(interpreted through the presence of coin moulds) and may have been a site of high-​sta-
tus occupation perhaps constructed and built by the same high-​status family/​individu-
als as the ‘proto-​villa’ at Gorhambury (see Neal 1990) and the ceremonial site at Folly
122

122   Lacey Wallace

Lane (see Niblett 1999). The ‘Central Enclosure’, although the course and nature of its
ditches are imperfectly known, formed an area of c. 2.4 hectares underneath the later
forum/​basilica complex (Niblett 2005). This feature, along with a 6-​metre-​wide bank
(possibly revetted with timbers), a pile-​supported timber causeway, and a major north-​
east–​south-​west ditch, pre-​dated the Flavian grid of streets and formed the planning of
the early ‘town’. The sub-​rectangular area formed by these ditches contained within it
the crossing of Watling Street (prior to its rerouting) and the main south-​west–​north-​
east road focused on the high-​status Folly Lane funerary/​ceremonial enclosure atop the
ridge across the stream. The four known roads within the area were lightly metalled,
and, although they were at roughly right angles, they did not enclose insulae in the way
of later roads and did not extend beyond the ditched area.

Colonia Claudia
In Colchester, the fort (which would form the basis for the colonia) was built within a
larger oppidum area delineated by major dykes. The relationship between the ‘town’ at
Colchester and the indigenous local group was, therefore, entirely different from that
at Verulamium. The area of the fort and additional areas constructed when it was trans-
formed into a colonia covered c. 36 hectares, forming a grid-​planned area far larger than
that of Londinium. We expect that those allocated land in the colonia were veterans,
but their families and households could well have been of indigenous British descent.
Labelling Colonia Claudia as a ‘military’ community is therefore difficult, although it is
likely that those British people who chose to live and work in the town were considered
‘sympathizers’ by traditionalists like Boudica and her army of confederated tribes.

Londinium
The earliest date for the construction of Watling Street through Londinium is ad 47/​48,
derived from the felling date of a tree used for its timber to line a drain underneath the
road (Tyers 2011). That date would be consistent with the beginning of development in
the town c. ad 49 (see also Wallace 2014).
In Londinium, the core Cornhill settlement was delineated with a grid of streets cov-
ering c. 18 hectares, containing six insulae, a gravelled ‘proto-​forum’ area, as well as rela-
tively ephemeral waterfront/​port structures, all of which were constructed slowly over
the first decade (see Wallace 2014). The other areas of the settlement were not ‘planned’
in this way, but rather formed in a linear development along the roadsides outside of
the centre; this other occupation covered a further 22 hectares and may have actually
housed more people than the central part of the town. The settlement was, therefore,
divided into three areas separated by streams: west of the Walbrook stream lay Ludgate
Hill, where indigenous families lived, probably engaging in scavenging and reclama-
tion to survive in the early days, which rapidly became a densely occupied mercantile
 The Early Roman Horizon    123

and production area focused on the London–​Verulamium road; east of the Walbrook
on Cornhill lay the planned core, probably created and inhabited by immigrant citizens
and merchants; south of the Thames, the London–​Canterbury stretch of Watling Street
crossing two low-​lying eyots was lined with smaller houses, shops, and workshops
inhabited by those immigrants who were perhaps slightly later in coming to Londinium,
of lower status than those on Cornhill, and probably culturally distinct from the popula-
tion on Ludgate Hill (Wallace 2014). Analysis of these data shows that the town was not
just founded, planned and expanding in a simple pattern of growth, even in this short
time period. The processes and decisions that affected its earliest years were mutable
and piecemeal. For many different reasons (availability of materials, need to build rap-
idly, rejection or devaluation of certain cultural values), nearly all of the early buildings
in Londinium were constructed of ephemeral materials. Buildings with posts and tim-
bers in direct contact with the ground would be susceptible to rot and would have to be
repaired and replaced at intervals. Even in the short period before the Boudiccan Revolt
fire, many buildings were torn down and replaced. Equally, the planning and construc-
tion of roads were not even across the town—​buildings were destroyed to make way for
roads in some places, while existing roads were blocked by new buildings in others. This
dynamic and changing urban landscape makes identifying broad patterns a complex
endeavour.

‘Things Called Towns’

The most significant change in settlement patterns at these three early urban centres is,
of course, that a condensed area was chosen in which to construct houses, workshops,
and public spaces. Although these new ‘towns’ (see Reece 1988) did not replace the set-
tlements surrounding Verulamium and Colonia Claudia, their existence suggests that
people did alter their lifestyles in either or both work and domestic activities. At the
outset of ‘planning’ of these towns, a significantly smaller area was selected than would
eventually be occupied at the height of their expansion.

Lack of Public Buildings

The small scale of each of these towns is not just in their restricted area, but also in the
general lack of public buildings in each. Neither Londinium nor Verulamium had a
monumental forum complex until the Flavian period (Verulamium’s is dated by a dedi-
catory inscription of ad 79 (see Frere 1983: 55, 69); and London’s, less specifically, was
constructed c. ad 80 (see Marsden 1987: 22)), despite the fact that fora arguably formed
the heart of any ‘Roman’ city and made it navigable and recognizable to those famil-
iar with urban space on the continent. The space where the forum would later be built
in each was a reserved central location in the pre-​Boudiccan period. In Verulamium,
124

124   Lacey Wallace

the forum overlay the LPRIA ‘Central Enclosure’; in Londinium, the gravelled ‘proto-​
forum’ and its large storage/​distribution/​administrative buildings may have served a
very different function from the later structures in the same area, but clearly show that
there was a power structure in place that could organize the construction of a large pub-
lic space. During the construction of Colonia Claudia, the area of the military principia
was possibly cleared to make way for the forum, but it is not proven.
The Temple of the Divine Claudius in Colonia Claudia is, however, an example of
a major early urban monument unparalleled elsewhere in Britain in this period—​and
may itself have been placed at the head of the forum of Colonia Claudia. In Verulamium,
the macellum was comprised of two wings of rooms (workshops? shops?) enclosing a
hall or courtyard with a conduit supplying water and may be the only pre-​Boudiccan
‘public’ building from Verulamium, but its date, as well as its function, are unclear. A
bathhouse comprising a mortar-​lined sunken ‘pool’ surrounded by limestone columns
with evidence for painted wall plaster need not have been public, but may, rather, have
been associated with elite occupation (that is, the so-​called proto-​forum within the
‘Central Enclosure’) (Niblett 2005; Niblett et al. 2006). In Londinium, the large build-
ings associated with the proto-​forum (one of which may have been a grain storage and
distribution centre) and a masonry-​founded aisled hall across the main street repre-
sent the only probable public buildings, but are not of the form or function typical of a
Roman town (Wallace 2014).
None, however, had the full suite of public buildings that we traditionally associ-
ate with urban centres in Roman provinces on the continent (baths, theatre, forum,
basilica, temples). The absence of such public buildings is surprising in the light of how
we generally think of the development of planned ‘Roman’ towns. For such crucial ele-
ments of the urban landscape not to have been extant during the first decade (or so) of
occupation in well-​dated urban centres in the south-​east—​well situated on the road
and communication/​trade networks, with excellent access to continental goods and
ideas and bearing the ‘trademark’ of at least a limited area of orthogonal streets (see,
e.g., MacDonald 1986: 17)—​presents a stark contrast to the foundation stories of cities
at the centre of the empire. The idea that a grid of streets only served to provide a shape
to the urban space and was not connected to what we might consider ‘urban planning’
in a modern sense (Laurence et al. 2011: 116) certainly applies to at least Verulamium
and Londinium, but underplays the significance of how people would have made the
decision to build a grid and organize labourers and space, which would have been a
significant and symbolic resolution integrally related to their dedication to urbanize
and plan for a town.

Adoption of New Material Culture

The early Roman levels of each of these settlements are characterized by a sharp increase
in the adoption of imported goods and objects, which also suggest the adoption of new
behaviours. In Londinium, the proportion of imported ceramics was at its height in the
 The Early Roman Horizon    125

pre-​Boudiccan period (Davies et al. 1994; Wallace 2014), which demonstrates the weak
ties the earliest population there had to local trade networks and craftsmen (This can be
compared to Waldgirmes, a town probably constructed by the military for civilian use
that was strongly situated within local trade networks (see Schnurbein 2003)). Coupled
with the cultural package of writing equipment, glass vessels, coins, lamps, toilet equip-
ment, and continental-​style adornment and body-​preparation objects, it is clear that
the population either arrived in Londinium with the knowledge and habits common to
other Roman territories on the continent or rapidly adapted to the use of new material
culture. Certain of these objects—​lamps and lamp-​holders, for instance—​are found in
far smaller numbers in pre-​Boudiccan Londinium than in Colonia Claudia (see Eckardt
2002), which may be related to the more widespread use of lamps by veterans than civil-
ians. Indeed, the relative quantities of military equipment bear the same pattern as
lamps in these two settlements (see Wallace 2014). Only a statistically sound methodol-
ogy for comparing quantities of artefacts in the three towns (for example, by density of
artefacts per volume of excavated sediment) could truly clarify the differences in the
adoption of new forms of material culture among the different groups of people who
chose to live in these early urban centres.

So Why Did People Choose to Urbanize?

Why did indigenous British people choose to move to these new settlements, so unlike
those they had lived in before? Why did people from the continent choose to come
to such a new and unstable, unproven place as Britannia? Recent work by Creighton
(2006), Britannia: The Creation of a Roman Province, explored the LPRIA-​to-​Roman
transition in several towns and proposed a variety of motivations for urbanization and
factors that affected the forms it took, such as memory, symbolic landscapes, and re-​
creation of the familiar. He provides an insightful and multifaceted view that draws
directly on the evidence and presents robust hypotheses. Much of the evidence suggests
that these very different people may have had similar motivations—​namely, opportu-
nity for social and economic improvement. Diverse circumstances undoubtedly drove
people to make the leap; some were probably dispossessed of land and status by war,
some may have found that their familiar systems of exchange and production were so
disrupted that they could not continue to survive without choosing to change, others
sought out the military life of stable pay and excitement followed by the reward of land,
while there were undoubtedly those who seized upon a new market and made wide-​
reaching arrangements for developing their hold in it.
Change certainly was more rapid (and presumably also experienced as accelerated) in
urban centres where trade, communication routes, and people came together intensely.
It is clear that certain groups of people chose to adopt new forms of material culture and
to take advantage of the opportunities provided by a closer relationship to the economy
and society of the continent, while others (especially those living in rural areas) contin-
ued life in a relatively unaltered fashion. The elite families that had made use of imported
126

126   Lacey Wallace

goods and continental-​style symbols (for example, wine, coinage, new burial rites, and
so on) to construct their identities in the LPRIA may have been at the centre of driving
change in some instances (for example, at Verulamium), while the early Roman admin-
istrators and military did, in one instance (Colonia Claudia), set up a group of veterans
to live in a town, and in others it appears that ‘regular’ people were on board with the
notion of moving up the social ladder by taking up a new place in a dynamic hierarchy,
especially that afforded by urban centres (for example, at Londinium).

Recognizing Continuity

The ‘early Roman’ period cannot be studied without understanding the preceding
period, the LPRIA (see, for example, the classic works of Cunliffe 1974 and Haselgrove
1984)—​the artificial division can illogically dissociate the archaeological record of
the two. The arrival of the Roman army and administrators did not put an end to kin-
ship, kingship, and indigenous culture; the significance of ‘tribal’ groups in Britain and
the systems of exchange, political alliance, and power in the LPRIA and early Roman
period have been explored through coinage and extrapolation/​comparison with the
histories of other provinces (Creighton 2001). Among those people who lived under
the rule of a British ‘friendly king’, the changes associated with the Claudian invasion
may have affected them little more than had previous treaties with the Roman Empire
(for example, those with Caesar and Augustus) and may have seemed more in the
way of another step in an already-​extant process, rather than the commencement of a
new era.
There are several examples of how assumptions based upon historical frameworks
affect archaeological interpretations. Separating archaeological features into LPRIA
and Roman-​period on the basis of morphology can be misleading and both devalue the
agency of the indigenous groups as catalysts of change as well as mask possible conti-
nuities. For example, at Skeleton Green (Partridge 1981), an Iron Age and Roman road-
side settlement compound in Hertfordshire, a stratum of alluvial sediment covered the
post-​built structures of the LPRIA levels and contained material dating it to c. ad 30–​40;
the subsequent rebuilding was of a timber-​framed construction type and featured right-​
angled roads. Although no artefacts were found in the post-​flood buildings and the dat-
ing evidence suggests that the structures from this phase went out of use before c. ad 50
(Partridge 1981: 49, 51–​52), the author still felt confident to put the early date of ad 43 on
this phase and to consider it ‘Roman’:

the relative neatness of the buildings and the regularity of the road-​system seems to
imply the presence of more than just Natives. The few scraps of military metalwork
may be significant, in that they point to a military presence in the early years follow-
ing the conquest in A.D. 43. (Partridge 1981: 52)
 The Early Roman Horizon    127

A similar dating interpretation was presented a decade earlier, in the assignment of

ad 43 to the earliest structures at Fishbourne (Cunliffe 1974). At Fishbourne, however,
the incongruity between the dating evidence (pottery from c. 10 bc–​ad 25) and the
improbability that the earliest sturdy and substantial ‘military supply-​base’ granary
was standing for just a few years before being razed and replaced was swept under the
rug (see Manley et al. 2005; also Creighton 2006: 54–​58). Cunliffe’s novel interpreta-
tion of ‘supply bases’, at both Fishbourne and Richborough, fit into a historical frame-
work that explained how the military organized the invasion and eventual control
of Britain. Once this historical hypothesis had been decided upon, the dating prob-
lems were explained under the same rubric—​that is, that the pottery was from an old
store that the Claudian-​era military had brought with them. If we follow the notion
of Occam’s razor, this explanation appears inventive and certainly worth questioning.
Most archaeologists would now agree that we cannot point to an archaeological feature
of broadly mid-​first-​century-​ad date and assume it has anything to do with an event as
specific as the Claudian invasion.
Equally important to recognizing that major changes were already happening
in the LPRIA is the acceptance that other changes that occurred after the Claudian
invasion did not necessarily have a direct relationship to it (although, of course, it
is true that some archaeological remains, especially rural sites with little datable
material culture, are notoriously difficult to date with any precision). In rural areas
especially, but also in suburban areas near to early Roman towns, continuity is char-
acteristic of the early Roman period. Niblett (2001: 60), for example, noted ‘there
is little evidence from the Verulamium area as a whole for disruption at the time
of the conquest. Indeed, as far as Verulamium is concerned, were it not for refer-
ences in classical texts, we might not realize that Claudius conquered the area at all.’
Such continuity, especially in close proximity to a municipium, demonstrates that
the experience of empire in the early years after the Claudian invasion was vastly dif-
ferent and dependent on economic status, place in the social hierarchy, occupation,
motivations, and needs.
The most conservative human activities—​such as gastronomy and language—​might
have taken at least a generation (that is, more than twenty years) to begin to change
among certain groups, even where their access to and potential benefit from choosing to
change was strong. The use (and loss/​disposal) of ‘typical’ Roman-​period objects, such
as writing implements, coins, imported ceramics, glass vessels, lamps, brooches, horse-​
tack ornaments, toilet equipment, and so on, may not have formed part of the lives of
ordinary farmers, shepherds, and rural craftsmen. People did not immediately abandon
their homes and dash for the new towns or raze their round-house farmsteads to the
ground after the Claudian invasion. Life continued much as it had previously for a huge
proportion of the population. Buildings would not be replaced just to suit the fashion
where surplus money was not rattling around, although when people did finally rebuild,
they may well have chosen a type of building and techniques of construction that were
fashionable—​and continental—​in origin.
128

128   Lacey Wallace

Exploring Continuity at Colchester

Within the wider oppidum area of Camulodunum/​Colonia Claudia (meaning the c. 28-​
square-​mile Camulodunon area, not a condensed continental-​style, proto-​urban oppi-
dum), LPRIA and early Roman-​period features can be difficult to differentiate where
continuity in form exists. The overall picture at many sites across the oppidum is one
of little change in the early Roman period; the scattered settlements continued to be
inhabited in much the same way after the construction of the legionary fortress. There is
dating evidence that demonstrates that some dykes (large ditches with U-​or V-​shaped
profiles with the excavated material used to create a bank) and non-​rectilinear enclo-
sures, which would traditionally be considered pre-​Roman based on their morpholo-
gies, were constructed across the ad 43 divide. It is worth noting the limitations of the
dating evidence: ‘it is generally difficult to distinguish between pre-​and post-​conquest
occupation because the late Iron Age grog-​tempered pottery continued in use into the
Flavian period’ (Radford 2013: 45). Continuity, however, has been identified into the late
first century in the system of dykes, at the Gosbecks farmstead and sanctuary, at the
Sheepen industrial/​production settlement, and in the burial area at Stanway (Niblett
1985; Crummy et al. 2007).
The material found in the dykes of Camulodunum has been difficult to date, and it
is evident that they were built over a period of time; for example, of the two dykes with
secure termini post quem, the Sheepen Dyke produced only two types of LPRIA ceram-
ics and the assemblage as a whole dates to c. ad 5–​61 (Hawkes and Hull 1947) and the
Gryme’s Dyke is thought to be post-​ad 43 (Hawkes and Crummy 1995: 115). The securely
Roman-​period changes to the dyke system included the construction of a new one in
the form of a triple ditch, while certain other dykes were filled in (Gascoyne 2013: 64,
73). The notion that these dykes were defensive in nature, to create obstacles in open
land rather than to enclose a defensive site, has been the traditional interpretation (e.g.
Hawkes and Hull 1947: 15) and is linked to a reading of Caesar’s Gallic Wars (V.21) as
well as the etymology of the name Camulodunon. Clearly, they demonstrate a stratified
society where labourers could be organized to undertake this monumental work, and
the way that they represent the power of those in authority is a far more poignant aspect
than their possible defensive or deterrent properties.
Sheepen, in the LPRIA and early Roman period (say, c. 15 bc–​Claudian period and con-
tinuing beyond), was a trading and industrial settlement and mint. The Sheepen settle-
ment site was located adjacent to the Sheepen Dyke, between the River Colne and a small
stream, west of the later Roman fortress and town. Within it, evidence of social stratifi-
cation and craft specialization was found: round and sub-​rectangular structures, one of
which was larger and possibly of higher status, evidence of industrial/​work areas (pot-
teries, metal-​working areas) and storage, as well as two coin-​mould sites (of Cunobelin?)
(Hawkes and Hull 1947; Niblett 1985). This settlement was probably controlled and/​or
overseen by the ruling elite (who would have controlled coin production and, possibly,
other industries), and that it was not affected by the construction of the fort and colonia
demonstrates that the local elite held the power to continue to administer their holdings.
 The Early Roman Horizon    129

The best hypothesis as to where this elite family was based is at Gosbecks, where there
is a large trapezoidal farmstead enclosure (Hull 1958:  259), very different in character
from Sheepen, with associated field systems around it visible as cropmarks (Hawkes and
Crummy 1995). It is usually considered to be a ‘royal’ farmstead occupied from the LPRIA
into the Roman period. A monumental enclosure with enormous ditches, adjacent to the
Gosbecks settlement enclosure, may have been the location of an LPRIA shrine or mortu-
ary enclosure (similar to Folly Lane, Verulamium?), which later became a Romano-​British
religious site, indicating that its significance and sanctity continued, possibly under the
patronage of the ruling family. Abutting a nearby LPRIA dyke, a fort thought to be early
Roman has been identified in aerial photographs near to the settlement enclosure. Based
on this fort’s apparent similarity to other mid-​first-​century forts, it has been suggested
that it either immediately pre-​dates or post-​dates the main fortress that underlay the town
(because duplication of military establishments in such close proximity would not be
sensible or beneficial) (Hawkes and Crummy 1995: 101). The location of the fort could be
interpreted as a way that the invaders (whether under Caligula or Claudius) exerted con-
trol over the indigenous population by placing the installation adjacent to the Gosbeck’s
enclosure or it could be interpreted as a British-​built fort modelled after Roman plans—​a
not-​inconceivable idea that is supported by continental/​Roman-​style military artefacts in
high-​status indigenous graves, such as the Lexden tumulus (Creighton 2001: 9).
Continued power probably held by the same people in Camulodunon/​Camulodunum
from the LPRIA through the Roman period can also be found in the elite burial area at
Stanway. The earliest enclosure there dates to the middle Iron Age, the second and larg-
est was late first century bc, the third enclosure was built c. ad 25–​50, the fourth and
fifth were securely post-​invasion (Hawkes and Crummy 1995: 170; Crummy et al. 2007).
Continuity in burial rites, location, and status at Stanway suggests either that there was
no significant break in the power structure, or that the outward expression of status
and power was important to communicate to the local population despite the changes
brought about by the construction of the fort and colonia.
Camulodunon/​Camulodunum is perhaps better understood as an elite-​controlled
agricultural holding than as an oppidum (Radford 2013: 42). Other stock enclosures,
droveways, and field systems within the area enclosed by dykes and streams were con-
structed and used across the LPRIA and Roman periods, indicating a dynamic land-
scape of agriculture and animal husbandry. Various fields, droveways and trackways,
and burial areas might have been associated with the dispersed settlement areas. These
other settlements and land uses did not come to an abrupt end with the construction
of the fort and, later, of the colonia. Continued agriculture and animal husbandry in
the oppidum would have been necessary to support the population, but it is likely that
unskilled and lower-​ranking veterans were given land to farm that was taken from the
indigenous population, although there is evidence from crop marks that LPRIA field
systems remained in use (there was no ‘centuriation’, which is uncommon in Britain
anyway (see Esmonde Cleary 1987); for further discussion of agriculture and animal
husbandry see also Millett and Van der Veen, this volume). Confiscated land for veteran
use would have been the most significant change to indigenous farming and, although
130

130   Lacey Wallace

that is how we understand the benefit to veterans to have taken shape, there is little
archaeological evidence for it.

Conclusions

At the time of the Claudian invasion of Britain, the vast majority of people in Britain
were living in small rural farmsteads or compounds constructed of ephemeral materi-
als, which are difficult for archaeologists to locate, identify, and study. Many of these
rural (often poor) families possessed neither the means nor the motivation to change
their home lives in any material way in the decade or two following the Claudian inva-
sion. Many more people did, of course, embrace the changing power structures, new
fashions, and connection to the imperial network—​no doubt some with a positive out-
look and others with ambition and cynicism—​and a great deal of rebuilding, redesign-
ing, expansion, and taking advantage of new opportunities for social and economic
advancement occurred. New urban centres and military settlements had their effects
on the surrounding territory, including supply demands. A sliding scale of the degree
and velocity of change in the Claudian/​Neronian period can be observed, with greater
changes visible in urban centres and high-​status sites and lesser changes in rural areas
disconnected from the network of trade and communication. An approach that singles
out the Julio-​Claudian/​Neronian periods for examination would probably find little
change or adoption of new material culture on most rural sites. Nonetheless, it would be
of great interest to determine the difference between rural sites that do produce material
evidence of these periods and those that do not.
From the theoretical paradigms (albeit explicitly not discussed here), we must move
forward to robust investigations of the evidence, especially the ‘grey literature’ (e.g.
that from London, which is incorporated into Wallace 2014) that is far more data-​rich
than traditional publications. The post-​excavation work of commercial archaeologi-
cal specialists is often funded, and their reports can be used to identify and analyse the
early Roman levels, even where they are not combined into a finished piece of research.
Rigorously comparing the features, deposits, and artefacts from this decade-​or-​so of
early urban development in Londinium to that of a contemporaneous site with a large
and powerful LPRIA society (for example, Verulamium) or a known military popula-
tion (for example, Colonia Claudia) would also be of great import. Such work is gener-
ally difficult, because such temporal specificity is so rare. Such comparisons would not
be unproblematic, of course. For example, Londinium had no local coin production,
cemeteries, elite burials, enclosures, or high-​status farmsteads dating to the earlier first
century ad; Verulamion had all of these. Although the people of Verulamion did not
adopt imported pottery styles, coin use, and burial behaviours as early as indigenous
people further north and east (for example, in east Hertfordshire), the explanation
probably lies in cultural differences between indigenous groups, rather than in supply or
economic obstacles. The society of Verulamion/​Verulamium was socially stratified, and
 The Early Roman Horizon    131

resources were directed to large ‘public’ works, such as the dykes and ‘Central Enclosure’.
Londinium had no such indigenous population—​indigenous British residents of the
town would have had to move there. Any comparison with a site like Verulamion/​
Verulamium would have to encompass the LPRIA levels.
Despite Verulamium’s pre-​Roman credentials, it—​like Londinium—​was included in
the wide-​sweeping fort-​to-​town urbanization hypothesis. Verulamium has the added
benefit of a recent reappraisal and broad reinterpretation (Niblett and Thompson 2005;
Niblett et al. 2006). Comparing and contrasting the patterns of structural development
and activities between these sites would produce an excellent context within which to
situate future discussions of the early Roman horizon. Changes in Verulamium and
Londinium were driven by indigenous and immigrant groups, but very little by military
or imperial administration. Showing how this process occurred in such different cul-
tural contexts would be illuminating.
We have a tendency to identify with the ‘Romans’ in a way that we do not with indig-
enous Britons—​we think we understand the Romans, that they are not mysterious and
superstitious, but rather that they acted with a dispassionate precision and planning for
the control of the province. Even if one can presume to understand the character and deci-
sions of the governor and his staff, it was the ‘regular’ people who made up the bulk of
the population and who were responsible for most of the archaeological record. When
studying humans, we must remember that pride, personalities, stubbornness, greed, lazi-
ness, ambition, weakness, devotion, sentimentality, and a whole host of other factors affect
the decisions that people make. There is no reason to assume that the ‘obvious’ or most
socially or economically ‘beneficial’ reactions to the Roman invasion were observed. Far
better for us to begin with the evidence and challenge hypotheses until we find one that
cannot be disputed easily. The re-​examinations of evidence (e.g. Creighton 2006; Niblett
et al. 2006; Wallace 2014) and narratives constructed around more recent early Roman
finds (e.g. Manley et al. 2005) show that large compilations of data, examined without pre-
determined patterns in mind, reveal that people of all kinds drove change for a variety of
reasons, producing complex relationships between authority, power, and the archaeologi-
cal record.

References
Primary Sources
Caesar, The Gallic War, trans. H. J. Edwards (London: Heinemann, 1917).
Cassius Dio, Roman Histories, trans. E. Cary et al. (London: Heinemann, 1960–​70).
Tacitus, Agricola, trans. M. Hutton, rev. R. M. Ogilvie (London: Heinemann, 1970).
Tacitus, Annals, trans. J. Jackson (London: Heinemann, 1925–​37).

Secondary Sources
Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge
University Press.
132

132   Lacey Wallace

Creighton, J. (2001). ‘The Iron Age–​Roman Transition’, in S. James and M. Millet (eds), Britons
and Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council
for British Archaeology, 4–​11.
Creighton, J. (2006). Britannia:  The Creation of a Roman Province. London and
New York: Routledge.
Crummy, P., Benfield, S., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An Élite
Burial Site at Camulodunum. SPRS Volume 24. London: Society for the Promotion of
Roman Studies.
Cunliffe, B. (1974). Iron Age Communities in Britain:  An Account of England, Scotland, and
Wales from the Seventh Century bc until the Roman Conquest. London: Routledge.
Davies, B., Richardson, B., and Tomber, R. (1994). A Dated Corpus of Early Roman Pottery from
the City of London. CBA Research Report 5. York: Council for British Archaeology.
Eckardt, H. (2002). Illuminating Roman Britain. Montagnac: Editions Monique Mergoil.
Esmonde Cleary, A. S. (1987). Extra-​Mural Areas of Romano-​British Towns. BAR British Series
169. Oxford: Archaeopress.
Frere, S. (1967). Britannia: A History of Roman Britain. London: Routledge & Kegan Paul.
Frere, S. (1972). Verulamium Excavations: Volume I. London: Society of Antiquaries of London.
Frere, S. (1983). Verulamium Excavations:  Volume II. London:  Society of Antiquaries of
London.
Gascoyne, A. (2013). ‘The Roman Legionary Fortress, ad 43–​49’, in A. Gascoyne and D. Radford
(eds), Colchester: Fortress of the War God. An Archaeological Assessment. Oxford: Oxbow
Books, 59–​76.
Gascoyne, A., and Radford, D. (2013). Colchester: Fortress of the War God. An Archaeological
Assessment. Oxford: Oxbow Books.
Haselgrove, C. (1984). ‘Romanisation before the Conquest’, in T. F. C. Blagg and A. C. King (eds),
Military and Civilian in Roman Britain. BAR British Series 136. Oxford: Archaeopress, 5–​63.
Hawkes, C. F.  C., and Crummy, P. (1995). Camulodunum 2. Colchester:  Colchester
Archaeological Trust.
Hawkes, C. F. C., and Hull, M. R. (1947). Camulodunum: First Report on the Excavations at
Colchester 1930–​1939. London: Society of Antiquaries.
Hull, M. R. (1958). Roman Colchester. London: Society of Antiquaries of London.
Laurence, R., Esmonde Cleary, A. S., and Sears, G. (2011). The City in the Roman West: c.250
bc–​c.ad 250. Cambridge: Cambridge University Press.
MacDonald, W. (1986). The Architecture of the Roman Empire. New Haven:  Yale
University Press.
Manley, J., Rudkin, D., Sykes, N., Lyne, M., Dannell, G., Scaife, R., Somerville, E., Barber, L.,
Allen, D., Williams, D., Pelling, R., and Clegg, S. (2005). ‘A Pre-​ad 43 Ditch at Fishbourne
Roman Palace, Chichester’, Britannia, 36: 55–​99.
Marsden, P. (1987). The Roman Forum Site in London: Discoveries before 1985. London: Museum
of London.
Merrifield, R. (1965). The Roman City of London. London: Ernest Benn.
Neal, D. S. (1990). Excavation of the Iron Age, Roman, and Medieval Settlement at Gorhambury,
St Albans. London: Historic Buildings and Monuments Commission for England.
Niblett, R. (1985). Sheepen: An Early Roman Industrial Site at Camulodunum. CBA Research
Report 57. London: Council for British Archaeology.
Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium.
London: Society for the Promotion of Roman Studies.
 The Early Roman Horizon    133

Niblett, R. (2001). Verulamium: The Roman City of St Albans. Stroud: Tempus.

Niblett, R. (2005). ‘Roman Verulamium’, in R. Niblett and I. Thompson (eds), Alban’s Buried
Town: An Assessment of St Alban’s Archaeology up to ad 1600. Oxford: Oxbow Books, 41–​165.
Niblett, R., Manning, W., and Saunders, C. (2006). ‘Verulamium:  Excavations within the
Roman Town 1986–​88’, Britannia, 37: 53–​188.
Niblett, R., and Thompson, I. (2005). Alban’s Buried Town:  An Assessment of St Alban’s
Archaeology up to ad 1600. Oxford: Oxbow Books.
Partridge, C. (1981). Skeleton Green: A Late Iron Age and Romano-​British Site. London: Society
for the Promotion of Roman Studies.
Perring, D. (2011). ‘Two Studies on Roman London: A. London’s Military Origins—​B.
Population Decline and Ritual Landscapes in Antonine London’, Journal of Roman
Archaeology, 24/​1: 249–​282.
Philp, B. J., Merrifield, R., Dannell, G., Couldrey, P., Stant, M. Y., and Williams, W. (1977). ‘The
Forum of Roman London: Excavations of 1968–​9’, Britannia, 8: 1–​64.
Radford, D. (2013). ‘Camulodunon in the Late Iron Age, c.50 bc–​ad 43’, in A. Gascoyne and
D. Radford (eds), Colchester:  Fortress of the War God, An Archaeological Assessment.
Oxford: Oxbow Books, 33–​57.
Reece, R. (1988). My Roman Britain. Cotswolds Studies Volume 3.  Cirencester:  Cotswold
Studies.
Schnurbein, S. von (2003). ‘Augustus in Germania and his New “Town” at Waldgirmes East of
the Rhine’, Journal of Roman Archaeology, 16: 93–​108.
Tyers, I. (2011). ‘Tree-​Ring Analysis’, in J. Hill and P. Rowsome (eds), Roman London and
the Walbrook Stream Crossing:  Excavations at 1 Poultry and Vicinity, City of London.
London: Museum of London, 562–​567.
Wallace, L. (2014). The Origin of Roman London. Cambridge: Cambridge University Press.
Webster, G. (1966). ‘Fort and Town in Early Roman Britain’, in J. Wacher (ed.), The Civitas
Capitals of Roman Britain: Papers Given at a Conference Held at the University of Leicester,
13–​15 December 1963. Leicester: Leicester University Press, 31–​45.
134

Chapter 7

Britain at t h e E nd
of Empi re

Simon Esmonde Cleary

Introduction

This section of the Handbook is concerned with Britain in the Roman Empire, help-
ing us to combat the tendency to scholarly insularity that is particularly pernicious
when dealing with the Roman period, since Britain formed part of a major empire
and thus was directly influenced by decisions or trends originating outside the
island. This chapter will seek to counter that risk effectively, by looking at Britain
from across the Channel from Gaul, to assess to what extent late Roman Britain was
part of and influenced by wider developments. This is not to say these wider trends
and developments explain all that was going on in Britain; they do not, since Britain
was a region (or regions) within the wider western provinces with its (or their) own
particularities. But if we can identify and factor in these more general develop-
ments, then it may be easier to look in new ways at some of what was going on in
Britain while at the same time defining features of the archaeology that do need
to be looked at on the insular level. I want to look at four major sets of interrelated
developments occurring on the continent in the fourth and fifth centuries of which
Britain formed part, and which may help to explain why aspects of the archaeol-
ogy of the island were the way they were. These are: the withdrawal of imperial
power; the effects of this on the economic structures of the west; the effects of this
on aspects of elite culture and auto-​representation for the ‘civilian’ aristocracies; the
development of new forms of elite display linked to military-​style identities along
with changes in settlement types.
 Britain at the End of Empire    135

The Withdrawal of
the Imperial Presence

One of the defining features of the later fourth and fifth centuries across the west was the
withdrawal of the imperial presence and of the exercise of imperial power from much
of Gaul, the Rhineland, and Britain. Since the mid-​third century, the provinces north
of the Alps had more often than not had an emperor (legitimate or ‘usurper’) resident,
from the end of the third century mostly at Trier. This meant that northern Gaul, the
Rhineland, and Britain were firmly on the agenda of a member of the imperial college.
After the suppression of Magnus Maximus in 388 ad, Milan then Ravenna became the
principal imperial seat in the west. Trier’s manifold administrative functions likewise
moved south at the turn of the century, with the Praetorian Prefecture transferred to
Arles. To emperors from the end of fourth century, northern Gaul, the Rhineland, and
Britain were no longer close by and high priorities; they were far away and lower priori-
ties. The Mediterranean mattered more.
From the third century the region between the Rhine and the Loire had become
increasingly militarized (Esmonde Cleary 2013a: ch. 2). In part this was militarization
in the strict sense, with the presence of military garrisons and installations not just on
the old frontier of the Rhine but also along the new maritime frontier of the southern
North Sea and Channel. But the military were far more present than they had been in
the interior at a range of new fortifications, including the fortified road linking Cologne
and Boulogne and thus Britain, and with the settling of non-​Roman formations (laeti,
foederati) in the interior. As well as garrisons, there were elements of the state ‘infra-
structure’, such as the fabricae across northern Gaul. The Rhine–​Loire region was also
the heartland of the heavily fortified ‘fortress cities’ behind their thick, high walls bris-
tling with towers. Excavation and re-​evaluation have led to the old idea of these being
hasty responses to the ‘barbarian invasions’ of the later third century being abandoned;
the excavated dates range from the later third, through the early fourth, and down to the
second half of the fourth century, and the volume of material that went into them and
the care with which they were constructed makes it clear these were major monumental
undertakings. Since these were sited at what under the High Empire had been cities,
civitas-​capitals, the assumption is often that they remained urban in the late period. For
some this is demonstrably untrue: sites such as Bavai or Jublains had become small forti-
fied strong-​points. Even sites where the defences enclosed a larger area, such as Amiens,
exhibit little evidence that could not be interpreted in a military sense. Another class of
fortifications was the regional tradition of hilltop fortifications in north-​eastern Gaul
between the Meuse and the middle Rhine. Because excavated examples have some-
times been associated with burials containing weapons and ‘Germanic’ objects, these
have been seen as strong-​points with ‘barbarian’ garrisons. The ‘Germanic’ objects form
part of a larger complex of late Roman dress fittings: for men primarily belt fittings and
brooches (Böhme 1974; Swift 2000: chs 2, 5, and refs.), for women principally brooches
136

136   Simon Esmonde Cleary

along with beads and bracelets (Swift 2000: chs 3, 4). Long considered as markers of eth-
nicity (e.g. Böhme 1974), much of this material is now recognized to be the markers of
late Roman officialdom, military or civilian, and thus to say more about the relationship
of the wearer to the Roman state (see also Allason-​Jones, this volume). The latter part of
the title of Böhme (1974), ‘zwischen Elbe und Loire’ (between Elbe and Loire) makes the
point that this material is also a phenomenon of northern Gaul—​indeed concentrated
in the Seine–​Rhine region. Moreover, traditional markers of civilian elites over much of
northern Gaul, in particular the villa but also urban residences, had shrunk greatly in
both numbers and elaboration (Van Ossel 1992). It can be argued that the ‘militarization’
of northern Gaul went wider, with the vocabulary of military self-​representation hav-
ing been taken up by the local elites instead of the civilian markers of the High Empire.
Another way in which the region may have been ‘militarized’ was in the increasing
dominance of state and military demands on the economy in order to support the gar-
rison units, the fabricae, and their personnel.
What happened to the Roman military complex in the fifth century is hard to read
from the archaeology, partly because of the lack of good archaeological sequences. A
key site on the middle Rhine is that of the fort of Alzey, constructed under Valentinian
I (364–​75 ad), partially burnt down early in the fifth century, but then in part rebuilt
before being destroyed by fire in the middle of the century (Oldenstein 1986). The mate-
rial culture from the first half of the fifth century is more ‘Germanic’ in that its origins lie
east of the Rhine. Another key site lies lower down the Rhine at Krefeld-​Gellep, where
the extensive excavations of the cemeteries dating from the second century onwards
(Pirling 1986; Pirling et al. 2000) have shown that from the turn of the fourth and
fifth centuries male and female burials grouped in discrete areas contained increasing
amounts of ‘Germanic’ material, not just dress accoutrements but also pottery of non-​
Roman fabrics and forms. A similar situation seems to obtain on the upper Rhine at
Kaiseraugst (Drack and Fellmann 1988: 3–​42, 411–​414); again extensive excavation in
the cemeteries has shown that from the end of the fourth century ‘Germanic’ grave-​
goods increasingly make their appearance. Because of the textual master narrative for
the period of the replacement of Roman imperial power by the nascent ‘Germanic’ suc-
cessor states, this evidence has been assimilated to that framework. Even if one were to
accept (which I do not) an easy equivalence between object and ethnicity, the presence
of ‘German’ men with their womenfolk does not of itself mean the garrisons were no
longer part of a standing Roman army—​the Romans had employed non-​Roman troops
since the early Republic. Moreover, one must beware of decontextualizing ‘Germanic’
burials from the larger majority of burials in the continuing Roman-​provincial rite.
Apart from these key sites, there are few places where the Roman army of the fifth cen-
tury in northern Gaul and the Rhineland can be pursued by archaeology. The lesson of
Krefeld and Kaiseraugst seems to be that cemeteries may be the most revealing sites to
excavate.
If the fate of the standing Roman army in northern Gaul and the Rhineland is dif-
ficult to apprehend through archaeology, there is another class of material that has
important lessons: the coinage. The current model for the striking and distribution of
 Britain at the End of Empire    137

the tri-​metallic coinage of the fourth century is closely linked to the army in that its
primary purpose is seen as the discharge of state obligations through gold and silver and
the recapturing of the precious metals in part through the bronze coinage. In the west,
army pay was probably the largest single head of expenditure by the fiscal authorities.
Of course, pay in coinage was not the only benefit the soldiery received, and we shall
look at some effects of the supply of materiel to the troops. At the turn of the fourth
and fifth centuries the part of the fiscal cycle involving the distribution of coin seems
to break down (Kent 1994; cf. Guest 2005). Put briefly: at this time gold was struck only
sporadically in the west; silver was struck in quantity (if not consistently), with a major
issue struck at Milan from 397–​402 ad, but the later issues of 408–​25 ad from Milan and
Ravenna remained largely distributed within Italy; bronze was issued in declining quan-
tity after the reorganization of the mints in 395 ad, until 402 ad, after which the issues
from Rome were also rare north of the Alps. From 402 ad there was little coin penetrat-
ing north of the Alps, and the occasional striking at Trier did not correct this. The logic
of this would seem to have to be that the Roman state no longer remunerated its troops
with elements of pay in precious metals, even if this was an unforeseen consequence of
the invasions of Gaul and Italy at the time. What happened to the proportion of remu-
neration in kind—​foodstuffs, animals and clothing, equipment and weaponry supplied
from the fabricae—​is very difficult to tell, though the dating of some of the ‘official-​issue’
brooches and belt fittings suggests they were being produced into the first half of the
fifth century (Böhme 1986). But by the second half of the century the regional variations
indicate local rather than centralized production, suggesting that, though such metal-
work retained ideological significance, official production of it had by then ceased. If
the Roman state had de facto ceased to pay and increasingly not to supply its army from
the start of the fifth century, then the likelihood must be that a standing Roman army of
the type with which we are familiar ceased to exist, or transmuted into something else.
The literary sources can be read in this way, with the military history of the west in the
first half of the fifth century increasingly becoming a matter of strong military com-
manders, with personal followings (bucellarii) putting together coalitions in order to
fight campaigns (Whittaker 1994: ch. 7). Some of these commanders were of Roman ori-
gin, but the army assembled by Aetius to confront Attila at the Catalaunian Fields in 451
ad was a coalition of a range of ethnicities, with very few troops that could be described
as ‘Roman’ save that they were in the military retinue of Aetius himself. Military force
seems effectively to have passed from the public power of the state to the private power
of war-​leaders; some of these fought for the emperor, some were ‘Germanic’ kings. Such
a series of developments would accord with the premiss put forward at the start of this
section that the withdrawal of the imperial presence meant that northern Gaul, the
Rhineland, and Britain were now low on the list of imperial priorities.
The picture proposed here is one that is not unlike that proposed for the military
zones of Britain. Fourth-​century Britain had two major defensive systems, Hadrian’s
Wall and its hinterland (Collins 2012: ch. 3) and the Saxon Shore (Pearson 2002). The
latter related to the fortifications on the other side of the Channel. It has also yielded
a very considerable number of brooches and belt-​fittings of ‘official-​issue’ type (Swift
138

138   Simon Esmonde Cleary

2000). Interestingly, the distribution of these latter is almost entirely southern, with a
marked absence from Trent to Tees and west of the Wye, the supposed ‘military zone’,
but a presence along the Wall. On the analogy with Gaul it could, therefore, be argued
that southern Britain also demonstrates a degree of ‘militarization’, though, as we shall
see, to nothing like the same extent. In fact the distribution of this material shows a pref-
erence for the Saxon Shore forts and urban sites, so may be related to the garrisoning
of such sites, or at least some form of military or official presence at the cities. From the
turn of the fourth and fifth centuries the archaeology at military sites starts to change
significantly. At the forts of Hadrian’s Wall and its hinterland there is a series of changes,
suggesting a transition from a relatively coherent and organized situation to one where
individual forts were adapted piecemeal, suggesting a transition to something much less
regular than a ‘Roman’ army (Collins 2012: ch. 4). The suggestion is that instead we are
seeing a series of smaller-​scale military groups becoming increasingly self-​reliant, both
in terms of their identities and in terms of their economic base. This pattern would fit the
wider evidence from the west already outlined, aligning the situation in Britain with the
continent. It is perhaps worth noting that one of the building types regularly modified in
the Wall zone was the horreum (the store building); that these buildings were no longer
needed would also fit with the idea that the state had withdrawn from provisioning its
troops. The evidence from the Saxon Shore is less good because of lack of excavation, but
it is worth noting that the occupation sequence at Portchester runs through from the
fourth century, when the site is supposedly a fort of the regular Roman army, into the
fifth, when the material culture starts to change from Roman-​provincial to ‘Germanic’
(Cunliffe 1976: 121–​127, 301), echoing sites such as Alzey. It would be fascinating to see
excavated a large late cemetery attached to a late fort in Britain. The wider pattern of the
withering of coin supply north of the Alps is reflected in Britain with the likely impacts
of this on the nature of the army; the smattering of coins post-​388–​402 ad from Britain
would also reflect wider patterns (e.g. Collins 2008).

The Disintegration of
the Economic Systems

The disengagement of the state from the military sphere at the turn of the fourth and
fifth centuries had implications not just for the army itself but also for the systems of sup-
ply that the state had engendered in order to sustain the army and thus of the other eco-
nomic activities that had depended from these systems (Wickham 2005: ch. 11). It is clear
that the ‘imperial’ economy, a system of redistribution by the state from the taxpayer to
politically crucial groupings (the army, the administration, the city of Rome), although
powered by taxation, involved the movement of material at state expense, allowing a
range of other goods to travel ‘piggy back’, with their movement effectively subsidized
by the state. This related to the operations of the ‘market’ economy, the movement
 Britain at the End of Empire    139

and exchange of goods for economic profit, where the general stability of the empire
encouraged longer-​distance links, and the ‘imperial’ economy promoted the distribu-
tion of some goods. Reciprocity, the mutual provision of goods and services for social
reasons and profit, remained the bedrock of much economic activity in the late Roman
world, most usually at the level of the ‘peasant’ economy, with goods moving over rela-
tively short distances, but there was also the ‘prestige’ economy of reciprocity, causing
high-​value items to move long distances. None of these forms of economic activity or
the participants in them was sealed off from the others; the operations of the ‘imperial’
economy impacted on the ‘market’ economy, and individuals might swap between the
various forms either as suppliers or as consumers. The discussion here will focus on the
‘imperial’ and the ‘market’ economies, partly because they are the most identifiable in
the archaeological record, partly because they were clearly profoundly affected by the
events of the fifth century. It will be based on the ceramic evidence because of its ubiq-
uity and the amount of study it has attracted—​as so often, ceramics that survive must
stand proxy for perishables. The industries with a regional or supra-​regional reach are
most clearly exemplified by the ‘Argonne’ wares produced in north-​eastern Gaul and
the Paris basin and distributed over the northern half of Gaul, with lesser amounts in
southern Gaul, the upper Danube provinces, and Britain (Van Ossel 2011). It is found in
quantity at military sites and the ‘fortress cities’, suggesting it may have been travelling
in part ‘piggy-​back’ on the movement of supplies for the army and state servants. It was
probably also being marketed, with the transport network and privileges of state sup-
ply allowing it to compete on favourable terms over long distances. Another industry
that seems to have been attuned to military supply was Eifelkeramik/​Mayen ware, with
its distribution in the Rhineland and south-​eastern Britain. There were other industries
with a regional distribution that cannot be seen to be so dependent on the state nexus,
such as the developing céramiques rugueuse and grumuleuse centres of the Île de France
and or the Jaulges/​Villiers-​Vineux complex in northern Burgundy, which seem to have
served more civilian markets. Other industries such as the céramique craquelée bleutée
of the Champagne achieved subregional distribution (for these industries, cf. Brulet et
al. 2010: §§29, 9, 28) Overall, the existence of the late Roman state seems to have pro-
moted the levels of integration represented by the ‘imperial’ and ‘market’ economies.
Turning to Britain, we have already seen that some continental products such as
Argonne or Mayen entered Britain; equally, some British products went the other way,
with surprising quantities of Black-​Burnished I now recognized along the Channel
coasts and penetrating inland into Picardy and the Paris basin, and some Hadham,
Oxfordshire, and New Forest products having made the crossing (Adrian 2006).
Within Britain, it is possible to recognize the effects of the levels of integration already
discussed for Gaul. Some industries seem to have achieved wide distributions partly
on the back of ‘official’ requirements such as foodstuffs. This would seem to be the
case in the north for the products of the Crambeck/​Malton kilns, with their wide dis-
tribution in the ‘military zone’ up to and including the Wall (cf. Evans 2000; see also
Evans, this volume). Further south, the Oxfordshire industry could almost be seen as
the British Argonne in terms of products and spread, with its distribution down the
140

140   Simon Esmonde Cleary

Thames seemingly preferentially linked to military sites and London (Fulford and
Hodder 1974). London was also a major destination for the Alice Holt industry. Some
industries may have existed primarily to supply particular installations. Equally, many
other aspects of the Oxfordshire distributions can be explained more in ‘market’ terms,
as can industries such as the Nene Valley or the New Forest. Other, smaller-​scale indus-
tries probably served small catchments, perhaps through cities or ‘small towns’, or else
may have been linked to major rural landholdings. One may argue for Britain, as for
the west more generally, that its membership of the Roman Empire at this date had pro-
found consequences for the way the island’s economy was structured and the degree
and levels of integration.
What then of developments on either side of the Channel in the fifth century?
In northern Gaul in particular the fifth century sees continuing evidence from the
ceramics (also other classes of material such as glass and metalwork) for continued
production and distribution at regional, subregional, and local levels. At the regional
level, Argonne wares continued to be distributed across much of the northern half of
Gaul, with a diminution in quantity and in reach over time, so that by the end of the
century it was at best a subregional industry. Other industries such as the céramiques
grumuleuse/​rugueuse also continued to have significant distributions in the Paris
basin and neighbouring areas in the first half of the century, less so thereafter; inter-
estingly at the end of the fifth century vessels in ‘Germanic’ forms and decoration start
to be produced in these fabrics. The Jaulges/​Villiers-​Vineux industry seems to have
continued in production through the first half of the century, though probably on a
diminishing scale. South of the Loire, ceramic productions continued uninterrupted,
with some of them such as DSP (dérivées des sigillées paléochrétiennes—​a stamp-​
decorated repertoire) perhaps attaining their maximal spread in the first half of the
century (Rigoir 1968). It is clear that the scale and reach of production, particularly
in the northern half of Gaul, diminished as the fifth century advanced, with even the
few remaining major industries, such as Argonne, a shadow of their fourth-​century
selves by the end of the fifth century. By then other industries had either ceased pro-
duction or become small-​scale local enterprises. The economic integration of the
fourth century had undergone significant disintegration; nevertheless, there were still
functioning ceramic industries, and more desirable commodities such as glass con-
tinued to be traded over considerable distances (Foy 1995). The contrast with Britain
is clear: there the Roman-​style pottery industries seem to have ceased production and
distribution in the first half of the fifth century. Indeed, some industries seem already
to have been in decline in the latter part of the fourth century. The dating of this is
debated, because of the dependence of pottery dating on the availability of coin as a
dating medium. Nevertheless, the debate centres on how far into the first half of the
fifth century pottery production continued rather than whether it continued to the
end of that century. If we accept ceramics as a proxy of the levels of economic activity
and integration, then the island suffered a breakdown and disintegration in economic
activity of a depth and geographical spread that Gaul simply did not experience (cf.
Esmonde Cleary 2013b).
 Britain at the End of Empire    141

Elite Auto-​R epresentation: The

‘Civilian’ Tradition

The consideration of Gaul north of the Loire in the fourth century proposed that it
became a ‘militarized’ region in the later Roman period. But Gaul south of the Loire was
a very different world, one that in many ways is more easily related to southern Britain.
Apart from some cities defended under the High Empire and a small number of the ‘for-
tress cities’ that characterized the north of Gaul, the cities of the southern half of Gaul
remained undefended. So far as can be seen, the street grids continued to articulate the
urban landscapes, although at many cities where there has been sufficient excavation
it is clear that the area occupied had contracted from its late-​second-​century peak to
more restricted areas in the fourth century. The public buildings and monuments pro-
vided under the High Empire do not seem to have been to be subjected to any general
programme of disuse or suppression, though there was variation by monument type
(Heijmans 2006a). At these south Gaulish cities, therefore, the layout and monumen-
tal buildings inherited from the earlier empire seem to have been preserved into the
later period. It can be argued that this was a question not just of the passive reception of
these earlier buildings but of their active curation under the late empire and thus self-​
representation that perpetuated civic ideologies inherited from the earlier empire.
Another major difference between Gaul north and south of the Loire was the pres-
ence of large numbers of villas, many of them extensive and luxuriously appointed, in
southern Gaul. The heartland of this villa culture was in the south-​west, with fewer in
the south-​east and with only a scatter in west-​central Gaul (Balmelle 2001). A common
feature of the larger villas was the elaboration of their plan and architecture, making
widespread use of the apsidal form for the principal room, and also on occasion the tri-
conch plan that allows the identification of the main dining suite. In addition, elaborate
bath-​suites were a regular feature of such complexes, with ambitious three-​dimensional
structures. The major ‘public’ areas of these villas were the setting for the decor of dis-
play with the use of marble (available from the Pyrenees) for architectural detailing
and the use of mosaic for flooring. The example of the villa at Chiragan south-​west of
Toulouse makes clear the importance of sculpture as a status display medium; evidence
from other villas in Gaul and Spain show that sculpture was more widespread and sig-
nificant than has perhaps been appreciated. The subject matter of the mosaics and the
sculpture was firmly in the Mediterranean traditions of divinities, heroes, and famous
men and women, along with the mythology of which they formed part.
This was not the only area of the west that exhibited such a concentration; others lay
in Spain, on the Catalonian littoral, on the northern meseta, and in southern Lusitania
(Chavarría 2007). Another small but important group lay in north-​eastern Gaul and the
Rhineland, in the hinterlands of Cologne and especially Trier (Van Ossel 1992). What
we are seeing is a major, alternative form of aristocratic self-​representation, contrasting
with the militarized one dominant in northern Gaul. This aristocratic vocabulary was
142

142   Simon Esmonde Cleary

one related to two hugely important social groupings in the late Roman world, the impe-
rial court and the senatorial order; to an extent the elaborate villas of Gaul and Spain can
be read as scaled-​down reflections of imperial grandeur and of senatorial opulence. It is
also the aristocracy of education and learning, of paideia, whereby regional aristocracies
were subsumed into a common framework of education and thus cultural formation
and expectation, within which the display of the traditional mythologies and personali-
ties operated.
It is clear that in certain important respects southern Britain better matches southern
than northern Gaul as regards the evidence of the cities and villas along with aspects
of their material culture. The cities of late Roman Britain parallel their south-​Gaulish
cousins in retaining their gridded layouts, many of their inherited public buildings and
monuments, and in remaining centres of elite residence and economic activity. The
retention, indeed curation, of earlier buildings and monuments varies by city and by
class of monument, but as with their southern equivalents there is no evidence for a con-
certed plan of suppression of these structures, though again with variation by city and
by monument class. A major difference at the cities of Britain is the presence of walls.
This was a historical peculiarity of the island and had resulted in the majority of the cit-
ies of Britain being enclosed from the later second century by defences enclosing the
then inhabited area—​with even those cities defended later on (for example, Canterbury,
Chichester) enclosing larger areas than the ‘fortress cities’ of Gaul. Moreover, the British
walls were less often provided with regular projecting towers (‘bastions’) than the north-​
Gaulish circuits.
The similarities between elite residences and their appointments in the villa-​rich
regions of southern Gaul and Spain, on the one hand, and southern Britain, on the
other, is striking, in particular for villas, but also for urban residences, where our infor-
mation for Britain is much better than for the continent. Comparison of the plans and
layouts of the larger villas and town houses, types of principal rooms, types of deco-
ration (such as mosaic and sculpture), along with their subject matter, with those of
south-​western Gaul and Spain show the elites of Britain to have been well aware of fash-
ions in these matters and to have participated in a wider common vocabulary of elite
auto-​representation. The elaboration of the decor of these residences is well known,
and the corpus of mosaics from Britain reveals the range of subjects of figured mosa-
ics (Neal and Cosh 2002–​10). Like their counterparts in Gaul and Spain, they display
their commissioners’ knowledge of Mediterranean culture and their internalization of
the linguistic, literary, and philosophical markers of paideia. So at one level southern
Britain forms one of a series of regions in the late Roman west exhibiting this sort of
civilian elite display, but, as has been noted, even within Britain such elite display is
regionalized. One region was East Anglia, where villas were architecturally modest but
which nevertheless contained the largest and richest concentration of silver plate, coin,
gold jewellery, and semi-​precious stones in the empire—​deposited it seems in the first
half of the fifth century but probably manufactured and in circulation in the course of
the fourth (Hobbs 2006).
 Britain at the End of Empire    143

One area where southern Britain seems to diverge markedly from the continent is in
its expressions of religious activities. In the northern half of Gaul the established narra-
tive has been the destruction of the temples of the traditional religions either by barbar-
ians in the later third century or by militant Christians such as Martin of Tours in the
second half of the fourth. Britain has long been known for the way in which its temples
continued to be used through the fourth century, along with other manifestations of cult
or ritual practice such as the wet-​place deposition (Henig 1984: ch. 10; Woodward 1992).
There is increasing evidence that the picture for Gaul may be overdrawn and that sanctu-
aries and cults of the traditional religions may have lasted longer than has been thought
(Esmonde Cleary 2013a: 19–​197). On the other hand, the evidence for Christianity in
Britain is not as well developed as that for Gaul (Petts 2003; see Petts, this volume, for
further discussion of Romano-​British Christianity). Clearly there are key Christian sites
in Britain, such as the Lullingstone wall-​paintings or Water Newton Christian plate,
but as yet there is no good evidence for the sort of urban church-​building that was tak-
ing place from the second half of the fourth century in some of the cities of Gaul, and
there are only one convincing and one probable late Roman Christian tombstones from
Britain (RIB 955 Carlisle; RIB 787 Brougham, both interestingly in the ‘military zone’)
compared with the burgeoning of such commemoration in Gaul (Handley 2003). Nor
are there any sarcophagi of continental Christian type. These two absences suggest that
in Christianity, as in the traditional religions, Britain differed from continental practice.
The demise of the ‘civilian’ forms of elite auto-​representation in southern Gaul and
Spain has been written largely from the settlement rather than the cemetery evidence,
in particular the villas. It was the fifth century that saw the ‘militarization’ of the cities of
southern Gaul by the construction of wall circuits, many of them very limited in extent
(Heijmans 2006b). The form and dating of the latest occupation at many villas in Gaul
and Spain are difficult to reconstruct—​partly because many of the major sites were exca-
vated some time ago, partly because of difficulties in the dating. But recent surveys of
the evidence for south-​western Gaul (Balmelle 2001: 115–​123) and for Iberia (Chavarría
2007: 114–​156) concur in seeing the middle of the fifth century as a horizon after which
the high-​quality residences were no longer maintained and progressive degradation
of the structures and their decor ensues—​with subdivision of space by timber struc-
tures (‘squatters’), new timber structures, or sometimes the installation of cemeteries.
The reasons for this abandonment of the villa culture are as yet obscure. It was probably
in part economic, with a combination of the reduction of resources consequent upon
politico-​legal fragmentation and a reduction of the specialists required to sustain such
a culture. But as well as the change in landholdings and the organization and social rela-
tions of production, it clearly signals the progressive collapse of the ‘civilian’ elite culture
of paideia and its material correlates. The other component of the classic late Roman city
was, of course, the church, which became a more general feature of the Gaulish urban
landscape in the fifth century and preserved many of the features of ‘civilian’ prestige
architecture and a Christianized paideia. This summary of developments in south-
ern Gaul should be familiar in outline to those studying the progressive dereliction of
144

144   Simon Esmonde Cleary

Romano-​British cities and villas, save that the dating for this in Britain is the first rather
than the second half of the fifth century and for Britain there is not the displacement to
prestige church buildings and decoration on the continental model. However, the chro-
nology for the remaining villas of northern Gaul fits rather better with that for Britain
than with their cousins south of the Loire (Van Ossel 1992: 8–​82). The demise of Britain’s
cities and the villas may, in part, have been due to the politico-​military withdrawal of the
imperial presence early in the fifth century. It may also have been, in part, a consequence
of the economic dislocations that ensued (Esmonde Cleary 1989: ch. 4), but what it also
represents is the collapse of the ‘civilian’ modes of elite self-​representation, probably, in
large part, because they no longer had validity without the socio-​political system that
gave them a frame of meaning. What may have replaced them we shall now consider.

The Archaeology of the Fifth Century:

Settlements, Cemeteries,
and ‘Militarization’

If the fifth century witnessed the demise of the ‘civilian’ forms of status display, then what
replaced them? In the first section of this chapter it was suggested that in Gaul north of
the Loire the later Roman period was characterized by the adoption of military-​derived
vocabularies of power and representation, which contrast with the vocabularies we
have just been looking at south of the Loire and southern Britain. It will be the conten-
tion of this section that in the course of the fifth century the ‘civilian’ elite vocabularies
were progressively replaced by the ‘military’ ones. This is a trend that has traditionally
been subsumed within another narrative—​that of the spread of the historically attested
‘Germanic’ peoples across Gaul, Spain, and Britain. This has been based in particular
upon burial evidence and the supposed ‘Germanic’ style both of the burial rite, espe-
cially ‘weapon graves’, and of the grave-​goods, above all the metalwork and, to a lesser
extent, the pottery. Here a certain separation of the two narratives will be attempted.
The demise of the tradition of construction in stone (churches apart) and the (re-​)
emergence of traditions of building in timber is one of the widespread archaeologi-
cal signals of the transition from the Roman to the early medieval. It has often been
seen as an index of ‘Germanization’, especially where particular building forms such as
the ‘hall’ or the sunken-​featured building (SFB) are present. More recent evidence and
analysis suggests proceeding with more caution. The resurgence in timber building was
already under way in northern Gaul in the fourth century, including at villa sites (Van
Ossel 2003, 2010), and continued into the fifth, with both ‘halls’ and SFBs in associa-
tion with former villas as well as on new sites, but always on a more restricted scale than
their villa predecessors (Peytremann 2003: 318–​319). The presence of structures of the
‘long house’ or Wohnstallhaus as early as the fourth century in the far north of Gaul
 Britain at the End of Empire    145

(Toxandria) as well as occasionally further south has been linked to the narratives of the
settlement of the Salian Franks in the former and of ‘Germanic’ troops and others in the
latter. Although these building types do have strong geographical links with areas either
side of the lower Rhine and beyond, their origins, especially of the SFBs, are not always
clear (Hamerow 2002: 31–​35). The presence in them from the fifth century of metalwork
and ceramics of ‘Germanic’ types has been read likewise as evidence of the spread of
Germanic, more specifically Frankish or Anglo-​Saxon, peoples. A problem here is that
this material is often decontextualized from the larger quantities of Roman-​provincial
material culture at these sites, risking giving a misleading impression by focusing on
the ‘Germanic’ material. A major recent consideration of early medieval settlements in
northern Gaul (Peytremann 2003) carefully eschewed any ‘ethnic’ interpretation and
instead discussed such sites in terms of agrarian and artisan production and social
structures; this allows us an alternative explanation of changes of society and produc-
tivity at and after the end of the Roman period without any particular need for ‘ethnic’
change. One result of such a reading is the move to an archaeology of settlements that
shows far less social gradient than that of the later Roman period, something that could
also be argued for the archaeology of the dead.
A similar problem of decontextualization affects the study of cemeteries, particularly
rural ones, in the northern Gaul of the fifth century. Metalwork, ceramics, and other
features such as burial rite or cranial deformation have been read in an ‘ethnic’ way. In
addition, the burial of some males with weapons has also been seen as ‘Germanic’. In
fact, the tradition of burial with weapons can be detected from what appear to be other-
wise Gallo-​Roman burial grounds in northern Gaul from the later fourth century on (cf.
Theuws 2009), so we may be seeing further change in the indigenous status-​markers.
From the fifth century, more in northern than southern Gaul, there was a quantity of
material whose stylistic origins lay outside what had been imperial territory, but it is
now recognized that seeing objects as direct ethnic indicators is an approach fraught
with imponderables. It is important to recall that such burials and material usually form
a minority, sometimes a small minority, of burials in a cemetery—​for instance, in the
well-​known site of Frénouville in Normandy (Pilet 1980). In such cemeteries the pre-
dominant rite is the unfurnished inhumation, characteristic of the late Roman provin-
cial population, and such material culture as is present is likewise of provincial-​Roman
type. The risk is clearly that, by decontextualizing certain types of burial and classes of
material, a misleading impression is given of the composition of the burying popula-
tion. Recent appreciation of the very debatable relationship between object and ‘ethnic’
or other identity compounds the problems. Instead it is possible to read the ‘weapon
burials’ not in ethnic terms but in terms of the increased importance of the vocabulary
of military power as markers of social status for ‘Romans’ and ‘Germans’ equally. One
other point worth making about some of these cemeteries, such as Frénouville, is that
they demonstrate considerable stability and longevity, mirroring some of the settlement
evidence for continuity from the fourth century into the fifth and sixth, again raising
questions as to what extent we are looking at ‘incomers’.
146

146   Simon Esmonde Cleary

Fifth-​century settlements and cemeteries in southern Britain have been analysed

almost exclusively within an ‘ethnic’ framework, above all as evidence for the arrival of
the Anglo-​Saxons. Do the ideas outlined above relating to changes in building technol-
ogy, settlement types, and the militarization of the elites have anything to contribute
this side of the Channel? The shift to timber building has long been acknowledged as
a marker of the end of the Roman order as much as, or more than, of the arrival of new
peoples. Even the types of structure found in eastern and southern Britain at ‘Anglo-​
Saxon’ sites sometimes have an equivocal relationship with ‘Germanic’ building types,
and the Wohnstallhaus is barely represented this side of the North Sea (Hamerow 2002).
As with the discussion of comparable settlements in northern Gaul, consideration is
now focusing more on questions of the social formation they represent and the agrar-
ian and artisan practices they embody rather than simply ethnic identity. Likewise, the
presence of weapons and militaria in graves of this period has been discussed almost
exclusively in terms of Anglo-​Saxon social structures (e.g. Böhme 1986). But with the
increasing appreciation of the part indigenous British people may have played in popu-
lating ‘Anglo-​Saxon’ cemeteries, the presence of such material might also be looked at
in terms of the ‘militarization’ of male, particularly elite male, identity more generally.
Linked with this is the question of the significance of burials containing ‘Quoit Brooch’
material (Suzuki 2000), with its evidently Roman-​style decoration and significance and
therefore its presence in supposedly Anglo-​Saxon cemeteries. This is emphasized by the
amount of such material from the continental coasts of the Channel (Soulat 2009), sug-
gesting a trans-​Manche region of prestige metalwork that does not fit well into standard
narratives about ‘Anglo-​Saxon’ and ‘Frank’.

Epilogue

As stated in the preamble, given this chapter’s place in the volume, its purpose has been
briefly to try to set developments in the archaeology of Britain in the context of develop-
ments on the continent and to identify those aspects of fourth-​and fifth-​century Britain
where trends in the island derived from wider trends within the late empire and thus
should be interpreted in that context. Equally, it has tried to identify those ways in which
practices in the regions of Britain deviated from those across the Channel, so requiring
explanations to be sought closer to home. What I hope this has demonstrated is that late
Roman Britain was more integrated with other regions of the west than has often been
appreciated. This is not to say that Britain was identical with the continent; the archae-
ology both of the end of imperial power in the island and of the relationship between
the indigenous population and incomers in the fifth century strongly suggests insular
particularities, but at least we should, in due course, be better able to separate the insular
from the wider processes. One might close by citing the growing evidence, particularly
from stable isotope analysis (see the chapters in Eckardt 2010; see also Eckardt, this vol-
ume), for the cosmopolitan nature of the population of later Roman Britain, at least in
 Britain at the End of Empire    147

its urban centres, with substantial percentages of incomers from the continent in the
fourth century, let alone the question of incomers in the fifth. We may also be beginning
to see the corresponding development of this, with some evidence identifying fourth-​
century Britons on the continent (Swift 2010; see also chapters by Eckardt and Ivleva,
this volume, for further discussion of population movements). Along with the more
conventional archaeological evidence considered above, this shows how dangerous it
would be to continue to try to consider late Roman Britain in isolation from its neigh-
bours in the west.

Abbreviation
RIB  R
 . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I:
Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Adrian, Y.-​M. (2006). ‘Céramiques et verreries des IVe et Ve s. dans la basse vallée de la Seine;
Les Exemples de Rouen, Lillebonne, Caudebec-​lès-​Elbeuf, Tourville-​la-​Rivière (Seine-​
Maritime)’, in P. Ouzoulias and P. Van Ossel (eds) Les Céramiques de l’Antiquité tardive en
Île-​de-​France et dans le Bassin parisien Volume I. Ensembles régionaux. Diœcesis Galliarum
Document de travail 7. Paris: Diœcesis Galliarum, 331–​389.
Balmelle, C. (2001). Les Demeures aristocratiques d’Aquitaine: Société et culture de l’Antiquité
tardive dans le Sud-​Ouest de la Gaule. Aquitania Supplément 10. Bordeaux:  Fédération
Aquitania.
Böhme, H. W. (1974). Germanische Grabfunde des 4. bis 5. Jahrhunderts zwischen unterer Elbe
und Loire: Studien zur Chronologie und Bevölkerungsgeschichte. Munich: C. H. Beck’sche.
Böhme, H. W. (1986). ‘Das Ende der Römerherrschaft in Britannien und die angelsäch-
sische Besiedlung Englands im 5.  Jahrhundert’, Jahrbuch des Römisch-​ Germanischen
Zentralmuseums Mainz, 33/​2: 469–​574.
Brulet, R., Vilvorder, F., and Delage, R. (2010) (eds). La Céramique romaine en Gaule du
nord: La Vaisselle à large diffusion. Turnhout: Brepols.
Chavarría Arnau, A. (2007). El Final de las Villae en Hispania, (Siglos IV–​VII d.C.), Bibliothèque
le l’Antiquité Tardive. Turnhout: Brepols.
Collins, R. (2008). ‘The Latest Roman Coin from Hadrian’s Wall: A Small Fifth-​Century Purse
Group’, Britannia, 39: 256–​261.
Collins, R. (2012). Hadrian’s Wall and the End of Empire: The Roman Frontier in the 4th and 5th
Centuries. London: Routledge.
Cunliffe, B. (1976). Excavations at Portchester Castle: Volume II Saxon. Reports of the Research
Committee of the Society of Antiquaries of London 23. London: Society of Antiquaries of
London.
Drack, W. and Fellmann, R. (1988). Die Römer in der Schweiz. Stuttgart: Theiss.
Eckardt, H. (2010) (ed.). Roman Diasporas: Archaeological Approaches to Mobility and Diversity
in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth,
RI: Journal of Roman Archaeology.
148

148   Simon Esmonde Cleary

Esmonde Cleary, S. (1989). The Ending of Roman Britain. London: Batsford.

Esmonde Cleary, S. (2013a). The Roman West, ad 200–​ 500:  An Archaeological Study.
Cambridge: Cambridge University Press.
Esmonde Cleary, S. (2013b). ‘Southern Britain in the Fifth Century:  A  Collapsed State?’
in F. Hunter and K. Painter (eds), Late Roman Silver:  The Traprain Treasure in Context.
Edinburgh: Society of Antiquaries of Scotland, 49–​53.
Evans, J. (2000). ‘The End of Roman Pottery in the North’, in A. Wilmott and P. Wilson (eds),
The Late Roman Transition in the North: Papers from the Roman Archaeology Conference
Durham 1999. BAR British Series 299. Oxford: Archaeopress, 39–​46.
Foy, D. (1995) (ed.). Le Verre de l’Antiquité tardive et du haut Moyen Âge: Typologie, chronologie,
diffusion. Guiry-​en-​Vexin: Association Française pour l’Archéologie du verre.
Fulford, M., and Hodder, I. (1974). ‘A Regression Analysis of Some Late Romano-​British
Pottery: A Case Study’, Oxoniensia, 39: 26–​33.
Guest, P. (2005). The Late Roman Gold and Silver Coins from the Hoxne Treasure.
London: British Museum.
Hamerow, H. (2002). Early Medieval Settlements: The Archaeology of Rural Communities in
North-​West Europe 400–​900. Oxford: Oxford University Press.
Handley, M. (2003). Death, Society and Culture: Inscriptions and Epitaphs in Gaul and Spain,
AD 300–​750. BAR International Series 1135. Oxford: Archaeopress.
Heijmans, M. (2006a). ‘La Place des monuments publics du Haut-​Empire dans les villes de la
Gaule méridionale durant l’Antiquité tardive (IVe–​VIe s)’, Gallia, 63: 25–​41.
Heijmans, M. (2006b). ‘La Mise en défense de la Gaule méridionale au IVe–​VIe s’, Gallia,
63: 59–​74.
Henig, M. (1984). Religion in Roman Britain. London: Batsford.
Hobbs, R. (2006). Late Roman Precious Metal Deposits c. ad 200–​700: Changes over Time and
Space. BAR International Series 1504. Oxford: Archaeopress.
Kent, J. (1994). Roman Imperial Coinage Volume X:  The Divided Empire and the Fall of the
Western Parts. London: Spink.
Neal, D., and Cosh, S. (200–​10). Roman Mosaics of Britain. 4 vols. London: Society of
Antiquaries of London.
Oldenstein, J. (1986). ‘Neue Forschungen im spätrömischen Kastell von Alzey: Vorbericht über
die Ausgrabungen 1981–​1985’, Bericht der Römisch-​Germanischen Kommission 67: 290–​356.
Pearson, A. (2002). The Roman Shore Forts:  Coastal Defences of Southern Britain.
Stroud: Tempus.
Petts, D. (2003). Christianity in Roman Britain. Stroud: Tempus.
Peytremann, E. (2003). Archéologie de l’habitat rural dans le nord de la France du IVe au XIIe
siècle, Mémoires de l’Association Française de l’Archéologie Mérovingienne XIII. 2  vols.
Condé-​sur-​Noireau: Association Française de l’Archéologie Mérovingienne.
Pilet, C. (1980). La Nécropole de Frénouville: Étude d’une population de la fin du IIIe à la fin du
VIIe siècle. BAR International Series 83. Oxford: British Archaeological Reports.
Pirling, R. (1986). Römer und Franken am Niederrhein. Mainz: Philipp von Zabern.
Pirling, R., Siepen, M., Noeske-​Winter, B., and Tegtmeier, U. (2000). Das römisch-​fränkische
Gräberfeld von Krefeld-​Gellep 1989–​2000. Stuttgart: Steiner.
Rigoir, J. (1968). ‘Les Sigillées paléochrétiennes grises et oranges’, Gallia, 26: 177–​243.
Soulat, J. (2009). Le Matériel archéologique de type saxon et Anglo-​Saxon en Gaule Mérovingienne.
Mémoires publiés par l’Association Française de l’Archéologie Mérovingienne XX. St-​
Germain-​en-​Laye: Association Française de l’Archéologie Mérovingienne.
 Britain at the End of Empire    149

Suzuki, S. (2000). The Quoit Brooch Style and Anglo-​Saxon Settlement. Woodbridge: Boydell.
Swift, E. (2000). Regionality in Dress Accessories in the Late Roman West. Monographies
Instrumentum 11. Montagnac: Éditions Monique Mergoil.
Swift, E. (2010). ‘Identifying Migrant Communities:  A  Contextual Analysis of Grave
Assemblages from Continental Late Roman Cemeteries’, Britannia, 41: 237–​282.
Theuws, F. (2009). ‘Grave Goods, Ethnicity and the Rhetoric of Burial Rites in Late Antique
Northern Gaul’, in T. Derks and N. Roymans (eds), Ethnic Constructs in Antiquity: The Role
of Power and Tradition. Amsterdam Archaeological Studies 13. Amsterdam:  Amsterdam
University Press, 293–​319.
Van Ossel, P. (1992). Établissements ruraux de l’Antiquité tardive dans le nord de la Gaule, (51e
Supplément à Gallia). Paris: Centre National de la Recherche Scientifique.
Van Ossel, P. (2003). ‘De la villa au village: Les Prémices d’une mutation’, in J.-​M. Yante and
A.-​M. Bultot-​Verleysen (eds), Autour du ‘village’: Établissements humains, finages et com-
munautés rurales entre Seine et Rhin (IVe–​XIIIe siècles) Actes du colloque international de
Louvain-​la-​Neuve, 16–​17 mai 2003. Turnhout: Brepols, 1–​19.
Van Ossel, P. (2010). ‘Transformations et continuités aux contacts de trois cultures:  Les
Campagnes de la Gaule septentrionale au Ve siècle’, in P. Delogu and S. Gasparri (eds), Le
trasformazioni del V secolo:  L’Italia, I  barbari e ‘Occidente romano, Seminari del Centro
Iteruniversitario per la storia e l’archeologia dell’alto medioevo 2.  Turnhout:  Brepols,
579–​600.
Van Ossel, P. (2011). ‘Les Sigillées du groupe Argonne dans le Bassin parisien au Bas-​
Empire: Caractérisation, production et diffusion’, in P. Van Ossel (ed.) Les Céramiques de
l’Antiquité tardive en Île-​de-​France et dans le Bassin parisien. Volume II. Synthèses. Diœcesis
Galliarum Document de travail no. 9. Paris: Diœcesis Galliarum, 231–​254.
Whittaker, C. (1994). Frontiers of the Roman Empire: A Social and Economic Survey. Baltimore
and London: Johns Hopkins University Press.
Wickham C. (2005). Framing the Early Middle Ages: Europe and the Mediterranean 400–​800.
Oxford: Oxford University Press.
Woodward, A. (1992). Shrines and Sacrifice. London: Batsford/​English Heritage.
150

Chapter 8

Britain be fore
the Roma ns

Timothy Champion

Introduction

In 1849 the historian Thomas Macaulay (1849: i. 4) wrote that the inhabitants of ancient
Britain were ‘little superior to the natives of the Sandwich Islands’ and ‘received only
a faint tincture of Roman arts and letters’. Other authors in this volume will debate
whether ‘faint tincture’ is an appropriate description of what happened after the con-
quest. Here I would like to consider the state of Britain immediately prior to that event.
Macaulay undoubtedly intended to be uncomplimentary, but he may have made a
serious misjudgement. The Sandwich Islands was the name given by Captain Cook
and still in use in the 1840s for the islands now known collectively as Hawaii, which
have become a classic case study of the development of a complex chiefdom society in
archaeological and anthropological textbooks (Kirch 2010, 2012). At the time of Cook’s
first visit in 1778 the islands were characterized by dense population, efficient irriga-
tion and rain-​fed agriculture, elite control of surplus production, divine kingship, and
wealth finance based on prestige goods. In the 120 years between Cook’s first visit and
the annexation of the islands by the empire of the United States in 1898, they experi-
enced growing contact with the European world through military and naval expedi-
tions, politicians, traders, arms dealers, and missionaries; under such external military,
economic, and religious influence from the USA, Britain, and Russia, they saw the
emergence of a unified political structure, largely through external support for one
particular chief, new economic structures, especially for exports, as well as new forms
of architecture, religion, language, and literacy (Kirch and Sahlins 1992). The parallel
with the situation in Britain in the century between Caesar and Claudius is not exact,
but it serves as a clear and well-​documented example of how a society can be trans-
formed by contact with very different cultural groups even before formal conquest and
annexation.
 Britain before the Romans    151

Macaulay had little sympathy for pre-​Roman Britain and certainly no notion of sig-
nificant changes in the centuries before the conquest. Archaeology developed rapidly in
the following decades, and by the early twentieth century an orthodoxy had emerged.
Sir John Evans’ Coins of the Ancient Britons (1864), a work that laid the foundations for
all future numismatic study of the Iron Age (de Jersey 2008), demonstrated clearly that
coins had been minted from at least the middle of the second century bc, and that some
of them showed evidence of a knowledge of Latin and the existence of the institution of
kingship. When Evans and his son Arthur visited a sand quarry at Aylesford in Kent in
1888, they found evidence for cremation burials of the Late Iron Age (LIA), one of which
contained an imported Italian bronze jug and pan. Arthur Evans’ eventual publica-
tion (1890: 386) of the finds traced the continental origins of LIA culture in Britain, and
attributed these innovations to an ‘invasion’ by ‘an intrusive Gaulish tribe’. The crema-
tion tradition was confirmed by the excavation of a small cemetery at Swarling in Kent
(Bushe-​Fox 1925), while connections with the classical world were again demonstrated
by the discovery of two rich graves at Welwyn, Hertfordshire, which contained not only
Roman bronze vessels and amphorae, but also a pair of Augustan silver cups (Smith
1912). Excavations in the 1930s at major sites such as St Albans (Wheeler and Wheeler
1936) and Colchester (Hawkes and Hull 1947) clearly demonstrated the Iron Age precur-
sors of some important Roman towns and the extent of contact with the Roman world,
seen especially in the presence of imported pre-​conquest pottery. By that time the cross-​
Channel connections had been explored in detail by Hawkes and Dunning (1930), and
Hawkes (1931) had published the original version of his influential tripartite scheme for
the British Iron Age, later to be revised and embellished (1959); in this scheme the final
stage, Iron Age C, was represented by two waves of immigration from the continent,
the first into Kent and Essex and the second further west, into Hampshire, Berkshire,
and Wiltshire. Though the so-​called second Belgic invasion in the west never won much
support, the idea of an Iron Age C culture centred on Kent, Essex, and Hertfordshire,
brought by immigrants from Belgium or northern France, remained the prevailing
theory for the explanation of LIA innovations in burial practice, pottery, coinage, set-
tlement, and social organization. Behind these upheavals in north-​western Europe was
the expansion of the Roman Empire, first into southern France in the late second cen-
tury bc, followed in the 50s by Caesar’s conquest of Gaul. The impact of Rome in the
post-​Caesarian period was now clear: ‘Romanization had struck deep roots before the
Claudian conquest’ (Kendrick and Hawkes 1932: 207).
Much of this orthodoxy is now questioned, if not yet actually discarded: the vision of
Britain as affected only by externally generated factors, the role of immigration, the geo-
graphical extent of major social changes in the later Iron Age, the chronology and pace
of social change, the nature of social organization immediately prior to the conquest,
and Romanization whether before or after Claudius. Some of this revision has been due
to the explosion of archaeological evidence since the 1990s, some to changing modes of
interpretation.
The changing modes of interpreting the evidence for the LIA will be reviewed in more
detail at the end of this chapter. There is now a growing literature of critical assessment
152

152   Timothy Champion

(e.g. Gardner 2013) and historiographical review (e.g. Hingley 2000, 2008; Hoselitz
2007) of the treatment of developments both before and after the conquest, but these
studies have been conducted mainly by specialists in Roman archaeology and have
focused on the dominant Roman side of the clash. By contrast, little (e.g. Webster 1996)
has been written by scholars specializing in the Iron Age or from the point of view of
the subaltern British. One notable exception to this is the topic of the Druids, who have
been the focus for several recent studies (Piggott 1985; Webster 1999; Hutton 2007, 2009;
Aldhouse-​Green 2010). Smiles’ scholarly study (1994) of the ancient Britons ranges far
more widely than the ‘images’ of the ‘romantic imagination’ that its title suggests, but its
chronological range stops before the rise of archaeology. We still lack a thorough histo-
riographical review of the study of the Iron Age.

The Evidence

With Caesar’s narrative of his cross-​Channel adventures in 55 and 54 bc, Britain enters
the written record of the classical world. His account (De Bello Gallico IV. 2–​36; V. 8–​23)
is fairly brief and not without problems, but it is fuller and more detailed than the patchy
record that survives for the following century before the Claudian invasion. This evi-
dence, primarily short passages in Strabo and Diodorus Siculus and even briefer refer-
ences in Suetonius and Dio Cassius, as well as some enigmatic references in the Roman
poets and one incomplete paragraph in Augustus’ Res Gestae, has been reviewed by sev-
eral authors (e.g. Braund 1996). There have been no additions to this corpus of sources,
though new editions, especially of the Res Gestae (Cooley 2009), have clarified some
issues. More important has been Woolf ’s exploration (2011) of the cultural and political
context in which knowledge of the west was created by the expanding Roman Empire.
By contrast, the archaeological evidence for Iron Age Britain has increased dramati-
cally since the last decades of the twentieth century (Haselgrove and Moore 2007a;
Haselgrove and Pope 2007). Some parts of this evidence will be discussed in detail later
in this chapter, but particularly noteworthy has been work on some of the major centres
of the LIA, such as Silchester (Fulford and Timby 2000) and Canterbury (Blockley et al.
1995), though the organization of the landscape and the pattern of ordinary rural set-
tlement remain poorly understood. Knowledge of the burial tradition in south-​eastern
England, mainly cremations, has been advanced by excavation of important cemeteries
(Parfitt 1995; Fitzpatrick 1997), warrior burials (Stevenson 2013), and elite cremations
(Niblett 1999; Crummy 2007). The Portable Antiquities Scheme has led to the record-
ing of a vast number of new finds of metalwork; the significance of these finds has not
been fully explored, though it is clear that there is great regional variation in the types of
material deposited in the Iron Age (Worrell 2007) and that detailed analysis of the finds
from a specific region can produce important evidence for the circulation and deposi-
tion of coins and other metal objects (Hutcheson 2004; Farley 2012). There has been
particular progress in the study of coinage: as well as more general reviews (Haselgrove
 Britain before the Romans    153

1987; Creighton 2000) and catalogues (van Arsdell 1989; Hobbs 1996; Rudd et al. 2010),
there have been important studies of some of the regional series, including the north-
eastern (Farley 2012), the East Anglian (Chadburn 2006), the south-​eastern or Kentish
(Holman 2000, 2005), the South Thames (Bean 2000), and the western (van Arsdell
1994).
The following sections will review how the new evidence and changing modes of
interpretation have altered perceptions of the later Iron Age, with particular attention
to southern and eastern England, where the first effects of contact with Rome were felt
most strongly. There were important developments in the preceding centuries, which
drew parts of Britain into a network of relationships with their continental neighbours,
so that, when the Roman Empire eventually extended into northern France, it was more
a case of transforming existing power structures than establishing completely new ones.

Changing Perceptions of the Iron Age

Since Arthur Evans’ publication (1890) of the rich LIA graves from Aylesford in Kent,
it has been recognized that there were significant cultural changes in the century or
two immediately preceding the Roman conquest. A distinctive late Iron Age as the cul-
mination of the insular Iron Age sequence has been identified predominantly in the
south-​east of England, though the idea has also found wider application in southern
and eastern England. More recently, however, archaeologists working in regions other
than the south-​east have tended to prefer a simpler division into earlier and later Iron
Age (Haselgrove et al. 2001; Moore 2006: 214–​224; Haselgrove and Moore 2007b: 1–​5),
with the dividing point somewhere around 40–​300 bc, suggesting that the changes seen
in the south-​east were not reflected elsewhere, but also that internal processes of change
were important over a much longer period. This vision of Iron Age developments has
not gone completely unchallenged, and the reality of widespread changes in the first
century bc has been reasserted; instead of steady social evolution, a pattern of punctu-
ated equilibrium, with long periods of comparatively static social organization inter-
rupted by brief episodes of change, has also been suggested (Barrett et al. 2011).
The question of the nature of Iron Age social organization has been the topic of recent
debate, though much of the discussion has revolved around the specific issue of the
extent to which society was hierarchically organized, especially in the earlier parts of the
Iron Age. In much early writing about the Iron Age there is little, if any, explicit discus-
sion of the nature of social organization or social relations, though there is an implicit
concept of a hierarchically organized society, presumably derived from the authority of
the classical references to chiefs and kings in the LIA, projected backwards into earlier
periods. Pictorial representations of ancient Britons in the eighteenth and nineteenth
centuries were also dominated by warriors and chiefs, as well as Druids (Smiles 1994).
More explicit discussion of the nature of Iron Age society began to appear in the lat-
ter part of the twentieth century. One influential model was derived from Cunliffe’s
154

154   Timothy Champion

discussion of his excavations at the hillfort of Danebury. Early attempts at explaining

its social role (Cunliffe 1984: 549–​562) included a comparison of Iron Age society with
the better-​documented historical societies of early medieval Ireland, underpinned by an
implicit assumption of widespread and long-​term homogeneity in the nature of ‘Celtic’
society; a concept of the hillfort as the possible residence of kings or nobles; a vision of
the role of the hillfort as an emerging centre of regional power and of redistribution,
drawing on neo-​evolutionary ideas of the role of a chiefdom; and the progression from
hillfort to developed hillfort to oppidum as the material expression in settlement terms
of the social evolutionary trajectory from tribe to chiefdom to state (Cunliffe 1976).
These interpretations have been the subject of sustained critique (e.g. Hill 1995, 1996),
as much on the grounds of the archaeological evidence as from a theoretical considera-
tion of the appropriateness of the analogies. Various attempts have been made to draw
on other models of pre-​industrial social organization: Crumley’s idea (1995) of ‘heter-
archy’, in which power is derived from a variety of different structures within society,
has attracted some writers, such as Armit (2007) and Cripps (2007); others, looking to
structural Marxist anthropology, have explored the concept of the Germanic Mode of
Production (e.g. Hingley 1984; Hill 1995). In an attempt to move beyond the concepts
of neo-​evolutionary anthropology and structural Marxism, Sharples (2010: 292–​294)
has adopted Mary Douglas’ grid and group matrix for social analysis. Hill (2011) has
explored non-​hierarchical structures for Iron Age society, citing the concept of segmen-
tary societies (Fortes and Evans-​Pritchard 1940). He also analysed some of the possibili-
ties for exploring elements of social organization such as households, communities, and
tribes, and, as the focus of attention has moved on from the debate about hierarchical
or non-​hierarchical social organization in the earlier periods of the Iron Age, so these
themes have been receiving more attention.
Much of this recent critique has been aimed at the concept of a hierarchical structure
in earlier Iron Age societies, and the attempt to find ways of discussing other forms of
society in which there may have been leaders rather than rulers. There has been surpris-
ingly little reconsideration of social organization in the final centuries of the Iron Age,
especially in the south-​east, when the classical authors refer to chiefs, military leaders,
and kings, supported by the evidence of some of the inscribed coinage, considered fur-
ther below. Hill (2007: 30) refers to kingship as ‘a novel experiment’ in the LIA of south-​
eastern England, but there has been little detailed consideration of the extent or the
basis of such ‘royal’ power. Caesar regularly describes the political groupings of south-
ern England in the mid-​first century bc by using the word civitates, usually translated as
‘tribes’, a term that Moore (2011) has argued to be so ideologically laden as to be unhelpful
and even misleading. It has also been shown that the distributions of the regional series
of coins in southern and eastern England, formerly attributed to such tribes and thought
to mark their distinct territories (Williams 2003, 2005a), are much more complex, reveal-
ing patterns at a variety of scales and making such an attribution very problematic (Leins
2008). Like the nature of political authority, the nature of political groupings in the LIA
has become much more difficult to discern, and almost certainly much more varied; at
the very least, projecting post-conquest structures back into prehistory is unwise.
 Britain before the Romans    155

Settlement and Economy

There is clear evidence for close contact between Britain and the continent throughout
the Iron Age (Webley 2015), though the precise nature of the social relationships repre-
sented by the material is often unclear. Much writing about the Iron Age has been based
on assumptions that such contacts declined after the end of the Bronze Age, that interac-
tion across the North Sea and the English Channel was intermittent until the LIA, that
these waters represented some form of cultural boundary, and that cultural innovations
began on the continent and crossed to Britain. All of these propositions should now be
regarded as doubtful.
Britain shared a number of significant cultural practices with the near continent at
a variety of scales, such as an architectural tradition of predominantly round rather
than rectangular structures, as well as the production and use of utilitarian artefacts
such as triangular clay loomweights (Champion 1975; Wilhelmi 1977, 1987), bone weav-
ing combs (Tuohy 1999), or, at a more localized scale, pottery (Leman-​Delerive 1984;
Hurtrelle et al. 1990; Blancquaert and Bostyn 1998). In such cases, the items form a uni-
fied cultural zone in which the Channel and North Sea did not constitute any form of
cultural boundary; it is more appropriate to talk of regions of shared cultural practice
than of points of innovation and areas of expansion or diffusion. In the case of other
items, such as weaponry or art, discussed further later in the chapter, it may make more
sense to think of a point or area of origin, and diffusion from there. The direction of such
diffusion, however, was not necessarily always in a north-​westerly direction from the
continent to Britain: some of the earliest evidence for rotary querns currently available
in Europe comes from southern England (Peacock and Cutler 2011; Wefers 2011). A clear
distinction between indigenous processes of change and externally generated innova-
tion is therefore problematic.
Indications of significant change in Iron Age economy and society can be seen from
perhaps 400 or 300 bc. The patterns of change vary considerably in chronology and are
highly variable regionally, being especially marked in southern and eastern England.
One obvious area of innovation is in the production and usage of material artefacts. The
sheer volume of such material items increased greatly, and there were important devel-
opments in the scale and specialization of production. Pottery manufacture was, at least
in some areas, increasingly in the hands of more specialized producers, especially for
finer wares, and their products were more standardized (Morris 1994, 1996). Quern pro-
duction also became more centralized, probably through the work of specialist produc-
ers working from major quarry sites such as Lodsworth (Peacock 1987). Production of
iron also increased, with the emergence of some major centres of smelting that distrib-
uted their iron for smithing in characteristically shaped bars of standardized size and
weight, the so-​called currency bars (Allen 1967; Hingley 1990, 2005; Crew 1994). Heavy
stone weights are known from a number of sites in southern England—​for example, at
Danebury (Cunliffe and Poole 1991: 383); though their function is uncertain, it is likely
156

156   Timothy Champion

that they were used for weighing out heavy or bulky items, suggesting the exchange of
agricultural produce. More specialized production would have been accompanied by
more complex relationships for the distribution and acquisition of products, and con-
trol over these processes of production and distribution would have provided new
opportunities for the accumulation of wealth and status.
There are also developments in the nature of settlement sites, though their relation-
ship to other changes is not always clear, and there is great regional variation. In the
regions of Wessex with a high concentration of hillforts, many sites were abandoned,
while a few were enlarged or elaborated to form the ‘developed hillforts’ in Cunliffe’s
terminology (2005: 388–​396). The term indicates a suggested role as emerging centres of
regional power and economic organization, but they may alternatively be the result of a
process of nucleation of a scattered population (Davis 2013). There are different histories
of hillfort construction and usage in other areas: in Sussex, hillforts were built in a vari-
ety of locations and for a variety of reasons from the late Bronze Age onwards, while in
Surrey and Kent the first forts were being built around the third century (Hamilton and
Manley 2000). In several regions, recent research has revealed a switch from open to
enclosed settlement forms (Knight 2007; Moore 2007). As well as changes in settlement
types, there is growing evidence for fluctuating densities of population; areas previously
densely occupied show much less evidence of activity, or less intensive usage, as activity
moved into new zones.
New types of sites also began to appear, especially from the second century onwards.
There is considerable variation in location, size, form, and function, as well as chronol-
ogy of foundation. Some of these sites have been grouped together and classed as oppida;
the term, originally applied by Caesar to some of the major defended sites in France, is
applied in continental, especially German, archaeology to some very large nucleated sites
of the LIA, but its usage in Britain is ‘more confused’ (Collis 1984: 6). Cunliffe (1976) has
suggested a typology of oppida: his scheme includes enclosed, undefended, and territo-
rial oppida, seen as types of nucleated centres that are different from the earlier hillforts.
Nevertheless, the range of sites included in this category makes it highly problematic:
some are indistinguishable from hillforts, while others that are in many ways similar to
oppida are excluded. It is doubtful whether the term is of much use for understanding the
actual processes of settlement development: Woolf ’s critique (1993a) of the term in con-
tinental archaeology could apply equally to its use in Britain (Bryant 2007; Pitts 2010:
34–​35). The enclosed oppida play an important role in Cunliffe’s model of social evolu-
tion (2005: 402–​406), replacing the developed hillforts as the settlement evidence for an
enhanced level of social organization and political centralization, and representing set-
tlements that were urban or of an urban character, while the territorial oppida are seen
as the centres of LIA power and authority. The fact that some of these sites later became
the location for Roman towns may indicate that they were places of importance in the
pre-​conquest period, but cannot be used to indicate their function at an earlier date; the
Roman Empire had very different structures of administrative organization.
One important development was the appearance of major sites on the coast, with a
significant role in cross-​Channel exchange. By the second half of the second century,
 Britain before the Romans    157

links between south-​western England and north-​western France, though perhaps

never totally absent, become more visible (Cunliffe and de Jersey 1997), articulated
through ports such as Hengistbury Head in Dorset (Cunliffe 1987) and Mount Batten
near Plymouth in Devon (Cunliffe 1988); though imported items such as coins, fine pot-
tery, and Roman amphorae, discussed further later in this chapter, did percolate into
the hinterlands, concentrations of such items at the ports suggest that they may have
been places with a distinct culture. Somewhat later, exchange between the continent and
south-​eastern England became more prominent; sites on the east coast of Kent, known
mostly from surface coin collections and small-​scale excavation (Holman 2005), may
have played a major role in this connection, a process beginning before the end of the
second century.
Elsewhere, there are signs of significant nucleation of settlement, though the chronol-
ogy is often uncertain. In eastern England, occupation at large sites such as Dragonby
(May 1996) and Sleaford (Elsdon 1997) had begun before the first century bc. At
Winchester a large rectangular enclosure had been occupied in the second and first cen-
turies (Qualmann and Whinney 2004). At Baldock, Hertfordshire, the main occupation
had begun ‘at least by the middle of the first century bc’ (Stead and Rigby 1986: 84),
and at Braughing, Hertfordshire (Partridge 1981; Potter and Trow 1988) and Heybridge,
Essex (Atkinson and Preston 1998) probably soon after that; at other sites, however,
such as Dyke Hills, Dorchester (Hingley and Miles 1984: 65 and fig. 4.9) and Abingdon
(Lambrick 2009: 362), both in Oxfordshire, the chronology of occupation is obscure,
though probably in the second or first centuries bc. These processes had clearly begun
in at least some regions by the second century, but were intensified in the first, especially
in the years after 50 bc.

Wealth, Power, and Ritual

As well as the similarities in architecture, pottery, and some artefact types noted above,
Britain also shared in more widespread fashions for prestigious or symbolic items (Joy
2015). There are very few objects that could be actual imports, but indigenous British
imitations and developments of continental styles suggest widespread familiarity with
continental practice, whether in weaponry such as daggers (Jope 1961) or swords (Stead
2006), in art (Jope 2000), or in bodily adornment and clothes fastening, with the brooch
replacing the pin (Adams 2014).
From the second century onwards, however, there are significant changes in the evi-
dence. Perhaps the most important of these is the reappearance of gold in the archaeolog-
ical record, especially in the form of coinage. Gold had been comparatively plentiful in
the late Bronze Age, but after about 800 bc it disappears from the archaeological record
in much of Britain, though recent metal-​detector finds suggest it may not have been
totally absent. In many parts of the central and western regions of continental Europe
gold became rarer, but continued to be deposited primarily in the richly furnished
158

158   Timothy Champion

graves of the elite in the form of personal ornaments or occasional vessels (Eluère 1987).
In the fourth and third centuries large quantities of gold flowed northwards from the
Mediterranean world as a result of barbarian raiding and the role of Celtic warriors as
the favoured mercenaries in the Hellenistic world. This influx of wealth played a major
role in the restructuring of social relationships in the barbarian world and the creation
of a new set of usages for wealth; gold was preferred in the west, though silver was more
common in some areas further east. From the late fourth century gold appears less and
less frequently in graves and more regularly as hoards or votive deposits, most often in
the form of coinage, though neck-​rings, or torques, of gold are also found, sometimes
in association with coins (Furger-​Gunti 1982; Fitzpatrick 2005). Celtic coins were being
minted from the early third century, copying Macedonian and other Mediterranean
prototypes. Gold was now used not for lavish display in the form of ornaments and ves-
sels, though torques may have been worn as a symbol of male prowess, but more com-
monly in the form of coins.
The earliest coins in western Europe were of precious metal, mostly gold, and were
of high-​value denominations. They were not used for everyday commercial exchange,
but could have been used as a medium for accumulating, assessing, and storing wealth,
as well as being ideally suited to use in the transfer of wealth. They are the material rep-
resentation of transactions involved in a limited set of social relationships (Allen 1976;
Nash 1987); these might have included payment for services rendered or tribute due to a
superior, dowry or bride price, gift exchange, and other acts of generosity, including the
deposition of votive offerings, which is the predominant archaeological context for their
discovery. Polybius, writing in the second century, describes Celtic society thus:

Their possessions consisted of cattle and gold, since these were the only objects which
they could easily take with them whatever their circumstances and transport wher-
ever they chose. It was of the greatest importance to them to have a following, and the
man who was believed to have the greatest number of dependents and companions
about him was the most feared and most powerful member of the tribe. (Hist. II. 17,
trans. I. Scott-​Kilvert)

At a later date, Caesar provides good evidence for the existence of such patron–​client
relationships in Gaul, especially in his account of the internal politics of the Helvetii (De
Bello Gallico I. 4): Orgetorix escaped trial when he turned up in court with his 10,000
followers. Some powerful men maintained their own permanent retinue of armed and
mounted men, such as Dumnorix of the Aedui (De Bello Gallico I. 18) or Commius of
the Atrebates (De Bello Gallico VIII. 23). Internal politics was dominated by factional
rivalry within the polities, often between brothers or other close male relatives in a
ruling lineage, such as the brothers Dumnorix and Diviciacus of the Aedui (De Bello
Gallico I. 18), and by alliances between factions within different polities, such as that
between Orgetorix, Dumnorix, and Casticus of the Sequani (De Bello Gallico I. 3). Such
patron–​client relationships and inter-​tribal alliances would have been maintained by a
variety of transactions, possibly including dynastic marriage alliances and the exchange
 Britain before the Romans    159

of gifts; in extreme cases, where alliances were called on to provide support in warfare,
large quantities of gold would have been needed. In these circumstances, coinage would
have provided a useful means for the new uses to which wealth could be put, especially
for its transfer.
Parts of eastern and southern Britain were clearly drawn into this network of social
relationships and wealth transfer, and into shared practices in the circulation and depo-
sition of wealth. Continental coins began to appear in Britain perhaps before 200 bc, and
more frequently in the first half of the following century. Some of the earliest coin types,
found on both sides of the English Channel, are usually thought to have been struck in
France, but it is not impossible that some were already being produced in England by the
middle of the second century. Local British coin series with imagery derived from the
Gaulish prototypes were being produced in southern and eastern England from the sec-
ond half of the second century, but some of the earliest coins certainly produced locally
were of potin, a high-​tin alloy of copper, which were cast rather than being struck; the
alloy and the production technology as well as the designs were copied from coin issues
of the Marseilles region (Haselgrove 1988, 1995, 2006). Though not made of precious
metal, they were hoarded and deposited in the same way as gold, and seem to have been
used and valued equivalently. Gold also circulated and was deposited in the form of tor-
ques, especially in parts of East Anglia, as at Snettisham (Stead 1991; Hutcheson 2003,
2004), though individual items have been found much more widely throughout the
country.
As well as the precious metals, other items began to appear in southern England.
Among the most significant are objects ultimately from the Roman world, transmit-
ted through late La Tène Gaul. Wine from Italy came in amphorae of Dressel 1 type:
the earlier form, Dressel 1A, arriving from around 125 bc, is found in southern central
England, while the later form, Dressel 1B, dating from around 8–​15 bc, is more clustered
in the south-​east (Peacock 1971, 1984; Fitzpatrick 1985; Carver 2001). Bronze vessels
from Campania (Boube 1991; Feugère and de Marinis 1991), though difficult to date pre-
cisely but quite possibly pre-​Caesarian, were also found widely in late La Tène Europe
(Werner 1954), and reached Britain, as shown by the jugs from Welwyn (Smith 1912) and
the jug and pan from Aylesford (Evans 1890).
Britain also adopted other practices from the late La Tène world, including new
styles of clothes fastening (Haselgrove 1997: 56–​57); from c. 120 bc brooches became
much more common, and new continental forms replaced indigenous ones. An impor-
tant technological innovation was the first production of wheel-​thrown pottery in
some parts of the south-​east of England, though much pottery was still hand-​made
(Hill 2002). The new technology was accompanied by new forms and a wider range of
more specialized vessel shapes, with close similarities to those in use in northern Gaul
(Hawkes and Dunning 1930; Tyers 1980). The new ceramic repertoire suggests new
modes of preparing, serving, and consuming food and drink, as well as perhaps a new
social setting and significance for these activities.
One context where wheel-​thrown pottery is found was in cremation burials, a new
rite also adopted from the continent (Collis 1977). The rite, often called the Aylesford
160

160   Timothy Champion

or Aylesford–​Swarling tradition after two of the earliest sites excavated (Evans 1890;
Bushe-​Fox 1925), is found most frequently in Kent, Essex, and Hertfordshire, though it
extends beyond this region. It involved the burial of a portion of the cremated remains
of the deceased in a pot (Fitzpatrick 2000), though occasional inhumation burials are
known, as well as regional variation in the details of deposition (Fitzpatrick 1997; Hill et
al. 1999). The absence of other artefacts in most of the graves makes precise dating dif-
ficult, but a small number of burials contain brooches, and the earliest such associations
would now be dated to the first half of the first century bc (Stead 1976; Fitzpatrick 1997:
95–​96) and the practice may have begun even before 100 bc.
Among these burials are a small number that are clearly much more richly furnished
than the rest. As originally identified (Stead 1967), this group was characterized by the
presence of a large rectangular burial chamber, and the inclusion of pottery, one or more
imported Roman amphorae and usually some glass or metal vessels, and was distinct
from another series of burials that contained metal-​bound wooden buckets (Stead
1971). Subsequently discovered burials (e.g. Dorton:  Farley 1983; Baldock:  Stead and
Rigby 1986) blurred the distinction between the groups and widened the range of grave
goods that occurred. Attempts to group the burials into categories defined on the basis
of wealth (Haselgrove 1982, 1984) have not been wholly persuasive. Though there is con-
siderable variation in the actual selection of goods for deposition in the graves, there are
predominant themes of feasting and drinking, with a mixture of indigenous items such
as buckets, cauldrons, and firedogs with exotics such as a bear-​skin rug or cloak, and
imports from the Roman world, including silver cups, bronze jugs, and pans, as well as
amphorae and other pottery vessels. Weaponry is notably absent, as are items decorated
with so-​called Celtic art; such decoration was most common in the second and the first
half of the first centuries bc, and there are few dates after about 50 bc, before a revival in
the middle of the first century ad (Garrow et al. 2009).
A very small number of graves are distinguished by their extreme wealth and a dif-
ferent burial rite, best documented at Folly Lane, St Albans (Niblett 1999), but also seen
at Lexden (Foster 1986) and Stanway (Crummy 2007), both at Colchester: the deceased
person was laid out in a sunken burial chamber, with a rich array of grave goods; the
grave goods were then smashed, and some placed on the pyre together with the body
and some animal parts; the chamber was demolished and the shaft backfilled, while a
portion of the cremated remains was placed in a separate burial pit. The ritual can be
paralleled in some of the richest graves of northern France (Niblett 1999: 394–​404) and
must represent the appropriate funerary treatment for some of the members of the most
important lineages in LIA Britain. The selection of grave goods is also different from
the feasting-​related emphasis of the other rich graves: Folly Lane contained local and
Roman horse harness, a tunic of iron chain mail, bronze and ivory furniture fittings,
and many other bronze and silver items melted beyond recognition, while Lexden also
included chain mail, furniture including a folding stool, and a denarius of Augustus.
The adoption of new burial rites closely modelled on those prevailing in north-
ern France would have served to underline the social and political relationships that
extended across the Channel, and to express social differences in a very visible way;
 Britain before the Romans    161

they may also have helped to establish new relationships if the settlement evidence is
rightly interpreted as indicating significant shifts in population density and new forms
of landscape organization. Burial, however, was only one way in which social pre-​emi-
nence could be expressed, and the dominant theme of feasting is echoed in non-​funer-
ary contexts, suggesting its significance in LIA social relations. Feasting may have been
an important feature of earlier periods in the first millennium, as at late Bronze Age
ringworks (Champion 2014: 291–​292) or early Iron Age hillforts (Jones 2007), based on
evidence for the storage and preparation of food. Joy (2014) has also argued that caul-
drons were used for the preparation of food or drink for feasts throughout the period.
Much has been made of the significance of imported wine in LIA Britain (e.g. Carver
2001), but the cauldrons, firedogs, and buckets suggest an indigenous element to feast-
ing rituals. In a survey of possible evidence for feasting in LIA East Anglia, Ralph (2007:
108) commented on the lack of evidence for drinking; but the absence of amphorae on
such sites does not mean an absence of drink, only an absence of Roman drink. Pitts
(2005) has argued that, in the fifty years before the Roman conquest, the regular pres-
ence of large drinking vessels, such as butt beakers, in ceramic assemblages implies
communal drinking of beer rather than wine. Whatever was consumed, feasts played a
major role in many societies, and the role of feasting may have changed greatly over time
(Hayden 2014). Dietler (1998, 2001) has suggested several different categories of feasting
that might be appropriate for LIA Britain; Fitzpatrick (2009: 397) has argued that the
rich burials of the period represent the burials of the organizers of feasts of his ‘patron-​
role’ type, acts of conspicuous entertainment designed to attract and retain a follow-
ing of clients, though Ralph (2007) has suggested a wider range of social contexts. The
Greek grammarian Athenaeus, writing about ad 200 (Deipnosophistae IV. 37), quotes
Posidinius’ much earlier description of the Arvernian nobleman Louernios, who rode
in a chariot distributing gold and silver, and prepared a feast with a huge quantity of
food and drink, to win popular support. This anecdote encapsulates some of the major
themes of LIA society, if social organization was broadly similar in central France and in
Britain: social difference, patron–​client relationships, circulation of wealth in fulfilment
of these relationships, feasting.
One other practice involving many of these elements was the ritual deposition of
wealth. Coins, whether found singly or in hoards, as well as other wealthy items such
as the gold jewellery in the Winchester hoard (Hill et al. 2004), are very rarely dis-
covered in settlement sites. More often they are from isolated findspots where little is
known about their context, though those places may have been significant in the Iron
Age landscape; nor, of course, is there any evidence of the people who carried out the
deposition or the rites associated with the act. Occasionally, it is possible to see that
there were sacred places or even shrines, as at Snettisham in Norfolk (Stead 1991). At
Hallaton, Leicestershire (Score 2011), a hill-​top shrine of the conquest period was the
site of multiple deposits including over 5,000 coins and a Roman silver-​covered cav-
alry helmet; large quantities of pig bones suggest feasting on the site. At a few of these
sacred sites a structure interpreted as a temple was built (Haselgrove 2005). The best
documented is at Hayling Island, Hampshire, where an Iron Age predecessor underlay a
162

162   Timothy Champion

later Romano-​Celtic temple and was the focus for deposits of coins, brooches, and other
ornaments, horse gear and chariot parts, as well as spearheads and chain mail (King and
Soffe 1994, 2008). One other frequent context for the ritual deposition of artefacts was
water, continuing a tradition with a long history in Britain (Bradley 1998): in eastern
England from the Thames northwards the predominant context for Iron Age swords
is rivers such as the Thames and the Trent (Stead 2006), while the Thames was also the
context for items such as the Battersea shield (Stead 1985). Springs and ponds may also
have been the location for deposits. Ritual activity, including deposition of coins, had
begun at the source of the Ebbsfleet river in Kent before the conquest, foreshadowing
the later religious complex at Springhead (Andrews et al. 2011).
By the middle of the first century bc, therefore, large parts of Britain had been incorpo-
rated into the west European network of wealth circulation and deposition, and a smaller
part of south-​eastern England had started to produce its own wealth items, with some
regional variation in preference for coins or torques, gold or bronze. An even smaller
region had begun to bury its dead according to continental practices, with some marked
evidence of social differences. Quite how this had happened is still in need of detailed
explanation, but it should be borne in mind, as noted above, that close relationships
across the North Sea and English Channel had existed earlier in the Iron Age, so that it
was more a case of transforming existing links than creating new ones. Though the histor-
ical record is minimal, Caesar’s account of Britain suggests that, at least in south-​eastern
England, similar political structures may have existed to those he had encountered in
Gaul and that political alliances or dependencies may have linked Britain and Gaul. An
enigmatic reference to Diviciacus of the Suessiones (De Bello Gallico II. 4), who is said
to have exerted some form of political authority across the water into England, suggests
the possibility of political alliances, whether between partners of nominally equal status
and authority or between those in a relationship of dominance and dependence. It is also
clear that such alliances could be called on to provide military support; again Caesar pro-
vides the evidence, with references to support from Britain for his opponents in almost
all his battles in Gaul (De Bello Gallico IV. 20). The leaders of the Bellovaci fled to Britain,
suggesting some form of real or potential alliance (De Bello Gallico IV. 20). The numerous
coins of Gallo-​Belgic E type appearing in Britain in the Caesarian period are generally
interpreted as payment for British involvement in the opposition to Rome. This repre-
sents not so much an unusual event as an extreme example of the operation of social and
political relationships through which people and goods regularly crossed the Channel.

From Caesar to Claudius: Britain

and Rome in the Final Iron Age

When Caesar first set foot in Britain in the late summer of 55 bc, it was Rome’s first
direct encounter with Britain, but perhaps not with Britons. Rome had entered into a
 Britain before the Romans    163

world where there was an existing network of frequently shifting political relationships,
whether alliances or antagonisms, and proceeded to play its part as a powerful new force,
with a mixture of diplomatic and military strategies. By the time of Caesar’s second inva-
sion in 54 bc, Rome was seen as a potential ally in the political rivalries of south-​eastern
England; Cassivellaunus had killed the king of the Trinovantes and Mandubracius, the
son of the dead king, had crossed the Channel to seek Caesar’s support (De Bello Gallico
V. 20).
Caesar’s expeditions across the Channel undoubtedly boosted his reputation in
Rome, even if they did not result in formal conquest and incorporation of any part of
Britain. Despite some doubts about the effectiveness or permanence of Roman control,
it does look as though the arrangements that Caesar made in 54 bc, including a demand
for annual payments and the taking of hostages, did establish the principle of Roman
authority in Britain. The historical and archaeological evidence gives some insight into
how this may have worked. Rome continued its policy of recognizing client kings out-
side the empire, though the details of any formal alliance are sparse. Caesar had installed
Mandubracius as ruler of the Trinovantes after his father’s death (De Bello Gallico V. 22),
and he may also have set up Commius as a ruler in central southern England. The his-
torical record is problematic (Bean 2000: 115–​117; Williams 2005a: 74–​75): Caesar made
Commius king of the Atrebates in northern France in 57 bc, and he served Caesar well
in his expeditions to Britain, but subsequently joined in the Gallic revolt of 52 bc, before
moving back to Britain; the name Commius appears on coins of southern England, and
Creighton (2000: 59–​64) has argued that Caesar established him as a king in southern
England, despite his rebellion. These Caesarian arrangements may have laid the foun-
dations for the emergence of two major polities, termed the Eastern Kingdom and the
Southern Kingdom in southern East Anglia and southern central England respectively.
At a later date, Strabo (Geog. 4.5.3) refers to British rulers sending embassies to
Augustus and making dedications on the Capitol in Rome; there are no details of the
date or the circumstances, but Braund (1996:  85)  has argued that one likely context
could have been the formal recognition of client kings. Whether or not they were for-
mally recognized as client kings, some British rulers, like Mandubracius before them,
regarded the Roman emperor as a powerful ally, especially at times of factional com-
petition: Augustus (Res Gestae 32) records that two British rulers, Dumnobellaun[us]
and Tincom[arus], fled to him, while Adminius, son of Cunobelin, was banished by
his father and fled to Rome (Suetonius, Caligula 44.2) and Verica did the same shortly
before the Claudian invasion, ‘having been driven out by an uprising’ (Dio Cassius, Hist.
lx. 19).
Despite Caesar’s demand for annual payments, the archaeological evidence suggests
that wealth was flowing from the Roman world to Britain. In the decades after his expe-
ditions, there are significant changes in the coinage of south-​eastern England. The gold
coins of the pre-​Caesarian period are rarely found in association with later issues, and
may have been deliberately withdrawn from circulation. Instead, new series began to
be produced from a different stock of metal, made from a different gold alloy and vis-
ible in a change of colour from a yellow gold to a red gold: the source was no longer the
164

164   Timothy Champion

wealth of the LIA barbarian world, but that of the Roman world (Creighton 2000: 68–​
74). Large quantities of silver also began to appear, again deriving from Roman bullion
and coinage, and this was the source of new silver coin issues in many parts of south-​
eastern England (Farley 2012: 34–​130); some silver circulated in other forms, such as
the Augustan silver cups found in the Welwyn and Welwyn Garden City burials (Smith
1912; Stead 1967).
From c. 20 bc there were two other significant changes in the coinage of the Southern
and Eastern Kingdoms. The imagery on the coins adopted a new set of motifs, closely
following the coinage of Augustus (Creighton 2000: 80–​125). Some issues also began
to carry inscriptions: the meaning of all the inscriptions is far from clear, but some are
the names of rulers known from other sources. A small number of issues bear the name
with the Latin title rex. Others have the Latin formula for claimed ancestral descent,
F (for filius) and a personal name in the genitive. Thus, in the Eastern Kingdom, some
coins bear the inscriptions cunobelinus and tasciovani f:  Tasciovanus is known
only from such coins, but Cunobelinus is the person known to us from Suetonius as
rex Britannorum (Caligula 44.2) and from Dio Cassius (Hist. lx. 20). In the Southern
Kingdom several rulers are described as sons of Commius. Whether such relationships
were actual biological ones or more like claims to ideologically important ancestry is not
clear, but these inscriptions show a familiarity with Latin and with Roman usage. These
developments in the imagery and inscriptions of the coinage of the south-​east demon-
strate a close connection to Rome, and Creighton (2000: 117–​124) has suggested that this
was due to the recognition of the rulers of the region as client kings and the practice of
members of those ruling dynasties spending time in Rome as guests or hostages.
The historical sources, together with the archaeological evidence of the coins and the
burials, clearly demonstrate the existence of powerful ruling lineages in south-​eastern
England, but the precise nature of that power is unclear. The classical sources natu-
rally concentrate on their relationships with Rome, and so we know that they were able
to raise and lead armies, to negotiate peace and give and take hostages; power could
be inherited within the lineage and was the subject of factional competition, so Rome
could be a powerful ally for those under internal pressure. What is not clear is how
these lineages had achieved such positions, or the real extent of their powers: the use
of the title rex and the occasional reference to them as kings in the historical sources
says more about the classical perception of such rulers than about the reality of their
rule. The concept of the Eastern and Southern Kingdoms is rooted in the distribution
of coins, but the relationship between coins and rulers, or between coins and polities, is
uncertain.
Some of the coins also bear inscriptions that can be interpreted as the names of
places, whether of mints or of other significant locations. In the Eastern Kingdom,
Camulodunum (Colchester) and Verulamium (St Albans) are named, and in the
Southern Kingdom, Calleva (Silchester). These were unlike the nucleated centres of
population such as Braughing. They were important new sites, apparently founded in
the second half of the first century bc, apparently in areas with little previous occupa-
tion. At Colchester (Gascoyne and Radford 2013), large parts of the enclosed area were
 Britain before the Romans    165

unoccupied; there was a probable area of industrial production at Sheepen (Hawkes and
Hull 1947), and rich burials at Lexden (Foster 1986) and Stanway (Crummy 2007), but
the function of the Gosbecks complex in the Iron Age is unknown, despite its signifi-
cance in the Roman period (Creighton 2006: 61–​64). Verulamium was very different
(Haselgrove and Millett 1997): activity around a marshy river valley included the mint-
ing of coins and the deposition of burials, especially the large cemetery at King Harry
Lane (Stead and Rigby 1989) and the rich Folly Lane cremation of the conquest period
(Niblett 1999).
In the Southern Kingdom, an area around Chichester was enclosed by a series of
embankments that may have been begun in the middle Iron Age (Bradley 1971), but
despite an assemblage of fine pre-​conquest ceramic imports there is no structural evi-
dence for occupation (Davenport 2003); a suggested focus of activity at Selsey is more
likely to be the result of multiple ritual deposits than domestic occupation (Bean 2000:
269–​271). The case is very different at Silchester, where excavation has produced unique
evidence for the adoption of new forms of settlement and living (Fulford and Timby
2000). Below the later forum was part of a rectilinear street grid laid out in the late
first century bc, flanked by rectangular structures. With imported amphorae that had
brought wine and olive oil from Italy and Spain, as well as food remains including oys-
ters and a high percentage of pig, it shows a remarkable contrast with contemporary
sites in southern England and much closer similarities to Gallo-​Roman sites in northern
France (Fulford and Timby 2000: 545–​564).
Other sites were founded at this period and went on to be important later, though
their status in the LIA is still obscure. At Canterbury structural evidence is fragmentary
but includes a triple-​ditched enclosure (Blockley et al. 1995), but, on the basis of coins
(Haselgrove 1987: 139–​145) and amphorae (Arthur 1986), the site may have been founded
in the middle of the first century bc. The coin evidence, however, now suggests that it
was one of several important sites in East Kent in the pre-​conquest period (Holman
2005). Leicester too was an important centre before the conquest, with imported pottery
and evidence of coin production (Clay and Mellor 1985).
Apart from the gold and silver, other items from the classical world also reached
Britain. Italian wine had been imported from the late second century, as noted earlier,
but the volume of imported ceramics grew considerably towards the end of the first
century bc. Wine from Italy and Spain, and olive oil from Spain, were reaching south-​
eastern England (Carver 2001; Fitzpatrick and Timby 2002). The silver cups from the
Welwyn and Welwyn Garden City burials may have been used for drinking such wine.
Other imports included flagons from central Gaul and Italian-​type terra sigillata, but
the largest volume of imports came from the Gallo-​Belgic industries founded in north-
ern Gaul after about 15 bc. These products represented a fusion of Roman and late La
Tène types, such as platters and beakers; they were startlingly different from the prod-
ucts of the indigenous tradition and were widely copied in Britain. The distribution of
Gallo-​Belgic imports shows marked concentrations in central southern and in south-​
eastern England, roughly coincident with the Southern and Eastern Kingdoms defined
by the coinage, but with other clusters in south-​western and eastern England, where the
166

166   Timothy Champion

imports may have arrived independently or have been distributed onwards from the
south-​east (Fitzpatrick and Timby 2002: figure 14.4).
Though most attention has been paid to the amphorae and other ceramics, other
innovations suggest a more complex pattern of interactions across the Channel. The
inscribed coinage shows a familiarity not only with Latin (Williams 2007), but also with
literacy more generally. Other signs of literacy in pre-​conquest contexts exist, though
they are not frequent, including styli and graffiti on pottery, and a probably just post-​
conquest grave at Stanway, Colchester, which contained an inkwell (Crummy 2007);
Williams (2001) has argued that literacy and its uses must have been familiar to Britain
through contact with the literate societies of Roman Gaul, but that it was adopted only
on a limited basis. Another innovation in Britain was the adoption of Roman-​style toilet
equipment; this suggests a new concern with the adornment of the body, though the
precise form that that took is uncertain (Hill 1997).

Changing Models of Interaction

The evidence cited demonstrates clearly that there was close contact between Britain
and the continent during the two centuries or more before the Roman invasion of ad
43. The nature of that contact, and the resulting processes of cultural change in Britain,
have been conceptualized in changing ways since the later nineteenth century. The idea
that prevailed then, and well into the twentieth century, of migration or invasion from
the continent, has now been largely rejected. Too much attention has been seduced
by Caesar’s reference to a migration from Belgium (De Bello Gallico V. 12: ‘qui … ex
Belgio transierant’) as a historical event in need of archaeological substantiation or
as an explanation for cultural change; there is much to be said for Williams’s sugges-
tion (2005a: 75) that it is an echo of an indigenous tradition, perhaps an origin myth.
Invasion hypotheses in general have gone out of favour, but there are more detailed
objections in the case of the LIA. The long chronology of cultural change that is now
evident is not compatible with a single causal event, or even with a small number of
such events; in any case, the idea of a migration does not of itself explain the adoption
in Britain of new cultural practices, and the rejection of others. A much longer history
of interaction is required, and a scenario that envisages regular cross-​Channel commu-
nication, with people crossing the water for very different reasons. Some people may
have crossed and returned, perhaps on many occasions, but others may have crossed
and settled. It is the nature of these crossings and their cultural consequences that is at
the heart of the debate.
Subsequent explanations have, to some extent, followed contemporary archaeo-
logical fashions. The heyday of processual archaeology can be seen in studies of LIA
Britain, emphasizing trade as the mechanism of change (Haselgrove 1982, 1984). These
were embedded in a spatial model of core and periphery, where the core (Rome) was
the active force in cultural change in an expanding periphery (the provinces), driven
 Britain before the Romans    167

by trade in prestige goods that allowed the indigenous elites to accumulate wealth and
power. The problems with this vision were that it placed too little weight on the political
connections that were implied by the historical evidence and the coinage, and gave an
exaggerated role to trade, both in its social significance (Woolf 1993b) and in its volume
(Willis 1994). The early trade was mostly confined to amphorae of wine, and did not
include other items such as Campanian ware, and only comparatively modest quanti-
ties reached Britain, to be used and deposited in a variety of ways appropriate to indig-
enous society. Nor were there clear signs of direct trade with Rome; without a historical
record or an agreed archaeological methodology for recognizing Roman traders in
Britain, claims for their presence, as at Braughing (Partridge 1981: 351), are not totally
convincing.
Nor did the concept of Romanization before the conquest (Kendrick and Hawkes
1932: 207; Haselgrove 1984) stand up to scrutiny. It became increasingly clear that many
of the innovations in LIA Britain were derived from contacts with the late La Tène world
of continental Europe, whether coinage, burials, or pottery. As already argued, even the
objects of Italian manufacture such as amphorae and bronze vessels were common in
the late La Tène world and could have been transmitted from there (Millett 1990: 38–​39).
When, after 15 bc, larger quantities of ceramic finewares were imported, they were from
the Gallo-​Belgic industries of northern Gaul, very different from the wares that were
supplied to Roman sites after the conquest (Pitts 2005). As Gosden (2004: 109) wrote,
‘South-​eastern Britain was as much Gallicised as Romanised in this period, with the
potentates of northern Gaul vital intermediaries for Mediterranean influences.’
More recent discussion has moved away from large-​scale models of interaction to
consideration of the processes of cultural change involved, with particular attention
to the active decisions of the indigenous populations as conscious agents, and to the
meaning of the material culture involved. It is clear that there was regular traffic across
the Channel and that people in south-​eastern England would have been well aware of
social practices and material culture on the continent. The nature of the connections
are inevitably somewhat speculative, but political alliances, marriage ties, and military
support are likely, first with the polities of northern Gaul and then with Rome. There
are some hints of the possibilities: a late La Tène helmet from a cremation burial near
Canterbury (Farley et al. 2014) and the Roman military horsegear in the Folly Lane
burial. The magnificent Roman cavalry helmet from Hallaton, Leicestershire, may have
been captured after the invasion, but could equally well have been the result of pre-​
conquest service in the Roman army or a diplomatic gift (Score 2011). Many people
may have crossed and returned, others may have settled permanently: settlers from
Gaul have been suggested for the foundation of Silchester (Fulford and Timby 2000:
563–​564).
These contacts would have been accompanied by the exchange of material items,
most importantly the Roman gold and silver that was transformed into insular coins;
they would also have given the opportunity of knowledge of continental practices, some
of which were adopted or imitated in Britain. Such adoption, however, was very variable.
It varied geographically: coins, for example, were widely distributed in Britain, but only
168

168   Timothy Champion

regularly used in southern and eastern England; cremation burial was adopted in an
even more restricted region, and perhaps by only part of the population. Other things,
such as Latin and literacy, were known, but used only in minimal ways in restricted
regions. Even where there are definite imports, their cultural significance and the social
context of their use may have been shaped by indigenous customs. Wine was certainly
imported, but its acceptability may have been because it fitted local practices of feasting,
alongside other drinks such as beer (Williams 2005b); its inclusion in burials in Britain
was most un-​Roman.
To try to explain this clash of cultures, a number of scholars have found an instruc-
tive parallel for the interaction of Rome and the non-​Roman societies of north-​western
Europe in the confrontation of European and non-​European societies in the early mod-
ern period. They have looked not to Macaulay’s Sandwich Islands, but to the east coast
of North America, and in particular to the work of Richard White (1991) and his con-
cept of the ‘middle ground’ to describe the relatively stable world that was formed for a
century or more in the Great Lakes region as the British and French penetrated deeper
into the territories of the native Americans. It was a world in which the old cultural
norms had been fatally disrupted, but neither side had established a new permanent
social or political order; in which conflicting languages, religions, technologies, modes
of exchange, and patterns of social relationships were accommodated. The exchange of
material items such as furs and iron axes was an important part of this interaction. In a
brief discussion, Gosden (2004: 110) envisaged the interaction of the Roman and non-​
Roman worlds as a series of middle grounds beyond the contemporary frontier, ‘bring-
ing new sets of cultural resources … which could be used, refused or subverted’. Woolf
(2011) also drew on the concept of the middle ground to explore how the classical world
tried to come to terms intellectually with its barbarian neighbours in Spain, Gaul, and
Britain. Within a relatively short period, new stories were created, drawing on classical
and native sources, to create a new hybrid body of historical and ethnographic knowl-
edge. We may never know how the pre-​conquest inhabitants of Britain conceptualized
their new Roman neighbours, but the classical literary tradition illuminates the reverse
process.
Williams (2005b: 37–​38) has also looked to North America as a comparative case, in
particular to the successful integration of imported alcohol into indigenous social ritu-
als. In the case of LIA Britain, he emphasized that, if we are looking for the meaning
of innovations, we must look ‘within the changing circumstances of Britain itself ’. The
most extended discussion of North American parallels has been by Farley (2012: 131–​
183), who has contrasted different phases of European and native American interac-
tion: an earlier phase where attempts to understand each other led to confusion and
frequent misunderstandings, followed by a phase in which a hybrid world of meaning
had been negotiated.
The problem of understanding this clash of cultures in LIA Britain can be exempli-
fied by the imported Italian bronze vessels, such as the jug and pan first identified by
Evans at Aylesford that began this debate. In the classical world such vessels are inter-
preted as parts of a set for the washing of hands (Nuber 1972; Bolla 1991; Feugère and de
 Britain before the Romans    169

Marinis 1991), but their association with evidence of feasting in elite burials in England
has suggested to some (e.g. Carver 2001; Ralph 2007) that they were used for the service
of wine. What did they mean in LIA Britain? Were they simply strange exotics, valued
for being exotic? Or did they signify the adoption of Roman dining rituals among the
British elites? Or were they incorporated into local feasting practices, and perhaps used
in very different ways from their original function?

References
Primary Sources
Athenaeus, Deipnosophistae.
Augustus, Res gestae.
Caesar, De Bello Gallico.
Dio Cassius, History of Rome.
Polybius, Histories.
Strabo, Geography.
Suetonius, Caligula.

Secondary Sources
Adams, S. A. (2014). ‘The First Brooches in Britain: From Manufacture to Deposition in the
Early and Middle Iron Age’. Ph.D. thesis, University of Leicester.
Aldhouse-​Green, M. J. (2010). Caesar’s Druids:  An Ancient Priesthood. New Haven and
London: Yale University Press.
Allen, D. F. (1967). ‘Iron Currency Bars in Britain’, Proceedings of the Prehistoric Society,
33: 307–​335.
Allen, D. F. (1976). ‘Wealth, Coinage and Money in a Celtic Society’, in J. V. S. Megaw (ed.), To
Illustrate the Monuments: Essays on Archaeology Presented to Stuart Piggott. London: Thames
and Hudson, 199–​208.
Andrews, P., Biddulph, E., Hardy, A., and Brown, R. (2011). Settling the Ebbsfleet Valley: High
Speed 1 Excavations at Springhead and Northfleet, Kent: The Late Iron Age, Roman, Saxon,
and Medieval Landscape, Volume 1:  The Sites. Oxford and Salisbury:  Oxford Wessex
Archaeology.
Armit, I. (2007). ‘Social Landscapes and Identities in the Irish Iron Age’, in C. Haselgrove and T.
Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow, 130–​139.
Arthur, P. (1986). ‘Roman Amphorae from Canterbury’, Britannia, 17: 239–​258.
Atkinson, M., and Preston, S. J. (1998). ‘The Late Iron Age and Roman Settlement at Elms Farm,
Heybridge, Essex, Excavations 1993–​1995: Interim Report’, Britannia, 29: 85–​110.
Barrett, J., Bowden, M., and McOmish, D. (2011). ‘The Problem of Continuity: Reassessing the
Shape of the British Iron Age Sequence’, in T. Moore and X.-​L. Armada (eds), Atlantic Europe
in the First Millennium bc: Crossing the Divide. Oxford: Oxford University Press, 439–​448.
Bean, S. C. (2000). The Coinage of the Atrebates and the Regni. Oxford:  Oxford University
School of Archaeology.
Blancquaert, G., and Bostyn, F. (1998). ‘L’Âge du fer à Coquelles et Fréthun (Pas-​de-​Calais)
(Fouilles du Transmanche 1986–​1988)’, Revue du Nord, 80: 109–​137.
170

170   Timothy Champion

Blockley, K., Blockley, M., Blockley, P., Frere, S., and Stow, S. (1995). Excavations in the Marlowe
Car Park and Surrounding Areas. The Archaeology of Canterbury 5. Canterbury: Canterbury
Archaeological Trust.
Bolla, M. (1991). ‘Considerazioni sulla funzione dei vasi in bronzo tardorepubblicani in
Italia settentrionale’, in M. Feugère and C. Rolley (eds), La Vaisselle tardo-​républicaine en
bronze: Actes de la table-​ronde CNRS organisée à Lattes du 26 au 28 avril 1990 par l’UPR 290
(Lattes) et le GDR 125 (Dijon). Centre de recherches sur les techniques gréco-​romaines,
no.  13. Dijon. Université de Bourgogne, Centre de recherches sur les techniques gréco-​
romaines, 143–​153.
Boube, C. (1991). ‘Les Cruches’, in M. Feugère and C. Rolley (eds), La Vaisselle tardo-​
républicaine en bronze: Actes de la table-​ronde CNRS organisée à Lattes du 26 au 28 avril 1990
par l’UPR 290 (Lattes) et le GDR 125 (Dijon). Centre de recherches sur les techniques gréco-​
romaines, no. 13. Dijon: Université de Bourgogne, Centre de recherches sur les techniques
gréco-​romaines, 23–​45.
Bradley, R. (1971). ‘A Field Survey of the Chichester Entrenchments’, in B. Cunliffe (ed.),
Excavations at Fishbourne. Research Report 26–​27. London:  Society of Antiquaries of
London, 17–​36.
Bradley, R. (1998). The Passage Of Arms: An Archaeological Analysis of Prehistoric Hoards and
Votive Deposits. 2nd edn. Oxford: Oxbow.
Braund, D. (1996). Ruling Roman Britain: Kings, Queens, Governors and Emperors from Julius
Caesar to Agricola. Abingdon: Routledge.
Bryant, S. (2007). ‘Central Places or Special Places? The Origins and Development of “Oppida”
in Hertfordshire’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and
Beyond. Oxford: Oxbow, 62–​80.
Bushe-​Fox, J. P. (1925). Excavations of the Late-​Celtic Urnfield at Swarling, Kent. Society of
Antiquaries of London Research Report. London: Society of Antiquaries.
Carver, E. (2001). The Visibility of Imported Wine and its Associated Accoutrements in Later Iron
Age Britain. British Archaeological Reports, British Series 325. Oxford: Archaeopress.
Chadburn, A. D. B. (2006). ‘Aspects of the Iron Age Coinage of Northern East Anglia with
Especial Reference to Hoards’. Ph.D. thesis, University of Nottingham.
Champion, T. C. (1975). ‘Britain and the European Iron Age’, Archaeologia Atlantica, 1: 127–​145.
Champion, T. C. (2014). ‘Food, Technology and Culture in the Late Bronze Age of Southern
Britain: Perforated Clay Plates of the Lower Thames Valley’, Proceedings of the Prehistoric
Society, 80: 279–​298.
Clay, P., and Mellor, J. (1985). Excavations in Bath Lane, Leicester. Archaeological Report 10.
Leicester: Leicestershire Museums, Art Galleries and Record Service.
Collis, J. R. (1977). ‘Pre-​Roman Burial Rites in North-​Western Europe’, in R. Reece (ed.), Burial
in the Roman World. Research Report 22. London: Council for British Archaeology, 1–​13.
Collis, J. R. (1984). Oppida: Earliest Towns North of the Alps. Sheffield: Department of Prehistory
and Archaeology, University of Sheffield.
Cooley, A. (2009). Res gestae divi Augusti:  Text, translation, and commentary.
Cambridge: Cambridge University Press.
Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge
University Press.
Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge.
Crew, P. (1994). ‘Currency Bars in Great Britain:  Typology and Function’, in M. Mangin
(ed.), La Sidérurgie ancienne de l’Est de la France dans son contexte européen: Archéologie
 Britain before the Romans    171

et archéométrie: Actes du colloque de Besançon, 10–​13 novembre 1993. Annales littéraires de

l’Université de Besançon 536. Paris: Belles Lettres, 340–​350.
Cripps, L. J. (2007). ‘Re-​Situating the Later Iron Age in Cornwall and Devon: New Perspectives
from the Settlement Record’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in
Britain and Beyond. Oxford: Oxbow, 140–​155.
Crumley, C. L. (1995). ‘Heterarchy and the Analysis of Complex Societies’, in R. M. Ehrenreich,
C. L. Crumley, and J. E. Levy (eds), Heterarchy and the Analysis of Complex Societies. Ann
Arbor: American Anthropological Association, 1–​5.
Crummy, P. (2007). Stanway: An Élite Burial Site at Camulodunum. London: Society for the
Promotion of Roman Studies.
Cunliffe, B. (1976). ‘The Origins of Urbanisation in Britain’, in B. Cunliffe and T. Rowley (eds),
Oppida: The Beginnings of Urbanisation in Barbarian Europe. Oxford: British Archaeological
Reports, Supplementary Series 11, 135–​161.
Cunliffe, B. (1984). Danebury: An Iron Age Hillfort in Hampshire. Volume 2: The Excavations
1969–​1978: The Finds. Research Report 52. London: Council for British Archaeology.
Cunliffe, B. (1987). Hengistbury Head Dorset: Volume 1: The Prehistoric and Roman Settlement,
3500 bc–​ad 500. Monograph 13. Oxford:  Oxford University Committee for Archaeolgy,
Institute of Archaeology.
Cunliffe, B. (1988). Mount Batten Plymouth: A Prehistoric and Roman Port. Monograph 26.
Oxford: Oxford University Committee for Archaeology, Institute of Archaeology.
Cunliffe, B. (2005). Iron Age Communities in Britain:  An Account of England, Scotland and
Wales from the Seventh Century bc to the Roman Conquest. 4th edn. London: Routledge.
Cunliffe, B., and de Jersey, P. (1997). Armorica and Britain:  Cross-​Channel Relationships in
the Late First Millennium BC. Monograph 45. Oxford: Oxford University Committee for
Archaeology.
Cunliffe, B., and Poole, C. (1991). Danebury:  An Iron Age hillfort in Hampshire. Volume
5: The Excavations 1979–​1988: The Finds. Research Report 73. London: Council for British
Archaeology.
Davenport, C. (2003). ‘The Late Pre-​Roman Iron Age of the West Sussex Coastal Plain:
Continuity or Change?’, in D. Rudling (ed.), The Archaeology of Sussex to ad 2000. King’s
Lynn: Heritage Marketing and Publications, 101–​110.
Davis, O. (2013). ‘Re-​ Interpreting the Danebury Assemblage:  Houses, Households, and
Community’, Proceedings of the Prehistoric Society, 79: 353–​375.
de Jersey, P. (2008). ‘Evans and Ancient British Coins’, in A. MacGregor (ed.), Sir John
Evans 1823–​ 1908:  Antiquity, Commerce and Natural Science in the Age of Darwin.
Oxford: Ashmolean Museum, 152–​172.
Dietler, M. (1998). ‘Feasts and Commensal Politics in the Political Economy: Food, Power and
Status in Prehistoric Europe’, in P. Wiessner and W. Schiefenhövel (eds), Food and the Status
Quest: An Interdisciplinary Perspective. Providence: Berghahn Books, 87–​125.
Dietler, M. (2001). ‘Theorizing the Feast: Rituals of Consumption, Commensal Politics, and
Power in African Contexts’, in M. Dietler and B. Hayden (eds), Feasts: Archaeological and
Ethnographic Perspectives on Food, Politics and Power. Washington: Smithsonian Institution
Press, 65–​114.
Elsdon, S. (1997). Old Sleaford Revealed: A Lincolnshire Settlement in Iron Age, Roman, Saxon
and Medieval Times: Excavations 1882–​1995. Monograph 78. Oxford: Oxbow.
Eluère, C. (1987). L’Or des Celtes. Paris: Bibliothèque des Arts.
Evans, A. J. (1890). ‘On a Late-​Celtic Urn-​Field at Aylesford, Kent’, Archaeologia, 52: 178–​192.
172

172   Timothy Champion

Evans, J. (1864). The Coins of the Ancient Britons. London: J. Russell Smith.

Farley, J. M.-​A. (2012). ‘At the Edge Of Empire: Iron Age and Early Roman Metalwork in the
East Midlands’. Ph.D. thesis, University of Leicester.
Farley, J. M.-​A., Parfitt, S. A., and Richardson, A. (2014). ‘An Iron Age Helmet Burial from
Bridge near Canterbury, Kent’, Proceedings of the Prehistoric Society, 80: 379–​388.
Farley, M. (1983). ‘A Mirror Burial at Dorton, Buckinghamshire’, Proceedings of the Prehistoric
Society, 49: 269–​302.
Feugère, M., and de Marinis, R. (1991). ‘Les poêlons’, In M. Feugère and C. Rolley (eds), La
Vaisselle tardo-​républicaine en bronze: actes de la table-​ronde CNRS organisée à Lattes du 26
au 28 avril 1990 par l’UPR 290 (Lattes) et le GDR 125 (Dijon). Centre de recherches sur les
techniques gréco-​romaines; no 13. Dijon: Université de Bourgogne, Centre de recherches
sur les techniques gréco-​romaines, 97–​112.
Fitzpatrick, A. P. (1985). ‘The Distribution of Dressel 1 Amphorae in North-​West Europe’,
Oxford Journal of Archaeology, 4: 305–​340.
Fitzpatrick, A. P. (1997). Archaeological Excavations on the Route of the A27 Westhampnett
Bypass, West Sussex, 1992. Volume 2: the Late Iron Age, Romano-​British and Anglo-​Saxon
cemeteries. Salisbury: Trust for Wessex Archaeology.
Fitzpatrick, A. P. (2000). ‘Ritual, Sequence and Structure in Late Iron Age Mortuary Practices
in North-​Western Europe’, in J. Pearce, M. Millett, and M. Struck (eds), Burial, Society and
Context in the Roman World. Oxford: Oxbow, 15–​29.
Fitzpatrick, A. P. (2005). ‘Gifts for the Golden Gods: Iron Age Hoards of Torques and Coins’,
in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Studien zu
Fundmünzen der Antike. Mainz: von Zabern, 157–​182.
Fitzpatrick, A. P. (2009). ‘The Champion’s Portion:  Feasting in the Celtic Pre-​ Roman
Iron Age’, in G. Cooney, K. Becker, J. Coles, M. Ryan, and S. Sievers (eds), Relics of Old
Decency: Archaeological Studies in Later Prehistory. Dublin: Wordwell, 389–​403.
Fitzpatrick, A. P., and Timby, J. (2002). ‘Roman Pottery in Iron Age Britain’, in A. Woodward
and J. D. Hill (eds), Prehistoric Britain: The Ceramic Basis. Oxford: Oxbow, 161–​172.
Fortes, M., and Evans-​Pritchard, E. E. (1940) (eds). African Political Systems. London: Oxford
University Press.
Foster, J. (1986). The Lexden Tumulus: A Re-​Appraisal of an Iron Age Burial from Colchester,
Essex. British Archaeological Reports, British Series 56. Oxford.
Fulford, M. G., and Timby, J. R. (2000). Late Iron Age and Roman Silchester: Excavations on the
Site of the Forum-​Basilica 1977, 1980–​86. Britannia Monograph Series 15. London: Society for
the Promotion of Roman Studies.
Furger-​Gunti, A. (1982). ‘Der “Goldfund von Saint-​Louis” bei Basel und ähnliche keltische
Schatzfunde’, Zeitschrift für Schweizerische Archaeologie und Kunstgeschichte, 39: 1–​47.
Gardner, A. (2013). ‘Thinking about Roman Imperialism: Postcolonialism, Globalisation and
Beyond?’, Britannia, 44: 1–​25.
Garrow, D., Gosden, C., Hill, J. D., and Bronk Ramsey, C. (2009). ‘Dating Celtic Art: A Major
Radiocarbon Dating Programme of Iron Age and Early Roman Metalwork in Britain’,
Archaeological Journal, 166: 79–​123.
Gascoyne, A., and Radford, D. (2013). Colchester: Fortress of the War God: An Archaeological
Assessment. Oxford: Oxbow Books.
Gosden, C. (2004). Archaeology and Colonialism: Cultural Contact from 5000 bc to the Present.
Cambridge: Cambridge University Press.
 Britain before the Romans    173

Hamilton, S., and Manley, J. (2000). ‘Hillforts, Monumentality and Place: A Chronological

and Topographic Review of First Millennium bc Hillforts in South-​East England’, European
Journal of Archaeology, 4: 7–​42.
Haselgrove, C. (1982). ‘Wealth, Prestige and Power:  The Dynamics of Late Iron Age
Political Centralisation in South-​East England’, in A. C. Renfrew and S. J. Shennan (eds),
Ranking, Resource and Exchange:  Aspects of the Archaeology of Early European Society.
Cambridge: Cambridge University Press, 79–​88.
Haselgrove, C. (1984). ‘“Romanization” before the Conquest: Gaulish Precedents and British
Consequences’, in T. F.  C. Blagg and A. C. King (eds), Military and Civilian in Roman
Britain: Cultural Relationships on a Frontier Province. British Archaeological Reports, British
Series 136. Oxford: British Archaeological Reports, 5–​63.
Haselgrove, C. (1987). Iron Age Coinage in South-​East England: The Archaeological Context.
British Archaeological Reports, British Series 174. Oxford: British Archaeological Reports.
Haselgrove, C. (1988). ‘The Archaeology of British Potin Coinage’, Archaeological Journal,
145: 99–​122.
Haselgrove, C. (1995). ‘Potin Coinage in Iron Age Britain, Archaeology and Chronology’,
Gallia, 52: 117–​127.
Haselgrove, C. (1997). ‘Iron Age Brooch Deposition and Chronology’, in A. Gwilt and C.
Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow, 51–​72.
Haselgrove, C. (2005). ‘A Trio of Temples: A Reassessment of Iron Age Coin Deposition at Hayling
Island, Harlow and Wanborough’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage
and Ritual Practices. Studien zu Fundmünzen der Antike. Mainz: von Zabern, 381–​418.
Haselgrove, C. (2006). ‘Early Potin Coinage in Britain: An Update’, in P. de Jersey (ed.), Celtic
Coinage:  New Discoveries, New Discussion. British Archaeological Reports, International
Series 1532. Oxford: Archaeopress, 17–​28.
Haselgrove, C., and Millett, M. (1997). ‘Verlamion Reconsidered’, in A. Gwilt and C. Haselgrove
(eds), Reconstructing Iron Age Societies. Oxford: Oxbow, 282–​296.
Haselgrove, C., and Moore, T. (2007a) (eds). The Later Iron Age in Britain and Beyond.
Oxford: Oxbow.
Haselgrove, C., and Moore, T. (2007b). ‘New Narratives of the Later Iron Age’, in C. Haselgrove
and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow, 1–​15.
Haselgrove, C., and Pope, R. E. (2007) (eds). The Earlier Iron Age in Britain and the Near
Continent. Oxford: Oxbow.
Haselgrove, C., Armit, I., Champion, T. C., Creighton, J., Gwilt, A., Hill, J. D., Hunter, F.,
and Woodward, A. (2001). Understanding the British Iron Age:  An Agenda for Action.
Salisbury: Trust for Wessex Archaeology.
Hawkes, C. F. C. (1931). ‘Hillforts’, Antiquity, 5: 60–​97.
Hawkes, C. F. C. (1959). ‘The ABC of the British Iron Age’, Antiquity, 33: 170–​182.
Hawkes, C. F. C., and Dunning, G. C. (1930). ‘The Belgae of Gaul and Britain’, Archaeological
Journal, 87: 150–​335.
Hawkes, C. F. C., and Hull, M. R. (1947). Camulodunum: First Report on the Excavations at
Colchester 1930–​1939. Research Report XIV. Oxford: Society of Antiquaries of London.
Hayden, B. (2014). The Power of Feasts: From Prehistory to the Present. Cambridge: Cambridge
University Press.
Hill, J. D. (1995). ‘How Should we Understand Iron Age Societies and Hillforts? A Contextual
Study from Southern Britain’, in J. D. Hill and C. G. Cumberpatch (eds), Different Iron
174

174   Timothy Champion

Ages:  Studies on the Iron Age in Temperate Europe. British Archaeological Reports,
International Series 602. Oxford: Tempus Reparatum, 45–​66.
Hill, J. D. (1996). ‘Hillforts and the Iron Age of Wessex’, in T. C. Champion and J. R. Collis (eds),
The Iron Age in Britain and Ireland: Recent Trends. Sheffield: J. R. Collis Publications, 95–​116.
Hill, J. D. (1997). ‘“The End of One Kind of Body and the Beginning of Another Kind of Body”?
Toilet Instruments and “Romanization”’, in A. Gwilt and C. Haselgrove (eds), Reconstructing
Iron Age Societies. Oxford: Oxbow, 96–​107.
Hill, J. D. (2002). ‘Just about the Potter’s Wheel? Using, Making and Depositing Middle
and Later Iron Age Pots in East Anglia’, in A. Woodward and J. D. Hill (eds), Prehistoric
Britain: The Ceramic Basis. Oxford: Oxbow, 143–​160.
Hill, J. D. (2007). ‘The Dynamics of Social Change in Later Iron Age Eastern and South-​Eastern
England’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond.
Oxford: Oxbow, 16–​40.
Hill, J. D. (2011). ‘How did British Middle and Late Pre-​Roman Iron Age Societies Work (if
they did)?’, in T. Moore and X.-​L. Armada (eds), Atlantic Europe in the First Millennium bc:
Crossing the Divide. Oxford: Oxford University Press, 242–​263.
Hill, J. D., Evans, C., and Alexander, M. (1999). ‘The Hinxton Rings: A Late Iron Age Cemetery
at Hinxton, Cambridgeshire, with a Reconsideration of Northern Aylesford–​Swarling
Distributions’, Proceedings of the Prehistoric Society, 65: 243–​273.
Hill, J. D., Spence, A. J., LaNeice, S., and Worrell, S. (2004). ‘The Winchester Hoard: A Find of
Unique Iron Age Gold Jewellery from Southern England’, Antiquaries Journal, 84: 1–​22.
Hingley, R. (1984). ‘Towards Social Analysis in Archaeology: Celtic Society in the Iron Age of
the Upper Thames Valley’, in B. Cunliffe and D. Miles (eds), Aspects of the Iron Age in Central
Southern England. Oxford: Oxford University Committee for Archaeology, 72–​88.
Hingley, R. (1990). ‘Iron Age “Currency Bars”:  The Archaeological and Social Context’,
Archaeological Journal, 147: 91–​117.
Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman
Archaeology. London: Routledge.
Hingley, R. (2005). ‘Iron Age “Currency Bars” in Britain:  Items of Exchange in Liminal
Contexts?’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices.
Studien zu Fundmünzen der Antike. Mainz: von Zabern, 183–​205.
Hingley, R. (2008). The Recovery of Roman Britain:  1586–​ 1906:  A  Colony so Fertile.
Oxford: Oxford University Press.
Hingley, R., and Miles, D. (1984). ‘Aspects of Iron Age Settlement in the Upper Thames
Valley’, in B. Cunliffe and D. Miles (eds), Aspects of the Iron Age in Central Southern Britain.
Oxford: University of Oxford Commitee for Archaeology, 52–​7 1.
Hobbs, R. (1996). British Iron Age Coins in the British Museum. London: British Museum Press.
Holman, D. (2000). ‘Iron Age Coinage in Kent’, Archaeologia Cantiana, 120: 205–​234.
Holman, D. (2005). ‘Iron Age Coinage and Settlement in East Kent’, Britannia, 36: 1–​54.
Hoselitz, V. (2007). Imagining Roman Britain:  Victorian Responses to a Roman Past.
Woodbridge: Boydell.
Hurtrelle, J., Monchy, E., Roger, F., Rossignol, P., and Villes, A. (1990). Les Débuts du second âge
du fer dans le Nord de la France. Les Dossiers de Gauheria. 1. Givenchy-​en-​Goelle: Gauheria.
Hutcheson, N. (2003). ‘Material Culture in the Landscape: A New Approach to the Snettisham
Hoards’, in J. Humphrey (ed.), Re-​ Searching the Iron Age. Leicester:  University of
Leicester, 87–​97.
 Britain before the Romans    175

Hutcheson, N. (2004). Later Iron Age Norfolk:  Metalwork, Landscape and Society. British
Archaeological Reports. British Series 361. Oxford: Archaeopress.
Hutton, R. (2007). The Druids: A History. London: Hambledon Continuum.
Hutton, R. (2009). Blood and Mistletoe: The History of the Druids in Britain. New Haven and
London: Yale University Press.
Jones, M. K. (2007). ‘A Feast of Beltain? Reflections on the Rich Danebury Harvests’, in C.
Gosden, H. Hamerow, P. de Jersey, and G. Lock (eds), Communities and Connections: Essays
in Honour of Barry Cunliffe. Oxford: Oxford University Press, 143–​153.
Jope, E. M. (1961). ‘Daggers of the Early Iron Age in Britain’, Proceedings of the Prehistoric
Soceity, 27: 307–​343.
Jope, E. M. (2000). Early Celtic Art in the British Isles. Oxford: Oxford University Press.
Joy, J. (2014). ‘“Fire burn and cauldron bubble”: Iron Age and Early Roman Cauldrons of Britain
and Ireland’, Proceedings of the Prehistoric Society, 80: 327–​362.
Joy, J. (2015). ‘Connections and Separation? Narratives of Iron Age Art in Britian and its
Relationship with the Continent’, in H. Anderson-​Whymark, D. Garrow, and F. Sturt (eds),
Continental Connections: Exploring Cross-​Channel Relationships from the Mesolithic to the
Iron Age. Oxford: Oxbow, 145–​165.
Kendrick, T. D., and Hawkes, C. F.  C. (1932). Archaeology in England and Wales 1914–​1931.
London: Methuen.
King, A. C., and Soffe, G. (1994.) ‘The Iron Age and Roman Temple on Hayling Island’, in A. P.
Fitzpatrick and E. L. Morris (eds), The Iron Age in Wessex: Recent Work. Salisbury: Trust for
Wessex Archaeology, 114–​116.
King, A. C., and Soffe, G. (2008). ‘Hayling Island: A Gallo-​Roman Temple in Britain’, in D.
Rudling (ed.), Ritual Landscapes of Roman South-​ East Britain. Great Dunham and
Oxford: Heritage Marketing & Publications and Oxbow, 139–​151.
Kirch, P. V. (2010). How Chiefs Became Kings: Divine Kingship and the Rise of Archaic States in
Ancient Hawai’i. Berkeley: University of California Press.
Kirch, P. V. (2012). A Shark Going Inland Is My Chief: The Island Civilization of Ancient Hawai’i.
Berkeley: University of California Press.
Kirch, P. V., and Sahlins, M. D. (1992). Anahulu: The Anthropology of History in the Kingdom of
Hawaii. Chicago: University of Chicago Press.
Knight, D. (2007). ‘From Open to Enclosed: Iron Age Landscapes in the Trent Valley’, in C.
Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow,
190–​218.
Lambrick, G. (2009). The Thames through Time:  The Archaeology of the Gravel Terraces
of the Upper and Middle Thames:  The Thames Valley in Late Prehistory, 1500 bc–​ad 50.
Oxford: Oxford University School of Archaeology for Oxford Archaeology.
Leins, I. (2008). ‘What can be Inferred from the Regional Stylistic Diversity of Iron
Age Coinage?’, in D. Garrow, C. Gosden, and J. D. Hill (eds), Rethinking Celtic Art.
Oxford: Oxbow, 100–​112.
Leman-​Delerive, G. (1984). ‘Céramique laténienne domestique de la région lilloise (Nord)’,
Gallia, 42: 79–​95.
Macaulay, T. B. (1849). The History of England from the Accession of James II. London: Longman,
Green, Brown and Longmans.
May, J. (1996). Dragonby: Report on Excavations at an Iron Age and Romano-​British Settlement
in North Lincolshire. Oxford: Oxbow.
176

176   Timothy Champion

Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.

Cambridge: Cambridge University Press.
Moore, T. (2006). Iron Age Societies in the Severn-​Cotswolds: Developing Narratives of Social and
Landscape Change. British Archaeological Reports, British Series 421. Oxford: Archaeopress.
Moore, T. (2007). ‘The Early to Late Iron Age Transition in the Severn–​Cotswolds: Enclosing
the Household?’, in C. C. Haselgrove and R. E. Pope (eds), The Earlier Iron Age in Britain and
the Near Continent. Oxford: Oxbow, 259–​278.
Moore, T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and
Roman Studies’, Journal of Social Archaeology, 11/​3: 334–​360.
Morris, E. L. (1994.) ‘Production and Distribution of Pottery and Salt in Iron Age
Britain: A Review’, Proceedings of the Prehistoric Society, 60: 371–​393.
Morris, E. L. (1996). ‘Iron Age Artefact Production and Exchange’, in T. C. Champion and
J. Collis (eds), The Iron Age in Britain and Ireland:  Recent Trends. Sheffield:  J R Collis
Publications, 41–​65.
Nash, D. (1987). Coinage in the Celtic World. London: Seaby.
Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium. Britannia
Monograph Series 14. London: Society for the Promotion of Roman Studies.
Nuber, H. U. (1972). ‘Kanne und Griffschale: Ihr Gebrauch im täglichen Leben und die Beigabe
in Gräbern der römischen Kaiserzeit’, Bericht der Römisch-​Germanischen Kommission,
53: 1–​232.
Parfitt, K. (1995). Iron Age Burials from Mill Hill, Deal. London: British Museum Press.
Partridge, C. (1981). Skeleton Green: A Late Iron Age and Romano-​British Site. Britannia
Monograph Series 2. London: Society for the Promotion of Roman Studies.
Peacock, D. P. S. (1971). ‘Roman Amphorae in Pre-​Roman Britain’, in D. Hill and M. Jesson
(eds), The Iron Age and its Hill-​Forts. Southampton: Southampton University Archaeology
Society, 161–​188.
Peacock, D. P. S. (1984). ‘Amphorae in Iron Age Britain: A Reassessment’, in S. Macready and F.
H. Thompson (eds), Cross-​Channel Trade between Gaul and Britain in the Pre-​Roman Iron
Age. Occasional Paper (ns) 4. London: Society of Antiquaries, 37–​42.
Peacock, D. P. S. (1987). ‘Iron Age and Roman Quern Production at Lodsworth, West Sussex’,
Antiquaries Journal, 67: 61–​85.
Peacock, D. P. S., and Cutler, L. (2011). ‘The Earliest Rotary Querns in Southern England’, in D.
Williams and D. Peacock (eds), Bread for the People: The Archaeology of Mills and Milling.
Oxford: Archaeopress, 70–​80.
Piggott, S. (1985). The Druids. London: Thames & Hudson.
Pitts, M. (2005). ‘Pots and Pits: Drinking and Deposition in Late Iron Age South-​East Britain’,
Oxford Journal of Archaeology, 24/​2: 143–​161.
Pitts, M. (2010). ‘Re-​Thinking the Southern British Oppida: Networks, Kingdoms and Material
Culture’, European Journal of Archaeology, 13/​1: 32–​63.
Potter, T. W., and Trow, S. D. (1988). ‘Puckeridge–​Braughing, Herts.:  The Ermine Street
Excavations, 1971–​1972’, Hertfordshire Archaeology, 10: 1–​191.
Qualmann, K. E., and Whinney, R. J.  B. (2004). Oram’s Arbour:  The Iron Age Enclosure at
Winchester. Winchester: Winchester Museums Service.
Ralph, S. (2007). Feasting and Social Complexity in Later Iron Age East Anglia. British
Archaeological Reports British Series 451. Oxford: Archaeopress.
Rudd, C., Cottam, E., de Jersey, P., and Sills, J. (2010). Ancient British Coins.
Aylsham: Chris Rudd.
 Britain before the Romans    177

Score, V. (2011). Hoards, Hounds and Helmets:  A  Conquest-​Period Ritual Site at Hallaton,
Leicestershire. Leicester Archaeology Monograph 21. Leicester:  University of Leicester
Archaeological Services.
Sharples, N. (2010). Social Relations in Later Prehistory: Wessex in the First Millennium bc.
Oxford: Oxford University Press.
Smiles, S. (1994). The Image of Antiquity:  Ancient Britain and the Romantic Imagination.
London: Yale University Press.
Smith, R. A. (1912). ‘On Late-​Celtic Antiquities Discovered at Welwyn, Herts’, Archaeologia,
63: 1–​30.
Stead, I. M. (1967). ‘A La Tène III Burial at Welwyn Garden City’, Archaeologia, 101: 1–​62.
Stead, I. M. (1971). ‘The Reconstruction of Iron Age Buckets from Aylesford and Swarling’,
in G.  d.  G. Sieveking (ed.), Prehistoric and Roman Studies. London:  British Museum,
250–​282.
Stead, I. M. (1976). ‘The Earliest Burials of the Aylesford Culture’, in G.  d.  G. Sieveking, I.
H. Longworth, and K. E. Wilson (eds), Problems in Economic and Social Archaeology.
London: Duckworth, 401–​416.
Stead, I. M. (1985). The Battersea Shield. London: British Museum Publications.
Stead, I. M. (1991). ‘The Snettisham Treasure: Excavations in 1990’, Antiquity, 65: 447–​465.
Stead, I. M. (2006). British Iron Age Swords and Scabbards. London: British Museum Press.
Stead, I. M., and Rigby, V. (1986). Baldock: The Excavation of a Roman and Pre-​Roman
Settlement, 1968–​72. Britannia Monograph Series 7. London: Society for the Promotion of
Roman Studies.
Stead, I. M., and Rigby, V. (1989). Verulamium:  The King Harry Lane Site. Archaeological
Report 12. London: English Heritage.
Stevenson, J. (2013). Living by The Sword:  The Archaeology of Brisley Farm, Ashford, Kent.
Portslade: SpoilHeap Publications.
Tuohy, C. (1999). Prehistoric Combs of Antler and Bone. British Archaeological Report British
Series 285. Oxford: British Archaeological Reports.
Tyers, P. A. (1980). ‘Correspondances entre la céramique commune La Tène III du sud-​est de
l’Angleterre et du nord de la France’, Septentrion, 10: 61–​70.
van Arsdell, R. D. (1989). Celtic Coinage of Britain. London: Spink.
van Arsdell, R. D. (1994). The Coinage of the Dobunni: Money Supply and Coinage Circulation in
Dobunnic Territory. Oxford: Oxford University Committee for Archaeology.
Webley, L. (2015). ‘Rethinking Iron Age Connections across the Channel and North Sea’, in
H. Anderson-​Whymark, D. Garrow, and F. Sturt (eds), Continental Connections: Exploring
Cross-​Channel Relationships from the Mesolithic to the Iron Age. Oxford: Oxbow, 122–​144.
Webster, J. (1996). ‘Ethnographic Barbarity: Colonial Discourse and “Celtic Warrior Societies”’,
in J. Webster and N. Cooper (eds), Roman Imperialism:  Post-​ Colonial Perspectives.
Leicester: School of Archaeological Studies, University of Leicester, 111–​123.
Webster, J. (1999). ‘At the End of the World: Druidic and Other Revitalization Movements in
Post-​Conquest Gaul and Britain’, Britannia, 30: 1–​20.
Wefers, S. (2011). ‘Still Using your Saddle Quern? A Compilation of the Oldest Known Rotary
Querns in Western Europe’, in D. Williams and D. Peacock (eds), Bread for the People: The
Archaeology of Mills and Milling. Oxford: Archaeopress, 67–​76.
Werner, J. (1954). ‘Die Bronzekanne von Kelheim’, Bayerische Vorgeschichtsblätter, 20: 43–​73.
Wheeler, R. E. M., and Wheeler, T. V. (1936). Verulamium: A Belgic and Two Roman Cities.
Research Report XI. London: Society of Antiquaries of London.
178

178   Timothy Champion

White, R. (1991). The Middle Ground: Indians, Empires, and Republics in the Great Lakes Region,
1650–​1815. Cambridge: Cambridge University Press.
Wilhelmi, K. (1977). ‘Zur Funktion und Verbreitung dreieckige Tongewichte der Eisenzeit’,
Germania, 55: 180–​184.
Wilhelmi, K. (1987). ‘Zur Besiedlungsgenese Englands und des nordwestlichen Kontinent von
1500 vor bis Christi Geburt’, Acta Praehistorica Archaeologica, 19: 71–​84.
Williams, J. H.  C. (2001). ‘Coin Inscriptions and the Origins of Writing in Pre-​Conquest
Britain’, British Numismatic Journal, 71: 1–​17.
Williams, J. H. C. (2003). ‘Iron-​Age and Roman Coins’, British Numismatic Journal, 73: 44–​57.
Williams, J. H. C. (2005a). ‘Coinage and Identity in Pre-​Conquest Britain: 50 bc–​ad 50’, in C.
Howgego, V. Heuchert, and A. Burnett (eds), Coinage and Identity in the Roman Provinces.
Oxford: Oxford University Press, 69–​78.
Williams, J. H. C. (2005b). ‘ “The newer rite is here”: Vinous Symbolism on British Iron Age
Coins’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices.
Studien zu Fundmünzen der Antike. Mainz: von Zabern, 25–​41.
Williams, J. H. C. (2007). ‘New Light on Latin in Pre-​Conquest Britain’, Britannia, 38: 1–​11.
Willis, S. (1994). ‘Roman Imports into Late Iron Age British Societies:  Towards a Critique
of Existing Models’, in S. Cottam, D. Dungworth, S. Scott, and J. Taylor (eds), TRAC
94: Proceedings of the Fourth Annual Theoretical Roman Archaeology Conference, Durham,
1994. Oxford: Oxbow, 141–​150.
Woolf, G. D. (1993a). ‘Rethinking the Oppida’, Oxford Journal of Archaeology, 12: 223–​232.
Woolf, G. D. (1993b). ‘The Social Significance of Trade in Late Iron Age Europe’, in C. Scarre
and F. Healey (eds), Trade and Exchange in Prehistoric Europe. Oxford: Oxbow, 211–​218.
Woolf, G. D. (2011). Tales of the Barbarians:  Ethnography and Empire in the Roman West.
Chichester: John Wiley & Sons.
Worrell, S. (2007). ‘Detecting the Later Iron Age:  A  View from the Portable Antiquities
Scheme’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and beyond.
Oxford: Oxbow, 371–​388.
 Chapter 9

Beyond Hadria n’ s  Wa l l

Fraser Hunter

Introduction

The edges of empires are fascinating places. Britannia’s northern frontier was a varied
one, with considerable fluidity in its location until the early third century. The vagar-
ies in frontier location and historical sources for conflict and turmoil indicate that the
relationship with Rome’s northern neighbours was far from easy. Yet archaeological evi-
dence can present other sides of this relationship, with Roman material culture finding
varied uses within these Iron Age societies, while the long and often difficult relation-
ship had a series of unexpected consequences on both sides. This chapter will start with
a pen portrait of Iron Age societies before the Roman invasion, consider the military
history and the wider picture of life on the frontier from the Roman perspective, before
turning once more to the indigenous population and their relationship with Rome over
four centuries.

The Iron Age before Rome

Recent Developments
The study of the Scottish Iron Age has been revolutionized over the past generation by
a series of factors. Increasing numbers of sites have reliable radiocarbon sequences,
especially with the advent of Bayesian statistical analysis, which has started to free the
chronology from a dependence on imports and stylistic parallels; examples include the
Traprain Law Environs Project (Haselgrove 2009), the debate over the dating of brochs
(Dockrill et  al. 2006) and patterns of hillfort development (the so-​called Hownam
sequence (Armit 1999)). The boom in developer-​funded archaeology, as elsewhere, has
180

180   Fraser Hunter

vastly increased the available evidence (e.g. Bradley 2007), much of it in areas such as
north-​east Scotland that had previously seen limited work, although there is still little
synthesis of the results. More generally, like the rest of the British Iron Age, it has ben-
efited from the application of an increasing range of theoretical views, not all necessarily
of lasting value but providing sets of alternative perspectives to challenge and develop
ways of thinking. There has been increasing concern over how exactly Iron Age societies
worked and whether typical models of chieftain-​based tribal societies are appropriate
(e.g. Hill 2006). This has been encouraged by the wider ‘Celts debate’, seeking to allow
archaeological evidence to speak for itself rather than incorporating it into interpreta-
tions derived from an assemblage of classical and insular historical sources about groups
termed ‘Celts’ by ancient or modern writers (e.g. Fitzpatrick 1996; Megaw and Megaw
1996, 1998; Sims-​Williams 1998; James 1999; Collis 2003). These debates continue and
what is sketched below will be an interim picture.
A more specifically Scottish debate has been over chronology and terminology.
Scholars working in the north and west of the country have pushed for a ‘long Iron Age’
from c. 800 bc–​ad 800 on Scandinavian models, rather than following the English Iron
Age chronology, which stops with Rome (Harding 2004: 3–​5). This has much to com-
mend it, not least in linking periods that are often studied separately. The concept of a
‘Roman Iron Age’—​common among scholars working beyond the frontier on the con-
tinent—​is a powerful and useful one, reflecting the relationships the area had with the
Roman world without treating this as a disciplinary boundary. For broader works on the
Scottish Iron Age, see Hingley (1992), Harding (2004), Armit (2016), and Hunter and
Carruthers (2012a).

Conceptualizing Iron Age Societies

By the first centuries bc/​ad, in lowland Scotland at least the picture is one of a densely
settled farming landscape, with woodland extensively cleared and remaining trees care-
fully husbanded (Tipping 1997; Tipping and Tisdall 2005). The settlement pattern and
social framework show marked regional variation. In the south of the country, while
hillforts may still dominate the landscape today, by the late pre-​Roman Iron Age the
age of these classic type-​fossils of the Iron Age was past—​there is very little evidence for
their construction at this time, with unenclosed clusters of houses overlying their ram-
parts. In the north and west, it seems that the main period of broch construction was
also over. These dramatic drystone round-houses, often tower-​like in form, remained as
focal points in the landscape, however, and many show long settlement sequences over a
thousand years or more (for example, Howe, Orkney, and Old Scatness, Shetland (Ballin
Smith 1994; Dockrill et al. 2010, 2015)).
The Geography of Ptolemy has been unduly influential in shaping our understanding
of Iron Age society. There have long been efforts to map his tribes onto the landscape
and the archaeology, efforts that have been contentious and largely unsuccessful (see,
e.g., the differing views of Richmond 1958: 131–1​46; Hind 1983; Mann and Breeze 1987;
 Beyond Hadrian’s Wall   181

Barrow 1989; Fraser 2005: 34–3​7). This should not surprise us: Ptolemy is a poor guide.
The sparsity of names for some areas must arise from an incomplete record, but there
are more general issues that pose wider concerns over its reliability. The information
is, at best, a snapshot of a complicated situation, moderated by unknown agents who
may have had little or no knowledge of local politics, trying to impose a Roman idea of
appropriate levels of social organization onto a situation that, from anthropological par-
allels elsewhere, was probably put in flux as a result of contact with empire. Such large-​
scale ‘peoples’ or ‘tribes’ are a poor basis for understanding the Scottish Iron Age; only
in extreme circumstances, such as the Roman invasion, is there evidence of large-​scale
collected action from the disparate units (Tacitus, Agricola 29). The emerging picture is
of small-​scale societies, with the extended household the key unit in most areas: single
houses or small clusters dominate our evidence. There is little evidence of larger-​scale
units or centres. The exception may be in south-​eastern Scotland, where larger commu-
nities may be represented in the clusters of houses over old hillforts.
This is not to see the late pre-​Roman Iron Age as a community of equals. Large parts
of the population may well be archaeologically invisible—​historical parallels from the
early Medieval period point to a critical difference between the free and the unfree, the
latter being almost impossible to detect, while among the free there would be varia-
tions according to ownership of and access to land (Fraser 2009: 34–​35). An increas-
ing concern with marking individual identities and differences is seen in the growing
use of personal ornaments, a phenomenon noted widely across Britain (Hill 1995: 85).
Emergent local and regional identities (for instance, the development of broch archi-
tecture in the Atlantic; the oblong forts of north-​east Scotland, or, later, the distinctive
late Iron Age metalwork of this area) suggest some broader-​scale patterns of community
from the later first millennium bc (Harding 2004: 85–​90, 108–​132; Hunter 2006a). But
the model does not seem to be the classic ‘Celtic chief and his tribe’: individual house-
holds or small communities seem broadly similar, with differences perhaps arising from
personal prowess or gained prestige and not necessarily persisting from one generation
to the next.
In southern and north-​eastern Scotland, a notable feature of the period is the increas-
ing use of decorated metalwork, normally termed Celtic or La Tène art. This reflects the
increasing concern with personal identity and status noted above, but may also be seen
as a direct response to the increasing power of Rome. Across much of Britain, the late
pre-​Roman and early Roman Iron Age saw an explosion of Celtic art (Garrow 2008: 30–​
36). These societies were coming under stress, with the presence or proximity of Rome
causing tensions and changes: Celtic art has been seen as a very visual, deliberately non-​
Roman response to this, perhaps emerging from the conflict situation itself, perhaps
from wider social tensions and shufflings caused by Rome’s proximity (Hunter 2007a:
289; Davis and Gwilt 2008).
The chronology of many changes in relation to the Roman conquest remains frus-
tratingly vague, as there are still too few secure radiocarbon-​dated sequences at this
critical time, but the move to an increasing concern with the individual starts in the
second or first century bc, while at least some of these styles of northern Celtic art
182

182   Fraser Hunter

pre-​date the Roman conquest, if only by a few decades (e.g. Fitzpatrick 1989). How
much impact did the Roman world have before conquest? There is none of the large-​
scale use of Roman imports or influence from Roman habits seen in south-​eastern
Britain in the period c. 50 bc–​ad 50 (e.g. Haselgrove 1984; Creighton 2000), nor
yet the equivalent of the privileged contacts to client kings as at Stanwick (North
Yorkshire (Fitts 1998)). But sparse evidence of early Mediterranean travellers to the
area indicates some pre-​military contacts (Breeze 2002: 11–​12), while evidence of links
to other areas of Iron Age Britain (e.g. Stevenson 1966, although within a diffusionist
paradigm) meant that at least some people in the north would have heard rumours of
change.

A Military History

The peoples of what is now Scotland had a very direct experience of the Roman mili-
tary. Motives behind Rome’s expansion have seen much discussion. Each of the various
moves into Scotland was imperially directed, and the desire to conquer the whole coun-
try was certainly on the agenda in the first century and, apparently, in the early third: the
second-​century invasion may have been more limited in its aims. Total conquest would
have made the occupation easier in the long run, with no ‘outside’ to continue resistance.
The reasons why this was not accomplished have long been debated and range from the
supposed ferocity of the local people and landscape, to the lack of Roman interest in an
area that did not fit strategic priorities, as well as the interplay with imperial politics.
The truth is likely to lie between these. Northern Britain was always expendable: on sev-
eral occasions, troops were withdrawn to deal with more pressing problems elsewhere
(such as Domitian’s Dacian wars or Antoninus Pius’ Moorish wars (Hanson 1987: 152;
Woolliscroft 2000; Breeze 2006a: 100–​101)). But the frontier was a restless one—​indeed,
the very presence of a frontier probably created tensions that led to conflict, and military
interventions were increasingly unsuccessful in dealing with the problem. This will be
considered later in the chapter.
Reconstruction of the history of Roman Scotland comes from a combination of
literary sources and detailed archaeological work. At times the latter has been rather
subservient to the former, but there is now more belief in (and argument over) the
interpretation of the archaeological remains in their own right, which is adding to and
questioning earlier models; some examples are mentioned later. It is worth noting,
however, that the overall picture of the occupations remains partial; for instance, the
distribution of Roman military sites is highly biased to eastern, arable-​producing areas
that provide good cropmark evidence (Figure 9.1). The distribution of sites in the wet-
ter, predominantly pastoral west of the country is clearly highly incomplete (e.g. Keppie
1990; Hunter and Carruthers 2012b: 10–​27). For general works on Roman Scotland see,
inter alia, Breeze (1982, 2006b), Maxwell (1989, 1998), Keppie (2004), and Hunter and
Carruthers (2012b).
(a)

(b)

Figure  9.1 Distribution of Roman military sites in Scotland. (a)  Flavian (c. ad 78–​86);
(b) Antonine (c. ad 139–​165).
Source: Reproduced by courtesy of Professor D. J. Breeze.
184

184   Fraser Hunter

The First Invasion: The Flavian Period

The governor Agricola is usually credited with leading the army into Scotland, thanks to
the testament of his son-​in-​law Tacitus, but recent years have seen considerable discus-
sion over this. Long-​known evidence for earlier contacts (Birley 1946) was brought back
to centre-​stage when dendrochronological dates revealed that the fort at Carlisle was
built in ad 72, a good six years before Agricola headed further north. Did the army really
stop here or did it move forward to conquer new land or soften the area up in the inter-
vening years? This has seen extensive debate, focusing on pre-​or early Flavian finds and
prolonged building sequences. While some of the marching camps in southern Scotland
could well relate to pre-​Agricolan activity, at the moment there is no clear evidence to
overturn Agricola as the prime mover (see papers in Breeze et al. 2009, with further ref-
erences; Fraser 2009: 17–​18).
This first invasion seems to have conquered southern Scotland quickly. The cam-
paign paused around the Tay, perhaps because the death of Vespasian, and then Titus,
required Agricola to await fresh orders (Hanson 1987: 107, 115), but Domitian ordered
the legions forward once more, with the north-​east of Scotland invaded and subdued
following a crushing victory at the battle of Mons Graupius. The quest for its site has
intrigued scholars for centuries, but, without firm evidence of a victory monument
or the discarded debris of battle, we are, in truth, no nearer to finding its location (see
Maxwell 1990; Fraser 2005).
The temporary camps built by troops on the move have long been a major piece of
evidence for those wishing to track the movements of Agricola and subsequent generals.
Recent synthesis has put this on a much firmer footing (Jones 2011: 99–​107). Likely cam-
paign lines can be identified north of the Forth—​although not all are closely dated—​
but, to the south, the camps resolutely refuse to fall into neat series. This reflects the
messier situation of large numbers of troops moving around southern Scotland over
two centuries in the business of subduing and controlling it (Jones 2011: 121–​123). A
striking development has been the extensive excavation of the Flavian camp at Kintore
(Aberdeenshire). Such camps were long dismissed as unworthy of large-​scale excava-
tion, but stripping of a substantial area uncovered latrine and rubbish pits, field ovens,
and a wide range of finds that breathe fresh light into our picture of the army on cam-
paign (Cook and Dunbar 2008). A notable finding was the evidence for multiple firings
of some ovens, implying this was far more than just an overnight stopping point, while
excavation of other sites has provided evidence of the army returning to an earlier camp
with little or no discernible reworking of its defences (Dunwell and Keppie 1995).
A network of control in the shape of forts and fortlets had already been placed over
southern Scotland (Tacitus, Agricola 20), and was extended north in the years imme-
diately before and after Mons Graupius. The linchpin was the legionary fortress at
Inchtuthil on the river Tay, well placed to control the newly conquered area and act
as a springboard for further campaigns (Pitts and St Joseph 1985; Woolliscroft and
Hoffmann 2006: 62–​72). It was supported by a range of auxiliary forts and by the Gask
Ridge—​a heavily supervised road variously seen as a frontier line, an element in a wider
 Beyond Hadrian’s Wall   185

frontier system, or a carefully controlled supply route (Hanson 1987: 153–​ 157;
Woolliscroft and Hoffmann 2006: 225–​34; Dobat 2009). Evidence for frequent struc-
tural changes on excavated sites indicates this was a rapidly evolving situation, with
garrisons shifted and installations modified as needs changed. Before it could settle
down—​indeed, before the Inchtuthil fortress was completed—​the decision was taken
to withdraw, as up to a quarter of the British garrison (Legio II Adiutrix and probably
some auxiliary units) was transferred to the continent as a result of the tense situation
on the lower Danube (Hanson 1987: 152). ‘Britain was conquered and immediately aban-
doned’ was Tacitus’ acid comment (Histories 1.2; trans. Ireland 2008: 92). Initially, parts
of southern Scotland were retained but, around ad 100, these forts were also abandoned
as the frontier fell back to the Tyne–​Solway line.

A Tale of Two Walls: The Efforts of Hadrian

and Antoninus Pius
The decision to withdraw from the north may have been related to wider politics, but there
were, clearly, local problems. Various sources indicate conflict during Hadrian’s reign—​
indeed, the building of Hadrian’s Wall may have exacerbated this, as it probably dissected
pre-​existing communities and restricted contact. This running sore of conflict probably
influenced Antoninus Pius’ decision to reconquer southern Scotland on his accession
and build another wall, along with wider motives such as his desire to show he was a wor-
thy commander in chief (Breeze 2006a: 12–​14). Both wall systems have seen enormous
amounts of detailed scholarship and, in recent years, innovative approaches and excit-
ing new results and debate over their planning, perceived role, and actual function (for
Hadrian’s Wall, see Hill 2004; Breeze 2006c; Hodgson 2009; Poulter 2009; Symonds and
Mason 2009; Graafstal 2012; for the Antonine Wall, Swan 1999; Robertson and Keppie 2001;
Breeze 2006a; Poulter 2009; Graafstaal et al. 2015). In the case of the Antonine Wall, not
only did its plan evolve but the building process was a protracted one (Gillam 1976a; Swan
1999; Breeze 2006a: 97–​102). Indeed, it may barely have been finished when the decision
was taken to abandon it. Evidence increasingly suggests that the Wall was under-​garrisoned
(Keppie 2009a), implying that the reconquest of Scotland had overstretched the army. The
issue of resistance remains under discussion. The old view of two separate Antonine occu-
pations, split by a revolt among the Brigantes, has been convincingly dismissed (Hodgson
1995), but there are signs of localized resistance, notably in Dumfriesshire (Hodgson 2009),
and this may have been another factor in the decision to return to Hadrian’s Wall.

A Restless Frontier
The sparse historical sources point to continuing troubles north of Hadrian’s Wall, lead-
ing eventually to the British expedition of Septimius Severus and Caracalla in ad 208–​11
(Breeze 1982: 128–​136; Birley 1988: 170–​187). These sources stress a military solution, but
186

186   Fraser Hunter

archaeological evidence can cast a different light on things. There is a strong cluster of sil-
ver coin hoards at this date in and around the troublesome areas, and work at the indige-
nous site of Birnie (Moray) has provided an excavated context: here, two such hoards came
to a local power centre in the 190s ad as part of a long-​running relationship to the Roman
world. This has been seen as a system of gifts, subsidies, or bribes, depending on one’s per-
spective (Hunter 2007b, 2007c, 2015a; Holmes 2014). The potential effects of this will be
considered in a later section. There is no further mention of conflict in the third century,
but the fourth century produced a litany of troubles and violence caused by a group now
termed ‘Picts’, with various military responses and expeditions to try to deal with them
(conveniently summarized in Maxwell 1987: 43). This issue too is considered later, but in
terms of Roman military archaeology it has left little or no recognizable trace, apart, argu-
ably, from a scatter of late Roman finds at earlier fort sites (Hunter and Carruthers 2012b:
25).

Frontier Lives

Until recently, research in the military zone could readily be caricatured as obsessed
with troop units, campaigns, and military strategy. A rare exception was the remark-
able excavation at Newstead, near Melrose (Borders) by James Curle. His report drew
together structures, finds, and environmental evidence to present a picture of ‘a Roman
frontier post and its people’, with the people a key part of the story (Curle 1911; Hunter
and Keppie 2012). Only recently has academic focus returned to such a broad under-
standing of life on the Roman frontier.
Newstead stressed the need to look beyond the fort walls in order to understand the
Roman period. Unusually large numbers of British fort sites are equipped with annexes:
there is an ongoing debate as to whether these should be seen as essentially military, for
instance, as supply depots and industrial zones, or as military vici for the camp followers
(Sommer 1984: 18–​22; 2006; Bailey 1994; Hanson 2007: 667–66​8). Only limited excava-
tions have yet been carried out in them. At Newstead, geophysical survey followed up by
targeted excavation produced a picture of a landscape rich in activity all round the fort,
with indications of different functions in different areas (e.g. Clarke 2000). Inveresk is
the only other Scottish fort to have seen substantial work beyond its walls (Figure 9.2).
It seems to have an unenclosed civil settlement rather than a defended annexe: exca-
vations have revealed a street grid, industrial activities, and indications of burials and
religious sites (e.g. Bishop 2002, 2004). Much of this work has been conducted through
developer-​funded archaeology; there has been little attempt yet to compare lifestyles
and activities in different excavated areas, but the potential is considerable.
Cemeteries and ritual sites, by contrast, are all but unknown. Only burial areas at
Camelon and Inveresk have seen some limited work (Breeze and Rich-​Gray 1980;
Chapman et al. 2012: 287–​288); otherwise the picture is dominated by a few stray finds
and sculpture (Collard and Hunter 2000). The distribution of altars beyond forts may
 Beyond Hadrian’s Wall   187

Figure 9.2  The Antonine military complex at Inveresk (East Lothian).

Source: Drawing by Alan Braby.

offer clues to the sacred landscape, but this was rather obscured by earlier scholarship’s
habit of gathering them under the fort’s umbrella (Collingwood and Wright 1965). The
enigmatic and long-​destroyed ‘Arthur’s O’on’ was probably a temple (Steer 1958), and
further examples must lurk around forts (as perhaps the now-​destroyed platform at
Easter Langlee, near Newstead (Steer 1966); a Mithraeum turned up unexpectedly in
development work near Inveresk in 2010 (Hunter et al. forthcoming)). Some evidence
of ritual deposits has been discussed. Evidence for on-​site deposition, for instance in
188

188   Fraser Hunter

the Newstead wells, remains hotly contested (Clarke and Jones 1994; cf. Manning 2006).
There were also deposits of valued material (such as bronze vessels) in wet locations,
some long hallowed by votive use, but whether these should be linked to beliefs of sol-
diers or habits of the locals remains for debate (Hunter 1997: 117–​118).
Evidence of craft activities taking place outside the fort and the field systems that sur-
round a number of them are reminders of the key role of supply and logistics. This has
traditionally been approached through the proxy of pottery, with evidence for a chang-
ing balance of imports and local production (e.g. Gillam 1976b; Hartley 1976; Breeze
1986; Swan 1988). An integrated, synthetic view of patterns of production and supply
would be highly desirable. In terms of the economic resources that the newly conquered
area could have provided, our evidence remains weak. ‘Invisible exports’ such as fur,
slaves, wild animals, and agricultural produce are the usual (and plausible) candidates
but hard to demonstrate, although stable isotope work on animal bones might ulti-
mately provide evidence of droving or similar extended chains of meat procurement
practices (Stallibrass 2009). Another key question is the role of metal resources. Within
the conquered area a number of sources of copper, lead, silver, and gold are known, but
with little direct evidence for their exploitation. Analysis of a lead ingot from Strageath
(Perth and Kinross) matched a Southern Uplands source (Hunter 2006b: 85), provid-
ing a tantalizing hint of Roman working of the Leadhills/​Wanlockhead area. Extensive
modern exploitation may have removed or concealed direct evidence, but proxy records
of smelting in peat cores (cf. Mighall 2003) offer a way forward.
The potential of the material culture and its distribution to explore different lifeways
on forts and their surrounding areas is largely untapped. Artefact distribution as a key
to social use of space has seen useful application on other frontiers (e.g. Allison 2006)
and could usefully be tackled with some of the Scottish data (see Giles 2012). Clarke
(2000) has begun to use the data from different parts of the Newstead complex to point
to different lifestyles reflected in differing access and approaches to material culture, and
there is enormous potential to explore this further, especially as developer-​funded work
offers ready samples from different areas of sites.
One way to investigate changing lifeways in this area is by studying the development
of hybrid forms of material culture, which is seen across the northern frontier zone
spanning Hadrian’s Wall. The development of Romano-​British brooches, often showing
so-​called Celtic design elements, has long been recognized (e.g. Collingwood 1930), and
the regional nature of such brooch distributions is increasingly clear thanks to data from
the Portable Antiquities Scheme (e.g. McIntosh 2011). Other object categories similarly
can offer insights into this, such as the wide range of copper-​alloy objects developing
from or influenced by indigenous traditions (for example, beaded torcs or horse harness
(Hunter 2008, 2010a, 2015b)), through the persistence of Iron Age types of worked bone
objects (such as long-​handled combs), to newly emerging forms such as glass bangles,
typical of the frontier zone in the Roman Iron Age. Standard works have focused on
typology and distribution (Kilbride-​Jones 1938; Stevenson 1956a, 1976; Price 1988), and
there is scope for much more consideration of how they were made and used across the
different frontier communities (e.g. Hoffmann 2003).
 Beyond Hadrian’s Wall   189

Such studies of material culture in context offer ways to study the different com-
munities of the frontier zone, how they perceived themselves and others and how they
shaped this—​a topic that has seen much interest in recent years on a broad level, but that
leaves lots of scope for detailed analysis (e.g. Goldsworthy and Haynes 1999; James 2002;
Mattingly 2006).

An Iron Age Perspective

The Romans did not march into an empty landscape. What impacts did the invasion and
proximity of the army have on the local populations? The answers are complex and var-
ied, but ongoing research is providing a better-​rounded picture.
The violence and resistance that feature in the historical sources are hard to trace
archaeologically, as is often the case. The much-​debated complex at Burnswark (Dumfries
and Galloway), where the Iron Age hillfort is threatened by two Roman camps, may pro-
vide one example of a violent climax to an act of resistance; opinion remains divided
(Jobey 1978; Campbell 2003; Keppie 2009b; Breeze 2011), though recent fieldwork seems
to support the siege theory (John Reid, pers. comm). Parallels elsewhere suggest we
should expect a range of responses to Rome, from support through acceptance to opposi-
tion, both between communities in different areas and within different segments of the
one community. This is hard to trace archaeologically. Our main evidence of interaction
comes from the discovery of Roman objects on indigenous sites, a topic with a long history
of study (e.g. Curle 1932; Robertson 1970; Macinnes 1984; Hunter 2001, 2007b; Bateson
and Holmes 2006, with previous references; Wilson 2010; Campbell 2012). These are not
without their problems. In contrast to areas beyond the German frontier, the general lack
of burials means that we are presented only with fragments from settlement sites, and their
value has often been overlooked because of this. There has been extended debate over the
significance of such fragments: did they come to the sites whole, or as tokens or talismans,
when did they arrive, and how long were they used for? (See Hunter 2007b: 11, 91 n. 3, for
references.) While there is clear evidence that some objects had an extended life, especially
samian sherds (which were prioritized for curation and reuse (Campbell 2012)), patterns
visible in the data (explored in the next section) make sense when the objects are consid-
ered as vessels rather than sherds, suggesting this is how they arrived and were initially
used. Improved understanding of the social roles of these objects will come from a broad,
life-​cycle approach (cf. Kopytoff 1986), which considers the object’s use-​lives (as whole
item and fragment) and deposition in the context of local material culture; the fruits of
such work are beginning to appear (Wallace 2006; Campbell 2012, 2015; Hunter 2013).
It is difficult to assess whether finds arrived predominantly when southern Scotland
was directly occupied (as Erdrich et al. 2000 argue) or whether there was a more con-
stant flow from the province lying to the south. This key question of whether and when
the frontier was open or closed for contact is currently hard to answer. We are caught
in a vicious circle, where the items that can be most closely dated (coins and samian)
190

190   Fraser Hunter

are the ones suspected to have long use-​lives, and thus the production date is a poor
reflection of the deposition date. As the data set grows and more independent dates are
obtained, this should become easier to analyse. This issue depends also on the mecha-
nisms and motives behind contacts, whether the results of violence (such as booty) or
peaceful contact (for example, trade, diplomacy, military service). The former seems
unlikely as a widespread process, especially when the material is contrasted to the much
more random and metal-​focused debris increasingly found beyond the German fron-
tier, where it is interpreted as loot for reuse (e.g. Becker 2006).

Selection Processes
Research has sought ways to alleviate the taphonomic bias imposed by this settlement
debris by looking at techniques such as presence/​absence data (rather than absolute
numbers) or the range of finds as an indication of variations in access (Hunter 2001).
This has revealed a strong selectivity in the material coming into indigenous hands, with
a marked preference for jewellery (especially brooches) and for feasting and drinking
gear—​in other words, material that could be used in local social displays.
Clear signs of differential access to this material, both regionally and socially, pro-
vide clues to the social networks of the period (Figure 9.3). In southern Scotland at least,
Roman material was not absolutely rare, but more is found on sites that are richer in other
senses as well—​for instance, in access to restricted craft skills such as bronze-​working, or
the presence of decorated metalwork or other imported goods. It suggests that Roman
contact was mediated through higher-​status individuals and groups. This is most marked
in south-​east and east-​central Scotland, suggesting the presence or development of more
hierarchical or competitive societies in these areas. Elsewhere there is a much more
equal pattern of access, matching the picture of relatively small-​scale societies already
discussed. Roman material was valued and used in regions far beyond the area of direct
occupation, reaching out to the Northern and Western Isles; indeed, small quantities may
have had disproportionate impact in areas of more restricted access, but this has seen lit-
tle detailed study.

Focal Sites
One set of sites that are particularly rich in Roman finds are the ‘lowland brochs’
(Macinnes 1984). The broch, a drystone tower-​like round-house, was typical of the
Atlantic north and west, but a small number are found in central and southern Scotland.
Most southern excavated examples are of first–​second-​century ad date, and many were
rich in finds, including a wealth of Roman material culture. The exotic architectural
form was a clear mark of social difference, and this is confirmed in the finds, with prefer-
ential access to a wide range of Roman objects as well as plentiful evidence of restricted
craft skills and rare objects. Most of these brochs were not new foundations, developing
(a) (b)

0 100 km 0 100 km
0 100 miles 0 100 miles

Birnie

Traprain Law

(c)

0 100 km
0 100 miles

Figure 9.3  Distribution of Roman finds from non-​Roman sites north of Hadrian’s Wall. (a) overall dis-
tribution; (b) middle Roman Iron Age c. ad 160–​250 (the sites of Birnie and Traprain Law, mentioned in
the text, are marked: *represents find-spots of denarius hoards); (c) late Roman Iron Age c. ad 250–​400;
*marks the major centres of Traprain Law, Edinburgh Castle, Eildon Hill, and Dumbarton Rock.
Source: © Fraser Hunter.
192

192   Fraser Hunter

on existing settlement sites. However, modern excavations suggest the brochs were a
boom-​and-​bust phenomenon with little sign of extended use, and a number show evi-
dence of deliberate destruction or demolition. This suggests a rather unstable situation
where the influence of competing power-​hungry groups could rise and fall rapidly; per-
haps because of changing relationships to the fickle Roman world, perhaps from chang-
ing local power politics and varying views on whether Rome was a friend or foe.
One site stands out for its long-​ term, sustained relationship with the Roman
world: Traprain Law in East Lothian (Jobey 1976; Hunter 2009). The defences of this
mighty hillfort were denuded by the Roman period, but the site was densely inhabited—​
indeed, current evidence suggests that it was a Roman-​period ‘boom town’ with little
immediately pre-​Roman settlement, and its rise to prominence was strongly connected
to the links its occupants built to the Roman world. These long-​lasting connections
brought a wealth of Roman material culture to the site from the first to the fifth century
ad. Along with the very light Roman military presence in the area, it suggests this area
acted as a client kingdom over a long period. Traprain stands alone at the moment, but
there may be other such sites, especially in southern Scotland, yet to be recognized.

The Impact of Frontiers

The differing impact of the Roman occupation in different areas is ripe for analysis, both
in comparison to the continental frontier and within Britain. For instance, it is striking
that the zone in front of Hadrian’s Wall, in north Northumberland, lacks indigenous
sites rich in Roman finds, in contrast to south-​east Scotland. It suggests that the occu-
pation had a dire effect on local societies in this area, especially the traditional power
structures, with the army sitting at the top of the local social tree (e.g. Hodgson et al.
2012: 211–220). In contrast, further from Hadrian’s Wall (or around the Antonine Wall,
according to period), more Roman material was more widely spread, with clearer indi-
cations of social hierarchies of access to it. The clearest examples are the lowland brochs,
suggesting groups growing powerful on the back of Rome. As noted, these may have
been a short-​lived phenomenon, but other sites show continuing, rich, Roman links
in places beyond Hadrian’s Wall, especially in south-​east and north-​east Scotland. A
regional picture of this varying impact of empire remains to be developed.

Diplomacy and Interference

Access to and use of Roman material was not a neutral phenomenon; it deeply affected
the people and societies involved. The precise sequence of cause and effect is rarely clear,
but a series of social phenomena of the Roman Iron Age can plausibly be traced to the
agency of Rome. Some may have been incidental byproducts of contact, but increasingly
it seems that the Roman world was deliberately creating and manipulating links with
local societies for political purposes, as it did on other frontiers. For north-​west Germany,
 Beyond Hadrian’s Wall   193

Erdrich (2000) has argued that the Romans sought to control and foment unrest within
local society by building some groups up at the expense of others, or by giving and then
withdrawing favours. There are signs of similar patterns in the Scottish record. The best
evidence comes from the mid–​late Roman Iron Age (c. ad 160–​400+; figure 9 b–c). In
the mid-​Roman Iron Age (c. ad 160–​250), the distribution of silver coin hoards concen-
trated strongly in central and north-​eastern Scotland has been argued to reflect delib-
erate targeting of troublesome tribes and their neighbours (Hunter 2007b: 23–2​7, 2015a;
Holmes 2014). The phenomenon is represented historically in references to unrest, espe-
cially among the Caledonians and Maeatae (e.g. Epitome of Dio Cassius LXXV, 5, 4; trans.
Ireland 2008: 112–​113). Subsidies of denarii may have been the main diplomatic vehicle,
but a range of other rich finds was targeted at the same area. Precise correlation with his-
torical events is hazardous, but, in general terms, these can be seen as attempts to solve
frontier unrest through diplomatic efforts; the subsequent Severan invasion suggests that
this was not entirely successful, although a few denarius hoards in central and southern
Scotland post-​dating the invasions indicate that the policy continued, targeted now per-
haps at a ‘buffer zone’ south of the unfriendly area (Holmes and Hunter 1997).
There are striking changes in patterns of access to Roman material in the succeeding
centuries. The north-​east, which had been highly favoured, was essentially cut off from
Roman contact in the later third and fourth centuries. Other areas, especially south-​east
Scotland, saw continuing access, but now much more focused on a small number of
sites. This perhaps suggests that social control over this material was much tighter—​and,
arguably, that a small number of groups were becoming pre-​eminent (Hunter 2007b:
32–3​6; 2010b; 2014). The effects of this are much debated. In the north-​east it is gener-
ally agreed that Roman interference was ultimately responsible for the emergence of the
‘Picts’, but opinion is divided on whether this represents a coalescence of existing groups
in the face of Roman threat (Mann 1974; Fraser 2009: 44–​49, 54–​61), or a more fun-
damental social collapse and the rise of altogether more hostile groups within society
(Hunter 2007b: 50–​53).
Roman contacts ran beyond the formal ‘end’ of Roman Britain. Traprain Law’s links
with Rome flourished in the late Roman period, and the spectacular hoard of fifth-​
century ‘Hacksilber’ shows the value of material to which its inhabitants had access
(Curle 1923). Earlier views of this as loot are now questioned, and the phenomenon is
better related either to diplomatic gifts or to the payment of mercenaries serving in the
late Roman army (Painter 2013).

The Long Term

What was the lasting effect of the Roman world on Scotland? Some have seen the period
as little more than ‘brief interludes’ (e.g. Hanson 2003), while others have argued for
much more fundamental effects (Fraser 2009: 116–​117). The latter seems closer to what
can be observed at a number of levels.
194

194   Fraser Hunter

There were practical benefits, in terms of raw materials: much of the non-​ferrous
metalwork of the Roman Iron Age reused Roman metal, for instance, while Roman
‘Hacksilber’ is argued to be the ultimate source of the wealth of early Medieval Scotland
(Stevenson 1956b; Youngs 2013). Something of the social effects have been highlighted
above: the emergence of the ‘Picts’ and, arguably, the development of larger-​scale social
units in southern Scotland, can both be connected with Roman interference and con-
tacts. The precise mechanisms merit more attention, but it seems that the ‘social shape’
of both Roman Iron Age and early Medieval societies was fundamentally affected by the
Roman world, whether by accident or by design.
An intriguing issue, but one that has received little attention, is a long-​term perspec-
tive on forts and their settings. How far did they influence the landscape in later cen-
turies? The temporary camp at Kintore seems to have disrupted landscape use, as the
interior was apparently left empty until the seventh century (Cook and Dunbar 2008:
354–​356). Surprisingly few fort sites have produced clear evidence for reuse before the
high Medieval period, although the frequent instances of Medieval churches positioned
on forts may have early Medieval roots (Hunter and Carruthers 2012b: 42; see also
Maldonado 2015). The topic merits more study.
Rome also became a model for the ‘barbarian world’ in the post-​Roman period, in
Scotland as in much of the rest of early Medieval Europe. This was tied up with the adop-
tion of Christianity, but Latin names of local leaders, the use of Latin in inscriptions, and
perhaps a continuing taste for imported Mediterranean goods in some areas point to a
lasting impact (Mann 1974; Campbell 2007: 134–​135; Fraser 2013).

Conclusions

The study of Roman Scotland has moved a long way from its earlier military focus. Many
aspects of older agendas remain valid, such as a more detailed understanding of the dispo-
sition of military sites, but recent work has begun to look much more broadly at the period.
Rome must be seen not just as a military force but in terms of supply needs, impact on the
landscape, and the different segments that made up a military community. The locals are
not just opponents or background, but a key element in studying the frontier, affected by
and affecting the military presence. The episodic nature of direct Roman occupation of
Scotland, and the restricted geographical area, mean that there is tremendous potential for
exploring complex issues of frontier control, supply, impact, and interaction.

Acknowledgements

Co-​chairing the development of a research framework into Roman Scotland, an ini-

tiative of the Society of Antiquaries of Scotland, was an opportunity that forced me to
 Beyond Hadrian’s Wall   195

think more broadly about the topic, and I am grateful to the colleagues who served on
the panel and others who commented on the document for a fascinating and thought-​
provoking process (see Hunter and Carruthers 2012b). It has influenced the contents of
this chapter. Earlier versions benefited greatly from the comments of David Breeze and
David Clarke.

References
Primary Sources
Tacitus, Agricola, trans. A. R. Birley. Oxford: Oxford University Press, 1999.
Tacitus, Histories, trans. K. Wellesley. Harmondsworth: Penguin, 1964.

Secondary Sources
Allison, P. (2006). ‘Mapping for Gender: Interpreting Artefact Distribution in Roman Military
Forts in Germany’, Archaeological Dialogues 13/​1: 1–​48.
Armit, I. (1999). ‘Life after Hownam: The Iron Age in South-​East Scotland’, in B. Bevan (ed.),
Northern Exposure: Interpretative Devolution and the Iron Ages in Britain. Leicester: School
of Archaeological Studies, University of Leicester, 65–​79.
Armit, I. (2016). Celtic Scotland: Iron Age Scotland in its European Context. 3rd edn. Edinburgh:
Birlinn.
Bailey, G. B. (1994). ‘The Provision of Fort-​Annexes on the Antonine Wall’, Proceedings of the
Society of Antiquaries of Scotland, 124: 299–​314.
Ballin Smith, B. (1994). Howe:  Four Millennia of Orkney Prehistory. Edinburgh:  Society of
Antiquaries of Scotland.
Barrow, G. W. S. (1989). ‘The Tribes of North Britain Revisited’, Proceedings of the Society of
Antiquaries of Scotland, 119: 161–​163.
Bateson, J. D., and Holmes, N. M. McQ. (2006). ‘Roman and Medieval Coins Found in
Scotland, 2001–​2005’, Proceedings of the Society of Antiquaries of Scotland, 136: 161–​198.
Becker, M. (2006). ‘Handelsgut, Sold und Beute’, in J. Stadler, B. Hanemann, M. Kolb, and R.
Petrovszky (eds), Geraubt und im Rhein versunken: Der Barbarenschatz. Speyer: Historisches
Museum der Pfalz, 219–​223.
Birley, A. (1988). The African Emperor: Septimius Severus. London: Batsford.
Birley, E. (1946). ‘Britain under the Flavians: Agricola and his Predecessors’, Durham University
Journal (June), 79–​84.
Bishop, M. C. (2002) (ed.). Roman Inveresk: Past, Present and Future. Duns: Armatura Press.
Bishop, M. C. (2004). Inveresk Gate: Excavations in the Roman Civil Settlement at Inveresk, East
Lothian, 1996–​2000. Edinburgh: STAR.
Bradley, R. (2007). The Prehistory of Britain and Ireland. Cambridge:  Cambridge
University Press.
Breeze, D. J. (1982). The Northern Frontiers of Roman Britain. London: Batsford.
Breeze, D. J. (1986). ‘The Manufacture of Pottery in Roman Scotland’, Proceedings of the Society
of Antiquaries of Scotland, 116: 185–​189.
Breeze, D. J. (2002). ‘The Ancient Geography of Scotland’, in B. Ballin Smith and I. Banks (eds),
In the Shadow of the Brochs. Stroud: Tempus, 11–​14.
196

196   Fraser Hunter

Breeze, D. J. (2006a). The Antonine Wall. Edinburgh: John Donald.

Breeze, D. J. (2006b). Roman Scotland: Frontier Country. 2nd edn. London: Batsford.
Breeze, D. J. (2006c). J. Collingwood Bruce’s Handbook to the Roman Wall. 14th edn. Newcastle
upon Tyne: Society of Antiquaries of Newcastle upon Tyne.
Breeze, D. J. (2011). ‘Burnswark: Roman Siege or Army Training Ground?’, Archaeological
Journal, 168: 166–​180.
Breeze, D. J., and Rich-​Gray, D. (1980). ‘“Fire Pits” at Camelon, Stirlingshire’, Proceedings of the
Society of Antiquaries of Scotland, 110 [1978–​80], 513–​517.
Breeze, D. J., Thoms, L. M., and Hall, D. W. (2009) (eds). First Contact: Rome and Northern
Britain. Perth: Tayside and Fife Archaeological Committee.
Campbell, D. B. (2003). ‘The Roman Siege of Burnswark’, Britannia, 34: 19–​33.
Campbell, E. (2007). Continental and Mediterranean Imports to Atlantic Britain and Ireland, c.
ad 400–​800. York: Council of British Archaeology Research Report 157.
Campbell, L. (2012). ‘Modifying Material: Social Biographies of Roman Material Culture’, in
A. Kyle and B. Jervis (eds), Make Do and Mend: The Archaeologies of Compromise. Oxford:
Archaeopress/​BAR International Series 2408, 11–​25.
Campbell, L. (2015). ‘Interfacing with the Empire: Materiality Matters’, in D. B. Breeze, R. H.
Jones, and I. A. Oltean (eds), Understanding Roman Frontiers: A Celebration for Professor Bill
Hanson. Edinburgh: Birlinn, 182–195.
Chapman, E. M., Hunter, F., Booth, P., Wilson, P., Worrell, S., and Tomlin, R. S. O. (2012).
‘Roman Britain in 2011’, Britannia, 43: 271–​421.
Clarke, S. (2000). ‘In Search of a Different Roman Period: The Finds Assemblage at the
Newstead Military Complex’, in G. Fincham, G. Harrison, R. Holland, and L. Revell (eds),
TRAC 99: Proceedings of the 9th Theoretical Roman Archaeology Conference, Durham, April
1999. Oxford: Oxbow Books, 22–​29.
Clarke, S., and Jones, R. F. J. (1994). ‘The Newstead Pits’, Journal of Roman Military Equipment
Studies, 5: 109–​124.
Collard, M., and Hunter, F. (2000). ‘A Roman Burial from Cramond (Edinburgh) Rediscovered’,
Proceedings of the Society of the Antiquaries of Scotland, 130: 525–​535.
Collingwood, R. G. (1930). ‘Romano-​Celtic Art in Northumbria’, Archaeologia, 80: 37–​58.
Collingwood, R. G., and Wright, R. P. (1965). The Roman Inscriptions of Britain. I: Inscriptions
on Stone. Oxford: Oxford University Press.
Collis, J. (2003). The Celts: Origins, Myths and Inventions. Stroud: Sutton.
Cook, M., and Dunbar, L. (2008). Rituals, Roundhouses and Romans: Excavations at Kintore,
Aberdeenshire 2000–​2006. Volume 1: Forest Road. Edinburgh: STAR.
Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge
University Press.
Curle, A. O. (1923). The Treasure of Traprain. Glasgow: Maclehose.
Curle, J. (1911). A Roman Frontier Post and its People: The Fort of Newstead in the Parish of
Melrose. Glasgow: Maclehose.
Curle, J. (1932). ‘An Inventory of Objects of Roman and Provincial Roman Origin Found on
Sites in Scotland not Definitely Associated with Roman Constructions’, Proceedings of the
Society of Antiquaries of Scotland, 66 [1931–​2], 277–​397.
Davis, M., and Gwilt, A. (2008). ‘Material, Style and Identity in First Century AD Metalwork,
with Particular Reference to the Seven Sisters Hoard’, in D. Garrow, C. Gosden, and J. D. Hill
(eds), Rethinking Celtic Art. Oxford: Oxbow Books, 146–​184.
 Beyond Hadrian’s Wall   197

Dobat, E. (2009). ‘The Gask “system” in Perthshire: The First Artificial Frontier Line of the
Roman Empire’, in D. J. Breeze, L. M. Thoms, and D. W. Hall (eds), First Contact: Rome and
Northern Britain. Perth: Tayside and Fife Archaeological Committee, 39–​48.
Dockrill, S. J., Outram, Z., and Batt, C. M. (2006). ‘Time and Place: A New Chronology for the
Origin of the Broch Based on the Scientific Dating Programme at the Old Scatness Broch,
Shetland’, Proceedings of the Society of Antiquaries of Scotland, 136: 89–​110.
Dockrill, S. J., Bond, J. M., Turner, V. E., Brown, L. D., Bashford, D. J., Cussans, J. E., and
Nicholson, R. A. (2010). Excavations at Old Scatness, Shetland. Volume 1: The Pictish Village
and Viking Settlement. Lerwick: Shetland Heritage Publications.
Dockrill, S. J., Bond, J. M., Turner, V. E., Brown, L. D., Bashford, D. J., Cussans, J. E., and
Nicholson, R. A. (2015). Excavations at Old Scatness, Shetland. Volume 2: The Broch and Iron
Age Village. Lerwick: Shetland Heritage Publications.
Dunwell, A. J., and Keppie, L. J.  F. (1995). ‘The Roman Temporary Camp at Dunning,
Perthshire: Evidence from two Recent Excavations’, Britannia, 26: 51–​62.
Erdrich, M. (2000). Rom und die Barbaren: Das Verhältnis zwischen dem Imperium Romanum
und den Germanischen Stämmen vor seiner Nordwestgrenze von der späten römischen
Republik bis zum Galischen Sonderreich. Mainz: Römisch-​Germanische Kommission.
Erdrich, M., Giannotta, K. M., and Hanson, W. S. (2000). ‘Traprain Law:  Native and
Roman on the Northern Frontier’, Proceedings of the Society of Antiquaries of Scotland,
130: 441–​456.
Fitts, L. (1998). ‘Inde opes et rerum secundarum luxus, Stanwick and Melsonby’, in J. Bird (ed.),
Form and Fabric: Studies in Rome’s Material Past in Honour of B. R. Hartley. Oxford: Oxbow
Books, 1–​7.
Fitzpatrick, A. P. (1989). ‘The Submission of the Orkney Islands to Claudius: New Evidence?’,
Scottish Archaeological Review, 6: 24–​33.
Fitzpatrick, A. P. (1996). ‘“Celtic” Iron Age Europe: The Theoretical Basis’, in P. Graves-​Brown,
S. Jones, and C. Gamble (eds), Cultural Identity and Archaeology: The Construction of
European Communities. London: Routledge, 238–​255.
Fraser, J. E. (2005). The Roman Conquest of Scotland: The Battle of Mons Graupius ad 84.
Stroud: Tempus.
Fraser, J. E. (2009). From Caledonia to Pictland:  Scotland to 795. Edinburgh:  Edinburgh
University Press.
Fraser, J. E. (2013). ‘St Patrick and Barbarian Northern Britain in the Fifth Century’, in
F. Hunter and K. Painter (eds), Late Roman Silver:  The Traprain Treasure in Context.
Edinburgh: Society of Antiquaries of Scotland, 15–​27.
Garrow, D. (2008). ‘The Time and Space of Celtic Art:  Interrogating the “Technologies of
Enchantment” Database’, in D. Garrow, C. Gosden, and J. D. Hill (eds), Rethinking Celtic Art.
Oxford: Oxbow Books, 15–​39.
Giles, R. D. (2012). Roman Soldiers and the Roman Army:  A  Study of Military Life from
Archaeological Remains. Oxford: Archaeopress/​BAR British Series 562.
Gillam, J. P. (1976a). ‘Possible Changes in Plan in the Course of Construction of the Antonine
Wall’, Scottish Archaeological Forum, 7: 51–​56.
Gillam, J. P. (1976b). ‘Coarse Fumed Ware in North Britain and Beyond’, Glasgow Archaeological
Journal, 4: 57–​80.
Goldsworthy, A., and Haynes, I. (1999) (eds). The Roman Army as a Community. Portsmouth,
RI: Journal of Roman Archaeology Supplementary Series 34.
198

198   Fraser Hunter

Graafstal, E. P. (2012). ‘Hadrian’s Haste: A Priority Programme for the Wall’, Archaeologia
Aeliana, 5th ser. 41: 123–​184.
Graafstaal, E., Breeze, D. J., Jones, R. H. and Symonds, M. F. A. (2015). ‘Sacred Cows in the
Landscape: Rethinking the Planning of the Antonine Wall’, in D. J. Breeze, R. H. Jones, and
I. A. Oltean (eds), Understanding Roman Frontiers: A Celebration for Professor Bill Hanson.
Edinburgh: Birlinn, 54–69.
Hanson, W. S. (1987). Agricola and the Conquest of the North. London: Batsford.
Hanson, W. S. (2003). ‘The Roman Presence: Brief Interludes’, in K. J. Edwards and I. B. M.
Ralston (eds), Scotland after the Ice Age: Environment, Archaeology and History 8000 bc–​ad
1000. Edinburgh: Edinburgh University Press, 195–​216.
Hanson, W. S. (2007). Elginhaugh:  A  Flavian Fort and its Annexe. London:  Society for the
Promotion of Roman Studies.
Harding, D. W. (2004). The Iron Age in Northern Britain. London: Routledge.
Hartley, K. F. (1976). ‘Were Mortaria Made in Roman Scotland?’, Glasgow Archaeological
Journal, 4: 81–​89.
Haselgrove, C. (1984). ‘“Romanization” before the Conquest: Gaulish Precedents and British
Consequences’, in T. F. C. Blagg and A. C. King (eds), Military and Civilian in Roman Britain.
Oxford: ​BAR British Series 136, 5–​63.
Haselgrove, C. (2009). The Traprain Law Environs Project: Fieldwork and Excavations 2000–​
2004. Edinburgh: Society of Antiquaries of Scotland.
Hill, J. D. (1995). ‘The Pre-​Roman Iron Age in Britain and Ireland (ca. 800 bc to ad 100): An
Overview’, Journal of World Prehistory, 9/​1: 47–​98.
Hill, J. D. (2006). ‘Are we any Closer to Understanding how Iron Age Societies Worked (or
did not Work)?’, in C. Haselgrove (ed.), Celtes et Gaulois, l’archéologie face à l’histoire 4: Les
Mutations de la fin de l’âge du Fer. Glux-​en-​Glenne: Bibracte, 169–​179.
Hill, P. R. (2004). The Construction of Hadrian’s Wall. Oxford: Archaeopress/​BAR British Series
375.
Hind, J. G.  F. (1983). ‘Caledonia and its Occupation under the Flavians’, Proceedings of the
Society of Antiquaries of Scotland, 113: 373–​378.
Hingley, R. (1992). ‘Society in Scotland from 700 bc to ad 200’, Proceedings of the Society of
Antiquaries of Scotland, 122: 7–​53.
Hodgson, N. (1995). ‘Were there Two Antonine Occupations of Scotland?’, Britannia,
26: 29–​49.
Hodgson, N. (2009). ‘The Abandonment of Antonine Scotland:  Its Date and Causes’, in
W. S. Hanson (ed.), The Army and Frontiers of Rome. Portsmouth, RI: Journal of Roman
Archaeology Supplementary Series 74, 185–​193.
Hodgson, N. (2009) (ed.). Hadrian’s Wall 1999–​2009. Kendal: Cumberland and Westmorland
Antiquarian and Archaeological Society/​Society of Antiquaries of Newcastle upon Tyne.
Hodgson, N., McKelvey, J., and Muncaster, W. (2012). The Iron Age on the Northumbrian
Coastal Plain: Excavations in Advance of Development 2002–2010. Newcastle: TWM
Archaeology/Arbeia Society.
Hoffmann, B. (2003). ‘Roman Glass from Newstead and Vindolanda’, Annales du 15e Congrès
de l’Association Internationale pour l’Histoire du Verre, 2001New York 2001. ​Nottingham:
Association Internationale pour l’Histoire de Verre, 41–​44.
Holmes, N. M. MCQ. (2014). ‘The Synton and Kippilaw Denarius Hoards: Further Numismatic
Evidence for Late Antonine and Severan Scotland,’ Proceedings of the Society of Antiquaries
of Scotland, 144, 138–168.
 Beyond Hadrian’s Wall   199

Holmes, N., and Hunter, F. (1997). ‘Edston, Peebles-​shire’, in R. Bland and J. Orna-​Ornstein
(eds), Coin Hoards from Roman Britain X. London:  Trustees of the British Museum,
149–​168.
Hunter, F. (1997). ‘Iron Age Hoarding in Scotland and Northern England’, in A. Gwilt and C.
Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow Books, 108–​133.
Hunter, F. (2001). ‘Roman and Native in Scotland:  New Approaches’, Journal of Roman
Archaeology, 14: 289–​309.
Hunter, F. (2006a). ‘New Light on Iron Age Massive Armlets’, Proceedings of the Society of
Antiquaries of Scotland, 136: 135–​160.
Hunter, F. (2006b). ‘Recent Finds from Strageath Roman Fort’, Tayside and Fife Archaeological
Journal, 12: 81–​88.
Hunter, F. (2007a). ‘Artefacts, Regions and Identities in the Northern British Iron Age’, in C. C.
Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow
Books, 286–​296.
Hunter, F. (2007b). Beyond the Edge of the Empire: C ​ aledonians, Picts and Romans. Rosemarkie:
Groam House Museum.
Hunter, F. (2007c). ‘Silver for the Barbarians: Interpreting Denarii Hoards in North Britain and
Beyond’, in R. Hingley and S. Willis (eds), Roman Finds: Context and Theory. Oxford: Oxbow
Books, 214–​224.
Hunter, F. (2008). ‘Celtic Art in Roman Britain’, in D. Garrow, C. Gosden, and J. D. Hill (eds),
Rethinking Celtic Art. Oxford: Oxbow Books, 129–​145.
Hunter, F. (2009). ‘Traprain Law and the Roman World’, in W. S. Hanson (ed.), The Army and
Frontiers of Rome. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 74,
225–​240.
Hunter, F. (2010a). ‘Changing Objects in Changing Worlds:  Dragonesque Brooches
and Beaded Torcs’, in S. Worrell, G. Egan, J. Naylor, K. Leahy, and M. Lewis (eds), A
Decade of Discovery:  Proceedings of the Portable Antiquities Scheme Conference 2007.
Oxford: Archaeopress/​BAR British Series 520, 91–​107.
Hunter, F. (2010b). ‘Beyond the Frontier: Interpreting Late Roman Iron Age Indigenous
and Imported Material Culture’, in R. Collins and L. Allason-​Jones (eds), Finds from the
Frontier: Material Culture in the 4th-​5th Centuries. York: Council for British Archaeology,
96–​109.
Hunter, F. (2013). ‘The Lives of Roman Objects beyond the Frontier’, in P. S. Wells (ed.), Rome
Beyond its Frontiers: Imports, Attitudes and Practices. Portsmouth, RI: Journal of Roman
Archaeology Supplementary Series 94, 15–​28.
Hunter, F. (2014). ‘Looking over the Wall: The Late and Post-Roman Iron Age North of
Hadrian’s Wall,’ in F. K. Haarer (ed.), AD 430: The History and Archaeology of Late and Post-
Roman Britain. London: Roman Society, 206–215.
Hunter, F. (2015a). ‘The Lure of Silver: Denarius Hoards and Relations across the Frontier,’ in
D. J. Breeze, R. H. Jones and I. A. Oltean (eds), Understanding Roman Frontiers: A Celebration
for Professor Bill Hanson. Edinburgh: Birlinn, 251–269.
Hunter, F. (2015b) ‘Interpreting Celtic Art on the Roman Frontier—The Development of
a Frontier Culture in Britain?’, in L. Vagalinski and N. Sharankov (eds), Proceedings of the
22nd International Congress of Roman Frontier Studies, Ruse, Bulgaria, September 2012. Sofia:
National Archaeological Institute and Museum, 721–727.
Hunter, F., and Carruthers, M. (2012a) (eds). ScARF Iron Age Panel Report <http://​www.scot-
tishheritagehub.com/​content/​iron-​age-​panel-​report> (accessed 21 January 2013).
200

200   Fraser Hunter

Hunter, F., and Carruthers, M. (2012b) (eds). ScARF Roman Scotland Panel Report <http://​
www.scottishheritagehub.com/​content/​roman-​scotland-​panel-​report> (accessed 21
January 2013).
Hunter, F., and Keppie, L. (2012) (eds). A Roman Frontier Post and its People: Newstead 1911–​
2011. Edinburgh: National Museums Scotland.
Hunter, F., Henig, M., Sauer, E., and Goodner, J. (forthcoming). ‘Mithras in Scotland: A
Mithraeum at Inveresk (East Lothian)’, Britannia 47.
Ireland, S. (2008). Roman Britain: A Sourcebook. 3rd edn. London: Routledge.
James, S. (1999). The Atlantic Celts:  Ancient People or Modern Invention? London:  British
Museum.
James, S. (2002). ‘Writing the Legions: The Past, Present and Future of Roman Military Studies
in Britain’, Archaeological Journal, 159: 1–​58.
Jobey, G. (1976). ‘Traprain Law:  A  Summary’, in D. W. Harding (ed.), Hillforts.
London: Academic Press, 191–​204.
Jobey, G. (1978). ‘Burnswark Hill, Dumfriesshire’, Transactions of the Dumfries and Galloway
Natural History and Antiquarian Society, 53 [1977–​8], 57–​104.
Jones, R. H. (2011). Roman Camps in Scotland. Edinburgh: Society of Antiquaries of Scotland.
Keppie, L. (1990). ‘The Romans in Southern Scotland:  Future Discoveries’, Glasgow
Archaeological Journal, 16 [1989–​90], 1–​27.
Keppie, L. (2004). The Legacy of Rome: Scotland’s Roman Remains. Edinburgh: John Donald.
Keppie, L. (2009a). ‘The Garrison of the Antonine Wall: Endangered Species or Disappearing
Asset?’, in A. Morillo, N. Hanel, and E. Martín (eds), Limes XX: XXth International Congress
of Roman Frontier Studies. Madrid:  Consejo Superior de Investigaciones Científicas,
1135–​1145.
Keppie, L. (2009b). ‘Burnswark Hill:  Native Space and Roman Invaders’, in W. S. Hanson
(ed.), The Army and Frontiers of Rome. Portsmouth, RI:  Journal of Roman Archaeology
Supplementary Series 74, 241–​252.
Kilbride-​Jones, H. E. (1938). ‘Glass Armlets in Britain’, Proceedings of the Society of Antiquaries
of Scotland, 72 [1937–​8], 366–​395.
Kopytoff, I. (1986). ‘The Cultural Biography of Things:  Commoditization as Process’,
in A. Appadurai (ed.), The Social Life of Things:  Commodities in Cultural Perspective.
Cambridge: Cambridge University Press, 64–​91.
Macinnes, L. (1984). ‘Brochs and the Roman Occupation of Lowland Scotland’, Proceedings of
the Society of Antiquaries of Scotland, 114: 235–​249.
McIntosh, F. (2011). ‘Regional Brooch-​Types in Roman Britain: Evidence from Northern
England’, Archaeologia Aeliana, 5th ser. 40: 155–​182.
Maldonado, A. (2015). ‘The Early Medieval Antonine Wall’, Britannia 46: 225–245.
Mann, J. (1974). ‘The Northern Frontier after ad 369’, Glasgow Archaeological Journal, 3:
34–​42.
Mann, J. C., and Breeze, D. J. (1987). ‘Ptolemy, Tacitus and the Tribes of North Britain’,
Proceedings of the Society of Antiquaries of Scotland, 117: 85–​91.
Manning, W. H. (2006). ‘The Roman Ironwork Deposits from the Fort at Newstead’, Bayerisches
Vorgeschichtsblätter, 71: 15–​32.
Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire 54 bc–​AD 409.
London: Penguin Books.
Maxwell, G. S. (1987). ‘Settlement in Southern Pictland: A New Overview’, in A. Small (ed.),
The Picts: A New Look at Old Problems. Dundee: Graham Hunter Foundation, 31–​44.
 Beyond Hadrian’s Wall   201

Maxwell, G. S. (1989). The Romans in Scotland. Edinburgh: Mercat Press.

Maxwell, G. S. (1990). A Battle Lost:  Romans and Caledonians at Mons Graupius.
Edinburgh: Edinburgh University Press.
Maxwell, G. S. (1998). A Gathering of Eagles: Scenes from Roman Scotland. Edinburgh: Birlinn.
Megaw, J. V. S., and Megaw, M. R. (1996). ‘Ancient Celts and Modern Ethnicity’, Antiquity,
70: 175–​181.
Megaw, J. V.  S., and Megaw, M. R. (1998). ‘The Mechanism of (Celtic) Dreams? A  Partial
Response to our Critics’, Antiquity, 72: 432–​435.
Mighall, T. (2003). ‘Geochemical Monitoring of Heavy Metal Pollution and Prehistoric
Mining: Evidence from Copa Hill, Cwmystwyth, and Mount Gabriel, County Cork’, in P.
Craddock and J. Lang (eds), Mining and Metal Production through the Ages. London: British
Museum, 43–​51.
Painter, K. (2013). ‘Hacksilber:  A  Means of Exchange?’, in F. Hunter and K. Painter (eds),
Late Roman Silver: The Traprain Treasure in Context. Edinburgh: Society of Antiquaries of
Scotland, 215–​242.
Pitts, L., and St Joseph, J. K. (1985). Inchtuthil: The Roman Legionary Fortress. London: Society
for the Promotion of Roman Studies.
Poulter, J. (2009). Surveying Roman Military Landscapes across Northern Britain.
Oxford: Archaeopress/​BAR British Series 492.
Price, J. (1988). ‘Romano-​British Glass Bangles from East Yorkshire’, in J. Price, P. R. Wilson, C. S.
Briggs, and S. J. Hardman (eds), Recent Research in Roman Yorkshire. Oxford: Archaeopress/​
BAR British Series 193, 339–​366.
Richmond, I. A. (1958) (ed.). Roman and Native in North Britain. Edinburgh: Nelson.
Robertson, A. (1970). ‘Roman Finds from Non-​Roman Sites in Scotland’, Britannia, 1: 198–​226.
Robertson, A. S., and Keppie, L. (2001). The Antonine Wall:  A  Handbook to the Surviving
Remains. Glasgow: Glasgow Archaeological Society.
Sims-​Williams, P. (1998). ‘Celtomania and Celtoscepticism’, Cambrian Medieval Celtic Studies,
36: 1–​35.
Sommer, C. S. (1984). The Military Vici in Roman Britain. Oxford: BAR British Series 129.
Sommer, C. S. (2006). ‘Military Vici in Roman Britain Revisited’, in R. J.  A. Wilson (ed.),
Romanitas: Essays in Roman Archaeology in Honour of Sheppard Frere on the Occasion of his
Ninetieth Birthday. Oxford: Oxbow Books, 95–​145.
Stallibrass, S. (2009). ‘The Way to a Roman Soldier’s Heart: A Post-​Medieval Model for Cattle
Droving to the Hadrian’s Wall Area’, in M. Driessen, S. Heeren, J. Hendriks, F. Kemmers,
and R. Visser (eds), TRAC 2008:  Proceedings of the 18th Theoretical Roman Archaeology
Conference, Amsterdam 2008. Oxford: Oxbow Books, 101–​112.
Steer, K. A. (1958). ‘Arthur’s O’on:  A  Lost Shrine of Roman Britain’, Archaeological Journal,
115: 99–​110.
Steer, K. A. (1966). ‘A Roman Building at Easter Langlee, Roxburghshire’, Proceedings of the
Society of Antiquaries of Scotland, 98 [1964–​6], 320–​1.
Stevenson, R. B. K. (1956a). ‘Native Bangles and Roman Glass’, Proceedings of the Society of
Antiquaries of Scotland, 88 [1954–​6], 208–​221.
Stevenson, R. B. K. (1956b). ‘Pictish Chain, Roman Silver and Bauxite Beads’, Proceedings of the
Society of Antiquaries of Scotland, 88 [1954–​6], 228–​230.
Stevenson, R. B. K. (1966). ‘Metal-​Work and Some Other Objects in Scotland and their Cultural
Affinities’, in A. L. F. Rivet (ed.), The Iron Age in Northern Britain. Edinburgh: Edinburgh
University Press, 17–​44.
202

202   Fraser Hunter

Stevenson, R. B. K. (1976). ‘Romano-​British Glass Bangles’, Glasgow Archaeological Journal,

4: 45–​54.
Swan, V. G. (1988). ‘Comments on Inveresk Ware’ [pp. 167–​171], in G. Thomas ‘Excavations at
the Roman Civil Settlement at Inveresk, 1976–​77’, Proceedings of the Society of Antiquaries of
Scotland, 118: 139–​76.
Swan, V. G. (1999). ‘The Twentieth Legion and the History of the Antonine Wall Reconsidered’,
Proceedings of the Society of Antiquaries of Scotland, 129: 399–​480.
Symonds, M. F. A., and Mason, D. J. P. (2009). Frontiers of Knowledge: A Research Framework
for Hadrian’s Wall. Durham: Durham County Council/​Durham University.
Tipping, R. (1997). ‘Pollen Analysis and the Impact of Rome on Native Agriculture around
Hadrian’s Wall’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies.
Oxford: Oxbow Books, 239–​47.
Tipping, R., and Tisdall, E. (2005). ‘The Landscape Context of the Antonine Wall: A Review of
the Literature’, Proceedings of the Society of Antiquaries of Scotland, 135: 443–​469.
Wallace, C. (2006). ‘Long-​lived Samian?’, Britannia, 37: 259–​272.
Wilson, A. (2010). Roman and Native in the Central Scottish Borders. Oxford: Archaeopress/​
BAR British Series 519.
Woolliscroft, D. J. (2000). ‘More Thoughts on why the Romans Failed to Conquer Scotland’,
Scottish Archaeological Journal, 22/​2: 111–​122.
Woolliscroft, D. J., and Hoffmann, B. (2006). Rome’s First Frontier. Stroud: Tempus.
Youngs, S. (2013). ‘From Chains to Brooches:  The Uses and Hoarding of Silver in North
Britain in the Early Historic period’, in F. Hunter and K. Painter (eds), Late Roman
Silver: The Traprain Treasure in Context. Edinburgh: Society of Antiquaries of Scotland,
403–​425.
 Chapter 10

Mobilit y, M i g rat i on,

and Diasp oras
i n Roman Bri ta i n

Hella Eckardt and Gundula Müldner

Introduction

Did the Roman period see unprecedented levels of mobility? In order to answer this
question, two main issues need to be addressed. The first is concerned with the various
forms of evidence for mobility, notably written epigraphy, material culture, and scien-
tific techniques such as isotopic analysis, and their respective strengths and weaknesses.
The second is about how we can model interactions and relationships between incomers
and locals, and indeed whether such distinctions mattered much in a cultural context
that appears to have been characterized by an emphasis on Romanitas rather than geo-
graphical origin.
This chapter builds on the results of a multidisciplinary AHRC-​funded project on
mobility in Roman Britain held at the University of Reading from 2007 to 2009 and,
in particular, on a paper contextualizing its results published in 2010 (Eckardt 2010).
There is little point in repeating here the detailed arguments made in 2010. Rather, key
issues will be summarized and only the evidence that has emerged since that publication
will be examined in detail. This is especially relevant in the fast-​moving field of isotope
analysis, where new data continue to challenge and help develop our original interpreta-
tions. However, there is also new research on material culture and epigraphy, and it is
worth re-​examining very briefly how thinking about diasporas and hybridization may
move us beyond established critiques of the ‘Romanization’ paradigm. There is also an
increased awareness of the political context in which research on emotive themes such
as migration is carried out, and this will be addressed by reviewing some recently devel-
oped educational resources.
204

204    Hella Eckardt and Gundula Müldner

Archaeological Evidence for Mobility

Conquest by Rome brought a wide range of people to Britain, through both volun-
tary and forced migration; the military, imperial administration, and trade are usually
thought to be the most important drivers, but the role of slavery also has to be con-
sidered (Eckardt 2010: 102–10​3; Webster 2010). Archaeologists have long attempted
to identify migrants and to estimate the proportion they make up within a given site
or region. Traditionally, this has been attempted through the study of written sources,
epigraphy, and material culture, but, more recently, isotope analysis has offered new
opportunities for research.

Epigraphy

Epigraphy identifies foreigners through individual names, stated places of origin, and
reference to regionally specific cults or calendars (e.g. Birley 1979; Noy 2000, 2010).
These data produce personal, individual histories and allow for the exploration of
wider patterns in terms of age, gender, and status. However, the evidence is unevenly
distributed, with relatively few inscriptions known from Britain in comparison with
other provinces, and inherently biased, for example, towards urban and military areas
(Rowland 1976; Jones and Mattingly 2002: map 5.10; see also Pearce 2010). Another
obvious proviso is that it is only those individuals stating their origins who are vis-
ible in the source material. Depending on the region and time, inscriptions may reflect
only a few per cent of the total population, making estimates of the percentage of for-
eigners within epigraphic populations fraught with difficulty (Eckardt 2010, 2012).
Nevertheless, epigraphy provides important insights and detailed information on
those few individuals, and it is therefore worth adding a number of important studies
that were missed, and new work that has been published since 2010. Thus Kakoschke
(2002) provides an extremely useful parallel to the Gaulish data by analysing the evi-
dence for incomers to the two Germanies (cf. Wierschowski 1995, 2001). Another
assessment of immigration comes from Roman Dacia, which apparently saw intensive
colonization after the Trajanic wars, with natives only very rarely attested (Mihailescu-​
Bîrliba 2011). Carroll (2006: 209–​232) provides an interesting comparison of migration
into military (Mainz and Chester) and civilian (Lyon and Cologne) sites, highlighting
the impact of the army, change through time, and again the diversity of people mov-
ing across the empire. Clay (2007) explores the epigraphic and linguistic evidence for
Germanic people on Hadrian’s Wall, discussing in particular how religious practices
may serve both to maintain relations with the homeland and form new ones with the
inhabitants of the local area. Another recent case study highlights the self-​representa-
tion of large numbers of Milesians in Hellenistic and Roman Athens, and places them
 Mobility, Migration, and Diasporas in Roman Britain    205

into the context of social relations and especially intermarriage between citizens and
non-​citizens (Gray 2011).
It is worth exploring one other new study of epigraphy and mobility in more detail, as
it provides a fascinating chronological contrast by focusing on the late antique period
(fourth–​eighth centuries ad). Handley recorded 623 examples of travellers or foreign-
ers on a total of 567 inscriptions, revealing ‘an intricate web of travel in all directions:
Spaniards in Italy, Pannonians in Dalmatia, Gauls in Africa’ (Handley 2011: 17). The study
uses only individuals who state their origin or use regionally specific calendars, and
demonstrates the unreliability of using names alone (Handley 2011: 21–​28). The advan-
tage of the epigraphic material is that it allows detailed analysis of travellers in terms of
gender, age, status, and profession, although numbers are low and the overall dataset is
biased (Handley 2011: 37–​51). For example, male migrants dominate with about 80 per
cent of the sample (while only making up 63 per cent of the whole epigraphic popula-
tion), but some women also travel long distances, and factors like intermarriage might
decrease the likelihood of their foreign origin being recorded. The age recorded in an
inscription is, of course, not necessarily the age at which an individual first travelled, but
the evidence from the late antique west shows more older male migrants leaving inscrip-
tions, while in the city of Rome younger males are more strongly represented in the epi-
graphic data (Figure 10.1). Recorded professions range from soldiers and merchants to
government office-​holders and bishops; there is also a lawyer and a stonemason. The
reasons for travel are diverse, including war/​soldiery, pilgrimage, ecclesiastical business,
exile, trade, and the emigration of individuals and families; there is also some evidence
that the dead were moved, sometimes over considerable time periods and distances
(Handley 2011: 53–​62). The study clearly illustrates that travel and mobility continued
into the late antique period, and that travel had considerable economic implications.

25 Ages of male travellers

20
Percentage

15

10

0
0–9 10–19 20–29 30–39 40–49 50–59 60–69 70–79 80–89 90–99

Males to Rome (Noy): N = 47

Males late antique west: N = 107

Figure 10.1  Epigraphically recorded age of travellers to Rome 300–​600 ad as recorded by Noy
(2000) and for the late antique west by Handley (2011).
Source: after Handley (2011: graphs 1 and 2). © Department of Archaeology, University of Reading.
206

206    Hella Eckardt and Gundula Müldner

Handley (2011: 114) uses ancient sources to highlight the numbers of people recorded
on ships and the potential financial value of such travellers. Thus an early sixth-​century
letter of Cassiodorus records that 75 per cent of the ship’s profit came from the passenger
fees, and only 25 per cent from the cargo, reversing the usual assumption about travel
‘piggy-​backing’ on the shipment of bulk goods (Handley 2011: 11).
In summary, the epigraphic evidence still has much to offer, providing rich detail and
a direct connection to people’s life stories; there is also much to learn about the rela-
tive proportions of migrants of different origin, gender, age, and status. However, it is
still impossible truly to estimate numbers of foreigners from this limited and biased
evidence.

Material Culture and Identity

In practice, particular objects associated with personal adornment and food ways have
long been used to identify ‘foreigners’. Thus certain beads, belt fittings, or brooches,
when found outside their ‘normal’ distribution, may indicate the presence of migrants
who carried small portable items with them, although the effects of trade and changing
fashions always have to be considered. Pots, even when made in local fabrics, may indi-
cate the movement of their makers and consumers, and this may be especially true for
vessels associated with particular ways of preparing and consuming food (Swan 1992,
1999, 2009; Cool 2006a: 172–​183; 2006b). There are significant provisos and problems:
for example, when attempting to establish the typological closeness of the forms found
in Britain to those found in the supposed regions of origin (Fulford 2010). Burial evi-
dence is of particular interest, as the rite may reveal ‘intrusive’ practices, although the
picture is much more complicated than early archaeologists thought (cf. Millett 1999:
194–19​6; Cool 2010; Pearce 2010).
More fundamentally, the very link between material culture and ethnicity, and indeed
the meaning of ethnicity as a concept, has seen a significant review in theoretical archae-
ology. Rather than ethnicity being viewed as fixed and given, it is now usually seen as
fluid, and there is an increased recognition that objects do not express just one aspect of
identity, but that their use and interpretation are context-​dependent and open to mul-
tiple interpretations (e.g. Jones 1997; Gardner 2007: 197–​203). Thus an object may relate
to age, gender, marital status, or religious beliefs, as well as act as an indicator of eth-
nicity (e.g. Meskell 2001), and, in some cases, an object type may appear intrusive or
unusual to a modern artefact specialist but other aspects of the wearer’s dress may have
been much more important signals of identity in the past. This is not the place to dis-
cuss these issues in detail, but merely to note that, in contrast to the post-​Roman period
(e.g. Brather 2000, 2004; Curta 2007; Härke 2007), the question of ethnicity has not,
until relatively recently, been a major point of discussion in Roman archaeology. This
seems paradoxical, given the evidence for high levels of mobility in the Roman period
and the fine-​grained and well-​published nature of the material culture, but this is now
 Mobility, Migration, and Diasporas in Roman Britain    207

beginning to be addressed (e.g. Pohl 1998). Work on the archaeology of the Rhine delta
in particular explored the nuanced ways in which Rome as a colonial power impacted on
the creation of the Batavians as an ethnic and political entity (Roymans 2004; Derks and
Roymans 2009). In other recent work, Swift (2010) examines a series of bracelets that
have restricted distributions in late Roman Britain, suggesting production in regional
workshops, but that also occur in limited numbers on continental sites. She argues that
such finds are indicative of migrants, who brought relatively low-​value items of per-
sonal adornment with them and that this is the case in particular where the type-​site (or
social) distribution of an object type is restricted to military and large urban sites. Using
a combination of artefact types and burial context in the well-​published, large cemetery
at Krefeld-​Gellep in north-​west Germany, Swift suggests the presence of migrants from
Roman Britain and the Upper Danube provinces.
Ivleva (2010) compares 31 epigraphic records (including military diplomas) and 241
British-​made brooches to trace the presence of British emigrants (and people of diverse
origins who may have lived in Britain for some time) on the continent. The argued
connection between British auxiliary soldiers, other British recruits, and returning
veterans and brooch distributions is not always convincing, and it is impossible to
know whether these brooches, thought to have been made in Britain, were worn as an
expression of ‘British’ identity or simply because they were available and convenient
personal possessions. Nevertheless, the paper does illustrate an interesting artefactual
distribution and, at the very least, highlights close connections between Britain and
the Rhineland in particular (Ivleva 2010: figure 2). What all these case studies share is
an attempt to recognize ‘intrusive’ material culture in the archaeological record but to
interpret these objects not simply as overt signals of ethnic identity but in their histori-
cal context and in relation to other aspects of identity such as gender or, for example,
military identities.

The Application of Archaeological

Science to the Question of Mobility

Isotopes and Mobility: The Background

The 2007–​9 Diasporas in Roman Britain project was the first to employ combined
strontium (87Sr/​86Sr) and oxygen (δ18O) isotope analysis on a large scale in order to dis-
tinguish between locals and foreigners in a Roman province. ‘Isotopic signatures’ are
incorporated into the body tissues of humans and animals through the food and drink
they ingested at the time of tissue formation (see Eckardt 2010: 112–​113 for a summary).
The strontium and oxygen isotope composition of tooth enamel, which is fixed in the
early years of life, allows us to characterize broadly an individual’s place of childhood
residence in terms of geology and climate. The ‘isotopic signature’ thus obtained can
208

208    Hella Eckardt and Gundula Müldner

then be compared to available background data for different geographical regions,

first in order to assess whether an individual’s isotope values are consistent with a local
upbringing (that is, at the place of burial). If they are, the most parsimonious explana-
tion is that this individual was indeed local, although it must always be borne in mind
that many other places may share a similar geology and climate. For example, typical
oxygen isotope values for Britain are shared by parts of western and southern Europe
and even the Mediterranean coast, while much of the presently available strontium iso-
tope data from Britain and mainland Europe appear to reflect the isotopic composition
of a relatively small number of soils that are particularly suitable for agricultural produc-
tion and are, therefore, sought out by humans everywhere (Montgomery 2010; Chenery
et al. 2011; Evans et al. 2012; Whittle and Bickle 2013).
Where the isotope values exclude a local origin, the information contained in the
isotopic signal can be used to narrow down potential places of childhood residence. It
should be emphasized, however, that it is usually impossible to assign an individual’s
origins to a particular locale based on isotope evidence alone, simply because too many
areas share the same strontium and oxygen isotope characteristics. Isotope analysis is
most valuable as an ‘exclusive technique’ (Montgomery 2010: 336)—​that is, it can be
used to eliminate local origin and other suggested homelands because they are incon-
sistent with the data, but, in order to make a case for one of the (usually) many possible
places of childhood residence that are compatible with the isotope data, the archaeol-
ogist has to rely on other contextual evidence (see, e.g., Fitzpatrick (2011) for a care-
fully constructed argument for an origin of the ‘Amesbury Archer’ in the Alpine region,
based on the combined analysis of isotopic and archaeological data).
Isotope analysis is a destructive technique, requiring a small sample often taken from
the second or third permanent molar, which represents an enamel development period
between c. 3 and 13 years of age, and this can only partly be compensated for by pho-
tographs and casts of the teeth made prior to sampling. For oxygen and—​especially—​
strontium isotope analysis, tooth enamel is the preferred analyte, not only because it
fixes geographical information from a well-​definable period in childhood, but also
because, unlike bone or tooth dentine, it has been shown to be largely resilient against
post-​mortem contamination by elements in the burial soil (e.g. Hoppe et al. 2003). A
detailed description of the methods used by the project can be found in Chenery et al.
(2010).
For the Roman world, isotope mobility studies conducted prior to the Diasporas
in Roman Britain project had focused on late Roman burials in southern Germany
(Schweissing and Grupe 2003) and on Portus near the city of Rome (Prowse et al. 2007),
although a number of studies have been added since (Perry et al. 2008, 2009, 2011;
Killgrove 2010b; Prowse et al. 2010) and there will likely be more by the time this chap-
ter is published. In a specifically Romano-​British context, a small study had previously
examined individuals from Mangotsfield near Bristol and from Winchester (Budd et al.
2004: 134), while a paper by Evans et al. (2006) had analysed eighteen individuals from
Lankhills, where Clarke (1979: 377–​389) identified a group of incomers on the basis of unu-
sual objects and burial rite (cf. Swift 2010: 237). Evans et al. (2006) compared the isotopic
 Mobility, Migration, and Diasporas in Roman Britain    209

signatures of nine individuals identified as ‘local’ and nine identified as ‘exotic’ on the
basis of these archaeological criteria. The results clearly showed that the starting assump-
tions about origin were oversimplified, and that the relationship between social identity,
as expressed in the burial rite, and geographical origin are much more complex. These
conclusions were very much borne out by our larger project, which analysed 155 skeletons
from five sites in Roman Britain (York, Catterick, Gloucester, Poundbury/​Dorchester, and
Lankhills/​Winchester (Eckardt et al. 2009; Leach et al. 2009, 2010; Chenery et al. 2010,
2011; Müldner et al. 2011)). The key result is that all sites showed considerable diversity,
with a range of values that may be compatible with local (within a 30 kilometre radius of
the site) or other British origins, but also several individuals who came from either cooler
and/​or more continental, or warmer, possibly more coastal, areas.
Some attempt was made in the 2010 paper to evaluate levels of mobility, comparing the
results of various epigraphic studies, which all suggest a proportion of c. 5 per cent stat-
ing foreign origin, with a major paper by Scheidel (2004), who argued that 40 per cent
of Italian Roman citizens over the age of 45 would not have lived in the place where they
were born (cf. Eckardt 2010: 103–​107). The isotope analyses of Prowse et al. (2007) and
Schweissing and Grupe (2003) suggest high levels (approximately 30 per cent) of migra-
tion, and our results also indicate a high proportion of incomers. While roughly half the
sampled individuals could be ‘local’ in the sense of consuming foods sourced within a
30 kilometre radius of the relevant site, and a significant proportion of the non-​locals are
likely to have come from elsewhere in Britain, the project also identified some individu-
als who are likely to have spent their childhood outside Britain (Eckardt 2010: table 7.2).
The small town and military fort of Catterick was the least diverse site, but nearby York
(especially the beheaded males from the unusual cemetery at Driffield Terrace), as well
as Lankhills (Winchester), showed larger numbers of suggested migrants (Eckardt et al.
2009; Leach et al. 2009, 2010; Chenery et al. 2010, 2011; Müldner et al. 2011). However, as
pointed out in the original paper, the proportions given in table 7.2 (Eckardt 2010) have
to be taken with a large pinch of salt. First, they are not based on a representative sam-
ple, since our selection of study sites was biased towards archaeologically visible, highly
‘Romanized’ urban populations and, within those, sample selection was often modified
to include ‘unusual’ individuals, whose origin the project aimed to explore. Additional
isotope data, especially from rural sites and from other parts of the Roman Empire, are
therefore required to confirm the wider relevance of the results. Below we will review
more recent work in what is a rapidly evolving area of research, concerning not just
methodological questions but also issues that are perhaps specific to the Roman world.

Recent Isotope Research: Challenges and Issues

Isotope analysis is an attractive tool to archaeologists and one that has had considerable
exposure in the media, but, because of this, there is clearly a risk that complex data are
oversimplified to the extent of creating the impression that strontium and oxygen iso-
tope analysis can provide ‘an unequivocal postcode for the domicile of that individual as
210

210    Hella Eckardt and Gundula Müldner

a child’ (Pollard 2011: 634). Although these dangers have already been pointed out above
and in our original papers, it is worth expanding on some of the issues here, in the light
of newly published data. In this respect, it is important to emphasize that isotope data
are just like other archaeological evidence in that they are subject to reinterpretation as
more information becomes available.
For both oxygen and strontium isotope studies the question of how to define the
‘local range’—​that is, the range of values that can be considered as consistent with
a local upbringing—​is a central one, and, while this range can be estimated using
a number of different indicators, it is important to realize that it is essentially the
result of culturally mediated behaviour (that is, food and drink acquisition strate-
gies) and that changes in food preferences and agricultural production, the move-
ment of foods, as well as the transport or differential treatment of drinking water,
could cause fluctuations in the local range of a given site over time (e.g. Bruun 2010;
Killgrove 2010a; Montgomery 2010; Chenery et al. 2011; Brettell et al. 2012). In this
context, the issue of conversion equations, which are used to relate human oxygen
isotope ratios to drinking water values (and thereby to certain geographical areas
through modern rainfall maps), should also be mentioned. While the possibility
of climate change is raised most often against these, this problem may be smaller
than often assumed, at least for European populations from the later Holocene (see
Chenery et  al. 2010:  153). However, recent discussions have also highlighted the
problem that different laboratories prefer different equations (computing differ-
ent drinking water values from the same skeletal δ18O) and that the often consider-
able statistical error inherent in these calculations is frequently understated (Daux
et al. 2008; Chenery et al. 2010; Pollard et al. 2011b). Interpretations based on this
approach may, therefore, not be as accurate as they might appear to a non-​specialist
reader. For this reason, direct comparisons with human oxygen isotope data, when-
ever these are available, are to be preferred.
One very significant advance since the start of the Diasporas in Roman Britain pro-
ject has been the large increase in the number of available human isotope data from
Britain, which can be used for making more robust local range estimates than was
previously possible. In 2009, we had only 57 humans to compute a likely range of
enamel oxygen isotope values (δ18Op) of individuals who spent their childhood in
Britain (see Leach et al. 2009; Chenery et al. 2010); subsequent publications—​espe-
cially the synthetic paper by Evans et al. (2012)—​have seen the number of data points
increase to more than 600. The revised range that can be calculated from these data
is somewhat wider than our earlier estimate (two standard deviations range of 16.3
to 19.1‰ compared to the previous estimation of 16.8–​18.6‰; see Evans et al. (2012:
figure 4). The implications of this are illustrated in Figure 10.2, which displays iso-
tope data from the Lankhills cemetery with both the earlier and the revised estimate
for a British skeletal oxygen isotope range. A number of individuals whose δ18Op had
previously seemed too high to be easily consistent with origins in Britain and for
whom a childhood in warmer, possibly more coastal, areas had to be considered (see
 Mobility, Migration, and Diasporas in Roman Britain    211

Eckardt et al. 2009) are now just inside the British range estimate, and local origins
for these now appear more plausible. It should be noted here that there is the danger
of a circular argument, since almost one quarter of the data set contained in Evans
et al.’s (2012) paper and from which the new estimate is computed are Diaspora pro-
ject results. The new (2012) estimate is also probably somewhat inflated, as the syn-
thetic data set is likely to include a number of non-​British migrants, even if some of
the more obvious outliers have been removed (see Evans et al. 2012). Nevertheless,
the suggestion that the skeletal oxygen isotope range for Britain is wider than had
been assumed previously appears to be also borne out by other new research (Parker
Pearson et al., forthcoming). Even if the new estimate is likely to be revised again in
the future, this exercise therefore demonstrates the fluidity (within reason) of local
range estimates in a rapidly expanding field of research, and we can safely say that
it will always be problematic, and to an extent nonsensical, to draw lines meant to
separate locals and non-​locals through a continuous field of data points. This is espe-
cially true for oxygen isotope values, which, unlike strontium, are subject to frac-
tionation and change through a large number of biological processes (but see Bentley
(2006) and Montgomery (2010) for some of the complexities of strontium isotope
data). Several recent publications have, for example, drawn attention to how the
oxygen isotope composition of drinking water can be modified by various cultural
practices (for example, prolonged cooking, brewing), causing a shift from the isotope
ratio of the local rainwater (which is usually assumed as baseline for local values in a
given region), to higher values (e.g. Daux et al. 2008; Knudson 2009; Price et al. 2010;
Brettell et al. 2012). Regular consumption of such culturally modified drinking water
could make individuals appear ‘too warm’ for a given locale. For the Roman period,
the regular consumption of wine, even in smaller quantities, might well have had
such an effect, since the isotopic shift between the (rain-​)water feeding the vine and
the finished product can be very considerable (see West et al. 2007).
However, it should be noted that there are next to no mechanisms that would cause a
shift to lower values—​only transport of drinking water from ‘cooler’ areas/​higher alti-
tudes would achieve this.
Despite these complexities, and although isotope analysis is therefore not the panacea
for archaeological migration studies that it is sometimes portrayed to be, the power of
isotopic methods is beyond doubt and should not be understated. They are particularly
effective when they can identify individuals who plot well outside the local range, ide-
ally as outliers from a main distribution of samples (Figure 10.2). In this way, isotope
data can provide an independent, ‘biological’ indication of mobility, which can be gain-
fully contrasted with other, more culturally mediated types of evidence relating to the
individual (such as burial rite or grave goods) in order better to understand the com-
plex ways by which identity can be expressed in material remains—​and the Diasporas in
Roman Britain project has produced a number of such examples.
Recent investigations have also highlighted the great advantage of multi-​isotopic
investigations in order to increase the resolution of the isotopic signal. Within the
212

212    Hella Eckardt and Gundula Müldner

‘Intrusive’ burials
0.7125
Other burials

0.7115

0.7105
87Sr/86Sr

0.7095

0.7085 Local Sr
isotope range

0.7075

0.7065
British O isotope range
Analytical error (2σ) (2009 and 2012 estimate)
0.7055
14.0 15.0 16.0 17.0 18.0 19.0 20.0
δ18Op(‰)VSMOW

Figure 10.2  Strontium and oxygen isotope data of humans from Lankhills/​Winchester, con-
trasting burials identified as ‘intrusive’ according to Clarke’s criteria (1979) with other buri-
als from the site; the two boxes indicate the local strontium isotope range for Winchester and
estimates for the range of oxygen isotope values consistent with a childhood in Britain based on
human skeletal phosphate (δ18Op) data available in 2009 (thin line) and 2012 (thick line).
Source: after Evans et al. (2006) and Eckardt et al. (2009). © Department of Archaeology, University of Reading.

Diasporas in Roman Britain project, we had particular success with integrating pal-
aeodietary (carbon and nitrogen isotope) data with the oxygen and strontium iso-
tope values. Even though carbon and nitrogen isotopes are not usually thought of as
useful for migration studies, especially in a European context, they reliably identified
a number of long-​distance migrants, probably because of the considerable environ-
mental and dietary differences that existed within the Roman Empire and its adjoin-
ing regions (Müldner 2013). Carbon and nitrogen isotope evidence can, therefore,
be used further to constrain possible areas of origin. In some cases, and especially
where individuals showed evidence of consumption of C4-​plants (millet)-​based pro-
tein, which was not available in Roman Britain in any quantity, carbon and nitro-
gen isotopes can even identify non-​locals when the strontium and oxygen isotope
data are ambiguous, as was the case for one of the ‘Headless Romans’ of York and
for a much less conspicuous male burial from Gravesend, Kent (Müldner et al. 2011;
Pollard et al. 2011a).
Lead isotopes may also add another dimension to the analysis. Montgomery et al.
(2010) demonstrate that during the Roman period—​when the geological informa-
tion normally contained in lead isotopes is lost owing to high levels of anthropogenic
 Mobility, Migration, and Diasporas in Roman Britain    213

lead pollution and individuals from Britain all exhibit similar values controlled by
an amalgamate of British ore lead—​lead isotopes can, nevertheless, be used to dis-
tinguish migrants from areas where lead from very different sources was used. For
example, they allowed for the high-​status female burial from Spitalfields, London,
to be identified as an incomer, despite her very generic strontium isotope values, as
her lead signature is clearly not British. Human lead isotope data from Rome itself,
which have recently become available and are very similar to those of the ‘Spitalfields
Lady’, give at least one possible area of origin (Montgomery et al. 2010: 217–​219).
Conversely, the lead isotope values for both the ‘Ivory Bangle Lady’ of York and
the Catterick ‘Eunuch’ are compatible with British ore lead (Eckardt and Müldner,
unpublished data), and, although it is as yet unclear which regions of the Roman
Empire shared similar values (an origin in Rome at least can apparently be excluded),
these two individuals may indeed have been of British origin, especially since none
of the other isotopic systems provided unequivocal evidence to the contrary (Leach
et al. 2010; Chenery et al. 2011). It is only by playing a host of analyses off against each
other that such patterns emerge, and it is probably that future investigations will help
further to refine interpretations.

Models for Interaction

The discussion of identities in Roman Britain and the Roman world has been domi-
nated by the ‘Romanization’ paradigm (Millett 1990; cf. Webster 2001: 209–​217),
which stressed the unifying aspects of the Roman Empire and focused on the ways
in which certain cultural traits were adopted and adapted by conquered popula-
tions. More recent models stress the varied responses to, and different experiences
of, empire (Woolf 1998; Webster 2001: 217–​223; Mattingly 2004, 2006). Given the
high levels of mobility within a complex colonial system, it has been suggested that
diaspora theory may be a useful model that allows us to focus on migrant interac-
tions with host communities and to deal with the forced and voluntary movement
of both groups and individuals (Eckardt 2010: 107–​109, with detailed references).
Lilley (2004) and Cohen (2008) provide useful summaries of the characteristics of
diasporic communities, which are defined not just by initial (forced or voluntary)
dispersal and a distinction from the host society, but also by a continuous social or
spiritual link to the homeland. Most of the initial work on diasporas has focused
on so-​called victim diasporas, in particular the Jewish and Afro-​Caribbean experi-
ence, but more recently scholars have begun to explore trade and colonial diasporas,
which both have potential relevance to the Roman world. Trade diasporas have been
described as a ‘nation of socially interdependent, but spatially dispersed, communi-
ties’ whose members are distinct from both the societies in which they originated
and those in which they live (Cohen 2008: 83). Archaeological case studies on trade
214

214    Hella Eckardt and Gundula Müldner

diasporas have largely focused on early modern, historical communities (e.g. Voss
2005), but the Roman world is full of examples of merchants setting up monuments
far from home, and it seems likely that in many cases their origin—​as in the case
of the famous Barates of Palmyra who set up the gravestone for his British wife in
Southshields—​continued to be important to them (Noy 2010: 20–​21).
Also of considerable relevance to the Roman period is the notion of colonial and
imperial diasporas. Cohen defines the imperial diaspora as ‘marked by a continuing
connection with the homeland, a deference to and imitation of its social and political
institutions and a sense of forming part of a grand imperial design’ (Cohen 2008: 69; cf.
Lilley 2004). Both indigenous groups and European/​British colonizers in Australia, the
United States, and India can be viewed as colonial and imperial diaspora communities
(e.g. Lawrence 2003; Casella 2005; Lilley 2006). Again, the applications to the Roman
world are obvious.
Finally, the theoretical and methodological development of diaspora studies in some
ways mirrors how ‘Roman-​ness’ has been studied in the provinces. For example, the
interactions between diasporic and host communities are clearly complex, and both are
potentially characterized by considerable diversity, and this has echoes in our changed
understanding of ‘Romans’ and ‘natives’, which has evolved from sharply defined
binary opposites to more nuanced appreciations of diversity (e.g. Mattingly 2004,
2006). Even more importantly, earlier studies have tended to focus on the identifica-
tion of cultural traits as a means to trace origin and to demonstrate cultural continu-
ity and continued connections with the homeland (Eckardt 2010: 107–​108). However,
such approaches make assumptions about the static nature of culture, oversimplify the
processes of cultural interaction, and are devoid of cultural context. Diaspora commu-
nities are different not just from the host population but also from that of their origi-
nal homeland. Thus Cool (2004, 2010) suggests that some of the practices observed
at Brougham may reflect not simply intrusive practices potentially reflecting a trans-​
Danubian homeland, but also traits ‘picked up’ from other areas such as the Rhineland,
which these individuals may have passed through on their way to Britain or which were
part of a military culture they encountered elsewhere. Humans continue to create their
own ‘correct’ ways of being, and their perceptions and practices will be affected by relo-
cations and by their continued interactions with various groups and with each other.
More recent work is thus increasingly concerned with the nature of culture change and
how, through processes of creolization and hybridization, diasporic identities are cre-
ated and expressed (cf. Meskell 2001: 286; Wilkie 2004: 113–​114). The concept of cre-
olization has been extensively explored in Roman archaeology (Webster 2001), while
that of hybridity is perhaps less well known. The latter is often associated with colonial
and postcolonial theory and the creation and maintenance of identities in contexts of
colonial oppression, but in its wider sense of ‘entanglement’ conveys the ‘creation of
something new which is more than just an addition of its origins’ (Stockhammer 2012:
47; Van Dommelen and Terrenato 2007: 8–​9).
 Mobility, Migration, and Diasporas in Roman Britain    215

Immigrants, Race, and the Politics

of Outreach

The identification of immigrants is an emotive topic in contemporary discourse, and a

careful engagement with the wider public is an important aspect of any research project
in this area.
In addition to using isotope analysis, the Diasporas in Roman Britain project also
applied craniomorphometric analysis (forensic ancestry assessment) to skeletons from
Roman York (Leach et al. 2009; see also Eckardt 2010: 111–​112). This led to the identifica-
tion of ‘mixed-​race’ individuals, most notably the so-​called ‘Ivory Bangle Lady’, a rich
female burial with unusual grave goods from Sycamore Terrace (Leach et al. 2010). The
identification of ‘race’ is a highly controversial and sensitive subject in the humanities,
and indeed ‘race’ is recognized as a social construct by the biological sciences; biological
anthropologists are, therefore, using racial categories mainly as convenient shorthand
for certain phenotypic attributes that vary in frequency in different populations around
the globe. Despite this, for some any attempt at measuring skulls is still strongly associ-
ated with historic attempts to demonstrate that variability in outward appearance cor-
relates with characteristics such as intelligence or moral character, an approach that was
not only racist but also scientifically deeply flawed (e.g. Arnold 2006; Gosden 2006).
When conducting these analyses within a Roman context, an added complication is
the question of how ethnicity and ‘race’ were viewed in antiquity and whether racism
towards, for example, Africans existed (Sherwin-​White 1967; Snowden 1983; Isaac 2004,
2006; McCoskey 2012).
Our project employed forensic ancestry assessment in an explicitly interdiscipli-
nary approach to examine the interplay of biological differences and social constructs.

Figure 10.3  Reconstruction of the so-​called ‘Ivory Bangle Lady’ from York.
Source: © Aaron Watson.
216

216    Hella Eckardt and Gundula Müldner

However, the more subtle points of an academic paper can get lost when reconstructions
such as that of the ‘Ivory Bangle Lady’ (Figure 10.3) clearly capture the public imagina-
tion and form part of a modern discourse about ‘black’ identity, in many ways crystalliz-
ing general debates about immigration. This was evident not only in the press coverage
of the ‘Ivory Bangle Lady’, which was dominated by ideas of exoticness and of identi-
fying ‘the first African’, but in the subsequent, often vitriolic, responses by readers of
certain newspapers (e.g. <http://​www.dailymail.co.uk/​news/​article-​1254187/​Revealed-​
The-​African-​queen-​called-​York-​home-​4th-​century.html>).
There is increasing awareness of the impact archaeological interpretations of the
past have in the present, as evident in recent calls for a ‘more multicultural’ approach
to British cultural heritage, where explicit links between the past and present are made
(Benjamin 2003, 2004). While some argue that reconstructions of the past should not be
influenced by contemporary politics, others stress the importance of postcolonial and
politically aware discourses. Thus Hingley (2010: 232–​236) has reviewed how contem-
porary anxieties of the nineteenth-​century British Empire influenced interpretations of
‘foreign’ troops on Hadrian’s Wall and, conversely, how in recent years the ‘diverse geo-
graphical origins of the soldiers, and the idea that they may have retained elements of
their original ethnic character, have been used to create an inclusive role for Hadrian’s
Wall’ (Hingley 2010: 238; cf. Hingley 2012). The multinational character of the Wall was
highlighted in a 2006 art project Writing on the Wall and an exhibition The Archaeology
of ‘Race’: Exploring the Northern Frontier in Roman Britain (Hingley 2010: 238–​240;
Tolia-​Kelly 2010). This exhibition, which was shown in Segedunum Museum and Tullie
House Museum in 2009 and is still on display at Durham University, aims to show the
impact that the diverse origins and mobility of the individuals living on Hadrian’s Wall
may have had on daily life. It also explores the complex role and representation of the
‘African’ emperor Septimius Severus: on the one hand, ‘celebrating black presence’ and,
on the other, interpreting him as a ‘violent imperialist’ (Tolia-​Kelly 2010: 85 <http://​
www.dur.ac.uk/​geography/​race/​exhibition>).
Bearing in mind these issues, the Diasporas in Roman Britain project has now devel-
oped a website for primary-​school children, working with the Runnymede Trust, a
race equality and educational charity. The website <http://​romansrevealed.com/​> is
designed to promote the project findings of diversity to Key Stage 2 learners, their teach-
ers and families, and to break down the astoundingly strong public perception of ‘the
Romans’ as uniformly ‘Italian’. Throughout, the emphasis is on questioning ‘how do we
know’, trying to teach children about archaeological techniques but also that knowledge
and interpretations can and do change.

Conclusion

The Roman Empire was characterized by high levels of mobility, and diaspora theory
may be a helpful way of thinking about the interactions of migrants and locals. The
 Mobility, Migration, and Diasporas in Roman Britain    217

concept of trade and imperial diasporas in particular has clear applications to the
Roman world. Attempting to assess numbers of incomers in a given site or region is
fraught with difficulty: the epigraphic data are heavily biased, and material culture is
clearly used in complicated, multi-​layered ways rather than as a simple indicator of eth-
nicity or origin. The combination of archaeological evidence with both isotopic and cra-
niometric data offers a new avenue for research, but, as with all the other approaches,
there are serious methodological and theoretical issues to consider. While some may
latch on to a headline figure of 20 or 30 per cent immigration, the sample studied by the
Diaspora Project is not representative of the Romano-​British population as a whole, and
the interpretation of the isotopic values may well change as more data become available
and the impact of various cultural practices on human isotope signals in different time
periods are better understood. To this end, there is also an urgent need for further study,
in particular of rural populations, which are expected to have been more stationary and
can therefore give a better estimate of typical local ranges than the highly diverse urban
populations on which research has tended to focus. It is also clear that a multi-​proxy
approach involving a combination of scientific techniques has the greatest potential,
and future studies should include multiple isotopes for the best possible results.
Ultimately, it is in any case not the identification of biological origin that is truly
of interest, but its interplay with social identity. As today, people make choices about
whether to cling to an old, national, or ethnic identity or to adopt a new one—​a deci-
sion that may be described as the ‘sport team choice’ (Handley 2011: preface). In many
societies, including quite possibly the Roman world, origin, in the sense of area of birth,
may not matter that much (Swift 2010: 268; Pollard 2011: 635) and we must avoid naive
attempts to create a multicultural past because we live (or want to live) in a multicultural
present. Nevertheless, it is important for archaeologists to be aware of the political con-
text of our work, and, in our attempts to communicate complicated scientific findings to
the wider public, and especially children, we have to make a deliberate effort to convey
the research process and its uncertainties. Hopefully, by combining a variety of scien-
tific approaches with theoretical awareness, we can begin to understand the historically
specific ways in which ‘otherness’ may have been perceived and expressed in Roman
Britain.

References
Arnold, B. (2006). ‘“Arierdämmerung”:  Race and Archaeology in Nazi Germany’, World
Archaeology, 38/​1: 8–​31.
Benjamin, R. P. (2003). ‘Black and Asian Representation in UK Archaeology’, The Archaeologist,
Institute of Field Archaeologists Newsletter, 48 (Spring), 7–​8.
Benjamin, R. P. (2004). ‘Building a Black British Identity through Archaeology’, Archaeological
Review from Cambridge, 19/​2: 73–​83.
Bentley, R. A. (2006). ‘Strontium Isotopes: From the Earth to the Archaeological Skeleton—​a
Review’, Journal of Archaeological Method and Theory, 13: 135–​187.
Birley, A. (1979). The People of Roman Britain. London: Batsford.
218

218    Hella Eckardt and Gundula Müldner

Brather, S. (2000). ‘Ethnische Identitäten als Konstrukte der frühgeschichtlichen Archäologie’,

Germania, 78: 139–​177.
Brather, S. (2004). Ethnische Interpretationen in der frühgeschichtlichen Archäologie: Geschichte,
Grundlagen und Alternativen. Berlin: Walther de Gruyter.
Brettell, R., Montgomery, J., and Evans, J. (2012). ‘Brewing and Stewing:  The Effect of
Culturally-​Mediated Behaviour on the Oxygen Isotope Composition of Ingested Fluids and
the Implications for Human Provenance Studies’, Journal of Analytical Atomic Spectrometry,
27/​5: 778–​785.
Bruun, C. (2010). ‘Water, Oxygen Isotopes and Immigration to Ostia-​Portus’, Journal of Roman
Archaeology, 23: 109–​132.
Budd, P., Millard, A., Chenery, C., Lucy, S., and Roberts, C. (2004). ‘Investigating Population
Movement by Stable Isotope Analysis: A Report from Britain’, Antiquity, 78: 127–​141.
Carroll, M. (2006). Spirits of the Dead: Roman Funerary Commemoration in Western Europe.
Oxford: Oxford University Press.
Casella, E. C. (2005). ‘Prisoner of His Majesty: Post-​Coloniality and the Archaeology of British
Penal Transportation’, World Archaeology, 37/​3: 453–​467.
Chenery, C. A., Müldner, G., Evans, J., Eckardt, H., Leach, S., and Lewis, M. (2010). ‘Strontium
and Stable Isotope Evidence for Diet and Mobility in Roman Gloucester, UK’, Journal of
Archaeological Science, 37: 150–​163.
Chenery, C., Müldner, G., and Eckardt, H. (2011). ‘Cosmopolitan Catterick? Isotopic Evidence
for Population Mobility on Rome’s Northern Frontier’, Journal of Archaeological Science,
38: 1525–​1536.
Clarke, G. (1979). Pre-​Roman and Roman Winchester. Part 2: The Roman Cemetery at Lankhills.
Oxford: Clarendon Press.
Clay, C. (2007). ‘Before there were Angles, Saxons and Jutes: An Epigraphic Study of the
Germanic Social, Religious and Linguistic Relations on Hadrian’s Wall’, in L. Gilmour
(ed.), Pagans and Christians:  From Antiquity to the Middle Ages. Papers in Honour
of Martin Henig, Presented on the Occasion of his 65th birthday. Oxford:  Tempus
Reparatum, 47–​63.
Cohen, R. (2008). Global Diasporas: An Introduction. London: Routledge.
Cool, H. (2004). The Roman Cemetery at Brougham, Cumbria:  Excavations 1966–​ 67.
London: Society for the Promotion of Roman Studies.
Cool, H. E.  M. (2006a). Eating and Drinking in Roman Britain. Cambridge:  Cambridge
University Press.
Cool, H. (2006b). ‘Sustenance in a Strange Land’, in P. Ottaway (ed.), A Victory Celebration:
Papers on the Archaeology of Colchester and Late Iron-​Age-​Roman Britain presented to Philip
Crummy. Colchester: Colchester Archaeological Trust, 75–​82.
Cool, H. E. M. (2010). ‘Finding the Foreigners’, in H. Eckardt (ed.), Roman Diasporas: Archaeological
Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology
Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 27–​44.
Curta, F. (2007). ‘Some Remarks on Ethnicity in Medieval Archaeology’, Early Medieval
Europe, 15/​2: 159–​185.
Daux, V., Lécuyer, C., Héran, M.-​A., Amiot, R., Simon, L., Fourel, F., Martineau, F., Lynnerup,
N., Reychler, H., and Escarguel, G. (2008). ‘Oxygen Isotope Fractionation between Human
Phosphate and Water Revisited’, Journal of Human Evolution, 55/​6: 1138–​1147.
Derks, T., and Roymans, N. (2009) (eds). Ethnic Constructs in Antiquity: The Role of Power and
Tradition. Amsterdam: Amsterdam University Press.
 Mobility, Migration, and Diasporas in Roman Britain    219

Eckardt, H. (2010). ‘A Long Way from Home: Diaspora Communities in Roman Britain’, in

H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in
the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth,
RI: Journal of Roman Archaeology, 99–​130.
Eckardt, H. (2012). ‘Foreigners and Locals in Calleva’, in M. Fulford (ed.), Silchester and the
Study of Romano-​British Urbanism. Journal of Roman Archaeology Supplementary Series
89. Portsmouth, RI: Journal of Roman Archaeology, 246–​256.
Eckardt, H., Booth, P., Chenery, C., Müldner, G., Evans, J. A., and Lamb, A. (2009). ‘Oxygen and
Strontium Isotope Evidence for Mobility in Roman Winchester’, Journal of Archaeological
Science, 36: 2816–​2825.
Evans, J., Stoodley, N., and Chenery, C. (2006). ‘A Strontium and Oxygen Isotope Assessment
of a Possible Fourth Century Immigrant Population in a Hampshire Cemetery, Southern
England’, Journal of Archaeological Science, 33: 265–​272.
Evans, J. A., Chenery, C. A., and Montgomery, J. (2012). ‘A Summary of Strontium and Oxygen
Isotope Variation in Archaeological Human Tooth Enamel Excavated from Britain’, Journal
of Analytical Atomic Spectrometry, 27/​5: 754–​764.
Fitzpatrick, A. P. (2011). The Amesbury Archer and the Boscombe Bowmen: Bell Beaker Burials at
Boscombe Down, Amesbury, Wiltshire. Salisbury: Wessex Archaeology.
Fulford, M. (2010). ‘Roman Britain: Immigration and Material Culture’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman
Archaeology, 67–​78.
Gardner, A. (2007). An Archaeology of Identity: Soldiers and Society in Late Roman Britain.
Walnut Creek, CA: Leftcoast Press.
Gosden, C. (2006). ‘Race and Racism in Archaeology: Introduction’, World Archaeology, 38/​1: 1–​7.
Gray, C. L. (2011). ‘Foreigners in the Burial Ground: The Case of the Milesians in Athens’, in
M. Carroll and J. Rempell (eds), Living through the Dead: Burial and Commemoration in the
Classical World. Oxford: Oxbow Books, 47–​64.
Handley, M. (2011). Dying on Foreign Shores: Travel and Mobility in the Late Antique West.
Portsmouth, RI: Journal of Roman Archaeology.
Härke, H. (2007). ‘Ethnicity, “Race” and Migration in Mortuary Archaeology:  An Attempt
at a Short Answer’, in S. Semple and H. Williams (eds), Early Medieval Mortuary Practices.
Oxford: Oxbow Books, 12–​18.
Hingley, R. (2010). ‘Tales of the Frontier: Diasporas on Hadrian’s Wall’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman
Archaeology, 227–​243.
Hingley, R. (2012). Hadrian’s Wall: A Life. Oxford: Oxford University Press.
Hoppe, K. A., Koch, P. L., and Furutani, T. T. (2003). ‘Assessing the Preservation of Biogenic
Strontium in Fossil Bones and Tooth Enamel’, International Journal of Osteoarchaeology, 13/​
1–​2: 20–​28.
Isaac, B. (2004). The Invention of Racism in Classical Antiquity. Princeton:  Princeton
University Press.
Isaac, B. (2006). ‘Proto-​Racism in Graeco-​Roman Antiquity’, World Archaeology, 38/​1: 32–​47.
Ivleva, T. (2010). ‘British Emigrants in the Roman Empire: Complexities and Symbols of Ethnic
Identities’, in D. Mladenović and B. Russell (eds), TRAC 2010: Proceedings of the Twentieth
Theoretical Roman Archaeology Conference. Oxford: Oxbow Books, 132–​153.
220

220    Hella Eckardt and Gundula Müldner

Jones, S. (1997). The Archaeology of Ethnicity: Constructing Identities in the Past and Present.
New York and London: Routledge.
Jones, B., and Mattingly, D. (2002). An Atlas of Roman Britain. Oxford: Oxbow Books.
Kakoschke, A. (2002). Ortsfremde in den römischen Provinzen Germania inferior und
Germania superior. Möhnesee: Bibliopolis.
Killgrove, C. (2010a). ‘Response to Bruun, 2010’, Journal of Roman Archaeology, 23: 133–​136.
Killgrove, K. (2010b). ‘Identifying Immigrants to Imperial Rome Using Strontium Isotope
Analysis’, in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and
Diversity in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78.
Portsmouth, RI: Journal of Roman Archaeology, 157–​175.
Knudson, K. J. (2009). ‘Oxygen Isotope Analysis in a land of Environmental Extremes: The
Complexities of Isotopic Work in the Andes’, International Journal of Osteoarchaeology,
19: 171–​191.
Lawrence, S. (2003). ‘Archaeology and the Nineteenth-​Century British Empire’, Historical
Archaeology, 37/​1: 20–​33.
Leach, S., Lewis, M., Chenery, C., Eckardt, H., and Müldner, G. (2009). ‘Migration and
Diversity in Roman Britain: A Multidisciplinary Approach to Immigrants in Roman York,
England’, American Journal of Physical Anthropology, 140: 546–​561.
Leach, S., Eckardt, H., Chenery, C., Müldner, G., and Lewis, M. (2010). ‘A “Lady” of
York: Migration, Ethnicity and Identity in Roman York’, Antiquity, 84: 131–​145.
Lilley, I. (2004). ‘Diaspora and Identity in Archaeology: Moving beyond the Black Atlantic’, in
L. Meskell and R. W. Preucel (eds), A Companion to Social Archaeology. Oxford: Blackwell,
287–​312.
Lilley, I. (2006). ‘Archaeology, Diaspora and Decolonization’, Journal of Social Archaeology, 6/​
1: 28–​47.
McCoskey, D. E. (2012). Race—​Antiquity and its Legacy. London: I. B. Tauris.
Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of
Roman Archaeology, 17: 5–​25.
Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409.
London: Penguin Books.
Meskell, L. (2001). ‘Archaeologies of Identity’, in I. Hodder (ed.), Archaeological Theory Today.
Cambridge: Polity, 187–​213.
Mihailescu-​ Bîrliba, L. (2011). Ex Toto Orbe Romano:  Immigration into Roman Dacia.
Leuven: Peeters.
Millett, M. (1990). The Romanization of Britain. Cambridge: Cambridge University Press.
Millett, M. (1999). ‘Coloniae and Romano-​British Studies’, in H. Hurst (ed.), The Coloniae of
Roman Britain: New Studies and a Review. Journal of Roman Archaeology Supplementary
Series 36. Portsmouth, RI: Journal of Roman Archaeology, 191–​196.
Montgomery, J. (2010). ‘Passports from the Past:  Investigating Human Dispersals Using
Strontium Isotope Analysis of Tooth Enamel’, Annals of Human Biology, 37/​3: 325–​346.
Montgomery, J., Evan, J., Chenery, C., Pashley, V., and Killgrove, K. (2010). ‘“Gleaming white
and deadly”: Using Lead to Track Human Exposure and Geographic Origins in Britain’, in
H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in
the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth,
RI: Journal of Roman Archaeology, 199–​226.
Müldner, G. (2013). ‘Stable Isotopes and Diet: Their Contribution to Romano-​British Research’,
Antiquity, 87/​335: 137–​149.
 Mobility, Migration, and Diasporas in Roman Britain    221

Müldner, G., Chenery, C., and Eckardt, H. (2011). ‘“The Headless Romans”:  Multi-​Isotope
Investigations of an Unusual Burial Ground from Roman Britain’, Journal of Archaeological
Science, 38: 280–​290.
Noy, D. (2000). Foreigners at Rome: Citizens and Strangers. London: Duckworth.
Noy, D. (2010). ‘Epigraphic Evidence for Immigrants at Rome and in Roman Britain’, in H.
Eckardt (ed.), Roman Diasporas:  Archaeological Approaches to Mobility and Diversity in
the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth,
RI: Journal of Roman Archaeology, 13–​26.
Parker Pearson, M., Richards, M. P., Chamberlain, A., and Jay, M. (forthcom-
ing) (eds). The Beaker People:  Isotopes, Mobility and Diet in Prehistoric Britain.
Oxford: Oxbow Books.
Pearce, J. (2010). ‘Burial, Identity and Migration in the Roman World’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman
Archaeology, 79–​98.
Perry, M. A., Coleman, D., and Delhopital, N. (2008). ‘Mobility and Exile at Second Century ad
Khirbet edh-​Dharih: Strontium Isotope Analysis of Human Migration in Western Jordan’,
Geoarchaeology, 23/​4: 528–​549.
Perry, M. A., Coleman, D. S., Dettmann, D. L., and Halim al-​Shiyab, A. (2009). ‘An Isotopic
Perspective on the Transport of Byzantine Mining Camp Laborers into South-​Western
Jordan’, American Journal of Physical Anthropology, 140: 429–​441.
Perry, M. A., Coleman, D. S., Dettman, D. L., Grattan, J. P., and Halim al-​Shiyab, A. (2011).
‘Condemned to Metallum? The Origin and Role of 4th–​6th Century ad Phaeno Mining
Camp Residents Using Multiple Chemical Techniques’, Journal of Archaeological Science, 38/​
3: 558–​569.
Pohl, W. (1998). ‘Telling the Difference: Signs of Ethnic Identity’, in W. Pohl and H. Reimitz
(eds), Strategies of Distinction:  The Construction of Ethnic Communities, 300–​800.
Leiden: Brill, 1–​69.
Pollard, A. M. (2011). ‘Isotopes and Impact: A Cautionary Tale’, Antiquity, 85: 631–​638.
Pollard, A. M, Ditchfield, P., McCullagh, J. S. O., Allen, T. G., Gibson, M., and Boston, C.
(2011a). ‘“These boots were made for walking”:  The Isotopic Analysis of a C4 Roman
Inhumation from Gravesend, Kent, UK’, American Journal of Physical Anthropology,
146: 446–​456.
Pollard, A. M., Pellegrini, M. and Lee-​Thorp, J. A. (2011b). ‘Technical Note: Some Observations
on the Conversion of Dental Enamel Delta O-​18(p) Values to delta O-​18(w) to Determine
Human Mobility’, American Journal of Physical Anthropology, 145/​3: 499–​504.
Price, T. D., Burton, J. H., Sharer, R. J., Buikstra, J. E., Wright, L. E., Traxler, L. P., and Miller, K.
A. (2010). ‘Kings and Commoners at Copan: Isotopic Evidence for Origins and Movement
in the Classic Maya Period’, Journal of Anthropological Archaeology, 29/​1: 15–​32.
Prowse, T., Schwarcz, H. P., Garnsey, P., Knyf, M., Macciarelli, R., and Bondioli, L. (2007).
‘Isotopic Evidence for Age-​Related Immigration to Imperial Rome’, American Journal of
Physical Anthropology, 132: 510–​519.
Prowse, T. L., Barta, J. L., von Hunnius, T. E., and Small, A. M. (2010). ‘Stable Isotope and
Mitochondrial DNA Evidence for Geographic Origins on a Roman Estate at Vagnari (Italy)’,
in H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity
in the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth,
RI: Journal of Roman Archaeology, 175–​197.
222

222    Hella Eckardt and Gundula Müldner

Rowland, R. J. (1976). ‘Foreigners in Roman Britain’, Acta Archaeologica Academiae

Scientiarium Hungaricae, 28: 443–​447.
Roymans, N. (2004). Ethnic Identity and Imperial Power: The Batavians in the Early Roman
Empire. Amsterdam: Amsterdam University Press.
Scheidel, W. (2004). ‘Human Mobility in Roman Italy, I: The Free Population’, Journal of Roman
Studies, 94: 1–​26.
Schweissing, M. M., and Grupe, G. (2003). ‘Stable Strontium Isotopes in Human Teeth
and Bone:  A  Key to Migration Events of the Late Roman Period in Bavaria’, Journal of
Archaeological Science, 30: 1373–​1383.
Sherwin-​White, A. N. (1967). Racial Prejudice in Imperial Rome. Cambridge:  Cambridge
University Press.
Snowden, F. M. (1983). Before Color Prejudice:  The Ancient View of Blacks. Cambridge,
MA: Harvard University Press.
Stockhammer, P. F. (2012). ‘Conceptualising Cultural Hybridization in Archaeology’, in P. F.
Stockhammer (ed.), Conceptualising Cultural Hybridization: A Transdisciplinary Approach.
Heidelberg: Springer-​Verlag, 43–​58.
Swan, V. G. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman
Pottery Studies, 5: 1–​33.
Swan, V. G. (1999). ‘The Twentieth Legion and the History of the Antonine Wall Reconsidered’,
Proceedings of the Society of Antiquaries of Scotland, 129: 399–​480.
Swan, V. G. (2009). Ethnicity, Conquest and Recruitment: Two Case Studies from the Northern
Military Provinces. Journal of Roman Archaeology Supplementary Series 72. Portsmouth,
RI: Journal of Roman Archaeology.
Swift, E. (2010). ‘Identifying Migrant Communities:  A  Contextual Analysis of Grave
Assemblages from Continental Late Roman Cemeteries’, Britannia, 41: 237–​282.
Tolia-​Kelly, D. (2010). ‘Narrating the Postcolonial Landscape:  Archaeologies of Race at
Hadrian’s Wall’, Transactions of the Institute of British Geographers, ns 36: 71–​88.
Van Dommelen, P., and Terrenato, N. (2007). ‘Local Cultures and the Expanding Roman
Republic’, in P. Van Dommelen and N. Terrenato (eds), Articulating Local Cultures: Power
and Identity under the Expanding Roman Republic. Journal of Roman Archaeology
Supplementary Series 63. Portsmouth, RI: Journal of Roman Archaeology, 7–​12.
Voss, B. (2005). ‘The Archaeology of Overseas Chinese Communities’, World Archaeology, 37/​
3: 424–​439.
Webster, J. (2001). ‘Creolizing Roman Britain’, American Journal of Archaeology, 105/​
2: 209–​225.
Webster, J. (2010). ‘Routes to Slavery in the Roman World:  A  Comparative
Perspective on the Archaeology of Forced Migration’, in H. Eckardt (ed.), Roman
Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
Journal of Roman Archaaeology Supplementary Series 78. Portsmouth, RI: Journal of
Roman Archaeology, 45–​65.
West, J. B., Ehleringer, J. R., and Cerling, T. E. (2007). ‘Geography and Vintage Predicted by a
Novel GIS Model of Wine δ18O’, Journal of Agricultural and Food Chemistry, 55: 7075–​7083.
Whittle, A., and Bickle, P. (2013) (eds). The First Farmers of Central Europe: Diversity in LBK
Lifeways. Oxford: Oxbow Books.
Wierschowski, L. (1995). Die regionale Mobilität in Gallien nach den Inschriften des 1.  bis
3. Jahrhunderts n. Chr. Stuttgart: Franz Steiner Verlag.
 Mobility, Migration, and Diasporas in Roman Britain    223

Wierschowski, L. (2001). Fremde in Gallien—​‘Gallier’ in der Fremde: Die epigraphisch bezeugte

Mobilität in, von und nach Gallien vom 1. bis 3. Jh. n. Chr. Stuttgart: Franz Steiner Verlag.
Wilkie, L. A. (2004). ‘Considering the Future of African-​American Archaeology’, Historical
Archaeology, 38/​1: 109–​123.
Woolf, G. 1998. Becoming Roman:  The Origins of Provincial Civilization in Gaul.
Cambridge: Cambridge University Press.
224

Chapter 11

Multicu ltu ra l i sm
on Hadrian’ s  Wa l l

Claire Nesbitt

Introduction

During much of the Roman occupation of Britain a great wall marked the northern
frontier of the Roman Empire. This frontier, running the width of the province across
the Tyne–​Solway isthmus, delimited the northern boundary of empire from ad 122 to
c. ad 142, when occupation was briefly pushed north to the central belt of Scotland and
demarcated by the Antonine Wall. The Wall became the frontier again from c. ad 160,
when the Roman military withdrew from Scotland, and it remained in use as a frontier
or barrier until the fifth century. Construction began in ad 122, during the reign of the
emperor Hadrian after whom the Wall is now named, with a turf and stone wall, which
was later replaced entirely with stone. The Wall has been a focus of study and a destina-
tion for antiquarians, archaeologists, and the public for around 200 years.
Throughout that time the monument has been appropriated by different groups
for different socio-​political and economic reasons (see Nesbitt and Tolia-​Kelly 2009;
Witcher et al. 2010; Hingley 2012). In the modern popular imagination, Hadrian’s Wall
represents a militarized zone stamped with the indelible impression of ‘Rome’ (see
Witcher 2010). Indeed, archaeological investigation of the Wall has focused heavily on
its military aspects (e.g. Divine 1969; Poulter 2009). However, as well as being a mili-
tary frontier, the Wall zone also represents a contested landscape of polyvocalities (see
Hingley and Hartis 2011). The Wall landscape has stories of a pre-​Roman heritage to
tell, biographies of peoples who were not Roman and, perhaps, not native to the region.
These individuals and groups are often eclipsed in archaeological studies by the shadow
of the Roman military machine. This chapter will consider the multicultural nature of
Hadrian’s Wall; that is to say, it will focus on the presence and influence of ‘otherness’
evidenced in the material culture and epigraphy left to us in the archaeological record.
It will investigate what Hingley (2010: 229) has described as the ‘dichotomy between the
 Multiculturalism on Hadrian’s Wall    225

Wall’s nature as a contested landscape and the inclusivity suggested by the material cul-
ture’ and it will focus on concepts of identity as expressed through materiality. Before
we turn to the archaeological evidence, however, I would like first to establish what we
mean by a ‘foreign’ presence in an empire as vast and diverse as Rome, and to spend
some time thinking through how we understand ‘multiculturalism’ and ‘identity’ in the
archaeological past.

What is Multiculturalism?

Multiculturalism is a word that is often used but seldom explained. Its definitions vary
considerably from Modood’s interpretation (2007) that ‘multiculturalism reflects the
cultural and racial diversity of a society and acknowledges the freedom of all members
of that society to preserve, enhance and share their cultural heritage’ (Modood 2007:
16), to Adler’s understanding (1976) that the multicultural citizen has indefinite bounda-
ries of the self, an identity that is in permanent flux and always being reformulated in
response to cultural input. Steinberg and Kincheloe (2009: 4) argue for several incar-
nations of multiculturalism. One extreme, taking a conservative perspective, expounds
occidental superiority, which entails an assimilation to western culture by other soci-
eties, and is essentially a promotion of a monoculture rather than multiculturalism;
another, a pluralist multiculturalism, focuses on race, class, and gender differences, pro-
motes pride in group heritage, and avoids the use of oppression.
So what do we mean when we speak of a multicultural society in the Roman Empire?
The variety of cultural and racial identities found across the Roman Empire has long
been recognized. Witcher has argued that ‘the key to any investigation of Roman Italy
must surely lie in the articulation of these multiple identities—​national, regional and
local; social, political and ethnic’ (Witcher 2000: 214; emphasis added). This might just
as effectively be applied empire-​wide. The crucial factor in discussing multiculturalism
in the Roman world is the definition of multiple individual and group identities within
the context of a unifying administration and the fluidity of cultural expression that this
allows.

Shaping Identities in a Frontier Zone

Identity has been defined as an ‘individual’s identification with broader groups on the
basis of differences socially sanctioned as significant’ (Diaz-​Andreu and Lucy 2005: 1).
Identity is made up of layers of different social signifiers including gender, religion, age,
and ethnicity. Here I follow Lucy’s premise (2005: 86) that some aspects of identity, such
as ethnicity, are ways of behaving rather than inherent qualities and, as such, must be
learned and can cross-​cut other aspects of identity. Ethnicity also ‘involves the social
226

226   Claire Nesbitt

negotiation of difference and sameness, and often entails larger tensions between indi-
viduals, the group, and the state’ (Meskell 2007: 25).
Who, then, do we expect to be able to identify in the archaeological record of the Wall
zone? There are several different groups that we might expect to recognize in the mate-
rial culture of Hadrian’s Wall. The indigenous communities who existed in the region at
the time of the Wall’s construction, for example; also the Roman military and perhaps
other groups and identities associated with the military. Alongside this we might expect
to see a range of individuals and communities who came to the region through trade,
employment, or slavery.
Traditionally the ‘Britons’ are the indigenous people who occupied the Wall zone
before the Roman conquest and probably continued to live there, though Hodgson et
al. (2012) have argued for a widespread dislocation and abandonment of sites on the
Northumberland coastal plain in the early Roman period. While the exact nature of
the identities of the pre-​Roman Iron Age peoples in the frontier zone has seldom been
explored in any detail, indigenous peoples have traditionally been assigned to tribal
groups based on the evidence from classical sources, most notably Ptolemy’s Geographia
and Tacitus’ Agricola (Breeze 1982: 28–​36). These have been identified as the Brigantes,
the Carvetii, and, more recently, the Anavionenses (Hunter 2007b: 7–​8).
Increasingly, there has been recognition that such tribal models potentially mask
a more complex socio-​political landscape and range of identities in the late Iron Age,
and that these were in a state of reconfiguration at the time of Roman conquest (Moore
2011). In his work on the expression of identity through material culture in frontier
zones, Hunter (2001, 2007a) has considered the adoption of Roman material culture by
native society as prestige objects or objects of power, particularly in relation to iden-
tity in northern Britain. Hunter (2001) concludes that social control of Roman artefacts
ensured that foreign or exotic objects had a complex role in social display, bribery, and
gift-​giving in the frontier zone. Hunter’s studies also demonstrate that identities in the
region were in a state of transformation at the time of the Wall’s construction and use
(Hunter 2007a). This process had distinct regional differences, however; in Scotland,
trends in material culture were towards a distinctly local style, seen very much in oppo-
sition to external forces. Meanwhile, in the frontier zone, material that began as an
expression of local identity was absorbed into a Romano-​British milieu and interpreted
differently by different people (Hunter 2007a: 293–29​4). Concepts of identity, then,
appear to have been fluid, and there may have been a blurring of cultural distinction in
areas like the frontier zone, as demonstrated by the inclusion of Celtic-​style motifs on
Roman objects (Hunter 2007a: 292). Cultural influence is not a one-​way process, and
the frontier zone can be seen as a melting pot of peoples, ideas, and cultures all in a
dynamic state of reshaping themselves into a Romano-​British ‘frontier’ culture. Moore
(2011) has argued that, in the later pre-​Roman Iron Age, rather than being a traditional
tribalized society, communities were reshaping their identities in response to the expan-
sion of Rome, perhaps in order to establish themselves in society in a new way through
their interactions with the invading peoples. In this way, groups along the frontier may
have re-​evaluated their own identity and perhaps even reshaped it in response to Rome’s
 Multiculturalism on Hadrian’s Wall    227

presence. Moore (2011: 347) cautions against confusing identities, as expressed in the
Roman province, with those situated in a pre-​Roman context because of the influence of
Roman power.
In addition, it seems more likely that most of these individuals and communities that
already existed in the Wall zone would have identified themselves more at the level of kin
group, settlement community, or sets of social and exchange networks. Archaeological
investigations of pre-​Roman sites in the vicinity of the Wall are relatively few in number,
reflecting the tendency to focus on military infrastructure, rather than evaluating the
social identity of the peoples who lived there. As a consequence, our understanding of
how pre-​Roman indigenous people in the region constructed their identities is unclear,
although almost certainly this was at a scale far smaller than at the level of a ‘tribe’ like
the Brigantes. Anderson’s analysis (2011) of regional ceramics from the later pre-​Roman
Iron Age and post-​conquest period suggests that, while there is an identifiable shift in
focus from preparation to consumption involving more ceramic vessels, overall the
ceramic record reveals that identity was expressed at the level of small-​scale groups. At
the same time, it is easy to underestimate the impact of Roman occupation on the exist-
ing population. Breeze (1990: 86) has noted the myriad ways in which society must have
been affected, through taxation, education, language, conscription, alternative farming
practices, redistribution of the population in terms of settlement patterns, the structure
of society, and, indeed, the size of the population. These factors could easily lead to a
complete transfiguring of a sense of individual, local, and regional identity, which may
have been played out in the abandonment and dislocation that Hodgson et al. (2012)
have suggested.

Military Identities: Join
the Army—​See the World?

By the end of the second century ad each legion was recruited mainly from the province
in which it was stationed (Dobson and Mann 1973: 193). That meant that recruitment of
legionaries from Britain was usual from the same time (Swan and Monaghan 1993: 26),
while auxiliary units were generally raised from single groups of peoples or from prov-
inces (Dobson and Mann 1973: 193). The civil settlements that developed around the
military garrisons probably gradually became pools of hereditary recruits who provided
most of the soldiery for the third century and later still on the frontier when heredi-
tary military service for the sons of servicemen became compulsory (Dobson and Mann
1973). The military is one of the more obvious and easily traceable ways to find a foreign
presence on the frontier, though what that means in terms of cultural identity is not
straightforward and will be discussed further below.
Book XL of the Notitia Dignitatum, originally compiled in the early fifth century, lists
the Item per lineam valli—​the route along the line of the Wall. The book lists the forts and
228

228   Claire Nesbitt

the units serving at them and this, together with epigraphic evidence, has enabled us to
people the landscape with the foreign military units. From this evidence we can identify
units from all over Europe, the Middle East, and North Africa (Figure 11.1 and Table
11.1). It would be easy to group the units by their nationality and look for a cultural iden-
tity in the materiality of their locations that reflects this. However, as has been argued,
identity is not as straightforward as that. Gardner (2002: 338) sees the boundedness of
the military communities as ‘blurred by the presence of people who were not members
in the organizational sense, but were nonetheless strongly tied to the soldiers, through
relations of friendship, marriage, enslavement, clientship or kinship’. This introduces a
tension between the complex and perhaps multiple identities that were enacted in the
frontier zone (Gardner 2002: 338). Mattingly has argued that ‘military identity was far
from static and was continually refashioned, inevitably influenced by the cultural back-
ground of the soldiery’ (Mattingly 2010: 167). James (2011: 28) notes that at the height of
empire around half of Rome’s soldiers were provincials who had acquired citizenship
by virtue of their military careers and argues for a particular military-​centred, frontier-​
focused ‘Romanness’ that was created out of and probably transcended the ethnic diver-
sity. Great cultural variability reflected the diversity and humanity of the army, but there
were also ‘significant common patterns of behaviour and social outlook that made the
army an empire-wide and self-​aware identity group of commilitones—​“fellow soldiers”’
(Mattingly 2010: 199). What this means, in terms of the identities of the soldiers, is that,
when there were familial or local connections with them in their day-​to-​day existence,
their unit identity may have been second to other complex identities shaped by their
interactions with other, non-​military companions. James (2002: 39) has viewed sol-
diers ‘not as automata, cogs in the machine, but as self-​aware social agents’ and chal-
lenges the military/​civilian dichotomy that is the established framework of research.
Gardner echoes this, suggesting that ‘in the daily lives of such people, unit, rank and
status identities are likely to have been more significant than an empire-​wide “militari-
ness”’ (Gardner 2002: 338). So though we can identify people from all over the empire,
their cultural identity may not have remained as rigidly military, national, or ethnic as
we may imagine.
It has been argued that, within a generation of a regiment leaving its area of origin, its
ethnic title became meaningless (Mann 1963). However, Haynes (1999: 165) understands
the military culture differently, preferring to see regimental communities as more than
simply militarized; rather, the units were exposed to a range of symbolism, experiences,
values, ideas, and artefacts that were not exclusively military. In fact, he, like Gardner,
argues that civilian followers probably countered the military culture with other local
traditions that were maintained. The dilution of a culture through connection with the
military does not necessarily indicate an abandonment of national or ethnic traditions.
Haynes (1999: 167) cites evidence for the survival of ethnic tradition in the vici from
Birdoswald, where a third-​century inscription commemorates Decebalus, a boy with
a distinctively Dacian name (RIB 1920). The fact that this name was still in use at the
site 50–​100 years after the Dacians serving in Cohors I Aelia Dacorum arrived there
suggests that at least some aspects of Dacian identity survived in the vicus beyond the
Figure 11.1  Map showing areas of the Roman Empire with military units on Hadrian’s Wall.
Source: © Rob Witcher, illustration by Christina Unwin.
230

Table 11.1 Foreign units serving on Hadrian’s Wall and

their provincial origins
Provincial origin of unit Unit name

Britannia I Cornoviorum
Hispania Ala I Hispanorum Asturum
Ala II Asturum
I Aelia Hispanorum equitata
II Asturum equitata
I Fida Vardullorum equitata
Gallia Ala Augusta Gallorum Petriana
I Aquitanorum equitata
IV Gallorum equitata
V Gallorum equitata
Gallia Belgica I Batavorum equitata
II Nerviorum
VI Nerviorum
I Tungrorum
II Tungrorum equitata
Germania I Baetasiorum
I Lingonum equitata
II Lingonum equitata
IV Lingonum equitata
I Nervana Germanorum equitata
Numerus Hnaudifridi
I Ulpia Traiana Cugernorum
I Vangionum equitata
Frisia Cuneus Frisionum Aballavensium
Cuneus Frisiorum Vercovicianorum
I Frisiavonum
Raetia Raeti gaesati
Cohors Raetorum
Vexillatio gaesatorum Raetorum
Pannonia Ala I Pannoniorum Sabiniana
(continued)
 Multiculturalism on Hadrian’s Wall    231

Table 11.1 Continued
Provincial origin of unit Unit name

I Pannoniorum equitata
II Pannoniorum equitata
Dalmatia I Delmatarum equitata
II Delmatarum equitata
Dacia I Aelia Dacorum
Thracia I Thracum
II Thracum equitata
Syria I Hamiorum sagittariorum
Mesopotamia Numerus barcariorum Tigrisiensium
Mauretania Numerus Maurorum Aurelianorum

Source: © Claire Nesbitt.

first generation of non-​native troops (Haynes 1999: 167). This should not be surprising:
belonging to a military cohort and the unity that this embodies can operate completely
independently of ethnic or cultural heritage, as James (2011: 28) has suggested. The regi-
mental customs of the unit, while having their origins in localized traditions, could be
adopted wholeheartedly by new (and foreign) recruits as part of a recognized unit cul-
ture and tradition to which they could aspire and be proud to belong. Perhaps, rather
than looking at what distinguished the multinational identities in the frontier zone, we
ought to consider what it was that united them, the sense of being ‘Roman’.

Shades of Romanness

Much consideration has been given to concepts of Romanization since Haverfield’s ini-
tial work (1912) on the subject. Millett (1990: 212) has conceived a framework within
which Romanization is seen ‘not as a passive reflection of change, but rather as an active
ingredient used by people to assert, project and maintain their social status’. Woolf
(1998: 7) argues that ‘there was no standard Roman civilization against which provincial
cultures might be measured’. He adds that Romanization did not culminate in a uniform
culture throughout the empire (Woolf 1998: 7). In Woolf ’s view, ‘becoming Roman did
not involve becoming more alike the other inhabitants of the empire, so much as par-
ticipating in a cultural system structured by systematic differences, differences that both
sustained and were a product of Roman power’ (Woolf 1998: 242).
232

232   Claire Nesbitt

The Rome of the ancient world can be described as an idea, a concept, rather than a
locus that can be pointed to or visited. As an empire Rome was necessarily multicultural,
absorbing so many nations and peoples. This meant, inevitably, that the Senate and
Imperial palace were themselves international, so that ‘Romanness’ was no longer deter-
mined by nationality or ethnicity but by citizenship and sharing in the ‘idea’ of Rome. In
this sense it is possible to be completely foreign and completely Roman at the same time,
to have multiple identities that cross traditional boundaries, because the growth of the
Roman Empire, as Woolf (1997: 347) has argued, was not ‘the expansion of one national
or ethnic culture at the expense of others’, but rather a new social formation that incor-
porated new cultural logic and new configurations of power.
Because of the ill-​defined and blurred nature of the cultural and ethnic implications
of the term ‘Roman’, the Roman attitude to other peoples focused more on civilization
and citizenship than it did on nationality or ethnicity. Citizenship was a crucial part of
‘Romanness’ or Romanitas; being a Roman citizen—​Civis Romanus—​meant belong-
ing to the Roman state, and it entailed certain rights and privileges. Citizenship, how-
ever, was a complex concept; there were several different levels of citizenship, and these
are not always clearly defined. Full Roman citizens (Cives Romani) were protected by
Roman law (ius civile). One of the best examples of the importance of citizenship can
be drawn from Christian scripture. The Acts of the Apostles describes the events that
unfolded when the Apostle Paul was preaching in Jerusalem. Paul was arrested by the
Roman authorities because the monotheistic doctrine of Christianity was incompatible
with the polytheism of the Roman Empire, meaning that Christians refused to under-
take the required sacrifice to Roman gods. He was to be detained and flogged for break-
ing the law. Paul’s father had been a Roman citizen and Paul inherited this status. This
posed a problem for his captors who had planned to flog him:

Paul said to the centurion who was standing there: ‘Can you legally flog a man who
is a Roman citizen, and moreover has not been found guilty?’ When the centurion
heard this, he went and reported it to the commandant. ‘What do you mean me to
do?’ he said. ‘This man is a Roman citizen.’ The commandant came to Paul. ‘Tell me,
are you a Roman citizen?’ he asked. ‘Yes,’ said he. The commandant rejoined, ‘It cost
me a large sum to acquire this citizenship.’ Paul said, ‘But it was mine by birth.’ (Acts
22: 22–​8)

Two things stand out from this passage. The first is the impact that the realization of
Paul’s citizenship had on his captors and the way it influenced their treatment of him.
The second thing that emerges is an insight into the nature of acquiring citizenship; we
have, on the one hand, Paul, who tells us citizenship was his by right of birth and, on
the other hand, the commandant, who bought his citizenship, no doubt at a high price.
Individuals could be granted citizenship for acts of outstanding service to Rome, and
provinces could be granted citizenship—​usually a limited version, such as Latini. Full
citizenship was often granted to auxiliary soldiers at the end of their military service of
twenty-​five years. Several military diplomas survive that set out the conditions of the
 Multiculturalism on Hadrian’s Wall    233

soldiers’ retirement. With citizenship, whether inherited, purchased, or granted, peo-

ples from all over the empire, of any ethnic background, could rise through the ranks of
Roman society. They were not classified by their nationality, their heritage, or the colour
of their skin; they were citizens of Rome, and this was the common factor that unified
the multicultural empire.
In terms of culture in the Roman Empire, then, Mann has argued that ‘individual
societies do not exist, instead we should think in terms of overlapping social networks of
varying scales relating to different types of social power, ideological, economic, military
or political’ (Mann 1986, cited in Shennan 1989: 11). However, equally it can be argued
that ‘the process of ethnic identity creation only comes to have its power in a situation in
which pre-​existing forms of identity creation and maintenance—​kinship for example—​
are being destroyed’ (Shennan 1989: 16). If Romanness was desired and welcomed, it had
the potential to dilute cultural traditions, to foster inclusivity and a cosmopolitan shared
identity. If, on the other hand, Rome was an unwelcome concept, it is more likely that
we would recognize traditional cultures being expressed, revived, and even reinvented.
It is clear, then, that identity can be a changing concept and one that can be manipu-
lated by individuals throughout life and in response to different circumstances. What
emerges from our understanding of identity is that multiculturalism may not be as easy
to see in the archaeological record as we may hope and is not as simple as identifying
well-​defined cultural or ethnic groups. If we are to seek to understand different cultures
and identities through material culture, it is important to keep in mind these fluid con-
cepts of identity, the projection of self-​identity, and how that is embedded in materiality.

Materiality of Multiculturalism

What, then, of other groups and cultural identities along Hadrian’s Wall and how might
we see identity in the materiality of the Wall zone? Lazzari (2005: 127) has argued that
materiality is a dialogue between people and objects that entails continual reflection, and
there is a cyclical relationship between the way people make things and the way things
make people. Miller suggests that ‘the importance of an engagement with the concept of
materiality for the archaeologist is that it helps us to grant to those we study a perspicac-
ity equivalent to that which we seek for ourselves’ (Miller 2005: 213). In terms of how
we understand multiculturalism, materiality is crucial, as it is often the lens through
which we see people as ‘other’. The difficulty in using materiality to understand the past
is that materiality ‘is not reducible to a set of given conditions or practices common to all
cultures and all times, it is a set of cultural relationships’ (Meskell 2005: 6). In attempt-
ing to disentangle the meaning in the materiality, we are trying to interrogate specific
moments of creation, exchange, use, and discarding of objects (Meskell 2005). Meskell
(2005) maintains that we conceive and construct the material world ourselves; within
this world we can be ‘constrained or triggered by objects and features, consciously or
unconsciously. They may be produced or appropriated with specific intentions and yet
234

234   Claire Nesbitt

influence future actions in an unpredictable way’ (Fahlander and Oestigaard 2008: 4).

In this way, material objects have the potential to be active in the sense of stimulating
or determining social action (Fahlander and Oestigaard 2008:  4). Through studying
objects that belonged to, were created by, or were buried with individuals, then, we can
begin to understand the creation and projection of social identities (Dellino-​Musgrave
2005: 220). The case studies below will focus on two different aspects of identity: the
identity that is lived through material culture during life and the identity that is pro-
jected by material culture at death.

Materiality of Life

The materiality of the Wall has suggested a diversity of groups were present there, rep-
resented by military units that included Frisians, Batavians, Dacians, Pannonians,
Raetians, and Thracians, to name a few (see Breeze 2006). One heavily contested ele-
ment of the multicultural nature of the Wall zone is the presence of communities and
individuals from North Africa. In ad 208, the emperor Septimius Severus arrived in
Britain to respond to troubles on the northern frontier. He was accompanied by a mili-
tary contingent from outside the province of Britannia, including Legio VI Victrix—​a
legion that almost certainly included soldiers from North Africa (Swan and Monaghan
1993: 26). The presence of North African individuals has also been claimed on the basis
of changes in material culture in parts of the province. Most notable is the appearance,
in York, of a new style of ceramic ware, known as ‘Late Ebor Ware’, manufactured in
York from local fabric but similar in style to pottery manufactured in North African
provinces (Africa Proconsularis, Numidia, and Mauretania: Swan 1992: 6). The vessels,
largely representing cookware, represent a departure from more common Romano-​
British cooking vessels, having uncharacteristic rounded bottoms and lids. Swan has
argued that the manufacture, in significant quantities, of North African-​style vessels
suggests that there were sufficient people who knew how to use them and desired to
use them (Swan 1999; 2009: 54); indeed she has gone so far as to see these as ceramics
made for Africans by Africans, probably soldiers who were garrisoned in the province.
Elsewhere, the presence of African-​style cooking vessels at Segedunum, Wallsend, has
also led to speculation that a unit of Moors may have been stationed there in the third
century ad (Breeze 2006: 134). The implications of such material may be more complex,
however. Fulford (2010: 77) has emphasized the need to be cautious in directly associat-
ing particular forms of pottery with particular ethnic groups; he argues that, while Ebor
ware has close parallels in Tunisian ceramic assemblages, the style was one commonly
seen around the Mediterranean. Its dissemination need not be seen as coming directly
from Tunisia and, even if we accept the presence of African potters, it does not follow
that we can assume an ethnic customer base (Fulford 2010: 75–7​7).
The possible presence in Britain of Africans is also indicated by other evidence: a
unit of North African soldiers (the Numerus Maurorum Aurelianorum) is attested to
 Multiculturalism on Hadrian’s Wall    235

at Burgh-​by-​Sands on Hadrian’s Wall in ad 253–​8 (RIB 2042), and a tombstone record-

ing a North African freedman, Victor (RIB 1064), has also been discovered on the Wall
(see section on Expressions of Individual Identity). Further evidence of a North African
influence, this time in building techniques, may be apparent in structures along the
Wall, with tubuli lingulati, ceramic tubing used in roof vaulting, discovered at Pardshaw,
Cumbria. This technique of roofing is common in North African building and is known
in Britain only from the baths at Caerleon and Chester and a sewer at York (Mason
1990: 221). These three sites are all known to have undergone refurbishments during the
Severan period, when the North African influence is believed to have arrived in Britain
with the emperor, and all are military in origin.
How far these trappings of North African material culture actually represent the
presence of North African individuals has been a matter of debate (Swan 1992; Fulford
2010), although it is notable that these appeared at the same time as historical records
tell us military units associated with North Africa arrived in Britain. We may be see-
ing North Africans units preferring the cookware or architectural forms that they were
familiar with. Alternatively, if we view armies as ‘major engines of integration’ forging
‘Romanness’ from ethnic diversity (James 2011: 28) there is evidence that military units
borrowed from a range of ethnic styles and created their own military culture, which
cannot be equated with a single geographical, ethnic, or cultural origin.
In their static role as material culture, they may be simply a different style of cook-
ing pot, but, as active materiality, these ceramics could represent part of a dialogue of
expressed identity for the people who owned and used them. It is, however, problematic
to equate specific cultural packages with ethnicity. Rather than simply reflecting ‘North
African’ origin, the use of new cultural packages may reflect these external groups’
need to express new identities in contrast to existing ones, perhaps marking their par-
ticular military unit or even groups within those units defining their social status and
group identity. Swan’s identification, therefore, of what seem to be North African or
Mediterranean eating and cooking habits emphasizes the complex set of identities (and
indeed contemporary multiple identities) being expressed even within a small area of
northern Britain; these individuals may have not seen themselves as African or even
come from North Africa, but food ways from that region had become integral to their
newly formed identity. Whether or not we can associate certain styles of pottery or
construction techniques with particular ethnic groups, these new forms of dining and
building methods emphasize the complex identities being formed in northern Britain,
particularly within the units on Hadrian’s Wall.
The diverse nature of the identities and cultural antecedents of many of the mili-
tary units in the Wall zone is also illustrated by a further example of ceramic evidence.
Ceramics known as ‘Housesteads ware’ were originally identified by Jobey (1979) as
probably being associated with Frisian units from north of the Rhine. Van Driel-​Murray
(2009: 818–​819), however, notes that examples of the pottery do not reflect the typical
Frisian domestic assemblage but, rather, are finer forms, which may reflect continued
use of strong Germanic ritual traditions centred around hospitality and guest-​friend-
ship. This could suggest that Housesteads ware represents not simply a community of
236

236   Claire Nesbitt

immigrant Frisians manufacturing and using objects that are familiar and remind them
of home, but also objects that have cultural symbolism that then become clear identity
signifiers.

Materiality of Death

In contrast to the accumulation of possessions and use of objects in life, which can be tran-
sient, expedient, and perhaps even mindless, the objects people are buried with represent
how they saw themselves in life, or perhaps more accurately, how they or their loved ones
wished them to be seen in death. Treatment of the body at death can, in some ways, be seen
as ‘the ultimate expressions of humans’ perceptions of themselves in society and cosmos’
(Fahlander and Oestigaard 2008: 1). In many ways, therefore, the expression of identity
can be seen more evidently in death and its memorialization than in life. This is evident in
the burial choices of individuals and communities in the frontier zone. Many burials have
been excavated along the Wall, but systematic excavation of cemeteries has been limited.
Of those that have been excavated, only Brougham provides sufficient material culture to
withstand broader analysis. Other cemeteries that have been examined, such as Beckfoot
in Cumbria (Caruana 2004), comprising mainly cremations with limited grave goods,
and the Botchergate cemetery in Carlisle (Breeze 2006), indicate little evidence of cultural
identities. In 2009, excavations at the cemetery at Birdoswald were undertaken. Initial
results from Birdoswald suggest that there is little evidence for cultural packages that may
reflect particular provinces or regions (see Haynes, this volume).
To date, only one cemetery in the frontier zone, Brougham, Cumbria (Cool 2004),
provides us with evidence concerning the expression of identity through the materiality
of death. The cemetery at Brougham was in use between c. ad 200 and 310, and there is
evidence of two distinct burial grounds; Cool has suggested that the second cemetery
could represent groups with a different set of identities establishing a new burial ground
if they felt that existing ones did not allow them to bury their dead in their traditional or
correct way (Cool 2004: 463). Many of the elements of the burial rite used in the cem-
etery have been argued as being Pannonian in style and as indicating the presence of
foreign groups (Cool 2004: 464). Though this is not attested by epigraphy, it is thought
that a new military unit that arrived at Brougham in the early third century ad (the
Numerus Equitum Stratonicianorum) may be the people represented by this change
in burial activity. To begin with it is known that, in Pannonia, as well as Noricum,
Dalmatia, Upper Moesia, and Dacia, graves were purified using a ritual burning event;
in the Brougham cemetery, dark staining in grave 227 may represent such a burn-
ing (Cool 2004: 465). Secondly, there are some finds in association with the burials at
Brougham, including iron bucket pendants and cremated horses, that are unparalleled
in funerary rites known elsewhere in Roman Britain. Cool (2004: 464) suggests that the
bucket pendants imply trans-​Danubian connections, while the horse cremations may
suggest contact with northern Germany. Other items of personal adornment discovered
 Multiculturalism on Hadrian’s Wall    237

at Brougham include red-​striped beads, also found at Augst and Avenches in modern
Switzerland. Cool (2004: 464) suggests that again these reflect a Danubian tradition
and are often a feature of the Maslomecz-​Gruppe type culture. The only object from the
cemetery that may offer us an individual identity is a jar that was included with a burial.
The jar is inscribed with a graffito of the name ‘Bata’, which is Illyrian in origin. Cool
(2004: 466) has argued that, if indeed the cemetery was that of the Numerus Equitum
Stratonicianorum and their families, it is likely the unit came from the Pannonian area of
the Danube and probably included cavalrymen from Barbaricum.
The drawback of burial archaeology is, of course, that the dead do not bury them-
selves. They are buried by others, and there is potential for great difference between their
view of the deceased and that of others (Fahlander and Oestigaard 2008: 11). While,
on the one hand, we can suggest that material culture is used in life and in death to
manipulate an individual’s persona (Cool 2010: 27), it is also true that the identity of
an individual is not straightforward, and each identity is a complex collection of vary-
ing situated roles and identities that transect one another in time and space (Fahlander
and Oestigaard 2008: 11). In many ways, epigraphic evidence, which was sometimes pre-
pared by the deceased before death, can offer us a more reliable interpretation, perhaps
not of who they were, but of whom they wanted us to believe they were.

Expressions of Individual Identity

The above examples illustrate the multicultural nature of the Army, but to what extent
can we determine the multicultural nature of individuals within the Wall zone?
Examples of individual identity that survive are almost exclusively epigraphic and from
tombstones. In Britain, the people most likely to record their life on a tombstone were
probably immigrants; the epigraphic habit was not a native British tradition, and it was
a trend that took off slowly among the civilian population, with most inscriptions repre-
senting the military and their families (Noy 2010: 18; see also Hope, this volume). Hope
(1997: 257) has argued that tombstones were a means by which foreigners or outsiders
could seek acceptance into a community. Derks has also argued ‘inscribed monuments
erected by individual members of local communities provide unparalleled access to
subjective feelings of belonging even at the level of the individual’ (Derks 2009: 240).
The tombstones appear also to transcend social differences with evidence of tomb-
stones for women (for example, that of Pussitta from Raetia, modern Switzerland (RIB
984)), children (for example, Salmanes, aged 15, commemorated on a tombstone from
Auchendavy on the Antonine Wall (RIB 2182)) and former slaves such as Victor, a Moor
found at South Shields (RIB 1064), whose short life was recorded by this inscription:

D M VICTORIS NATIONE MAVRVM ANNORVM XX LIBERTVS

NVMERIANI EQITIS ALA I  ASTVRVM QVI PIANTISSIME PROSEQVTVS
EST [RIB 1064].
238

238   Claire Nesbitt

To the spirits of the departed and Victor, of the Moorish nation, twenty years old,
freedman of Numerianus, a trooper of the First Wing of Asturians, who most devot-
edly conducted [his burial].

It was a common practice in the Roman world to express identity away from one’s place
of origin (Salway 1965). This even appears to have included those who came from else-
where in the province, some of whom were serving in units with origins supposedly
elsewhere in the empire—​for example, Nectovelius son of Vindex, serving in a Thracian
cohort, is described as nationis Brigans (RIB 2142), or Verecunda, commemorated by
her husband as cives Dobunna (RIB 621), and another woman from Ilkley remembered
as civis Cornovia (RIB 639) (see also Ivleva, this volume).
Burial was usually the duty of the deceased’s heir. A Roman tombstone fulfilled two
functions: it commemorated the dead by simply recording the name, sometimes with
his or her achievements, and it also stated in writing the commemorator’s discharge
of his duty. It was the Roman way of indicating the fulfilment of a particularly Roman
obligation (Meyer 1990: 78). In this sense, what may be reflected are the agendas of the
bereaved, rather than the deceased (Cool 2004: 437). However, tombstone evidence is
unique in its ability to reveal the individuals of the frontier landscape. Tombstones are
a specific form of evidence, which inevitably mean that our understanding of the indi-
vidual is led by how the individual wanted to be represented.
There are one or two epigraphic examples that provide detailed evidence of the com-
plex multiple identities individuals could have on the Wall. The first is the tombstone of
Regina discovered at South Shields (Figure 11.2). The tombstone reads:

D M REGINA LIBERTA ET CONIVGE BARATES PALMYRENVS NATIONE

CATVALLAVNA AN XXX. (RIB 1065)
To the spirits of the departed and Regina, freedwoman and wife of Barates of
Palmyra, a Catuvellaunian by race, thirty years old. (Noy 2010: 18)

Below this, in Palmyran script, is the following text: ‘Regina freedwoman of Barate(s),

alas!’ (Noy 2010: 18).
This stone is regularly held up as an example of a foreign presence on the Wall, of
intermarriage between native Briton and foreign immigrants, and the nature of ethnic-
ity and race in the Wall region (Hingley 2010: 235; Noy 2010: 18; Pearce 2010: 87). But how
much can we discern about multiculturalism on the frontier from the short inscription?
The lady in question, Regina, is described as a freedwoman and wife, suggesting that she
had been in captivity and been bought, perhaps by the husband, Barates. The inscrip-
tion describes her as a 30-​year-​old Catuvellaunian—​the civitas associated with the area
of modern Hertfordshire. The choice to include this regional heritage and the decision
to self-​identify with this region may reflect a desire to associate with a provincial status
conferred on this particular social group by the Roman administration, rather than to
hark back to the longevity of an Iron Age identity (Moore 2011). Mattingly (2010: 218)
has suggested that the inscription referring to her ethnic origin as Catuvellaunian would
Figure 11.2  Regina’s tombstone.
Source: © Claire Nesbitt.
240

240   Claire Nesbitt

be unusual in a civilian context but is more common in a military context, situating

her within the practices and beliefs of the military community. As Cool (2010: 29) has
argued, ‘being native in Roman Britain was no more straightforward than being Roman
was from an ethnic point of view’. Her status as a freedwoman also reveals the fluidity of
social standing in the Roman Empire. Whatever the motivation for the description of
Regina, the tombstone represents a mixed-​race couple from very different parts of the
world, unified in a frontier landscape through an empire that dissolved many socio-​cul-
tural barriers.
Equally interesting is Barates’ own heritage. A tombstone that may belong to the same
Barates was discovered at Corbridge. This inscription reads:

D(IS) M(ANIBUS) [BA]RATHES PALMORENVS VEXIL(I)A(RIUS) VIXIT

AN(N)OS LXVIII [RIB 1171].
To the Sprits of the departed and Barathes the Palmyrene, flag bearer/​maker lived
68 years. (Noy 2010: 19)

It is uncertain whether this individual was associated with a particular unit, although
the presence of the Numerus Barcariorum Tigrisiensium, a unit of Tigris bargemen,
from the Middle East, at Arbeia, South Shields, may suggest he was part of this unit. That
he chose to honour his wife’s memory in Palmyrene as well as Latin is revealing, mark-
ing a potentially strong statement of ethnic or cultural identity (Noy 2010: 22). Latin
was the usual epigraphic language in Roman Britain, and there is nothing to suggest
a large community of Palmyrans who would understand the inscription. Perhaps the
subtext represents a tender desire to say farewell in his own language, in the Palmyran
way. Despite the complexities of their origins, these inscriptions emphasize the multi-
cultural identities, even within a married couple, not only of the Hadrian’s Wall com-
munity but of the multi-​layered and complex nature of individual identities within the
broader Wall zone.

Conclusions

Hadrian’s Wall has often been seen as a barrier between peoples and cultures, and
a contested landscape. However, as Hingley (2010: 229) has suggested, the material
culture tells a different story, a story of inclusivity. In fact it was a place where people
of many different cultures came together, brought to the frontier by military service,
slavery, trade, or family ties; almost every corner of the empire can be seen repre-
sented on the Wall. The interesting aspect of the presence of these people is not so
much the culture they brought with them, but rather the culture they developed and
adopted, the culture they shared and the ways in which that culture was expressed
to form new identities. The military materiality of the Wall belies a greater, more
 Multiculturalism on Hadrian’s Wall    241

cosmopolitan reality, of a place where peoples of a diverse range of ethnic, cultural,

and religious backgrounds unified under a common Roman identity. The materi-
ality of multiculturalism on the Wall is rich and will no doubt continue to unfold
as further excavations happen and more discoveries are made. The archaeology
reveals quite clearly the presence of the ‘other’, of peoples far away from their homes
and perhaps determined to maintain a cultural identity. However, at the heart of
multiculturalism in the frontier zone lies one unifying factor, that of being, or aspir-
ing to be, Roman.

Abbreviation
RIB  R
 . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain Volume I:
Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Adler, P. S. (1976). ‘Beyond Culture Identity’, in L. Samovar and R. Porter (eds), Intercultural
Communication. Belmont, CA: Wadsworth Publishing, 362–​378.
Anderson, A. W. (2011). ‘Traditions and Transitions: Later and Roman Iron Age Communities
in the North-​East of England’. Unpublished Ph.D. thesis, Durham University.
Breeze, D. J. (1982). The Northern Frontiers of Roman Britain. London: Batsford.
Breeze, D. J. (1990). ‘The Impact of the Roman Army on the Native Peoples of North Britain’, in
H. Vetters and M. Kandler (eds), Internationalen Limeskongresses 1986 in Carnuntum: Der
römische Limes in Österreich, Heft 36. Berlin: Wein, 85–​97.
Breeze, D. (2006). Handbook to the Roman Wall. Newcastle: Society of Antiquaries.
Caruana, I. (2004). ‘The Cemetery at Beckfoot Roman Fort’, in R. J. A. Wilson and I. Caruana,
(eds), Romans on the Solway: Essays in Honour of Richard Bellhouse. Kendal: Cumberland
and Westmorland Antiquarian and Archaeological Society, 134–​173.
Cool, H. E. M. (2004). The Cemetery at Brougham, Cumbria: Excavations 1966–​67. Britannia
Monograph Series 21. London: Society for the Promotion of Roman Studies.
Cool, H. E. M. (2010). ‘Finding the Foreigners’, in H. Eckardt (ed.), Roman Diasporas:
Archaeological Approaches to Mobility and Diversity in the Roman Empire. JRA
Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 27–​44.
Dellino-​Musgrave, V. (2005). ‘British Identities through Pottery in Praxis: The Case Study of a
Royal Navy Ship in the South Atlantic’, Journal of Material Culture, 10: 219–​243.
Derks, T. (2009). Ethnic Constructs in Antiquity:  The Role of Power and Tradition.
Amsterdam: Amsterdam University Press.
Diaz-​Andreu, M., and Lucy, S. (2005). ‘Introduction’, in M. Diaz-​Andreu, S. Lucy, S. Babić, and
D. N. Edwards (eds), The Archaeology of Identity. London: Routledge, 1–​12.
Divine, D. (1969). The North-​West Frontier of Rome:  A  Military Study of Hadrian’s Wall.
London: MacDonald.
Dobson, B., and Mann, J. C. (1973). ‘The Roman Army in Britain and Britons in the Roman
Army’, Britannia, 4: 191–​205.
242

242   Claire Nesbitt

Fahlander, F., and Oestigaard, T. (2008). The Materiality of Death: Bodies, Burials, Beliefs. BAR
International Series 1768. Oxford: Archaeopress.
Fulford, M. (2010). ‘Roman Britain: Immigration and Material Culture’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 67–​78.
Gardner, A. (2002). ‘Social Identity and the Duality of Structure in Late Roman-​Period Britain’,
Journal of Social Archaeology, 2/​3: 323–​351.
Haverfield, F. (1912). The Romanization of Britain. Oxford: Clarendon Press.
Haynes, I. (1999). ‘Military Service and Cultural Identity in the auxilia’, JRA Supplementary
Series 34. Portsmouth, RI: Journal of Roman Archaeology, 165–​174.
Hingley, R. (2010). ‘Tales of the Frontier: Diasporas on Hadrian’s Wall’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 227–​243.
Hingley, R. (2012). Hadrian’s Wall: A Life. Oxford: Oxford University Press.
Hingley, R., and Hartis, R. (2011). ‘Contextualizing Hadrian’s Wall:  The Wall as “Debatable
lands”’, in O. Hekster and T. Kaizer (eds), Frontiers in the Roman World: Proceedings of the
Ninth Workshop of the International Network Impact of Empire (Durham, 16–​19 April 2009).
Leiden: Brill, 79–​96.
Hodgson, N., McKelvey, J., and Muncaster, W. (2012). The Iron Age on the Northumberland
Coastal Plain. Newcastle upon Tyne: TWM Archaeology and the Arbeia Society.
Hope, V. M. (1997). ‘Words and Pictures: The Interpretation of Romano-​British Tombstones’,
Britannia, 28: 245–​258.
Hunter, F. (2001). ‘Roman and Native in Scotland:  New Approaches’, Journal of Roman
Archaeology, 14: 289–​309.
Hunter, F. (2007a). ‘Artefacts, Regions and Identities in the Northern British Iron Age’, in C. C.
Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow
Books, 286–​296.
Hunter, F. (2007b). Beyond the Edge of the Empire:​Caledonians, Picts and Romans. Ross-​shire:
Groam House Museum.
James, S. (2002). ‘Writing the Legions: The Development and Future of Roman Military Studies
in Britain’, Archaeological Journal, 159: 1–​58.
James, S. (2011). Rome and the Sword:  How Warriors and Weapons Shaped Roman History.
London: Thames and Hudson.
Jobey, I. (1979). ‘Housesteads Ware:  A  Frisian Tradition on Hadrian’s Wall’, Archaeologia
Aeliana, 5 [ser. 7]: 127–​143.
Lazzari, M. (2005). ‘The Texture of Things: Objects, People, and Landscape in North-​West
Argentina (First Millennium ad)’, in L. Meskell (ed.), Archaeologies of Materiality. Oxford:
Blackwell, 126–​161.
Lucy, S. (2005). ‘Ethnic and Cultural Identities’, in M. Diaz-​Andreu, S. Lucy, S. Babić, and D. N.
Edwards (eds), The Archaeology of Identity. London: Routledge, 86–​109.
Mann, J. C. (1963). ‘The Role of the Frontier Zones in Army Recruitment’, in G. Novak (ed.),
Quintus Congressus Internationalis Limitis Romani Studiosorum 1961. Acta et Dissertationes
Archaeologicae, Arheološki Radovi i Rasprave 3. Zagreb: Jugoslavenska Akademija Znanosti
i Umjetnosti, 145–​150.
Mann, M. (1986). The Sources of Social Power, Volume I: A History of Power from the Beginning
to ad 1760. Cambridge: Cambridge University Press.
 Multiculturalism on Hadrian’s Wall    243

Mason, D. J.  P (1990). ‘The Use of Earthenware Tubes in Roman Vault Construction:  An
Example from Chester’, Britannia, 21: 215–​222.
Mattingly, D. J. (2010). Imperialism, Power and Identity. Princeton: Princeton University Press.
Meskell, L. (2005). ‘Introduction:  Object Orientations’, in L. Meskell (ed.), Archaeologies of
Materiality. Oxford: Blackwell, 1–​17.
Meskell, L. (2007). ‘Archaeologies of Identity’, in T. Insoll (ed.), The Archaeology of
Identities: A Reader. Abingdon: Routledge, 23–​43.
Meyer, E. A. (1990). ‘Explaining the Epigraphic Habit in the Roman Empire: The Evidence of
Epitaphs’, Journal of Roman Studies, 80: 74–​96.
Miller, D. (2005). ‘Afterword’, in L. Meskell (ed.), Archaeologies of Materiality. Oxford: Blackwell,
212–​219.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Modood, T. (2007). Multiculturalism. Cambridge: Polity Press.
Moore T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and
Roman Studies’, Journal of Social Archaeology, 11/​3: 334–​360.
Nesbitt, C., and Tolia-​Kelly, D. P. (2009). ‘Hadrian’s Wall: Embodied Archaeologies of the
Linear Monument’, Journal of Social Archaeology, 9/​3: 368–​390.
Noy, D. (2010). ‘Epigraphic Evidence for Immigrants at Rome and in Roman Britain’, in H.
Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity
in the Roman Empire. JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman
Archaeology, 14–​26.
Pearce, J. (2010). ‘Burial, Identity and Migration in the Roman World’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
JRA Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 79–​98.
Poulter, J. (2009). Surveying Roman Military Landscapes across Northern Britain: The Planning
of Roman Dere Street, Hadrian’s Wall and the Vallum, and the Antonine Wall in Scotland.
BAR British Series 492. Oxford: Archaeopress.
Salway, P. (1965). The Frontier People of Roman Britain. Cambridge:  Cambridge
University Press.
Shennan, S. (1989). Archaeological Approaches to Cultural Identity. London: Unwin Hyman.
Steinberg, S. R., and Kincheloe, J. L. (2009). ‘Smoke and Mirrors: More than One Way to be
Diverse and Multicultural’, in S. R. Steinberg (ed.), Diversity and Multiculturalism: A Reader.
New York: Peter Lang, 3–​22.
Swan, V. G. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman
Pottery Studies, 5: 1–​34.
Swan, V. G. (1999). ‘The Twentieth Legion and the History of the Antonine Wall Reconsidered’,
Proceedings of the Society of Antiquaries of Scotland, 129: 399–​480.
Swan, V. G. (2009). Ethnicity, Conquest and Recruitment: Two Case Studies from the Northern
Military Provinces. JRA Supplementary Series 72. Portsmouth, RI: Journal of Roman
Archaeology.
Swan, V. G., and Monaghan, J. (1993). ‘Head-​Pots: A North African Tradition in Roman York’,
Yorkshire Archaeological Journal, 65: 21–​38.
van Driel-​Murray, C. (2009). ‘Ethnic Recruitment and Military Mobility’, in Á. Morillo, N.
Hanel, and E. Martín (eds), Limes XXth International Congress of Roman Frontier Studies.
Madrid: Ediciones Polifemo, 813–​822.
244

244   Claire Nesbitt

Witcher, R. E. (2000). ‘Globalisation and Roman Imperialism: Perspectives on Identities in

Roman Italy’, in E. Herring and K. Lomas (eds), The Emergence of State Identities in Italy
in the First Millennium BC. London:  University of London Accordia Research Institute,
213–​225.
Witcher, R. (2010). ‘The Fabulous Tales of the Common People, Part 2’, Public Archaeology, 9/​
4: 211–​238.
Witcher, R., Tolia-​Kelly, D. P., and Hingley, R. (2010). ‘Archaeologies of Landscape: Excavating
the Materialities of Hadrian’s Wall’, Journal of Material Culture, 15: 105.
Woolf, G. (1997). ‘Beyond Romans and Natives’, World Archaeology, 28/​3: 339–​350.
Woolf, G. (1998). Becoming Roman. Cambridge: Cambridge University Press.
 Chapter 12

B ritons on t h e  Mov e
Mobility of British-​Born Emigrants in
the Roman Empire

Tatiana Ivleva

Introduction: Identifying
The Mobile Identities

The movement of people and objects is attested for all periods of human history. For the
Roman Empire the evidence for such movement is abundant, owing to the conquests
of various territories, which resulted in a wide range of individuals and communities
being on the move, with both voluntary and forced migration being common. Within
Roman Britain itself the movement of immigrants from the continent has been the
topic of a number of publications covering the origin of migrants, their distribution,
and the ways one might identify them (cf. Thompson 1972; Rowland 1976; Birley 1988;
Swan 1992; Wilmott 2001; Leach et al. 2009, 2010, to name but a few). This chapter seeks
to go beyond the issues of internal migration within the Roman province of Britannia
and avoids addressing the movement of foreigners within Britain. It focuses instead on
movement away from Britain and discusses the presence of Britons elsewhere in the
Roman Empire.
The movement to a new territory influences the ways individual migrants or com-
munities of settlers see themselves, (re)forming along the way the myriad of identities
that already existed within both newcomer and host societies. The way one perceives
the other in migrant groupings undergoes identity stress when new forms of identifica-
tion are constructed, manipulated, or adjusted to circumstances (Oltean 2009: 92–​93).
Consequently, mobility and the transformation of one’s identity go hand in hand, and
discussion of the changes within the personal self cannot be avoided in studying indi-
viduals on the move.
246

246   Tatiana Ivleva

The concept of ‘identity’ can usefully be considered in terms of two categories: uni-
versalization and duality. The first category is based on an Aristotelian approach,
whereby identity is defined according to the principle that ‘a thing is itself ’ (Aristotle,
Metaphysics VII. 17), emphasizing the universality and sameness in things. The second
category allows identity to be considered in the same terms as the theoretical concept
‘duality of structure’ (Giddens 1979): individual identity—​selfhood—​is formed within
the personal self and as an opposition to the perception of others. Here the emphasis
is placed on the duality of the nature of ‘identity’, where ‘selfhood’ is perceived by the
self and by others, allowing the understanding, formation, negotiation, fragmentation,
fluctuation, and so on of the self. This is where identity, better understood in its plural
form—​identities—​is categorized as fluid, dynamic, and unstable; it is constantly chang-
ing, depending on situations in which agents find themselves. While the first category
has one level, ‘sameness’, the second category implies various levels of identification—​
that is, an individual or a self has many identities, based on gender, ethnicity and culture,
age, status, class, and religion, opposing and contrasting them with definitions imposed
by ‘the others’.
One of the most studied levels of identity is ethnic identity. Ethnicity has usually
been considered to be based on ‘racial’ characteristics: the same origin, language, or
descent (cf. Brather (2004: 77–​88) on the notion of ‘race’ in discussions of ethnicity),
but it has now been widely acknowledged that ethnicity is more ‘an idea than a thing’,
based primarily on social relationships and similar ways of behaving, and is something
that can be learnt, rather than something one is born into (Lucy 2005: 86). Any ethnic
affiliations can be changed by agents through mobility or social associations; ethnic-
ity is, therefore, highly mutable and dependent on the contexts in which agents find
themselves (Brather 2004: 568; Lucy 2005: 97). As a result, ethnic identity is created and
(re)invented within a variety of cultural repertoires embedded within, and formed by,
social practices and formulated through dialectic opposition of the self and the other.
Ethnic realization is born within particular groups at the moment when cultural differ-
ences are recognized, providing the motive for the universalization of these differences
and making them into ‘practice’. Yet, there is a precondition that allows such groups to
justify their communal closure and to find a common ground for the group’s forma-
tion: the unifying principle (that is, the ‘sameness’), embedded within the familiarity
in the use of objects or interaction between agents known as habitus (Bourdieu 1998).
In other words, recognition of the similarities that allows a group to form ‘an ethnicity’
derives from a social practice based on ‘shared ways of doing things’ (Lucy 2005: 101).
Ethnic identity is, therefore, a product of understood differences and preconditioned
sameness.
Taking into consideration the explanation proposed here of the notion ‘ethnic iden-
tity’, this chapter seeks to outline, in general, the possible ways one can identify British
emigrants on the continent and to understand, in particular, how ethnicity was formed
and operated within the groups that moved away from Roman Britain. It attempts to
chart the changes over time in terms of the personal and communal identification and
to provide explanation for such changes. Furthermore, it explores the consequences that
 Britons on the Move    247

movement might have had on the use, as well as on the changing roles, of objects made
in Britain and brought over to the continent.

Multiplicity of Identities in 

the Roman Empire: Being Roman
and Being British

One cannot approach the study of mobility in the Roman Empire without asking ques-
tions relating to issues of identity and ethnic identification. The existence of a multi-
plicity of identities in, and the multiculturalism of, the Roman Empire has been widely
recognized; the majority of scholarship in this area has been devoted to showing the
fragmentation of Roman identity and has perceived the Roman Empire as a heteroge-
neous society containing a variety of individual and group responses to ‘being Roman’
(Wallace-Hadrill 2007; Hingley 2009; Revell 2009). ‘Roman’ cannot be regarded as a
fixed entity, since the different individuals and groups dwelling within the boundaries of
the Roman Empire may have understood and experienced ‘being Roman’ in a variety of
ways: ‘being Roman’ always meant something different (Revell 2009: p. xii).
In a similar vein, in thinking about ‘being British’ we might suppose that, for indi-
viduals or groups coming from Roman Britain and settling on the continent, this notion
was inverted. It has been proposed that within Roman Britain itself there were ‘no such
social groups as “Britons”, the peoples were an assortment of tribes’ (Mattingly 2004:
10). This seems indeed to have been the case, especially when we take into consideration
the epigraphic record of Roman Britain: a total of ten inscriptions have been recorded,
dating roughly to late first–​late second centuries, which mention the origin of an indi-
vidual from a particular British tribe or town. These ten individuals were interprovin-
cial migrants and belonged to various British tribes: six were citizens of the Canti (RIB
192), Cornovi (RIB 639, female), Dobunni (RIB 621, female), Dumnonii (RIB 188), and
Catuvellauni (RIB 1065, female) and the city of Lindum (RIB 250, female). Three indi-
cated their origin (natione) as belonging to the Belgae (RIB 156), Briganti (RIB 2142), and
Catuvellauni (RIB 1962) tribes, while one simply named his origin as (colonia) Victrix
(RIB 3005). These inscribed stones were erected either by the relatives of individuals
who had died somewhere other than in the territory of their tribe or by individuals who
were fulfilling vows in a foreign region of their home province. All of them found it
important to emphasize their origin—​an action that indicates both the significance of
tribal above provincial forms of identification and a possible continuation of tribal divi-
sions and differences in Britain under Roman rule.
By trying to understand the refusal to denote provincial origin and the continua-
tion in the use of tribal identification, one needs to explore the significance of the terms
Britannus and Britto—​labels associated with the pan-​tribal community. It is likely that
both terms were coined and artificially imposed by the new dominant power in Britain
248

248   Tatiana Ivleva

after ad 43 and actually do not derive from any self-​awareness on the part of the indig-
enous population (Matthews 1999). The labels were probably imposed by the Roman
administration for an administrative convenience or perhaps intended to speed up the
process of inclusion of the natives into the Roman orbit: a process called ‘superficial
homogenization’ (Matthews 1999: 29). Such homogenization is recorded in other com-
munities who supplied recruits for the Roman army. The main purpose was the pro-
motion of a special type of military identity—​a regional one. For instance, the Romans
continuously cultivated tribal associations among the Batavians, a tribe from Germania
Inferior, placing a particular emphasis on their militaristic nature (van Driel-​Murray
2003: 201; Roymans 2004: 223). The Romans might also have reinvented and manipu-
lated British ethnic identity by consistently referring to the people who originated from,
or were born in, the province of Britannia as ‘Britons’. For instance, Dio Cassius (62.4)
puts the following phrase in Boudicca’s mouth prior to the major battle between Roman
and British forces in ad 60/​61: ‘for I [Boudicca] consider you all my kinsmen inasmuch
as you inhabit a single island and are called by one common name.’ This is, clearly, an
example of Roman rhetoric and propaganda rather than an exhibition of pan-​tribal
British identity; such notions of artificial ethnicity may not have had much relevance
for the peoples of Britain. In this sense, Rome created a new ethnic unity among the
fragmented groupings (cf. Hingley (2009), commenting on the formation of artificial
Batavian ethnicity), but one may ask how successful this attempt was, taking into con-
sideration the promotion of tribal affiliations discussed above.

Britons Abroad: Forms of
Identification

If tribal affiliation was emphasized when an individual moved to another tribal region,
how did Britons settling elsewhere in the Roman Empire denote their origin? In a more
general sense, do we have any examples of such Britons and how can we identify them?
The theme of the presence of foreigners in the various provinces of the Roman
Empire has usually been tackled from an epigraphic perspective (cf. Noy 2001;
Wierschowski 2001; Kakoschke 2002, 2004; Oltean 2009). The epigraphic record is the
obvious source of evidence to turn to here: funerary, votive, and other types of inscrip-
tions can be seen to have played an important role in reflecting existing identities—​
social, cultural, and ethnic—​as well as having created new ones (Hope 2001). When
left by emigrants, inscriptions can indicate the choices they made when stating their
origin, the places they settled in, and their reasons for migration overseas. Together
with inscriptions, military diplomas (Roman citizenship certificates issued to auxiliary
soldiers who had completed twenty-​five years of military service) can be used to deter-
mine the ways in which Britons drafted into the Roman army indicated their origin,
as well as provide us with information on their status within a unit, and their social
 Britons on the Move    249

and family relations. While the epigraphic record provides us with glimpses of various
aspects of an emigrant’s life such as their age, occupation, or ethnic origin, it can rarely
be taken at face value; the information provided was often ‘cleaned up’. Inscriptions and
diplomas were a medium for invented identities: what was included, and in what form,
might have been determined by the circumstances and the desires of a client who made
the choices regarding what was appropriate to communicate (Bodel 2001: 34). Another
problem when dealing with inscriptions and military diplomas is dating: while some
can easily be dated by means of specific references in the texts, others can be dated
only approximately, on the basis of the known development of linguistic formulae (cf.
Holder 1980).
Another type of evidence that can be used to trace British emigrants is that of dress
accessories, because of their regionality and their ability to serve as a medium for the
expression and negotiation of a person’s various identities, not least their origin (cf. Swift
2000; Rothe 2009; see also Cool, this volume). The significance of brooches, the most
common and regionally specific dress accessory, as identity-​markers and their double
functionality (that is, being passive, functional tools, they also acted as active partici-
pants in constructing the identities of the wearer) have been considered elsewhere (cf.
Jundi and Hill 1998; Pudney 2011). Brooches were personal items used to secure clothing,
and, while crossing the Channel, emigrants from Britain most likely wore them or had
them with them among their personal belongings. Moreover, British-​made brooches
were distinctive in their design, decoration, and form compared to local products in
other parts of the empire, which makes them stand out within the homogenous material
culture of continental sites, as I have argued elsewhere (Ivleva 2011: 133). British-​made
brooches found on the continent have a similar dating problem as the epigraphic record:
the precise date range when brooches were in use will always be uncertain (Snape 1993:
6). Yet the contexts where these brooches were located can provide a relative time span
when particular types were in use (cf. dating of British-​made brooches in Bayley and
Butcher 2004; see also Snape 1993; Mackreth 2011). One needs, however, also to take
into account that brooches that are found in a context dating to a period when their
popularity was on the wane may represent ‘heirlooms’. This problem is undoubtedly
significant when one discusses a migrant population, considering that a family on the
move may have curated brooches over a long period of time as a reminder of home ties
(Revell, pers. comm; cf. also Gilchrist 2013). Moreover, these artefacts also have limita-
tions regarding how representative they are of the population and present specific prob-
lems. For example, an object without context does not allow any conclusions concerning
a person’s religious belief, status, or age. Personal accessories are particularly valuable as
sources to study the projection and negotiation of personal identities—​not just ethnic-
ity—​but such actions depend on the circumstances when a particular object was worn,
something that is not visible in archaeological record.
A total of 40 persons of British descent have been identified through the epigraphic
record yet only 27 mention the individual’s origins directly—​which is extremely low
in comparison to other ethnic groups such as Dacians (150 cases: Oltean 2009: 96) or
Germans (174 cases: Kakoschke 2004: 198). Such low numbers are not representative
250

250   Tatiana Ivleva

of the real level of mobility: not everyone was able to commission a funerary or votive
monument or received a military diploma. A second factor might be the irrelevance of
naming individual origins for soldiers serving among their own countrymen (Oltean
2009: 91): for a ‘Briton’ in a British auxiliary unit, it would have been unnecessary spe-
cifically to name his origin, whereas if he had served in another ethnic unit he would
most likely have wanted to emphasize his ethnic background.
A total of 242 British-​made brooches have been recorded from 102 sites across the
empire—​the majority being found in the western part of the Roman Empire, in the mili-
tarized areas of Germania Inferior and Superior and on civilian sites in Gallia Belgica;
the provenance of 19 brooches is unknown (Figure 12.1). That these brooches were actu-
ally ‘made in Britain’—​as opposed to the brooches reproduced by local craftspeople
from templates—​can be supported by the fact that they occur in too small a quantity on
the continent. While brooches with typical British characteristics appear to be relatively
numerous in Britain, overseas they are found in limited numbers: 1 or at most 3 per cent
of the total number found on any given site.

Figure 12.1  Distribution of British brooches.

Source: Brooches’ distribution partly after Morris (2010: 86, figure 4.35 and appendix 6); map by author. © Tatiana Ivleva.
 Britons on the Move    251

Identities in Words: Epigraphic
Narrative

A total of just twelve British soldiers have been identified who had served in British aux-
iliary and numeri units, which constitutes a small minority of all recorded soldiers who
are known to have served in these troops (Ivleva 2012). Of these twelve British soldiers,
the origin of five is recorded: a unit’s prefect and also legionary soldier from Lindum
(Lincoln) (AE 1973: 459 and CIL III. 6679); a Dobunnian infantryman (CIL XVI. 49;
Kennedy (1977) argued that he was Dobunnian through his mother, but Mullen and
Russell (2009) show that his name is well attested in British epigraphic record); two
footsoldiers from Ratae Corieltauvorum (Leicester), and the Belgae tribe (CIL XVI. 160
and AE 1944: 58 respectively), and, lastly, another infantryman who claimed to be Britto
(RMD I. 47). The origin of the other seven has been identified through (a) linguistic
analysis of their names (CIL V. 7717: Catavignus (his name has the suffix–​ign common
in Insular Celtic; cf. Evans (1967: 209) and Sims-​Williams (2004: 155, n. 921); CIL III.
3256: Virssuccius and Bodiccius (for discussion, see Birley 1980: 103)); (b) the recruit-
ment period (AE 1999: 1258 and CIL III. 10331: two ignotii, recruited at the same time
and to the same unit as foot-​soldiers who originated from the Belgae tribe and Ratae
Corieltauvorum); (c) the recorded recruitment pattern (AE 1994: 1487: possibly a son or
a grandson of a British soldier who had followed his father or grandfather into military
service (that is, following the pattern of hereditary military service whereby recruitment
was from among the sons of veterans who had settled in the proximity of a fort) (cf.
Dobson and Mann 1973: 202)); (d) the find-​spot of the military diploma (AE 2005: 954,
found in southern Britain and which probably records a British veteran returning from
Pannonia (cf. Tully 2005)).
Those who were born in Britain were also selected to fill gaps in the legionary and
auxiliary units stationed in the province and abroad. A variety of evidence comes from
different parts of the empire and records the existence of at least seventeen men who
emphasized their origin from Britain: five legionaries (CIL III. 11233: origin recorded
as Claudia Camulodunum (Colchester); CIL VI. 3594: origin recorded as cognomen
Britto; CIL VIII. 21669: from Lindum (Lincoln); CIL VI. 3346: originated from Glevum
(Gloucester); CIL VIII. 2877: considered to be of British descent owing to his service in
five British legions (cf. Malone 2006: 117)); three troopers in the Imperial horse guard in
Rome (CIL VI. 3279, 3301, and 32861: the origin of all three is recorded as natione Britto/​
Britan(n)icianus); one centurion in a British detachment in Mauretania Tingitana (AE
1920: 47 and 48: the linguistic analysis of his name shows his possible British origin (cf.
Raybould and Sims-​Williams 2009: 22)); eight auxiliaries in various cohorts and in fleets
(AE 1951: 47: recorded as ex Br(e)itonibus; AE 2003: 1218: Trinovantian by origin; CIL III.
14214: origin recorded as Britto; CIL XIII. 8314: recorded as civi Brittoni; AE 1956: 249:
citizen of Dumnonii; ILJug 02, 679: origin recorded as natione Britto; RMM 20: recorded
as Britto; AE 2007: 1772: from the Cornovi tribe). The origin of two legionary soldiers has
252

252   Tatiana Ivleva

been identified based on their religious beliefs: they made dedications to British Mother
Goddesses while on service in legio XXX Ulpia in Xanten (CIL XIII. 8631 and 8632; for
discussion see Ivleva 2011: 139).
The number of civilians of British descent known to have settled abroad is low in
comparison with the number of British servicemen: only seven are known—​one trader
(AE 1922: 116: priest of the Imperial cult at the coloniae at Eboracum (York) and Lindum
(Lincoln)); five whose occupation is unknown (CIL XIII. 1981 (a male): origin recorded
as natione Britto; CIL XIII. 6221 and AE 1915: 70: both from Deva (Chester); AE 1939: 53
(a male): described as Britannus natione; Martial 11.53 (a female): ‘born amongst the
woad-​stained Britons’; and one freedman in Rome, CIL VI. 2464: recorded as ‘taken
from Brittannia’ (sic)).
In total, thirty-​seven men and one woman have been identified, although two more
females can be added to this list: Catonia Baudia (CIL VI. 3594) and Lollia Bodicca (CIL
VIII. 2877), ‘travelling’ wives of the legionary soldiers who followed their partners to
their posts within the empire. Both have quite a remarkable cognomen, one that resem-
bles the name of British rebel Queen Boudicca. Considering that their husbands were
of British descent, it is plausible that these women hailed from one of the British tribes.
These forty people born in Britain had various professions, although the majority served
in the Roman army: in the legions and auxiliary units posted overseas, in the fleet gar-
risoned on the continent, and in Rome as the emperor’s bodyguards (Figure 12.2). Quite
surprisingly, only one British trader has been detected epigraphically, although there
must have been British-​born indigenous traders (as opposed to British-​born immigrant
traders) involved in cross-​Channel trade. It is unlikely that all trading activities between

17

4
3 3
2
1 1 1

Legionaries Auxiliaries Sailors Imperial Centurion Trader Civilians Freedman Women

horse (profession
guard unknown)

Figure 12.2  Britons abroad: profession and status.

Source: © Tatiana Ivleva.
 Britons on the Move    253

Britain and the continent lay in the hands of people born on the continent, as the epi-
graphic record would seem to suggest (Hassall 1978: 43).
Those who were born in Britain and later moved to the continent were not neces-
sarily of native British stock: at least three legionary soldiers might have been sons
or grandsons of immigrants to Britain in the mid and late first century ad—​namely,
Titus Statius Vitalis (CIL III. 11233), who hailed from Colchester, and Marcus Minicius
Marcellinus and Marcus Iunius Capito (CIL III. 6679; CIL VIII. 21669), who both came
from Lincoln. Both Colchester and Lincoln were colonies for retired legionary veter-
ans, and these men were probably descendants of legionary veterans who had settled
in Britain upon their retirement (Birley 1980: 104–​105). One legionary soldier, Marcus
Ulpius Quintus from Gloucester (CIL VI. 3346), might have been a son or grandson of
an auxiliary veteran who had either come from the continent or been drafted from a
British tribe to serve in Britain (Dobson and Mann 1973: 203; Birley 1980: 105).
The geographic spread of inscriptions mentioning British emigrants is not confined to
a particular province: they are distributed across the whole Roman Empire from North
Africa to Germania Inferior and from Gallia to the Roman frontiers on the Danube
(Figure 12.3). While the presence of some Britons in particular territories was due to the
orders of Roman officials, others seem to have settled in particular places in a search of
a better life. The example of the latter is Atianus, recorded as natione Britto, who settled
in Lyon (CIL XIII. 1981). Neither his profession or the reason for his presence in Lyon

Figure  12.3 Distribution of the military diplomas (star), funerary (circle), and votive
(diamond-​shape) inscriptions mentioning British emigrants.
Source: Map by author. © Tatiana Ivleva.
254

254   Tatiana Ivleva

is recorded on his funeral monument, but Lyon was a hub for commercial activity and
attracted wealthy merchants and craftsmen, so he might have gone there to open a ware-
house selling British goods or to help in establishing trading contacts between the two
provinces (for the existence of such contacts, see Morris 2010).
There is also evidence for female migration, with three women being identi-
fied: Catonia Baudia and Lollia Bodicca on the basis of their names and ‘British’ hus-
bands (see above), and Claudia Rufina, who was referred to as ‘being born among the
woad-​stained Britons’ by her friend, the poet Martial (11.53). It was relatively common
for women—​whether wives, partners, or sisters—​to follow their military husbands,
partners, or brothers to their postings (Allason-​Jones 1999:  48). These three British
women, therefore, conform to this picture and should not be seen as an exception. All
three enjoyed a privileged status: they were wives of legionary centurions. Moreover,
Catonia Baudia and Lollia Bodicca are recorded on the funerary inscriptions that they
themselves erected to commemorate their husbands, another hint as to their status and
wealth. Claudia Rufina is praised by Martial (4.13, 8.60, and 11.53) on numerous occa-
sions as an educated woman as well as adoptee of a Roman way of life.
The epigraphic material also shows a considerable degree of variation in the nomen-
clature of origin, which varied from naming a tribe (e.g. Dobunni, Belgae, Cornovi,
Trinovantes, Dumnonii) or specific place (e.g. Lindum, Ratae Corieltauvorum, Claudia
Camulodunum, Glevum, Deva) to the formula natione Britto/​Britan(n)icianus. I have
argued elsewhere (Ivleva 2011: 142–​144) that when these inscriptions and diplomas are
divided by century a pattern seems to emerge in terms of the changing way in which ori-
gin is referred to. Inscriptions dated to the late first century ad usually record the name
of the tribal and city origin, with an emphasis on the individual’s citizen status, which
might signify the importance of indicating that one was Roman (that is, having citizen-
ship) and at the same time belonging to a specific British tribe. Inscriptions dated to
the second century ad show a difference in the choices made when referring to origin:
while some individuals continued to name as their place of origin either a British city
or a tribe as their place of origin (that is, thereby emphasizing their tribal affiliation),
others preferred to identify themselves through geographical provenance as natione
Britto/​Britan(n)icianus on inscriptions or as Britto on military diplomas. Such changes
relate to a wider shift in forms of identification detected in the recording of affiliations
in the second century ad, where the tribal affiliations were being ‘replaced by formu-
lae using geographical provenance or political-​administrative inscription in a certain
civitas’ (Derks 2009: 269). While this process appears to be relatively common, one may
pose the question: what could have prompted some inhabitants of Britain settling on the
continent to choose the geographical over the tribal provenance? As has already been
mentioned, there were no such inter-​provincial grouping as ‘Britons’—​the term is likely
to have been a Roman construct designed to denote all the inhabitants of the newly
acquired province in ad 43 without paying attention to inner tribal divisions. By choos-
ing this Roman-​imposed label, Britons abroad may have been expressing this new form
of Roman-​imposed identity and constructed ethnicity. This invented label might have
been used by second-​generation emigrants—​those who were not born in Britain but
 Britons on the Move    255

whose parents belonged to one of the British tribes—​because they did not have a precise
ethnic identification: being born at particular place on the continent would not neces-
sarily have made them a member of a continental tribal entity (Ivleva 2011: 142–​143). The
use of the Roman-​imposed identification, therefore, became a necessity: by choosing to
refer to one’s origin as British, one distinguished oneself from other groups of migrants
or from the dominant group in the territory where British migrants and their families
settled down.
The epigraphic record left by migrants in the third century ad and later indicates that
the tendency for designating origins then shifted the other way: emigrants preferred to
name their province instead of their tribe or city. This situation may have resulted from
being incorporated into a new identity group in the third century ad as a result of eve-
rybody being given Roman citizenship by the edict of Caracalla in ad 212. This broke
tribal ties, and the supra-​regional identity suppressed the regional one, resulting in the
ultimate ‘e pluribus unum’ when, from a variety of tribes, one ‘province’ of emigrants
emerged.

Archaeological Narrative of Mobile

Identities: British-​made Objects
and Britons Abroad

The presence of 242 British-​made brooches on the continent can be related to a vari-
ety of activities of the people who brought these objects with them. Brooches ‘trav-
elled’ because of their function as clothes fasteners: after all, the people (whatever their
origin) who travelled from Britain to the continent needed something to hold their
clothes together. This makes brooches useful tools in determining the places where
such migrants settled down, be they traders, military men (veterans or soldiers), the fol-
lowers of the first two (households, slaves, partners, wives, and children) or craftspeo-
ple. Because brooches were brought overseas by various groups, emigrants from and
immigrants to Britain alike, this leads to the consideration that British brooches do not
provide evidence for the ethnicity of their users and wearers: other identities—​such as
status, gender, and, to a lesser extent, ethnicity—​might have projected through their use
(see also, Cool, this volume). Moreover, brooches may have changed meaning depend-
ing on their usage, on the context in which they were worn or discarded, and on their
viewers or admirers. Owners may also have had particular associations with them, pos-
sibly treating brooches as tourist knick-​knacks rather than objects of personal use.
In general, it is possible to associate particular sites with the presence of British-​born
emigrants through in-​depth analysis based on an object’s biography, site location, the
history of a settlement, epigraphic analysis, and the study of the context in which an
object has been found (Ivleva 2011: 137–1​42). It is likely that most British-​made brooches
were taken overseas by British recruits of British auxiliary units, or by recruits of other
256

256   Tatiana Ivleva

legionary and auxiliary forces of different ethnic origin who served in Britain, or by
veterans who, after being discharged, returned home from Britain (Ivleva 2011: 142).
Similarly, as the epigraphic record indicates that women were also on the move (often as
the partners of British recruits or the partners of returning veterans of different ethnic
origins), it is also probable that some of the British brooch types were taken to the con-
tinent on the clothes of such women (for how brooches were worn by Romano-​British
women, see Croom (2004); but see also Allason-​Jones (1995) for discussion of the ‘sex-
less’ nature of brooches). This conclusion is in many ways similar to that derived from
the epigraphic analysis of the presence of British-​born individuals, where the majority
had a military connection.
The variety of contexts in which British-​made brooches appear on the continent
reflects the diversity of their meanings and the associations that emanated through
their usage. This is an indication of the mobile identities of the brooches themselves.
The analysis of the sites where brooches were found in burials has shown that they were
probably brought by veterans returning from Britain to their own homelands (Ivleva
2011: 145). Within Britain, brooches were found in both inhumation and cremation bur-
ials, albeit in small numbers, and usually performed a double role: they were placed both
for their functionality (that is, they were used to fasten a piece of cloth containing the
remains of the deceased or to fasten a piece of clothing covering the deceased body) and
for their associations with the dead person (that is, they were placed in a wooden box
or a cloth or leather bag positioned next to the cremated remains) (Philpott 1991). As
for the continental examples—​where all burials were cremations—​some brooches were
positioned on top of the remains, which suggests that they were also used as cloth fas-
teners, while some appear to have been placed as votive offerings. Notably, most of the
brooches had their pins intact (slightly corroded but still with the spiral attached), sug-
gesting that they were deposited not as broken objects of no further use but as functional
items intended to secure pieces of clothing.
The brooches’ functionality was, therefore, an important factor; yet, one may ask why
these particular brooches were put into graves: that is, why the relatives of the deceased
chose British brooches to follow their beloved ones into the afterlife. The deliberate
inclusion of brooches suggests that they had important connotations for the deceased
whose remains they were supposed to secure as well as for the relatives, whose choice
of a particular brooch may have been a conscious act. While the evidence indicates that
brooches were rare as grave goods in Roman Britain, it does not mean that brooches
were not placed with the bodies of the deceased. Rather their absence as intact objects
may indicate that they were placed together with the body of the deceased and conse-
quently were completely burned and, therefore, did not survive to enter the archaeo-
logical record. However, for the users of British-​made brooches on the continent, these
objects may have had other associations: made in Britain, brought across the Channel
to the continent because of their functionality, not destroyed but kept intact, they could
have been used by other members of a family or community because of their limited
availability, exoticness, and uniqueness. Yet they officially ended their lives being buried
and being a protector of a dead individual’s remains. Therefore, it may not have been
 Britons on the Move    257

their precious looks or their functional value for the living, but their particular associa-
tions with the deceased that were important.
Brooches in continental burials are confined to areas where there is evidence for the
presence of veterans having returned from Britain. Brooches, therefore, could have been
valued by their owners and, later, by the relatives of the deceased for their associations
with the past, indicating the (dead) owner’s experience in Britain. If we think first about
British-​made brooches as the embodiment of a ‘British’ past and second about the care
taken to avoid them being destroyed and the deliberateness with which they were placed
in burials, it could follow that their inclusion in graves was a manifestation of memory
relating to the deceased’s connection with Britain, either as a soldier who had served in
Britain or as a Briton (be he or she male or female) who had died in a foreign land.
A similar conclusion can be proposed for the placement of British-​made brooches
in sanctuaries. Analysis has shown that these brooches too were brought by families of
returning veterans or by veterans themselves. While they wore British-​made brooches
in the setting of foreign cultures, the owners projected their past as people who had lived
in Britain. When they deliberately refused to use the brooches any more as clothes fas-
teners, the projection of a foreign past and the experience in a foreign land was brought
to an end. In this sense, brooches were subject to a twofold action: as personal offerings
to gods and as closures of past activities—​when such personal items were given away to
the gods, the past was symbolically buried and vows were fulfilled.
Notably, the majority of brooches recorded as rubbish and accidental losses were
found on sites where there is evidence of the stationing of troops coming from Britain.
The rarity of British-​made objects on the continent did not influence the decision of
some brooch owners intentionally to deposit these objects in rubbish pits, which meant
the functional death of the object. Such an action had consequences for the projec-
tion of any form of identity, be it gender, ethnic, or cultural; the intentional death of an
object stands for the death of meaning with which this item is associated, following up
on the death of identities desired or wished for or (un)intentionally projected. However,
such actions might have been influenced by the ready availability of brooches on the
site, which would have allowed owners to continue to transmit whatever identities they
wished and to serve as a reminder of an ethnic origin. One should take into account that
the presence of objects made in Britain on continental sites with a homogenous mate-
rial culture would have allowed them to stand out in the material record of that site. The
realization that a brooch was different and exotic might have provided the grounds for
the emergence of new meanings, possibly not existing in Britain itself, as is evident from
the utilization of British-​made brooches as embodiments of the past among the veter-
ans returning from Britain. Within British emigrant groups, the realization of brooches’
uniqueness could have reinforced the sense of being different, leading to the realization
of belonging to another culture, of being of different ethnic stock. Ethnicity, therefore,
becomes a by-​product of the relationship between the owner and the object: the particu-
larity of the artefact might enhance the expressions of ethnic identity.
The variety of treatment of British-​made brooches suggests that they were valued
by migrants for particular reasons and played an important part in the processes of
258

258   Tatiana Ivleva

remembrance and evocation of the past. That the idea of the past played a role when
brooches were put in specific contexts abroad indicates a desire to forget, to reinvent, to
evoke, or to project the past; it also emphasizes the value of memory within the groups
travelling from Britain.

Conclusions

Although the evidence is limited, it makes it possible to pinpoint the location of few
Britons abroad. It is clear that both the past and memory of the land of their birth were
important to these Britons, although it should be emphasized that for any moved indi-
viduals the past and homeland are important: an increase in the demonstration of one’s
origin is particularly noticeable in moved communities (Oltean 2009: 94–​95). Britons
were no different from any other migrants, and some were rather keen to make their
ethnic origin explicit through written language, whether the decision lay in naming
their tribe or in employing the adopted Roman construct Britonnes.
For most Britons wearing a British-​made brooch abroad would have been a necessary
and obvious thing to do, since it would have been brought among their personal pos-
sessions. Whether wearing a British brooch would reinforce the sense of ‘being’ from
Roman Britain is a difficult issue, since a variety of other identities and messages could
have been projected as well, yet the objects’ uniqueness and distinct style might have
provided ground for the growing realization of ‘being’ different.
At the beginning of this chapter I have stated that ethnic realization is born within
particular groups at the moment when cultural differences are recognized, providing the
motive for the universalization of these differences and making them into ‘practice’. In
both epigraphic and archaeological evidence, we see that universalization of ethnic con-
sciousness in a community living abroad might have taken place, because groups could
have realized their uniqueness through the use of different objects and through exploi-
tation of the imposed and invented label. In this sense, ‘being British’ abroad becomes
more of an invented identity, something that can be evoked and reinforced through the
use of brooches (and other British-​made objects) and when naming an origin.

Abbreviations
AE   L’Année Épigraphique
CIL     Corpus Inscriptionum Latinarum
ILJug 02  A
 . Šašel, and J. Šašel (1978). Inscriptiones latinae quae in Iugoslavia inter annos
MCMLX et MCMLXX repertae et editae sunt. Ljubljana: Narodni Muzej.
RIB      R
 . G. Collingwood and R. P. Wright, Roman Inscriptions of Britain. Oxford:
Clarendon Press, 1965–​to date.
 Britons on the Move    259

RMD   M. M. Roxan, Roman Military Diplomas. London: Institute of Archaeology,

1978–​to date.
RMM     B. Pferdehirt, Römische Militärdiplome und Entlassungsurkunden in der
Sammlung des Römisch-​Germanischen Zentralmuseums. Mainz:  Römisch-​
Germanisches Zentralmuseum, 2004.

References
Primary Sources
Aristotle, The Metaphysics, trans. H. Lawson-​Tancred. London: Penguin, 1998.
Dio Cassius, Roman History, trans. E. Cary. Bury St Edmunds: St Edmundsbury Press, 1989.
Martial,. Epigrams, trans. D. R.  Schackleton Bailey. Bury St Edmunds:  St Edmundsbury
Press, 1994.

Secondary Sources
Allason-​ Jones, L. (1995). ‘Sexing “Small Finds”’, in P. Rush (ed.), Theoretical Roman
Archaeology: Second Conference Proceedings. Aldershot: Avebury, 22–​32.
Allason-​Jones, L. (1999). ‘Women and the Roman Army in Britain’, in A. Goldsworthy and
I. Haynes (eds), The Roman Army as a Community:  Including Papers of a Conference
held at Birkbeck College, University of London on 11–​ 12 January, 1997. Journal of
Roman Archaeology Supplementary Series No 34. Portsmouth, RI:  Journal of Roman
Archaeology, 41–​51.
Bayley, J., and Butcher, S. (2004). Roman Brooches in Britain:  A  Technological and
Typological Study based on the Richborough Collection. London:  The Society of
Antiquaries in London.
Birley, A. (1980). The People of Roman Britain. Berkeley and Los Angeles:  University of
California Press.
Birley, E. (1988). ‘Pannonians in Roman Britain’, Zeitschrift für Papyrologie und Epigraphik, 73,
151–​155.
Bodel, J. (2001). ‘Epigraphy and the Ancient Historian’, in J. Bodel (ed.), Epigraphic
Evidence: Ancient History from Inscriptions. London: Routledge, 1–​57.
Bourdieu, P. (1998). Practical Reason: On the Theory of Action. Cambridge: Polity Press.
Brather, S. (2004). Ethnische Interpretationen in der frühgeschichtlichen Archäologie: Geschichte,
Grundlage und Alternativen. Berlin: De Gruyter.
Croom, A. (2004). ‘Personal Ornament’, in M. Todd (ed.), A Companion to Roman Britain.
Oxford: Blackwell, 288–​298.
Derks, T. (2009). ‘Ethnic Identity in the Roman Frontier: The Epigraphy of Batavi and Other
Lower Rhine tribes’, in T. Derks and N. Roymans (eds), Ethnic Constructs in Antiquity: The
Role of Power and Tradition. Amsterdam: Amsterdam University Press, 239–​283.
Dobson, B., and Mann, J. C. (1973). ‘The Roman Army in Britain and Britons in the Roman
Army’, Britannia, 4: 191–​205.
Evans, D. E. (1967). Gaulish Personal Names: A Study of Some Continental Celtic Formations.
Oxford: Clarendon Press.
260

260   Tatiana Ivleva

Giddens, A. (1979). Central Problems in Social Theory: Action, Structure and Contradiction in

Social Analysis. Houndmills: Macmillan.
Gilchrist, R. (2013). ‘The Materiality of Medieval Heirlooms:  From Biographical to Sacred
Objects’, in H. P. Hahn and H. Weiss (eds), Mobility, Meaning and the Transformations of
Things. Oxford: Oxbow Books, 170–​183.
Hassall, M. (1978). ‘Britain and the Rhine Provinces: Epigraphic Evidence for Roman Trade’,
in J. du Plat Taylor and H. Cleere (eds), Roman Shipping and Trade: Britain and the Rhine
Provinces. London: The Council for British Archaeology, 41–​48.
Hingley, R. (2009). ‘Cultural Diversity and Unity: Empire and Rome’, in S. Hales and T. Hodos
(eds), Material Culture and Social Identities in the Ancient World. New York: Cambridge
University Press, 54–​75.
Holder, P. A. (1980). Studies in the Auxilia of the Roman Army from Augustus to Trajan. BAR
International Series 70. Oxford: Archaeopress.
Hope, V. (2001). Constructing Identity: The Roman Funerary Monuments of Aquileia, Mainz
and Nimes. BAR International Series 960. Oxford: Archaeopress.
Ivleva, T. (2011). ‘British Emigrants in the Roman Empire: Complexities and Symbols of Ethnic
Identities’, in D. Mladenović and B. Russell (eds), TRAC 2010: Proceedings of the Twentieth
Theoretical Roman Archaeology Conference, Oxford 2010. Oxford: Oxbow Books, 132–​153.
Ivleva, T. (2012). ‘British Military Units and the Identity of British-​Born Recruits in the Roman
Army, between the First and Third Centuries ad’, Orbis Terrarum: Internationale Zeitschrift
für historische Geographie der alten Welt 10 [2008–​2011], 59–​92.
Jundi, S., and Hill, J. D. (1998). ‘Brooches and Identities in First Century ad Britain: More
than Meets the Eye?’ in C. Forcey et al. (eds), TRAC 97: Proceedings of the Seventh Annual
Theoretical Roman Archaeology Conference Nottingham 1997. Oxford: Oxbow Books, 125–​137.
Kakoschke, A. (2002). Ortsfremde in den römischen Provinzen Germania inferior und
Germania superior: Eine Untersuchung zur Mobilität in den germanischen Provinzen anhand
der Inschriften des 1. bis 3. Jahrhunderts n. Chr. Möhnesee: Bibliopolis.
Kakoschke, A. (2004). ‘Germanen’ in der Fremde:  Eine Untersuchung zur Mobilität aus den
römischen Provinzen Germania inferior und Germania superior anhand der Inschriften des
1. bis 3. Jahrhunderts n. Chr. Möhnesee: Bibliopolis.
Kennedy, D. (1977). ‘The ala I and cohors I Britannica’, Britannia, 8: 249–​255.
Leach, S., Lewis, M., Chenery, C., Müldner, G., and Eckardt, H. (2009). ‘Migration and Diversity
in Roman Britain:  A  Multidisciplinary Approach to the Identification of Immigrants in
Roman York, England’, American Journal of Physical Anthropology, 140: 546–​561.
Leach, S., Eckardt, H., Chenery, C., Müldner, G., and Lewis, M. (2010). ‘A Lady of
York: Migration, Ethnicity and Identity in Roman Britain’, Antiquity, 84: 131–​145.
Lucy, S. (2005). ‘Ethnic and Cultural Identities’, in M. Diaz-​Andreu and S. Lucy (eds),
The Archaeology of Identity:  Approaches to Gender, Age, Status, Ethnicity and Religion.
London: Routledge, 86–​109.
Mackreth, D. F. (2011). Brooches in Late Iron Age and Roman Britain. Oxford: Oxbow Books.
Malone, S. J. (2006). Legio XX Valeria Victrix: Prosopography, Archaeology and History. BAR
International Series 1491. Oxford: Archaeopress.
Matthews, K. J. (1999). ‘Britannus/​Britto:  Roman Ethnographies, Native Identities, Labels,
and Folk Devils’, in A. Leslie (ed.), Theoretical Roman Archaeology and Architecture: Third
Conference Proceedings. Glasgow: Cruithne, 14–​33.
Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of
Roman Archaeology, 17: 5–​25.
 Britons on the Move    261

Morris, F. (2010). North Sea and Channel Connectivity during the Late Iron Age and Roman
Period (175/​150 BC–​AD 430). BAR International Series 2157. Oxford: Archaeopress.
Mullen, A., and Russell, P. (2009). Celtic Personal Names of Roman Britain <http://​www.asnc.
cam.ac.uk/​personalnames/​> (accessed 3 June 2014).
Noy, D. (2001). Foreigners at Rome: Citizens and Strangers. London: Duckworth.
Oltean, I. A. (2009). ‘Dacian Ethnic Identity and the Roman Army’, in W. S. Hanson (ed.),
The Army and Frontiers of Rome:  Papers offered to David J.  Breeze on the Occasion of
his Sixty-​Fifth Birthday and his Retirement from Historic Scotland. Journal of Roman
Archaeology Supplementary Series 74. Portsmouth, RI: Journal of Roman Archaeology,
90–​103.
Philpott, R. (1991). Burial Practices in Roman Britain:  A  Survey of Grave Treatment and
Furnishing, AD 43–​410. BAR British Series 219. Oxford: Archaeopress.
Pudney, C. (2011). ‘Pinning down Identity:  The Negotiation of Personhood and the
Materialization of Identity in the Late Iron Age and Early Roman Severn Estuary’, in D.
Mladenović and B. Russell (eds), TRAC 2010: Proceedings of the Twentieth Annual Theoretical
Roman Archaeology Conference, Oxford 2010. Oxford: Oxbow Books, 115–​131.
Raybould, M. E., and Sims-​ Williams, P. (2009). Introduction and Supplement to the
Corpus of Latin Inscriptions of the Roman Empire Containing Celtic Personal Names.
Aberystwyth: CMCS.
Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge: Cambridge University Press.
Rothe, U. (2009). Dress and Cultural Identity in the Rhine–​Moselle Region of the Roman Empire.
BAR International Series 2038. Oxford: Archaeopress.
Rowland, R. J. (1976). ‘Foreigners in Roman Britain’, Acta Archaeologica Academiae Scientiarum
Hungaricae, 28: 443–​447.
Roymans, N. (2004). Ethnic Identity and Imperial Power: The Batavians in the Early Roman
Empire. Amsterdam: Amsterdam University Press.
Sims-​Williams, P. (2004). The Celtic Inscriptions of Britain: Phonology and Chronology, c. 400–​
1200. Oxford: Blackwell.
Snape, M. (1993). Roman Brooches from North Britain:  A  Classification and a Catalogue of
Brooches from Sites on the Stanegate. BAR British Series 235. Oxford: Archaeopress.
Swan, V. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman Pottery
Studies, 5: 1–​33.
Swift, E. (2000). Regionality in Dress Accessories in the Late Roman West. Montagnac: Monique
Mergoil.
Thompson, L. A. (1972). ‘Africans in Northern Britain’, Museum Africum, 1: 28–​38.
Tully, G. D. (2005). ‘A Fragment of a Military Diploma for Pannonia found in Northern
England?’, Britannia, 36: 375–​382.
van Driel-​Murray, C. (2003). ‘Ethnic Soldiers: The Experience of the Lower Rhine Tribes’, in T.
Grünewald and S. Seibel (eds), Kontinuität und Diskontinuität: Germania Inferior am Beginn
und am Ende der römischen Herrschaft. Berlin: De Gruyter, 200–​217.
Wallace-​Hadrill, A. (2007). ‘The Creation and Expression of Identity: The Roman World’, in S.
Alcock and R. Osborne (eds), Classical Archaeology. Oxford: Blackwell, 355–​380.
Wierschowski, L. (2001). Fremde in Gallien, ‘Gallier’ in der Fremde:  Die epigraphisch
bezeugte Mobilität in, von und nach Gallien vom 1. bis 3. Jh. n. Chr. (Texte, Übersetzungen,
Kommentare). Stuttgart: Steiner.
Wilmott, T. (2001). ‘Cohors I Aelia Dacorum: A Dacian Unit on Hadrian’s Wall’, Acta Musei
Napocensis, 38/​1: 103–​123.
262

Chapter 13

Britain , G au l ,
and Germa ny
Cultural Interactions

Tom Moore

Introduction

The changes that took place in Britain as a result of the influence of Rome, both before
and after the conquest, cannot be seen in isolation. Britain’s place on trade routes, its
position as a key focus for the Roman military, and its economic interaction with its
nearest neighbours meant it had close relationships with its adjacent provinces in Gaul
and Germany. Many of these links, particularly between southern Britain and Gaul,
stemmed from earlier interaction in the late Iron Age; indeed, many of the changes
that took place at this time underpinned the cultural changes that have in the past been
described as ‘Romanization’.
With widespread recognition that the term ‘Romanization’ poses a problematic idea
of ‘Roman’ versus native culture, inadequately explaining the nature of change in Roman
Britain, it has been suggested by some that interactions with Gaul and Germany were
more significant than those with the Mediterranean in defining cultural transformations
in Roman Britain. Some have gone so far as to suggest processes of ‘Gallicization’ in the
first century ad (e.g. King 1999: 189), with the changes at the end of Roman Britain part
of a corresponding ‘Germanization’ (Mattingly 2006: 224). Focusing on key changes in
diet, architecture, and burial rites, this discussion aims to assess the nature and extent
of cultural interactions between these provinces. In particular, it will be asked whether
terms such as Gallicization are helpful in reconceptualizing the processes of cultural
change before and after the Roman conquest, or whether these too present their own
problems in describing the changes that took place.
 Britain, Gaul, and Germany: Cultural Interactions    263

Cross-​C hannel Interactions

before the Roman Conquest

To assess the significance of interactions between Roman Britain, and Gaul and
Germany, we need first to assess the context and importance of these interactions prior
to Britain’s inclusion in the empire. Interaction between Britain and the near conti-
nent became increasingly pronounced from the second century bc onwards with the
appearance of Gallo-​Belgic pottery and Dressel 1 wine amphorae, first in central south-
ern Britain—​most notably at Hengistbury Head, Hampshire—​and later in south-​
east England. This interaction coincided with a range of cultural changes in southern
Britain. One of the most significant of these was the appearance of new burial rites in
the early first century bc, with a move away from excarnation and inhumation, to cre-
mation burial in southern and eastern England (Figure 13.1). These burial rites range
from the earlier, relatively simple cremations with urn and occasional grave goods, as
at Westhampnett, Hampshire (Fitzpatrick 1997: 201), to more elaborate ‘Welwyn-​type’
burials with a range of imported grave goods from the Mediterranean and Gaul, includ-
ing wine amphorae and metal drinking paraphernalia. These burials were normally sit-
uated in rectangular enclosures varying in size from a few metres to c. 50 metres across.

0 km 100
n
er
ev
rS
ve
Ri

River Thames

Figure 13.1  Distribution of cremation burials in late Iron Age Britain.

Source: after Fitzpatrick (1997); © T. Moore.
264

264   Tom Moore

Some were extremely elaborate, including wooden burial chambers (Folly Lane, Herts.),
tumuli (Lexden, Essex), and burials situated in large enclosures (Stanway, Essex).
Despite the presence of some high-​status imports from the Mediterranean, the
form and content of these burials emphasize a connection to the near continent rather
than influence directly from Rome. Similarly enclosed cremation burials are found in
areas of modern-​day Luxembourg and the Champagne region of France (Crummy et
al. 2007: 453), while similar burials and ceramic forms to those at Westhampnett also
occur in Normandy (Fitzpatrick 1997: 208). Some can be seen to be remarkably simi-
lar in form and content; for example, those at Welwyn and Folly Lane in Hertfordshire
and at Clemency and Goblange-​Nospelt in Luxembourg (Figure 13.2; Metzler et al.
1991; Niblett 1999: 397). The objects in these burials also emphasize continental links;
the ‘warrior burial’ from Stanway, for example, has brooches most closely paralleled in
central eastern France (Crummy et al. 2007: 446), while a mid-​first-​century bc inhuma-
tion burial from North Bersted, Sussex, has a helmet type normally found only on the
continent (Taylor et al. 2014). Even some of the Mediterranean imports in such graves
may have come not directly from Roman traders, but indirectly as prestige goods from
elites in Gaul (Millett 1990: 38). Changes in burial corresponded with other changing
attitudes to the body, including a marked increase in the use of toilet instruments (ear
scoops, nail cleaners, and tweezers) which also had their origins on the continent (Hill
1997; Eckardt and Crummy 2008).
To explain this phenomenon, from an early stage archaeologists noted Caesar’s claim
(Gallic War 5.12) that people from northern Gaul had previously migrated to south-
ern Britain. Allied to the archaeological evidence of the appearance of cremation burial,
archaeologists in the cultural–​historical tradition saw this as supporting evidence for
a large-​scale migration of people from the continent prior to Caesar’s invasion (Hawkes
and Dunning 1931; Cunliffe 1988: 147). The similarity in names of the social entities (or
‘tribes’) in central southern Britain and on the near continent, the Atrebates, in modern-​
day Hampshire and north-​eastern France and the Catuvellauni of Essex and Cataluni of
north-​eastern France (Crummy et al. 2007: 452), has also been used as evidence of move-
ment of peoples. More recently, most archaeologists would argue that many of the changes
witnessed in the late Iron Age took place around the time of, or after, Caesar’s invasion and
that, while burial rites changed, many other aspects of society remained similar to those
that had gone before, suggesting little evidence for a large-​scale movement of people.
Although we may not be witnessing mass migration prior to the Roman conquest,
there were clearly close relationships between communities in southern Britain and
northern France. Classical sources provide some evidence of the complex inter-​rela-
tions that existed across the Channel. Julius Caesar in his account of the Gallic War
observes, for example, that Diviciacus, king of the Suessiones, held land on both sides
of the Channel (Gallic War 2.4). It has also been suggested that Commius, king of the
Gallic Atrebates, a one-​time ally of Caesar, might have been installed in central south-
ern Britain to control the Atrebates there, as may have been the case with Tasciovanus
and the Catuvellauni (Creighton 2000: 64; 2006: 22). The appearance of coins bear-
ing the name ‘Commios’ in central southern Britain in the mid-​first century bc might
 (a)

(b)

Figure  13.2  Comparison of late Iron Age cremation burials from (a) Welwyn Garden City,
Hertfordshire, and (b) Clemency, Luxembourg.
Source: © British Museum and Musée National d’Histoire et d’Art, Luxembourg.
266

266   Tom Moore

corroborate this (Creighton 2000). These elite relationships before Rome’s final submis-
sion of Britain in the mid-​first century ad were part of a much wider interplay between
Rome and its neighbours in the manipulation of allies and client kings—​one in which
Gaul and southern Britain were inextricably linked (Creighton 2000, 2006).
Evidence for these complex political relationships is potentially present in the coinage
that appears in southern Britain from the second century bc. Gallo-​Belgic type coins,
which emerged on the continent imitating Greek coinage, were adopted and then imi-
tated in Britain from the early first century bc. These gold coins were unlikely to be for
everyday transactions, and their high value, alongside similarities of types across the
Channel, may indicate that they represented connections between elites both within and
beyond Britain (Cunliffe 2012: 17) or the movement of mercenaries between Britain and
Gaul (James 2001: 191). The distribution of imported coinage types in some areas, such
as the Isle of Wight (Wellington 2001), indicates that some communities were involved
in a range of connections along and across the Channel.
The adoption of coinage was only part of the broader changes in society. This included
the emergence of so-​called oppida, such as Verulamium and Camulodonum—​large dyke
complexes encompassing vast areas of landscape, largely lacking in dense settlement but
often displaying evidence of elite burial, ritual activity, and craft production (Moore
2012). The distinct differences between oppida in Britain and Gaul perhaps indicates
that, despite the political and cultural interactions evident above, there were significant
differences in social organization and power structures between the two areas. Only
one such site shares significant similarities with those on the continent. At Silchester,
Hampshire, an organized pre-​conquest street layout suggests relatively intense occupa-
tion in a defined enclosure more similar to the planned, nucleated settlements in north-
ern France—​such as Villeneuve-​Saint-​Germain in Picardy (Haselgrove 1996; Fulford
and Timby 2000: 549)—​than the sprawling dyke systems in Britain. At Silchester it has
even been suggested that the nature of the artefact assemblage, changes in dietary pref-
erences (including increasing consumption of cattle and pig), and the planned layout
indicate it was a planted settlement, with large numbers of settlers from northern Gaul
(Fulford and Timby 2000: 563). However, while the presence of Gaulish immigrants and
traders (for example, at Hengistbury: Fitzpatrick 2001: 89)—​particularly specialists or
elites—​is certainly a possibility (Millett 1990: 34), there seems no reason to argue for
wholesale migration or that it was only immigrants who were using new forms of mate-
rial culture (cf. Fitzpatrick and Timby 2002: 171). Many aspects of ceramic use and treat-
ment of the dead continued (as at other oppida such as Bagendon and Verulamium),
reflecting existing practices, suggesting the majority of people were indigenous and that
more complex social changes were underway than just immigration.

Explaining Interaction Before the Roman Conquest

Earlier explanations for the relationship between southern Britain and the continent
in the late Iron Age tended to emphasize the economic nature of these relations and
 Britain, Gaul, and Germany: Cultural Interactions    267

the expansion of Rome in this process as fundamental. Evidence of Dressel 1 ampho-

rae (imported first into central southern and then south-​eastern England), alongside
Strabo’s description (Geography 5.2.2) of the exports from Britain, indicated to many
that there was increased trade with the continent that was indirectly influenced by the
expansion of Rome into Gaul (Cunliffe 1988: 103). With indigenous elites in south-​east
England acting as ‘middlemen’, it was argued that this region became a core zone that
exploited the peripheries of eastern and western England for commodities—​such as
slaves, precious metals, and foodstuffs—​to supply Rome (Haselgrove 1982).
More recently, the economic nature of this interaction has been questioned, with a
recognition that the amount of imports arriving in Britain was relatively small and that
the connections across the Channel were related to social and cultural interaction rather
than trade (Hill 2007). Creighton (2000, 2006) has argued that the division between
pre-​and post-​conquest Britain overlooks the ongoing interaction between elites in
southern Britain and Rome from the time of Caesar’s invasion in the mid-​first century
bc, suggesting that much of the material record—​such as the elaborate burials—​reflects
the complex political and identity changes underway either side of the Channel.
There is also increasing recognition that interaction across the Channel preceded
the late Iron Age. For example, recognition that round-houses, often thought of as a
peculiarly British architectural tradition, did exist in northern France implies that cul-
tural interaction between communities on both sides of the coast had much longer
antecedents (Lefort and Marcigny 2009; Cunliffe 2012: 16; Webley 2015). This indi-
cates that the expansion of Rome was not necessarily the prime instigator of change:
Britain was already participating in a range of cultural zones. The nature of interaction
was also complex: burial practices across the Channel—​while often strikingly simi-
lar in particular instances—​obscure a variety of rites that existed, even in those areas
where cremation was adopted (Hamilton 2007). Despite their similarities, these burials
reflect a combination of both longer-​lived burial traditions and continental phenomena.
We should not underestimate the longer-​term impact of internal social and economic
changes (as a result of population increase, agricultural intensification, and the dynam-
ics of existing social systems), which were also important in stimulating the change in
social structures, forms of identity, and burial rites taking place at this time (Hill 2007).
What we are seeing, therefore, is not simply the adoption of foreign traditions, but
the creation of new practices. None of these can be put down to one particular source—​
external or internal, Rome or Gaul; rather, they are the result of a combination of ele-
ments adopted and manipulated by individuals and communities. The varied practices
and grave goods in the individual burials at Stanway, for example, emphasize that indi-
viduals were reconceptualizing identities at a range of levels: gender, status, ethnicity.
The use of toilet instruments reflects this process: a continental form of material cul-
ture was adopted in the late Iron Age, but then certain types continue to be used (often
in different ways) in Roman Britain—​in contrast to the areas in which they originally
emerged (Eckardt and Crummy 2008).
The identities being expressed appear to be focused on elite connections and social
networks, and there is no reason necessarily to see many of these individuals as
268

268   Tom Moore

immigrants from northern Gaul. We should be wary of seeing the practices adopted
as ethnic attributes, such as ‘Gallic’ (let alone ‘Roman’). Gaul and Germany were them-
selves artificial constructs, which only emerged from the works of Roman authors
and, later on, of Imperial administrators, which often had little to do with the realities
of pre-​Roman identities. Writers such as Julius Caesar, in his Gallic War, and Tacitus,
in his Germania, often defined the habits and nature of these regions in opposition to
each other, as civilized versus barbaric, for example (O’Gorman 1993; Wells 1999: 102).
Such concepts, and the regions they portrayed, often have little relationship to the evi-
dence from archaeology, which indicates the varied identities, settlement patterns, and
languages of these communities, and they often show no clear division between east
and west of the Rhine (Woolf 1998: 49; 2011; Wells 1999: 102). We might even ques-
tion whether identity was expressed at the level of the civitas or ‘tribe’ (for example,
Catuvellauni or Atrebates), when such labels were also largely a creation of the post-​
conquest era (Moore 2011). Individuals, like those at Stanway, therefore, may have been
imitating other elite individuals as part of social connections and expressions of sta-
tus, but this need not be seen as an expression of ethnicity. When discussing interaction
between Britain, Gaul, and Germany, we must recognize the complex variety of identi-
ties and range of communities involved in these areas (both before and after the Roman
conquest) and the dangers in defining attributes as reflecting ethnic groups—​many of
which are likely to be constructs of our own and Roman writers’ imaginations.

Continental Influence
in  Roman Britain

Villa Architecture
The close links between southern Britain and Roman Gaul continued after the conquest,
at least between certain sectors of the elite. Perhaps one of the clearest signs of these inter-
actions is in the architectural changes that took place in early Roman Britain. The row
houses and winged corridor villas, which developed in the late first and early second
century ad, are a shared style that occurs across Britain and northern Gaul. Many such
early villas, constructed in the late first century ad, such as Ditches, Gloucestershire, and
Gorhambury, Hertfordshire, developed along similar lines and relatively contemporane-
ously to examples in northern Gaul (Figure 13.3; Trow et al. 2009: 55, 70; Taylor 2012: 181).
As with their continental counterparts, they also often emerged from existing late Iron
Age settlements, potentially representing the continuity of those groups and individuals
who had participated in connections across the Channel prior to the conquest.
The origin of these building forms is uncertain, but they are clearly not a direct imita-
tion of Italian architectural forms and appear to have developed in Gaul (Haselgrove
1995: 71). The most likely explanation for their adoption in Britain is a process similar
 Mansfield Woodhouse, Notts. Herouville, Calvados

10m

Boos-Le Bois Flahault, Seine-Maritime Gayton Thorpe, Norfolk

Gadebridge Park, Herts.

Herouville, Calvados (late phase)

Behen, Somme
Ditches, Glos.

C
C

Boos-Le Bois Flahault, Seine-Maritime (late phase)

Shakenoak II, Oxon.

C C

Figure 13.3  Similarities in villa layouts from northern Gaul and Britain.

Source: © T. Moore.
270

270   Tom Moore

to the one that took place before the conquest. They were not necessarily the buildings
of immigrants from Gaul, rather, indigenous elites were attempting to imitate the trap-
pings of their counterparts on the opposite side of the Channel (Trow et al. 2009: 70).
The development of stone architecture in early Roman Britain was largely alien to indig-
enous building traditions, and it has been suggested that Gallic masons were imported
for such constructions (Blagg 1984: 260); there is certainly tantalizing evidence to sup-
port this, with epigraphic evidence of at least one stonemason at Bath (RIB 151) who
hailed from northern Gaul (Mattingly 2006: 299). Powerful individuals (and their archi-
tects) were likely to be far more familiar with the elite forms of architecture in northern
Gaul than the peristyle houses of Italy; it is natural that it is this template they should
imitate. At sites like Ditches, cultural connections may have been indirect through elites
located around more politically connected towns, such as Verulamium; but such indi-
viduals were potentially well aware of the wider cultural and social connections into
which they were embedded. By contrast, the owners of the palatial Italian-​style court-
yard villa of the late first century ad at Fishbourne, Sussex (Perring 2002: 33), were
clearly making more direct connections with the Roman imperial administration.
The adoption of such villa designs was not just one of functional adaptation to a
northern climate or imitation of Gallic examples. The layout of such villas also repre-
sented a mode of social display that emphasized power relations within and between
communities, perhaps harking back to indigenous modes of social relations, as well as
new ones (Smith 1997). The similarities in villa form may, therefore, reflect similarities in
social relations that had developed concurrently in southern Britain and northern Gaul
immediately before the Roman conquest. Despite these early similarities, however, villa
architecture in Britain did not continue to emulate those on the near continent (Taylor
2012: 182), which emphasizes the way villa development existed within its own pecu-
liarly British social and economic context, while also sitting within the broader context
of household organization seen elsewhere in the Roman world (Perring 2002: 206).

Religion and Romano-​Celtic Temples

Links between Britain and Gaul in the late Iron Age have also been suggested in ritual
practices. This has largely been on the basis of Caesar’s assertion that the Druids of Gaul
had their training in Britain (Gallic War 6.13). Whether there were close links in ritual
and religious practices between these areas is open to debate (Webster 1999); in reality,
the complex regional nature of the Iron Age archaeological record across north-​western
Europe implies quite varied and localized practices. However, the appearance of shrine
structures in late Iron Age Britain marked a significant change in religious practices—​
one that seems to have spread from the continent (see also Kamash, this volume).
Dedicated shrines emerged in Britain in the first century bc, with the most well-​
understood example at Hayling Island, Hampshire, on the south coast of England. This
consisted of a rectangular enclosure with circular inner structure, which saw the votive
deposition of coinage and metalwork (King and Soffe 2001). In similar fashion to its
 Britain, Gaul, and Germany: Cultural Interactions    271

counterparts of the continent, the Iron Age structure was replaced after the conquest
by a Romano-​Celtic temple (these are sometimes referred to as Gallo-​Roman temples:
e.g. Derks 1998). These temples were part of a common architectural form found across
the north-​western Roman provinces: they were normally a rectangular (sometimes cir-
cular) cella often situated within a larger temenos (Figure 13.4). As they were restricted
to Gaul, Germany, and Britain, it has been suggested that this form of temple represents
a fusion between ‘Celtic’ and Roman ritual practices (King 2007). In Britain, there has
been a attempt to draw the antecedents for these sanctuaries back to the Iron Age (e.g.
King 2007), although the evidence for Iron Age shrines in Britain continues to rely on
a handful of examples, many of which may not be shrines at all (Millett 1995). The idea

0 150 km

Figure 13.4  Approximate distribution of Romano-​Celtic temples.

Source: after Derks (1998); © T. Moore.
272

272   Tom Moore

that these structures represented a fusion of religious practices has also been critiqued,
not least on the lack of evidence for an overarching ‘Celtic’ religion in these areas.
Despite these problems, the reason behind the commonality of temple design requires
explanation (King 2007: 13). The lack of evidence for antecedents in Britain suggests the
inspiration for these came from elsewhere. The earliest securely dated and recogniz-
able temple at Hayling Island, in its form and practices, implies it was inspired by the
much longer tradition of rectangular sanctuary structures, which occurred in northern
Gaul from the third century bc, at sites such as Gournay-​sur-​Aronde and Ribemont-​
sur-​Ancre (King and Soffe 2001: 121). The more widespread adoption of Romano-​Celtic
temples after the conquest has been seen by some, therefore, as part of the adoption
of what were perceived by communities in Britain as ‘Romanized’ practices from else-
where, where they had already been hybridized in Gaul and Germany (Millett 1995:
98). The same may be true of another similarly restricted phenomenon: that of Jupiter
columns—​tall, decorated columns surmounted with a figure riding a horse over a
half-​human/​half-​snake creature and found predominantly in Germany, eastern Gaul,
and, to some extent, in Britain. Wells (1999: 220) has seen the iconography depicted on
these columns and their distribution as representing a fusion of indigenous and Roman
beliefs transformed into a peculiarly north-​western imagery (see also Woolf 2001).
We need to be wary, however, of seeing ritual practices either as imported by immi-
grant communities or as continuity from Iron Age practices. The varied and fluid nature
of Iron Age ritual practices is well recognized and the complex adoption and adaptation
of practices, despite overarching similarities, are increasingly recognized in the Roman
provinces (e.g. Goldberg 2009). Late Iron Age ritual practice across Gaul and Britain
was in a state of flux, and the appearance of aspects such as Romano-​Celtic temples was
part of widespread changes. Many of the gods depicted, the religious processes, and
the architectural forms—​such as the Hayling Island temple—​that developed as part of
the so-​called Romano-​Celtic beliefs owe as much to post-conquest developments and
importations from the continent as from indigenous belief systems (Webster 1995).
Indeed, Derks (1998: 243) regards the Romano-​Celtic temple itself as a new architectural
and ritual development, with many in Gaul, as in Britain, showing little evidence of con-
tinuity from the Iron Age. The adoption of the Romano-​Celtic temple in parts of Britain
might be better seen, therefore, as part of a wider repertoire of ritual architecture that
emerged in the north-​western Roman provinces as part of a dialogue between existing
Iron Age practices and a variety of external forces, including the Roman state. As with
the adoption of particular villa architecture, this particular form of temple may have
been regarded in some areas as expressions of personal benefaction and engagement
within wider religious spheres beyond the local.

Diet and FoodWays

Food ways are a particularly sensitive way in which groups differentiate themselves and
express new identities in colonial contexts, and it is through this arena that much of
 Britain, Gaul, and Germany: Cultural Interactions    273

what has, in the past, been regarded as early ‘Romanization’ can be seen (Meadows 1994:
136). However, like much else in early Roman Britain, the source and inspiration for
changing paraphernalia and food choices were not necessarily from the Mediterranean,
but from Britain’s nearest neighbours.
The importation of Gaulish ceramics began well before the Roman conquest, begin-
ning with Amorican vessels to sites like Hengistbury Head in the early first century bc
and then—​far more widely in the early first century ad—​imports of terra rubra and
nigra finewares from northern France, followed by the import of terra sigilatta via the
Rhine from central France. These Gallic ceramics had a far wider range of distribution,
before and after the Roman conquest, than imports direct from Italy (Millett 1990: 33).
These new forms of pottery included platters and cups that had not existed in the Iron
Age ceramic repertoire. This signalled new forms of eating and drinking that were more
personalized and indicate different foodstuffs being consumed. The adoption of new
dining wares was matched by changes in consumption patterns—​at least on some high-​
status sites in southern Britain—​with greater quantities of cattle, pig, and fowl being
eaten (for example, at Skeleton Green, Puckeridge (King 1988); Ditches (Reilly 2009);
Silchester (Fulford and Timby 2000; Albarella 2007: 395)). Such consumption prefer-
ences were more similar to those in Gaul than the continued focus on sheep on most
late Iron Age sites elsewhere in Britain (Albarella 2007). These changes have often been
interpreted as largely reflecting new markets open to the emergent pottery industries in
Gaul, who were now firmly ‘Romanized’ (Fitzpatrick and Timby 2002: 169). However,
this rather underestimates the choices being made by indigenous communities as part
of a broader adoption of new culinary habits by certain sectors of the community.
Major changes in diet appear to have continued on a more widespread basis after
the Roman conquest (see also Maltby and Van der Veen, this volume). King’s analysis
(1984, 1999) of faunal assemblages in Britain indicates an increasing proportion of cattle
being consumed, particularly on military and urban sites. Traditionally, this has been
argued as a process of ‘Romanization’ led by the elite and then imitated more widely in
society (King 1984: 193). This preference for beef in Britain corresponds with a similar
pattern across the northern provinces—​one that contrasts sharply with consumption
patterns in Italy and the Mediterranean. It also marks a major change from Iron Age
norms of a predominantly sheep-​based diet, although the latter appears to have contin-
ued in more rural areas (King 1984; 1999: 189). The parallel of these dietary patterns with
Gaul and Germany, rather than Italy, led King (1984: 198) to see this as a process of the
‘Gallicization’ or ‘Germanization’ of British diets, with the Roman army being the engine
for change, as many of its units had been based in Gaul and Germany, or contained
members who originated there. This dietary preference subsequently became regarded
as ‘Romanized’ after the conquest of Britain and was adopted by those of high status and
in the more ‘Romanized’ areas of the new province. As with all such patterns, there is
significant diversity within and between provinces (King 1999: 190), but the variation in
faunal assemblages emphasizes the changing nature of what it meant to be ‘Roman’ and
indicates that the continued influences—​particularly on the upper echelons of society in
Britain—​were often from military sources, and by extension from Gaul and Germany,
274

274   Tom Moore

rather than from the Mediterranean. Vivien Swan (2009) has also suggested that par-
ticular cooking vessels, and their associated culinary habits, were imported into Britain
in the first and second centuries ad via Gaulish influences, which themselves were a
‘Gallicized’ development of forms that had their origins further east. Like King (1984,
1999), Swan (2009: 51) sees the vector of this material culture and culinary change in
Britain as Gauls serving in the military, who brought their culinary preferences with
them.
There are potential problems in regarding the dietary changes and culinary habits in
early Roman Britain as a process of ‘Gallicization’. Such a term implies cultural imitation
of ethnic identity: that diet or food ways were defined by (and reflected) ethnic identi-
ties, which were then adopted wholesale by different communities. As King’s study (1999)
illustrates, changes in dietary practices varied within and between sectors of society, and
changes in diet were more complex than the adoption of, or resistance to, a cultural pack-
age. In both late Iron Age and Roman Britain, certain serving wares and types of food
were adopted but were often used in traditional ways, or foodstuffs were prepared in new
forms (Meadows 1994: 137). Food ways and ceramic traditions were increasingly diverse
across the Roman Empire and were themselves in a state of flux, through trade, adoption,
and fusion (Fulford 2010). What studies by King (1999) and Swan (2009) illustrate more
convincingly is that particular sectors in society could adopt elements of traditions from
other parts of the empire and utilize them within new contexts, as part of new identities—​
be they elites, the military, or a host of other social groups.

Migration and Cultural Change

Many of the changes regarded as ‘Romanizing’ in pre-​conquest Britain have been

explained in the past by the migration of already Romanized peoples from northern
Gaul, either on a large scale (Hawkes and Dunning 1931) or by smaller groups and indi-
viduals (Fulford and Timby 2000: 563; Crummy et al. 2007). The evidence for system-
atic migration prior to the conquest is largely lacking, but to what extent can some of
the cultural changes in Britain—​once it was part of the Roman Empire—​be explained
by migration from Gaul and Germany? It is increasingly recognized that the Roman
Empire saw significant population mobility and that the urban population (at least) of
Britain was relatively cosmopolitan (Eckardt 2010). It has been suggested that individu-
als from Gaul, because of the existing cultural and linguistic similarities with southern
Britain, were in the vanguard of exploiting opportunities in the new province (Mattingly
2006: 109; Fulford 2010: 69). There is certainly evidence for the presence of individuals
from Gaul, such as monuments to individuals from the Bituriges-​Cubi of central Gaul in
York (RIB 678; Cool 2010) and from the Bellovaci of northern Gaul in London (RIB 3014;
Mattingly 2006: 298), but assessing the number of immigrants is far harder. The army
too was clearly a significant agent of cultural influences from Gaul and Germany, be it
in the form of dress (Swift 2000), architecture (Creighton 2006: 91), or food ways (King
 Britain, Gaul, and Germany: Cultural Interactions    275

1999; Swan 2009). However, this was not necessarily a simple process of the transfer of
cultural signifiers with particular groups and units; many such elements had already
been incorporated into more complex fusions of military identity (Fulford 2010). The
influence of individuals cannot be underestimated, but migration as a cause of many
of these changes was only one element of interaction with Gaul and Germany; influ-
ence was far more complex than through people bringing identities with them (see also
Ivleva and Eckardt, this volume, for discussions on mobility in the empire).

Late Roman Britain

As with the claimed ‘Gallicization’ of identities in late Iron Age and early Roman Britain,
some have also perceived changes in dress in the late Roman western empire, largely
influenced from origins in the Germanic world (Swift 2000: 119), as a corresponding
‘Germanization’ of existing practices (Mattingly 2006: 228). One of the strongest routes
by which identities changed in the late Roman Empire was via the Roman army, with
the use of peoples with German origin as mercenaries. Significant movements of groups
of people—​particularly groups attached to the army but including women and chil-
dren—​occurred in the late empire, but there were also earlier movements in smaller
numbers. The cemetery at the fort at Brougham, for instance, with its unusual burial
rites and brooch types, has been used as an argument for the presence of people with
trans-​Danubian or German origins (Cool 2004). By the fourth century ad, a variety of
evidence indicates both incoming groups and individuals, alongside changes in local
forms of dress (see Cool, this volume, for a discussion of the role of dress and iden-
tity). At Lankhills, Winchester, some burials with particular types of bracelets in the late
Roman cemetery have been used to suggest a Pannonian origin for some individuals, or
at least the expression of foreign identities (Swift 2000: 75; Pearce 2010: 85). The process
was not one way: the occasional presence of belt fittings of ‘British type’ in Germany
emphasizes that, while such an item might have a symbolic meaning in one place, it
could have different meanings in other contexts (see Eckardt and Crummy 2006: 94).
Such interactions did not just happen in the late Roman period, however. We should not
underestimate the array of influences that took place across the Roman period, many of
which occurred before these specific Germanic influences in the late empire. For exam-
ple, Bruhn (forthcoming) has argued that the adoption of glass bangles in northern
Britain from the late first century ad could have been inspired by the particular pref-
erence for bangles in the lower Rhineland; the localized adoption of material culture
through specific channels of interaction (for example, the Roman army) could then
have been ‘reinvented’ in a new regional context and should not necessarily be regarded
as an ‘ethnic’ trait.
Many of these developments were as much about changes in the expression of mili-
tary identity and the emergence of localized identities, as adopting notions of ‘German’
identities. The adoption of the cross-​bow brooch, for example, which appears to have
276

276   Tom Moore

originated in Hungary, was a symbol of military or bureaucratic status rather than eth-
nic origin (Swift 2000). Notably, however, as before, such symbols were being adopted
from the peripheries of the empire, particularly Germany (Swift 2000; Pearce 2010: 85)
rather than from the Mediterranean. As with changing modes of dress and diet ear-
lier, we must remain cautious in seeing this process as ‘Germanization’. Many of these
changes in material culture, such as the supposedly Pannonian bracelets from burials
at the cemetery of Lankhills, often argued as representing ethnic identity, might bet-
ter be seen in terms of age or gender identities (Gowland 2007; Theuws 2009), while
many of the rites expressed appear to relate to a range of different practices, rather than
one definable cultural package (Pearce 2010: 90). Rather like earlier influences from
Gaul, many of the new items of dress adopted from the German world became part of
Romano-​British provincial dress, moving from association with the military and ‘for-
eignness’, to localized expressions of place and identity.

Conclusions: Romanization,
Gallicization, or Germanization?

The impact of interaction with Gaul and Germany has a long history in understand-
ing the changes from the late Iron Age in to the Roman period. However, it is no longer
regarded as a product of mass migration or just an economic relationship. As attitudes
to the complex nature of identity change in Roman Britain have emerged, scholars such
as Richard Reece (1988: 11) increasingly saw this as a process not of Romanization, but
of people adopting a multitude of identities and practices from elsewhere in the empire:
‘becoming more Gaulish, more Rhinelandish, more Spanish’. While helpful in recog-
nizing the varied external influences Britain was subject to from the first century bc
onwards—​and moving away from seeing these changes as ‘Romanizing’—​defining
these processes as ‘Gallicization’ is problematic. In using such terms we are in danger
of once again reducing the process of cultural change in the Roman period to a sim-
plistic process of acculturation: the transferring of one culture for another (James 2001:
191). Alternatively, and equally problematically, is to continue to see these changes as
processes of ‘hybridization’, the fusing of two discreet cultures: indigenous Iron Age
with either Roman or Gallic. Recognition by some that Gallic or German culture had
already changed and been informed by its earlier interaction with Rome goes some way
to accepting that these identities and influences were not, in themselves, reflective of
discreet indigenous practices. In both Roman Britain and in the late empire, particu-
lar material culture and architectural forms—​while originating in areas such as Gaul or
Germany—​could be perceived as part of a broader ‘Roman’ cultural experience or take
on connotations of ‘inclusivity’ in an exotic or high-​status set of practices—​even if much
of those practices might have been relatively alien to elites in Rome itself. However, we
need to be cautious in regarding ‘Gallicization’ (or even ‘Gallo-​Belgicization’ (Creighton
 Britain, Gaul, and Germany: Cultural Interactions    277

2000: 82) as potentially ‘perceived Romanization’ by people in Britain. In so doing we

are in danger of effectively seeing aspects such as dietary change or the adoption of
Romano-​Celtic temples as ‘Romanization by proxy’: forms of culture that were essen-
tially ‘Gallic’ but perceived and engaged with as ‘Roman’ by their adopters. We have to
be careful, therefore, in defining normative comparanda (such as villa or temple forms)
and regarding these as ‘the same’ outside of their local context; while forms of architec-
ture or types of grave goods may be similar between Britain and Gaul, the ways in which
they were used in everyday practice may have been significantly different.
The evidence discussed above suggests instead that cultural interactions and influ-
ences between Britain, Gaul, and Germany were not simply a process of adoption.
Links—​when they existed—​seem to be through particular media by particular groups
and often (as seen in the burial rites and dining practices of the immediate pre-​con-
quest period) adapted and utilized within an array of existing traditions and localized
requirements. For some periods this was largely restricted to emergent elites who had
close political (and perhaps economic) connections with Gaul, but not all elites engaged
with this process. As Creighton (2006: 77) recognizes, the province of Britain (and even
pre-conquest Britain) was ‘full of individuals with differing experiences’, based on back-
grounds from different areas of the world and different social agendas, ensuring this was
a dynamic set of relationships and not one done at the level of ethnic group. We should
also not necessarily regard this as a one-​way process of elite adoption. It is perhaps better
to regard (parts of) southern Britain and northern Gaul in what James (2001: 191) has
described as a ‘zone of cultural convergence’ (or, rather, a range of zones), which existed
both prior to, and after, Britain’s incorporation into the empire. Through cross-​pollina-
tion of cultural preferences and media for exchange and clientage, shared practices of
burial and villa architecture existed, but they were more complex than simply the elite
imitating their Gallic neighbours. Thus, while we might envisage certain individuals in
Britain adopting aspects of identities and expressions of power from late Iron Age and
early Roman Gaul, these were part of a range of identities being expressed and transfig-
ured in a colonial encounter. Later, in the Roman province, the adoption of ritual prac-
tices and architectural forms that owed their origins to Gaul may have been understood
and practised in quite different and locally oriented ways.
There is a danger in seeing this as merely an elite process that then ‘trickled down’
through emulation to the rest of the community. The adoption of aspects of material cul-
ture or architectural forms—​while often processes of elite emulation—​were also influ-
enced by other factors; for example, the discreet distribution of Romano-​Celtic temples
in Britain reflecting localized social and ritual practices rather than more Gallicized
practices. ‘Gallic’ and ‘British’ identities were far from static, and, in reality, it is ques-
tionable how much such identities meant to, or defined, inhabitants of these provinces.
The varied burial rites, settlement patterns, and material culture of southern Britain
and northern Gaul emphasize diverse and discrepant identities, cross cut by concepts
of status, gender, and age, on both sides of the Channel (Mattingly 2004). It is increas-
ingly recognized in Roman studies that identities—​including ethnicity—​are fluid and
multifaceted (e.g. Eckardt 2012: 251). Concepts of Gallic and Germanic identities were
278

278   Tom Moore

themselves constructs, and it is increasingly apparent that regional, so-​called ethnic

identities within the broader sphere of ‘Germany’ were being (re)conceptualized in the
context of Roman power and administration (Roymans 2004). Many identities, such
as ‘Batavian’, may have meant far more to people who claimed them (and those exter-
nal to them) than any notion of Germania; but such identities themselves also changed
their meaning, significance, and expression between the first century bc and fourth
century ad.
The relationships between Gaul, Germany, and Britain and how cultural transforma-
tions were played out, is likely to have been equally complex. Interactions and influences
from Gaul and Germany were just part—​if a significant and recurring one—​of the mul-
titudes of relationships in which different individuals, groups, and regions in Britain
were engaged from the first century bc to the fifth century ad and cannot be isolated
from the multitude of indigenous and external influences on expressions of power, iden-
tity, ritual practices, and social structure (see also Esmonde Cleary, this volume).

Acknowledgements

I am grateful to Richard Hingley, Rebecca Gowland, and Claire Nesbitt for their com-
ments on earlier versions of this chapter and to staff at the Musée national d’histoire et
d’art, Luxembourg, for their assistance in obtaining, and granting permission to use, the
image of the tomb at Clemency.

Abbreviation
RIB  R
 . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I:
Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Primary Sources
Caesar, The Gallic War, trans. C. Hammond. Oxford: Oxford University Press, 1996.
Tacitus, The Agricola and The Germania, trans. H. Mattingly. London: Penguin, 1948.
Strabo, Geography, trans. H. L. Jones. Cambridge, MA: Loeb, 1923.

Secondary Sources
Albarella, U. (2007). ‘The End of the Sheep Age: People and Animals in the Later Iron Age’, in
C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow
Books, 389–​402.
 Britain, Gaul, and Germany: Cultural Interactions    279

Blagg, T. F.  C. (1984). ‘An Examination of the Connexions between Military and
Civilian Architecture’, in T. F.  C. Blagg and A. C. King (eds), Military and Civilian in
Roman Britain:  Cultural Relationships in a Frontier Province. BAR British Series 136.
Oxford: Archaeopress, 249–​264.
Bruhn, J. (forthcoming). ‘Glass Bangles in Northern Britain: The Regional Uptake of Hybrid
Cultural Objects in the Roman Period’.
Cool, H. (2004). The Roman Cemetery at Brougham, Cumbria: Excavation 1966–​67. Britannia
Monograph Series 21. London: Society for the Promotion of Roman Studies.
Cool, H. (2010). ‘Finding the Foreigners’, in H. Eckardt (ed.), Roman Diasporas: Archaeological
Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology
Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 27–​44.
Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge
University Press.
Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge.
Crummy, P., Benfield, P., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An
Élite Burial Site at Camulodunum. Britannia Monograph Series 24. London: Society for
Promotion of Roman Studies.
Cunliffe, B. (1988). Greeks, Romans and Barbarians: Spheres of Interaction. London: Batsford.
Cunliffe, B. (2012). ‘Calleva in Context’, in M. Fulford (ed.), Silchester and the Study of Romano-​
British Urbanism. Journal of Roman Archaeology Supplementary Series 90. Portsmouth,
RI: Journal of Roman Archaeology, 14–​22.
Derks, T. (1998). Gods, Temples and Ritual Practices: The Transformation of Religious Ideas and
Values in Roman Gaul. Amsterdam: Amsterdam University Press.
Eckardt, H. (2010). ‘A Long Way from Home: Diaspora Communities in Roman Britain’, in
H. Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in
the Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth,
RI: Journal of Roman Archaeology, 99–​130.
Eckardt, H. (2012). ‘Foreigners and Locals in Calleva’, in M. Fulford (ed.), Silchester and the
Study of Romano-​British Urbanism. Journal of Roman Archaeology Supplementary Series
90. Portsmouth, RI: Journal of Roman Archaeology, 247–​256.
Eckardt, H., and Crummy, N. (2006). ‘“Roman” or “Native” Bodies in Britain: The Evidence of
Late Roman Nail Cleaners’, Oxford Journal of Archaeology, 25/​1: 83–​103.
Eckardt, H., and Crummy, N. (2008). Styling the Body in Late Iron Age and Roman
Britain:  A  Contextual Approach to Toilet Instruments. Monographies Instrumentum 36.
Motagnac: Instrumentum.
Fitzpatrick, A. (1997). Archaeological Excavations on the Route of the A27 Westhampnett Bypass,
West Sussex, 1992, ii. The Cemeteries. Wessex Archaeology Report 12. Salisbury:  Wessex
Archaeology.
Fitzpatrick, A. (2001). ‘Cross-​Channel Exchange, Hengistbury Head, and the End of the
Hillforts’, in J. Collis (ed.), Society and Settlement in Iron Age Europe. Actes du XVIIIe coloque
de l’AFEAF Winchester, April 1994. Sheffield: J. R. Collis Publications, 82–​97.
Fitzpatrick, A., and Timby, J. (2002). ‘Roman Pottery in Iron Age Britain’, in A. Woodward and
J. D. Hill (eds), Prehistoric Britain: The Ceramic Basis. Oxford: Oxbow Books, 161–​172.
Fulford, M. (2010). ‘Roman Britain: Immigrations and Material Culture’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman
Archaeology, 67–​78.
280

280   Tom Moore

Fulford, M., and Timby, J. (2000). Late Iron Age and Roman Silchester: Excavations on the Site
of the Forum Basilica 1977, 1980–​86. Britannia Monograph Series 15. London: Society for the
Promotion of Roman Studies.
Goldberg, M. (2009). ‘The Dichotomy in Romano-​Celtic Syncretism:  Some Preliminary
Thoughts on Vernacular Religion’, in M. Driessen, S. Heeren, J. Hendriks, F. Kemmers,
and R. Visser (eds), TRAC 2008: Proceedings of the Eighteenth Annual Theoretical Roman
Archaeology Conference, Amsterdam, 2008. Oxford: Oxbow Books, 187–​202.
Gowland, R. (2007). ‘Beyond Ethnicity: Symbols of Social Identity from the Fourth to Sixth
Centuries in England’, in S. Semple and H. Williams (eds), Early Medieval Mortuary Practices.
Anglo-​Saxon Studies in Archaeology and History 14. Oxford: Oxbow Books, 56–​65.
Hamilton, S. (2007). ‘Cultural Choices in the “British Eastern Channel Area” in the Late Pre-​
Roman Iron Age’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and
Beyond. Oxford: Oxbow Books, 81–​106.
Haselgrove, C. (1982). ‘Wealth, Prestige and Power: The Dynamics of Late Iron Age Political
Centralisation in South-​East England’, in C. Renfrew and S. Shennan (eds), Ranking,
Resource and Exchange. Cambridge: Cambridge University Press, 79–​88.
Haselgrove, C. (1995). ‘Social and Symbolic Order in the Origins and Layout of Roman Villas’,
in J. Metzler, M. Millett, N. Roymans, and J. Slofstra (eds), Integration in the Early Roman
West:  The Role of Culture and Ideology. Luxembourg:  Dossiers d’Archéologie du Musée
National d’Histoire et d’Art IV, 65–​76.
Haselgrove, C. (1996). ‘Roman Impact on Rural Settlement and Society in Southern Picardy’, in
N. Roymans (ed.), From the Sword to the Plough. Amsterdam: Amsterdam University Press,
127–​188.
Hawkes, C., and Dunning, G. (1931). ‘The Belgae of Gaul and Britain’, Archaeological Journal,
87: 150–​335.
Hill, J. D. (1997). ‘The End of One Kind of Body and the Beginning of Another Kind of Body?
Toilet Instruments and Romanization’, in A. Gwilt and C. Haselgrove (eds), Reconstructing
Iron Age Societies. Oxford: Oxbow Books, 96–​107.
Hill, J. D. (2007). ‘The Dynamics of Social Change in Later Iron Age Eastern and South-​Eastern
England c.300 bc–​ad 43’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain
and Beyond. Oxford: Oxbow Books, 16–​40.
James, S. (2001). ‘Romanization and the Peoples of Britain’, in S. Keay and N. Terrenato (eds),
Italy and the West: Comparative Issues in Romanization. Oxford: Oxbow Books, 187–​209.
King, A. (1984). ‘Animal Bones and Dietary Identity of Military and Civilian Groups in Roman
Britain, Germany and Gaul’, in T. F. C. Blagg and A. C. King (eds), Military and Civilian
in Roman Britain:  Cultural Relationships in a Frontier Province. BAR British Series 136.
Oxford: Archaeopress, 187–​218.
King, A. (1988). ‘A Comment on the Animal Bones’, in T. W. Potter and S. D. Trow (eds),
‘Puckeridge-​Braughing, Hertfordshire: The Ermine Street Excavations 1971–​2’, Hertfordshire
Archaeology, 10: 154–​155.
King, A. (1999). ‘Diet in the Roman World: A Regional Inter-​Site Comparison of the Mammal
Bones’, Journal of Roman Archaeology, 12: 168–​202.
King. A. (2007). ‘Romano-​ Celtic Temples in Britain:  Gallic Influence or Indigenous
Development?’, in R. Haeussler and A. King (eds), Continuity and Innovation in Religion
in the Roman West. Volume 1. Journal of Roman Archaeology Supplementary Series 67.
Portsmouth, RI: Journal of Roman Archaeology, 13–​18.
 Britain, Gaul, and Germany: Cultural Interactions    281

King, A., and Soffe, G. (2001). ‘Internal Organisation and Deposition at the Iron Age Temple on
Hayling Island, Hampshire’, in J. Collis (ed.), Society and Settlement in Iron Age Europe: Actes
du XVIIIe coloque de l’AFEAF Winchester, April 1994. Sheffield:  J. R.  Collis Publications,
111–​124.
Lefort, A., and Marcigny, C. (2009). ‘La Probable Agglomération portuaire du second Âge du
Fer d’Urville-​Nacqueville: Un état de la documentation’, Bulletin de l’AMARAI, 22: 39–​81.
Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of
Roman Archaeology, 17: 5–​25.
Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Penguin.
Meadows, K. (1994). ‘You are What You Eat: Diet, Identity and Romanisation’, in S. Cottam,
D. Dungworth, S. Scott, and J. Taylor (eds), TRAC 94:  Proceedings of the Fourth Annual
Theoretical Roman Archaeology Conference, Durham. Oxford: Oxbow Books, 133–​140.
Metzler, J., Waringo, R., Bis, R., and Metzler-​Zens, N. (1991). Clemency et les tombes de
l’aristocratie en Gaule Belgique. Luxembourg: Musée National d’Histoire et d’Art.
Millett, M. (1990). The Romanization of Britain. Cambridge: Cambridge University Press.
Millett, M. (1995). ‘Re-​Thinking Religion in Romanization’, in J. Metzler, M. Millett, N.
Roymans, and J. Slofstra (eds), Integration in the Early Roman West: The Role of Culture and
Ideology. Luxembourg:  Dossiers d’Archéologie du Musée National d’Histoire et d’Art IV,
93–​100.
Moore, T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and
Roman Studies’, Journal of Social Archaeology, 11: 334–​360.
Moore, T. (2012). ‘Beyond the Oppida: Polyfocal Complexes and Late Iron Age Societies in
Southern Britain’, Oxford Journal of Archaeology, 31/​4: 391–​417.
Niblett, R. (1999). The Excavations of a Ceremonial Site at Folly Lane, Verulamium. Britannia
Monograph Series 14. London: Britannia.
O’Gorman, E. (1993). ‘No Place like Rome: Identity and Difference in the Germania of Tacitus’,
Ramus, 22/​2: 135–​154.
Pearce, J. (2010). ‘Burial, Identity and Migration in the Roman World’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman
Archaeology, 79–​98.
Perring, D. (2002). The Roman House in Britain. London: Routledge.
Reece, R. (1988). My Roman Britain. Cirencester: Cotswold Studies.
Reilly, K. (2009). ‘Animal Bone’, in S. Trow, S. James, and T. Moore (eds), Becoming Roman,
Being Gallic, Staying British: Research and Excavations at Ditches ‘Hillfort’ and Villa 1984–​
2006. Oxford: Oxbow Books, 187–​210.
Roymans, N. (2004). Ethnic Identity and Imperial Power: The Batavians in the Early Roman
Empire. Amsterdam Archaeological Studies 10. Amsterdam: Amsterdam University Press.
Smith, J. T. (1997). Roman Villas: A Study in Social Structure. London: Routledge.
Swan, V. (2009). Ethnicity, Conquest and Recruitment:  Two Case Studies from the Northern
Military Provinces. Journal of Roman Archaeology Supplementary Series 72. Portsmouth,
RI: Journal of Roman Archaeology.
Swift, E. (2000). The End of the Western Roman Empire:  An Archaeological Investigation.
Stroud: Tempus.
Taylor, A., Weale, A., and Ford, S. (2014). Bronze Age, Iron Age and Roman Landscapes of the
Coastal Plain and a Late Iron Age Warrior Burial at North Bersted, Bognor Regis, West Sussex:
282

282   Tom Moore

Excavations 2007–2010. Reading: Thames Valley Archaeological Services Monograph

Volume 19.
Taylor, J. (2012). ‘The Idea of the Villa: Reassessing Villa Development in South-​East Britain’, in
N. Roymans and T. Derks (eds), Villa Landscapes in the Roman North: Economy, Culture and
Lifestyles. Amsterdam: Amsterdam University Press, 179–​194.
Theuws, F. (2009). ‘Grave Goods, Ethnicity, and the Rhetoric of Burial Rites in Late Antique
Northern Gaul’, in N. Roymans and T. Derks (eds), Ethnic Constructs in Antiquity: The Role
of Power and Tradition. Amsterdam: Amsterdam University Press, 283–​320.
Trow, S., James, S., and Moore, T. (2009). Becoming Roman, Being Gallic, Staying British: Research
and Excavations at Ditches ‘Hillfort’ and Villa 1984–​2006. Oxford: Oxbow Books.
Webley, L. (2015). ‘Rethinking Iron Age Connections across the Channel and North Sea’, in D.
Garrow, H. Lamdin-​Whymark, and F. Sturt (eds), Continental Connections: Cross-​Channel
Relationships from the Lower Palaeolithic to the Iron Age. Oxford: Oxbow Books [ebook].
Webster, J. (1995). ‘Interpretatio:  Roman Word Power and the Celtic Gods’, Britannia,
26: 153–​162.
Webster, J. (1999). ‘At the End of the World: Druidic and Other Revitalization Movements in
Post-​Conquest Gaul and Britain’, Britannia, 30: 1–​20.
Wellington, I. (2001). ‘Iron Age Coinage on the Isle of Wight’, Oxford Journal of Archaeology,
20/​1: 39–​57.
Wells, P. (1999). The Barbarians Speak. Princeton: Princeton University Press.
Woolf, G. (1998). Becoming Roman:  The Origins of Provincial Civilization in Gaul.
Cambridge: Cambridge University Press.
Woolf, G. (2001). ‘Representation as Cult: The Case of the Jupiter Columns’, in W. Spickermann,
H. Cancik, and J. Rüpke (eds), Religion in den Provinzen Roms. Tubingen: Mohr Siebeck,
117–​134.
Woolf, G. (2011). Tales of the Barbarians:  Ethnography and Empire in the Roman West.
London: Wiley-​Blackwell.
 PA RT  I I

SOCIETY AND
T H E  I N DI V I DUA L
284
 Chapter 14

Inscrip tions a nd
Identi t y

Valerie M. Hope

IMP CAES DIV[I NERVAE F] /​NERVAE TRAIA[NO AVG] /​GER

PONTIF MAXIM[O TRIB] /​POTEST P P /​COS III /​LEG II AVG
For the emperor Caesar Nerva Trajan Augustus, conqueror of Germany,
son of the deified Nerva, pontifex maximus, with tribunician power,
father of his country, consul for the third time, the Second Augustan
Legion made this.
(RIB 330; Caerleon, dedicatory panel for a building or structure)

DEA DIA/​NA SACRATI/​SSIMA VOTV/​M SOLVIT /​VETTIVS BE/​

NIGNVS LIB
To the most sacred goddess Diana, Vettius Benignus, a freed slave, ful-
filled his vow.
(RIB 138; Bath, small votive altar)

D M /​IVLIE BRICE AN XXXI /​SEPRONIE MARTINE AN VI /​

SEPRONIVS MARTINVS F C
To the spirits of the departed Julia Brica, aged 31, and Sepronia Martina,
aged 6, Sepronius Martinus had this set up.
(RIB 686; York, tombstone with full figure portrait images of
the deceased above the epitaph)
286

286   Valerie M. Hope

Inscriptions on Stone

Latin inscriptions capture the imagination by presenting us with the range of Roman
society: children and adults, men and women, soldiers and civilians, emperors and
slaves, gods and mortals. Inscriptions can record the names of the powerful, but more
often represent ordinary people doing relatively everyday things such as remembering
the dead or thanking a god. The term ‘inscription’ is wide ranging in its modern applica-
tion; it can incorporate words cut on stone, scratched or painted graffiti, labels on pots,
stamps on tiles, and the text of writing tablets. The intention here is to focus on inscrip-
tions on stone, sometimes termed monumental inscriptions (for a useful summary of the
distinctions between, and different types of, inscriptions, see Tomlin 2011: 135; Cooley
2012: 117–​126; RIB I and RIB III record inscriptions on stone). Arguably these monu-
mental inscriptions were far from mundane, since in Roman Britain they seem to have
been relatively unusual and thus were, and continue to be, special. Many of the surviving
inscribed stone objects were prestige items or associated with substantial monuments,
and the creation of these objects and buildings required considered decisions, as well as
expense. To inscribe on stone involved careful evaluation of what to say and how to say
it, and implied that the chosen words would have a claim on permanency. The inscriber
or dedicator was thinking about the past, present, and future, since they were recording
what was important in both their past and present and projecting it forwards in time.
When setting up, or commissioning, the inscriptions given ‘at the head of this chapter’,
Vettius Benignus, Sepronius Martinus, and the members of the Second Legion must have
been conscious of saying the right things, communicating a message, and publicly hon-
ouring commitments and relationships (equal and unequal, human and divine).
Inscriptions on stone were generally only a small part of a much bigger whole—​both
a bigger sequence of actions, rituals, or customs and a bigger monument, building, or
structure. To be understood in full, inscriptions cannot be isolated from these acts and
monuments or from the people who made them and the audience that viewed them.
In reality, reconstructing how inscriptions worked, were read, and reacted to, and how
these things changed across time, is not always straightforward owing to frequent loss
of context. Nevertheless we can still ask who commissioned inscriptions in Roman
Britain, where they were set up and why, and what collectively and individually these
inscriptions reveal about life in the province. This chapter looks briefly at where, and in
what numbers, stone inscriptions survive from Roman Britain and the nature of the ‘epi-
graphic habit’ in the province, before considering how and what the inscriptions com-
municated and their role in identity construction.

Numbers and Distribution

From Roman Britain there are, approximately, 2,800 surviving inscriptions on stone.
Many of these inscriptions are incomplete, rendering reading and full understanding
 Inscriptions and Identity    287

of the original text impossible. Where the purpose of the inscription can be established,
most fall into the broad categories of religious dedications (mainly to named gods and
goddesses on altars), epitaphs (mainly on stele, with a few from larger monuments),
building dedications (mainly inscribed panels recording the construction or restoration
of buildings), centurial stones (small panels indicating construction work completed by
centuries of legions and other military units, particularly common on Hadrian’s Wall),
and milestones (pillars recording the emperor’s titles and distances). There has been lit-
tle research on Romano-​British stone inscriptions as a complete body of evidence and
thus numerical (and other) comparisons between different types of inscriptions. Most
research focuses on specific categories of inscription—​for example, epitaphs or build-
ing dedications. In addition, attaching figures to these broad categories is problematic
owing to damage to stones and also the overlap between categories. Some very rough
numbers for these categories would be:  votive inscriptions—​well in excess of 1,000;
epitaphs—​in the region of 600; building dedications—​around 200; centurial stones—​
around 250; and milestones—​approximately 100; with the remaining fragmentary
inscriptions being of uncertain use.
In terms of find-​spots, the surviving inscriptions are not evenly distributed across the
province. There are marked concentrations of inscriptions at military bases, especially
in the north of the province (Biró 1975; Cepas 1989; Mattingly 2008). Most inscriptions
originate from the legionary garrisons at Chester, Caerleon, and York, from along the
length of Hadrian’s Wall, and from the provincial capital of London. The coloniae of
legionary veterans (Colchester, Lincoln, and Gloucester) and towns that had strong mil-
itary or official connections (Bath and Richborough) also produce a number of inscrip-
tions. The native civitates, by contrast, did not extensively set up inscriptions, though
small clusters are found at Silchester, Wroxter, and St Albans (Jones and Mattingly
1990: map 5.10). If inscriptions in other materials are included to create overall ‘epi-
graphic indexes’, these concentrations change little (Mattingly 2008; for mapping of the
epigraphy (of all types) at Roman London, see Holder 2007; find-​spots for sculpture
from across Roman Britain also suggest a similar distribution, with clear military and
urban biases; see Stewart 2010).
The chronological distribution of these inscriptions is also uneven, though it
needs to be noted that the precise dating of inscriptions is a precarious business.
Latin inscriptions rarely contained dates or other information (for example, refer-
ences to emperors or consuls or specific events) that allow for the allocation of a
certain date or date range. However, approximate dates (and periods) can be sug-
gested for some inscriptions by comparison with examples that can be dated (e.g.
RIB 331, 665, 667), reference to known troop movements (though these are often
supported or inferred by inscriptions, r​ isking circular arguments), and certain epi-
graphic and stylistic conventions. The approximate chronological distribution of
the inscriptions of Roman Britain indicates that production increased during the
later first century, was at its height during the second and early third centuries, with
a marked decline in numbers during the later third century; there may have been a
small resurgence in Christian epitaphs during the late fourth century (Hope 1997;
Handley 2001; Knight 2010).
288

288   Valerie M. Hope

The Epigraphic Habit

Compared to many other provinces of the Roman Empire, the number of inscriptions
that survive from Roman Britain is paltry. In terms of its epigraphy, Roman Britain
would seem to be the poor relation to the rest of the empire, an oddity and an epigraphic
backwater. Is this epigraphic difference between Britain and other provinces the mis-
fortune of what has survived or does it reveal important things about Britain’s place in
the Roman Empire? Are the number and distribution of inscriptions related to levels
of literacy, to the uptake of Latin, to urbanization and the extent of ‘Romanization’? Big
questions and issues such as these are often explored through epigraphy, and the act of
inscribing on stone in Latin can be viewed as a gauge (however crude) of acculturation
and the dialogue between Roman and indigenous cultures (Cepas 1989: 1; Cherry 1998;
Woolf 1998).
At the outset it needs to be acknowledged that chance and circumstance have played
their role. The vagaries of survival may account for low numbers of inscriptions from
some parts of Roman Britain, with many stones vanishing from view, owing to loss,
reuse, or destruction. For example, if so many inscribed stones had not been reused in,
and then recovered from, Chester’s late defensive walls, the state of Romano-​British
epigraphy would be all the poorer (Richmond and Wright 1956; Clay 2004). However,
this same factor of chance survival could hold for many other regions of the Roman
Empire, where only a small percentage of what once existed has probably survived (for
relative numbers, see Bodel 2001: 6–​10). No doubt many inscribed stones have been lost
from Roman Britain, but what now exists may still be roughly indicative of relative rates
of production. Another factor that may have impacted upon the creation of inscrip-
tions is the poor supply of suitable stone, especially in the south of the province (Mann
1985; see also Stewart 2010 for stone used in sculptures). However, some areas did over-
come supply issues, even importing stone from large distances. Inscribed imported
marble slabs are to be found, for example, at Caerleon (RIB 330) and London (Hassall
and Tomlin 1989: 326), while the early tombstone to the legionary centurion Marcus
Favonius Facilis (RIB 200) at Colchester was made from a Gallic limestone also used for
military tombstones at Mainz, Bonn, and Neuss (Hayward 2006; see also Isserlin 1998;
Hayward 2009). The absence of good stone must have had an impact upon the ability
to inscribe and monumentalize at some sites, but is not a sufficient explanation for the
overall low number of Romano-​British inscriptions.
Britain may have fewer inscriptions than many other provinces, yet it is not alone in
the unevenness (geographical and chronological) of what survives, which is important
for our understanding of the epigraphy of the province. In general, the act of inscribing
in the Roman world was an urban phenomenon, and, in many regions, it was closely
aligned with the military and centres of administrative power (Bodel 2001: 80). Large
areas of Gaul—​such as Brittany and western Gaul, for example—​produced no or very
few inscriptions (Woolf 1998:  maps 4.5 and 4.6); from the German provinces most
 Inscriptions and Identity    289

inscriptions have been found at military forts and veteran colonies (for example, Mainz
and Cologne). What survives from Britain, even in reduced numbers, follows this trend.
The use of epigraphy was also not chronologically static. A brief article published by
MacMullen (1982; building on earlier work by Mrozek 1973) has had a substantial impact
in identifying ‘the epigraphic habit’ and its inconsistencies. MacMullen noted that the
production of inscriptions during the Roman period rose and fell; inscribing was not
a constant phenomenon. In particular, MacMullen mapped how inscribing increased
in popularity during the first and second centuries ad and then rapidly fell during
the third century ad. We should not overlook that the accurate dating of inscriptions
is fraught with difficulty (see ‘Numbers and Distribution’, above), so such charting of
rises and falls can only ever be approximate. Nevertheless, even allowing for differences
between sites (and different types of epigraphy), MacMullen identified an empire-​wide
trend. It is clear that the act of inscribing in Latin, and setting up certain types of monu-
ments with which inscriptions were associated, went in and out of fashion. Explaining
‘the epigraphic habit’ has taxed a few minds, and explanations have noted regional vari-
ations and patterns and have highlighted factors such as status display, social mobility,
geographic mobility, elite emulation, and patronage networks (Meyer 1990, 2011; Woolf
1996; Häussler 1998). It may well be that there is no simple all-​encompassing explana-
tion for the rise and fall of the ‘epigraphic habit’ that suits all sites or parts of the empire
equally; and it is certainly the case that finding factors (for example, spread of citizen-
ship or status display) for the uptake of inscribing is often easier than accounting for its
decline (Meyer 2011).
For Roman Britain, with its comparatively low numbers of inscriptions, it is possible
that the ‘epigraphic habit’ had insufficient time to catch on, before it went into decline
among key groups (Hope 1997). This possibility becomes more marked when it is noted
that elsewhere in the Latin-​inscribing west most inscriptions were epitaphs (Saller and
Shaw 1984: 124; Bodel 2001: 30; Meyer 2011: 193). In the German provinces, from where
many of Britain’s ‘Roman’ soldiers first arrived, the number of epitaphs, and the ‘epi-
graphic habit’ in general, peaked in the late first century. By the time the troops were
securely established in Roman Britain, the relevance of inscribed tombstones may have
already been on the wane, for soldiers at least. In Roman Britain epitaphs are outnum-
bered by votive offerings to deities, and these also eclipse epitaphs at some other military
sites elsewhere in the empire during the second century ad, although never rivalling the
earlier popularity of epitaphs (Biró 1975: 27, 42; Decker and Selzer 1976; Cepas 1989: 56;
Forbis 1996; Derks 1998; Hope 2001: 93). In short, the Roman ‘epigraphic habit’ (at least
in the west), which was mainly an urban and military phenomenon, incorporated dif-
ferent types of epigraphy, and that associated with commemorating the dead, which was
often the most popular, did not substantially take hold in Roman Britain.
In Roman Britain the low level of epigraphy cannot be viewed as a general indicator
of poor literacy, low levels of urbanization, or some sort of rejection of ‘Roman’ display
(especially funerary) culture. Instead, epigraphy remained largely military, largely reli-
gious, and not of interest or relevance to the majority of the population, and, in this,
290

290   Valerie M. Hope

Roman Britain is not that different from some other parts of the empire. Elsewhere it
was epitaphs that most urban populations ‘bought into’, and the significance of these
epitaphs may have been declining (especially among the military that came to Britain)
by the time Roman forts and towns were securely established. In general among the
civil population, neither the poor nor the moderately well-​off nor the wealthy set up
inscriptions. The absence of the most economically and politically successful from the
epigraphic record is striking, and it seems likely that the indigenous British elite did not
compete within either the monumental or the epigraphic sphere, since their position
was embedded in native traditions not established or contested through urban archi-
tectural munificence (Biró 1975; Blagg 1990; Millett 1990: 105; Revell 2007; 2009: 183;
Eckardt et al. 2009). There is evidence, for example, for a range of tomb types, including
some substantial structures, with an eye to visibility and integration into the landscape,
in both rural and peri-​urban zones, even if these were, for the most part, un-​inscribed
(Esmonde Cleary 2000: 137; Strück 2000; Pearce 2011). The low level of epigraphy does
not indicate that the civil population was actively rejecting epigraphy, but that it may
simply have preferred other methods of display, such as mosaics (Mattingly 2008: 68),
and display locations other than just the main urban centres.
A range of factors was probably at play as to why Britons did not take up epigraphy
in any great numbers, but this makes those inscriptions that do survive, and those who
used them, all the more interesting.

Communicating to an Audience

Inscriptions, by their very nature, suggest social interaction and communication. The
commissioner decided that he or she had something—​a building, for example—​and
also the very act of creating (or restoring) that building that were worth recording in
verbal form. Initially the commissioner of the inscription entered into a dialogue with
the stonemason, engaging him and dictating, or creating with him, a suitable text. Once
set up, the inscription was poised to engage the passer-​by, to gain an audience, who, by
looking and reading, would continue to note the building and the actions of the people
named. It is possible that some inscriptions were concealed from view (inside a tomb or
shrine, for example), visible to only a select few, but the vast majority of Romano-​British
inscriptions on stone do appear to have been public statements, even if we cannot know
the exact identity of the target audience, or the extent to which that audience paused,
read, understood, and appreciated the information therein.
Inscriptions were not cut in native dialects. Latin was the language of inscriptions
(for the few exceptions see ‘Identity’). It was through Latin that the inscription com-
municated; through Latin that the commissioner of the inscription chose to speak. The
use of inscribed Latin suggests a certain level in both the use and the understanding of
Latin among those communities (military and civil) in which inscriptions were set up;
however, it is difficult to generalize about how many in the inscribing military–​urban
 Inscriptions and Identity    291

zones, or indeed those from non-​inscribing areas, spoke and read Latin. The presence or
absence of epigraphy, or other forms of the written word (for example, writing tablets),
is not a simple gauge of relative literacy (see, e.g., Raybould 1999; Pearce 2004; Adams
2007: 580–58​1; Parsons 2011: 117–​119). To set up a Latin inscription was, nevertheless, a
considered choice, and presumes an audience that would be able either to understand
the text or to grasp its symbolic significance or both. Inscriptions, and the monuments
they were associated with, were put up both to be read and to be seen. How individuals
responded to the texts depended on their level of literacy to be sure, but also on factors
such as status, personal background, and previous experiences (Barrett 1993: 237). Many
inscriptions may have been aimed at peers—​at a fellow Latin-​speaking audience of sol-
diers and officials—​but inscriptions were accessible and visible to others, who may have
responded in varied fashions. It is worth noting, for example, that inscriptions were fre-
quently riddled with abbreviations, and that we can create competing interpretations for
the use and impact of these abbreviations. There were, no doubt, cost implications here.
The stonemason charged by the letter, and the more letters cut the greater the expense.
Abbreviations saved money, yet can also reinforce the impression that inscriptions were
a rarefied means of communication, for the special few, for those in the ‘know’ who
could break the code. On the other hand, it could be argued that abbreviations made
inscriptions more accessible, since it gave them a pictorial or pictogram aspect that, due
to repetitive use, combined with the general simplicity of content, made inscriptions
easy to decipher and demanded only a low level of literacy (Woolf 1996: 28; Petronius,
Satyricon 58.7). Set in context—​a cemetery, an altar, a large building—​inscriptions
may have been readily understandable on a general level. You did not need to be able
to read Latin or, even if you could, stop and take the time to read it, to get the meaning
or construct some sort of message from what was seen, even if this, and the individual
response, diverged (in some respects) from the message of the original commissioner.
Names were a central part of all inscriptions, and again these were probably relatively
easy to spot and ‘read’ in terms of the inscription content. Inscriptions recorded the acts
of people, either individuals or groups, who set up buildings or honoured the emperor
or dedicated altars to the gods or remembered a dead person. The inscription com-
memorated these acts and named names, in the process revealing and/​or constructing
identities for those involved (see ‘Identity’). Some inscriptions were, in fact, just names.
Centurial stones were inscribed with the centurion’s name or other military unit titles,
but, by their placement, they affectively, if often with a subtle simplicity, communicated
the contribution of that unit to the construction of the building involved. Hadrian’s Wall
produces many such markers indicating the completion of stretches of the wall and its
structure, such as that carried out by the century of Julius Rufus (RIB 1357), whose men
worked on the eastern part of the wall.
Epitaphs were similarly centred on names, that of the dead person, and also that of
the person(s) who had set up the monument—​a clear indication that inscriptions were
about commemorating actions as much as marking places (such as graves or altars).
Epitaphs might also state the relationship between the deceased and the commemora-
tor, add an age at death (and or length of service and unit for a soldier), and employ the
292

292   Valerie M. Hope

opening phrase Dis Manibus (‘To the spirits of the departed’), often abbreviated to D.M.
At Chester, for example, Lucius Sempronius Probianus commemorated his young son:
D(is) M(anibus) /​L(ucio) F(e)STINIO PROBO F(i)L(io) /​VIX(it) AN(nos) II D(ies)
XXVIIII /​L(ucius) SEM(pronius) PROBIANVS /​PATER F(aciendum) C(uravit) (RIB
537). The epitaph started with the abbreviation D.M., then recorded the boy’s name
(Lucius Festinius Probus), before carefully noting his age of 2 years and 29 days, fol-
lowed by the name of the commemorator (Lucius Sempronius Probianus), his relation-
ship (pater) to the deceased and that he had made the monument (faciendum curavit).
Votive altars often placed a similar emphasis on names and relationships, though in
this case between the human and the divine. The god’s name was usually inscribed
first, followed by that of the dedicator. A statement that the vow had been willingly and
deservedly fulfilled (votum solvit libens merito—​often abbreviated to V.S.L.M.) might
be included, and sometimes the motivation behind the dedication specified. In County
Durham, for example, a cavalryman gave thanks to the god Silvanus for the capture of
a fine wild boar (RIB 104), although most votive inscriptions did not reveal the explicit
reason for the dedication (for an overview of the content of epitaphs and religious ded-
ications, see Raybould 1999: 10–​33, 94–​118; for lists of inscriptions with brief descrip-
tions, by type, including milestones, building dedications, and religious inscriptions, see
de la Bédoyère 1999).
Building dedications focused on the name of the dedicator(s), often linking this with
deities and or the name of the emperor. A reference to the type of structure was, some-
times but certainly not always, included. In Caerleon, an unspecified building (perhaps
the fort rampart) was marked by an imported inscribed panel dedicated by the Second
Legion to the emperor Trajan (RIB 330, see above). In York, the fortress gates were
similarly dedicated to Trajan, whose full name and honours were listed, by the Ninth
Hispanic Legion (RIB I. 665). Milestones, generally in the shape of pillars, also placed
emphasis on imperial titles, often giving only the name and titles of the emperor; the
distances that these milestones marked were not always inscribed on them (perhaps
they were added in paint), but they were affective territorial indicators and statements of
imperial authority, especially during the third and early fourth centuries ad.
Most inscriptions were then brief and formulaic, yet, because of their verbal dimen-
sion, they have been readily mined for insights into Roman society. Names can pro-
vide status indicators and suggest ethnic origins; recorded relationships highlight
family formation; ages at death hint at demographic profiles; unit titles indicate troop
movements; religious dedications suggest ritual practices; and so on. Because inscrip-
tions often survive in large numbers (at least from elsewhere in the empire), they have
formed the basis for statistical analysis on demography in particular (see, e.g., Saller
and Shaw 1984; Saller 1994; Scheidel 2012). Such approaches are not without problems,
since the population commemorated is a self-​selecting one or is skewed towards cer-
tain social groups (in particular, see Hopkins 1966). There have been some attempts
to employ such approaches with Romano-​British inscriptions, with epitaphs in par-
ticular, although the relative paucity of their numbers makes their use and interpreta-
tion particularly problematic (Raybould 1999; and, with caution, Tobler and Adams
 Inscriptions and Identity    293

2007). To note one example of the difficulties: Saller and Shaw (1984: 139–1​45) have
argued that the lack of familial references in Romano-​British epitaphs suggests that
there was little local recruitment into the army, contrasting this with provinces such
as North Africa and Spain, where soldiers were frequently commemorated by family
members, which, they believe, indicates local recruitment. Compared with many other
provinces, however, Roman Britain has few relevant epitaphs, and those that do not
mention military families mainly date to the first and early second centuries ad, with
familial references becoming more frequent in epitaphs dating to the second and early
third centuries—​the periods also best represented by the epitaphs from Africa and
Spain. In other words, the chronology of the epitaphs is key to the interpretation of the
available evidence, and the issue of local recruitment (or not) into the Romano-​British
army cannot be resolved using this epigraphic data (Alföldy 1987; Hope 2001: 70–​7 1;
2007: 116–​117; Phang 2001: 245–​246; note also Dobson and Mann 1972; Raybould 1999:
95). The ‘epigraphic habit’, as we have seen, was not chronologically or geographically
static, and it appealed to some groups more than to others.
In terms of their content, inscriptions were often standardized and repetitive (hence
their appeal for number crunching), but they were also only one part of a bigger whole.
A building inscription, for example, was a relatively small aspect of the temple, arch,
or baths to which it was attached. To catch the eye, inscriptions might be cut on large
panels and the latter might be of expensive stone and/​or framed with architectural or
sculptural details. Letter size could vary, with larger text size for key information; and
some letters may have been painted. The building dedication found at Caerleon to the
emperor Trajan by the Second Augustan Legion (RIB 330) was inscribed, and then the
letters picked out in red paint, on a large imported marble slab, with the first line of the
inscription—​the name of the emperor—​cut in the largest letters. The tomb of the procu-
rator Classicianus, from London, was a substantial funerary altar, the front surface of
which was dominated by the text of the epitaph, the opening line ‘DIS MANIBVS’ cut
in letters large enough to be read from a distance of hundred paces (Grasby and Tomlin
2002: 71; RIB 12). Several distance slabs from the Antonine Wall used sculptural motifs,
such as legionary emblems and deities, to create ornate panels often cleverly combin-
ing text and images. In one example a winged Victory holds up a wreath in the centre
of which is inscribed the unit title of the Twentieth Legion (Keppie 1998: 87–​89; RIB
220). Similarly epitaphs were often accompanied by images, most often placed above
the epigraphic panel; sculpted portraits, for example, drew the eye to the tombstone and
worked with the text to build up a picture of the deceased. A few inscriptions also stood
out by toying with the standardized content, deviating from the norm by adding a poem
or extra personal details (for example, verse epitaphs: RIB 265, 292, 684, 758, 1253). In
two examples the inscriptions literally spoke, giving the dead a voice and addressing the
traveller. A soldier of the Fourteenth Legion in Wroxeter describes his life in the under-
world (RIB 292) advising the reader to ‘live honestly while your star grants you time for
life’; while a Greek epitaph (RIB 758) from Brough-​under-​Stainmore, encourages the
traveller to call out the name of 16-​year-​old Hermes, for in doing so ‘you will do him a
good service’.
294

294   Valerie M. Hope

In many ways inscriptions played with the senses of the passers-​by, especially the
experiences of sight, hearing, and touch. Inscriptions (and associated sculpture, décor,
and buildings) were clearly visual; also we can imagine inscriptions being read aloud
and some being touched, as fingers traced letter forms, or when buildings, altars, and
graves were visited (especially in ritual contexts). Events (for example, feasting and sac-
rifices) associated with the places and structures that inscriptions marked may have
drawn on the other senses of taste and smell. All this suggests that inscriptions could
both be unusual (in Roman Britain in particular) and special, while simultaneously
being an integrated and unremarkable aspect of the lived environment. In the urban,
military, religious, and cemetery situations in which they were mainly found, inscrip-
tions were part of the landscape, entwined with the structures and patterns of daily life,
potentially long after those identified in them were forgotten. Setting up inscriptions
may have marked those structures, and the people named, as significant (in those set-
tings at least) and made a claim on posterity, yet the human capacity to remember can
be a fickle thing. Inscriptions could have longevity, but some also had a limited life span
(Barrett 1993). In Chester, it has been argued that tombstones of the Twentieth Legion
may have been deliberately defaced and reused in the North Wall of the fort (Clay 2004);
the distance slabs of the Antonine Wall may have been buried when the army drew back
(Keppie 1998: 52); inscribed votive altars found buried at Maryport, once thought to have
been concealed as part of religious ritual, probably made for convenient ballast for sup-
ports for a new building (Hill 1997; Hilts 2011). Individual inscriptions did not change
with the times, since they were rarely re-​cut or updated. Some inscriptions stayed in
place, continuing to speak when the inherent message ceased to have current relevancy,
while many other inscriptions were forgotten, defaced, prematurely silenced, reused, or
buried. It is difficult to reconstruct how people read and reacted to inscriptions, in their
original setting and across their lifespan, but, whether feted or ignored, inscriptions
were about communicating, and they now communicate to modern audiences.
Inscriptions are, to modern commentators, highly valuable as sources of informa-
tion on, for example, the use of Latin language, or demography, or religious practice.
Isolating certain elements within inscriptions, at the expense of others, however, can be
problematic and risks looking at words without the inscription, the inscription without
any associated images, the images without the monument, and the monument without
its wider physical and locational context. It is inevitable that we cannot access all these
dimensions, but we can prioritize the knowledge that inscriptions were a means, often
a multifaceted means, of telling a story, albeit a select and carefully edited story. In their
original context, inscriptions spoke to social groups to whom the dedicator or inscriber
was claiming membership, as well as to others (outsiders) who saw the monument and
reacted in their own individual way. There were, sometimes, other communicative
dimensions (even if apparently one-​sided) to these inscriptions, such as spiritual mes-
sages to the gods or the claim on a relationship with a higher authority, such as a patron
or the emperor. Inscriptions were clearly connected to memory—​put simply, they were
set up to commemorate, to record, and to preserve people’s names, and the lives and
actions of those people. Inscribed monuments memorialized names and acts for the
 Inscriptions and Identity    295

future, by employing a repertoire of established words, phrases, and images; they drew
on past practice, they looked to the future, and reflected (or constructed) their present.
Inscriptions, and the monuments with which they were associated, could give a sense of
rootedness, of continuity; they placed individual acts in a wider communal and societal
context (for communicative dimensions, see, in particular, Meyer 2011).

Identity

Latin inscriptions are centred on people: most often, named individuals and some-
times groups; for Roman Britain, military units and subunits are often recorded as act-
ing together. It is these names that bring inscriptions to life, especially the sense that
inscriptions often record the acts of relatively ordinary individuals (or groups of peo-
ple). Inscriptions do not state only the names of Roman emperors, governors, military
commanders, and the elite of Roman Britain, but also rank-​and-​file soldiers, children,
women, and even slaves. Latin names, as recorded in inscriptions, can be very revealing,
since they indicate gender and can provide clues about legal status and place of origin.
For example, Quintus Valerius Fronto, of the Second Legion, commemorated at Chester
(Q(uintus) VALERI/​VS Q(uinti) F(ilius) CLA(udia tribu) /​FRONTO CELE/​A MILES
LEG /​II AD(iutricis) P(iae) F(idelis) AN/​NORVM L /​STIPENDIORV/​M XXV … (RIB
479)), is revealed as a Roman citizen by his three names (tria nomina), as free-​born,
through his father’s praenomen (‘son of Quintus’) and his voting tribe (‘Claudian’), and
as originating from Celeia (in Noricum, now in Slovenia). The flipside of such a wealth
of information are those names that say frustratingly little. The single name of Vacia,
commemorated at Carlisle (DIS (Manibus) /​VACIA INF/​ANS AN(norum) III (RIB
961))], reveals only her gender, and not explicitly her legal status, place of origin, or par-
entage, but the accompanying portrait, and her description as ‘an infant of three years’,
powerfully evokes the sentiments caused by her death.
Because inscriptions are centred on people, and their names, they present and fashion
images of the self. Any commissioner, as noted above, could not control the audience
response to, and the ultimate fate of, an inscribed monument, but he or she did make
considered choices about what the inscription should say and how individuals and
groups should be represented. This active decision-​making is connected to why inscrip-
tions had relevance as a means of communication to some individuals, and groups, more
than to others. Romano-​British inscriptions do not include a random cross-​section of
people, as tempting as it might be to wish for this. Time, geography, and personal expe-
rience all played their part in where, when, and by whom inscriptions were set up. Those
who used inscriptions often had something particular to say, or present, and sometimes
an axe to grind, or a difference to announce. There were reasons why some people chose
to inscribe and equally why others did not.
Romano-​British inscriptions were largely military, and thus male, in character. The
army was a substantial presence, but still only ‘a minority group’ in terms of the overall
296

296   Valerie M. Hope

population of Roman Britain (Mattingly 2004: 15). Soldiers were the main inscribers,
or commissioners of inscriptions, and thus these inscriptions acted as a clear symbol of
military identity (Mattingly 2008: 68) and also military community. Inscriptions, and
the monuments associated with them, were aspects of the backdrop, routine, and con-
formity of military life, which helped forge a shared identity and commonality for sol-
diers who may have originated from across the empire (MacMullen 1984; James 1999;
Kampen 2006: 132). Inscriptions were a significant feature of how soldiers communi-
cated with each other, and also with others (or represented themselves to others) such
as enemies, civilians, authority figures (for example, the emperor), and even the gods.
We should not forget, however, that different types of inscription had different roles. In
the military context, milestones, centurial stones, building dedications, and some votive
offerings often had an official or semi-​official purpose, such as recording the completion
of designated acts and allegiance to the military unit and/​or the emperor. The content
of these inscriptions might vary little, but their impact, especially when taken together,
was one of organization, power, and control—​a symbol of both the Roman presence
and military authority. These inscriptions were part of how Rome was written onto the
British landscape. For ordinary soldiers, however, these same inscriptions were per-
haps less a symbol of power and more one of common acts, common allegiances, and a
shared military identity.
Inscriptions were not just unchanging signs of the Roman and military presence, since
many, especially epitaphs and votive offerings, incorporated more personal dimensions,
and, in addition, the role and impact of inscriptions also shifted with time. Inscriptions
could celebrate military identity, military unity, differences between soldiers and civil-
ians, and also differences between military units (see below), but the relevancy of some
of these messages would have waxed and waned as the army marched on, or established
permanent bases, or sent squadrons to serve elsewhere in Britain and the empire. In the
early years of the province, for example, building dedications within a fort, and the epi-
taphs and tombstones that ringed that fort, may have symbolized military unity to the
soldiers, their ‘Roman’ identity, and a territorial claim on the new province; while to the
recently subjugated population they may have been all these things with a more sinister
twist. As the years passed, however, and units relocated or became more permanently
stationed, or were even locally recruited, the messages inherent in, and to some extent
the relevance of, these inscriptions (both old and new) would have changed.
Identity could be highly personal and individualized, and to suggest that inscrip-
tions primarily promoted some sort of overarching military identity can overlook this.
Inscriptions often suggest diversity and difference and underline the complex and
interwoven aspects of individual identity in the Roman world (Hope 2000a; Mattingly
2004: 10–​11; 2006: 18). Military epitaphs and tombstones, for example, might suggest the
unity and conformity of the army, but many of these memorials marked the graves of
men of rank, or of men serving away from their home base (Hope 2003; see also Russell
2010). These people, and those who commemorated them, did not want the things that
made them different in the environments where they were commemorated, to be for-
gotten (Hope 2003: 135). Similarly, the most striking Romano-​British tombstones, those
 Inscriptions and Identity    297

decorated with portraiture, were promoting not just an un-​problematic Roman identity,
but often a sense of difference from the majority culture—​a sense, among both those
commemorating and commemorated, of being, in some respects, an outsider. These
decorated tombstones were often set up to soldiers in newly conquered areas or to aux-
iliary troops who were not Roman citizens and to women who were not fully enfran-
chised members of the communities in which they lived (Hope 1997; Stewart 2009).
The complexities of identity are well illustrated by auxiliary tombstones decorated with
detailed cavalry scenes, which were popular mainly during the first century. A recently
discovered tombstone from Lancaster commemorating Insus, son of Vodullus, a citizen
of the Treveri of the ala Augusta, depicts him in a splendid helmet, brandishing, by the
hair, a decapitated head, the corpse lying at the feet of his rearing mount (Hassall and
Tomlin 2006: 468–​471). On this and similar monuments, horsemen ride down fallen
barbarians, representing, it would seem, Roman victory and supremacy, yet the epi-
taphs reveal that those commemorated, including Insus, were members of non-​citizen
auxiliary units, who might be paid less well than their legionary counterparts, might be
regarded as more expendable in battle, and were often recruited from recently subju-
gated provinces (Hope 1997, 2000b; Stewart 2010). In addition to these funerary monu-
ments, we can also note that, in the north of the province, votive altars, including those
set up to Jupiter, were more often the product of auxiliary soldiers than of legionaries—​
which again has been related to their non-​citizen identity (Vanderspoel 2005: 46–​52).
There is an irony here that, on the surface, many Latin inscriptions present and symbol-
ize Roman (and military) identity, yet, if we dig a little deeper, those associated with
these inscriptions were sometimes individuals for whom this identity could be con-
tested. This is not to claim that all the inscriptions of Roman Britain were aberrations,
generated only by those with chips on their shoulders, but that identity could be com-
plex and multifaceted; in the Roman world one could both belong and not belong, and
the act of inscribing (with associated monuments and sculpture) could be a method of
presenting, editing, and negotiating that identity, as witnessed elsewhere in the Roman
world for groups such as freed slaves and gladiators (see, e.g., Kleiner 1977; Kockel 1993;
Hope 1998, 2001; Mouritsen 2005).
To end, we can note a group of Romano-​British inscriptions that stand out because
of a notable deviation to the norm. It was observed above that language united the
inscriptions—​the Latin in which they were composed was integral to the iden-
tity that was projected. This is not, however, strictly true, since there is a hand-
ful of Romano-​British inscriptions cut in Greek (RIB 461, 758, 808, 1072, 1124, 1129,
3151). The choice to inscribe in Greek was a considered one, suggesting the signifi-
cance of Greek culture and language to a few. Finding someone to inscribe in Greek
may have been challenging and emphasizes a choice ‘to make a strong statement of
ethnic and cultural identity’ (Noy 2010: 22). Others present in Roman Britain who
originated from Greece chose Latin not Greek for their inscriptions (RIB 251, 864,
955). Where Greek is used, it seems to be loosely connected to medicine (Noy 2000:
173) and trade—​once more a choice to emphasize difference and individuality. The
poem to young Hermes (as noted in ‘Communicating to an audience’ —​RIB 758) was
298

298   Valerie M. Hope

found at Brough-​under-​Stainmore, which may have served as a trade depot (Birley

1979: 111); two votive altars from Chester (RIB 461, 3151) were both set up by doc-
tors; a dedication at Maryport to Asclepius (god of healing) may also hint at medi-
cal connections for its dedicator Aulus Egnatius Pastor (RIB 808); and the same god
was honoured at Lanchester in a bilingual inscription (RIB 1072) set up by a soldier.
Perhaps the ultimate symbol of the complexities of identity, however, is the tomb-
stone of Regina, which proclaims her connections to Britain, Rome, and Palmyra
(RIB 1065). Regina may never have visited the latter two places, but had moved from
south-​east England (she is described as a Catuvellaunian) to South Shields (where
the tombstone was found), and the Latin of her epitaph, coupled with her Palymrene
husband (Barates), links her to the wider empire. The accompanying portrait, perhaps
carved by a Palmyrene hand (Phillips 1976), presents Regina, an ex-​slave, as a ‘Roman’
lady, although sporting apparently native British clothing (Noy 2010; Carroll 2012).
To her husband went the last words, cut in Palmyrene Aramaic: ‘Regina, freedwoman
of Barates, Alas!’ As with the Greek inscriptions, we are left to wonder who carved the
Palmyrene text. Was it Barates himself? Would anyone else in South Shields have been
able to understand it? Did that matter—​after all everyone would be able to tell that
there was something different about the tombstone, and about Regina and Barates,
without having to grasp all the details (Mullen 2012: 1–​5)? Regina’s short life, even if
confined to the shores of Britain, had been a diverse journey and experience, and at
the last that life was summed up and viewed through the prism of her commemora-
tor’s equally diverse experiences (for a discussion of multiculturalism, see Nesbitt, this
volume; see also Leach et al. 2010; Tolia-​Kelly 2010).
These Greek inscriptions, and Regina’s bilingual inscription, toy with identity, both
reflecting and creating it; they celebrate a sense of belonging and a sense of difference;
the inscriptions capture cultural diversity and mixed identities that characterized at
least some people’s experiences of living in the Roman Empire. Setting up inscriptions
in Roman Britain may never have been popular, or the norm, but, for a few, it was a pow-
erful medium for communicating, for recording important actions, and for telling sto-
ries. The inscriptions on stone that survive from Roman Britain are so low in numbers,
and so focused on certain areas and certain social groups, that it can be hard to general-
ize from them; nevertheless they do suggest aspects of varied identities and how these
were represented, negotiated, and constructed. The inscriptions suggest a sense of what
it was to be ‘Roman’ (and often a member of the army) in Britain, and simultaneously
the fluidity of identities in an outpost of empire.

Abbreviations
RIB   R. G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain:
Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date; Oxford: Oxbow
Books, 2009.
 Inscriptions and Identity    299

References
Primary Sources
Petronius. The Satyricon, trans. J. P. Sullivan. London: Penguin Classics, 2011.

Secondary Sources
Adams, J. N. (2007). The Regional Diversification of Latin 200 bc–​ad 600. Cambridge: Cambridge
University Press.
Alföldy, G. (1987). ‘Das Heer in der sozial Struktur des römischen Kaiserreiches’, in G. Alföldy
(ed.), Römische Heersgeschichte: Beitrage 1962–​1985. Amsterdam: J. C. Gieben, 33–​34.
Barrett, J. (1993). ‘Chronologies of Remembrance:  The Interpretation of Some Roman
Inscriptions’, World Archaeology, 25/​2: 236–​247.
Birley, A. (1979). The People of Roman Britian. London: B. T. Batsford.
Biró, M. (1975). ‘The Inscriptions of Roman Britain’, Acta Archaeoloica Academiae Scientiarum
Hungaricae, 27: 13–​58.
Blagg, T. F.  C. (1990). ‘Architectural Munificence in Roman Britain:  The Evidence of the
Inscriptions’, Britannia, 21: 13–​31.
Bodel, J. (2001) (eds). Epigraphic Evidence:  Ancient History from Inscriptions. London and
New York: Routledge.
Carroll, M. (2012). ‘“The Insignia of Women”:  Dress, Gender and Identity on the Roman
Funerary Monument of Regina from Arbeia’, Archaeological Journal, 169: 281–​311.
Cepas, A. (1989). The North of Britannia and the North-​West of Hispania:  An Epigraphic
Comparison. BAR International Series 470. Oxford: Archaeopress.
Cherry, D. (1998). Frontier and Society in Roman North Africa. Oxford: Clarendon Press.
Clay, C. (2004). ‘Iconoclasm in Roman Chester: The Significance of the Mutilated Tombstones
from the North Wall’, Journal of the British Archaeological Association, 157: 1–​16.
Cooley, A. (2012). The Cambridge Manual of Latin Epigraphy. Cambridge:  Cambridge
University Press.
Decker, K., and Selzer, W. (1976). ‘Mogontiacum: Mainz von Zeit des Augustus bis Ende der römis-
chen Herrschaft’, Aufstieg und Niedergang der Römischen Welt II.5. Berlin: de Gruyter, 457–​559.
de la Bédoyère, G. (1999). Companion to Roman Britain. Stroud: Tempus.
Derks, T. (1998). Gods, Temples and Ritual Practices: The Transformation of Religious Ideas and
Values in Roman Gaul. Amsterdam: Amsterdam University Press.
Dobson, B., and Mann, J. C. (1972). ‘The Roman Army in Britain and Britons in the Roman
Army’, Britannia, 4: 191–​205.
Eckardt, H., Brewer, P., Hay, S., and Poppy, S. (2009). ‘Roman Barrows and their Landscape
Context: A GIS Case Study at Bartlow, Cambridgeshire’, Britannia, 40: 65–​98.
Esmonde Cleary, S. (2000). ‘Putting the Dead in their Place: Burial Location in Roman Britain’,
in J. Pearce, M. Millet, and M. Strück (eds), Burial, Society and Context in the Roman World.
Oxford: Oxbow Books, 127–​142.
Forbis, E. (1996). Municipal Virtues in the Roman Empire: The Evidence of Roman Honorary
Inscriptions. Stuttgart and Leipzig: Teubner.
Grasby, R. D., and Tomlin R. S.  O. (2002). ‘The Sepulchral Monument of the Procurator
C. Julius Classicianus’, Britannia, 33: 43–​76.
300

300   Valerie M. Hope

Handley, M. (2001). ‘The Origins of Christian Commemoration in Late Antique Britain’, Early
Medieval Europe, 10/​2: 177–​199.
Hassall, M. W.  C., and Tomlin, R. S.  O. (1989). ‘Roman Britain in 1988, II:  Inscriptions’,
Britannia, 20: 326–​345.
Hassall, M. W.  C., and Tomlin, R. S.  O. (2006). ‘Roman Britain in 2005, III:  Inscriptions’,
Britannia, 37: 467–​488.
Häussler, R. (1998). ‘Resta, viator et lege: Thoughts on the Epigraphic Habit’, Papers from the
Institute of Archaeology, 9: 31–​56.
Hayward, K. M. J. (2006). ‘A Geological Link between the Facilis Monument at Colchester
and the First Century Army Tombstones from the Rhineland Frontier’, Britannia,
37: 359–​363.
Hayward, K. M.  J. (2009). Roman Quarrying and Stone Supply on the Periphery—​Southern
England:  A  Geological Study of First-​ Century Funerary Monuments and Monumental
Architecture. BAR British Series 500. Oxford: Archaeopress.
Hill, P. R. (1997). ‘The Maryport Altars:  Some First Thoughts’, in R. J.  A. Wilson (ed.),
Roman Maryport and its Setting: Essays in Memory of Michael G. Jarrett. Cumberland and
Westmorland Antiquarian and Archaeological Society, 92–​104.
Hilts, C. (2011). ‘Jupiter, Best and Greatest—​ Revisiting Maryport’s Ritual Pits’. Current
Archaeology, 259 <http://​www.archaeology.co.uk/​articles/​features/​revisiting-​maryports-​
ritual-​pits.htm> (accessed 24 September 2013).
Holder, N. (2007). ‘Mapping the Roman Inscriptions of London’, Britannia, 38: 13–​34.
Hope, V. M. (1997). ‘Words and Pictures: The Interpretation of Romano-​British Tombstones’,
Britannia, 28: 245–​258.
Hope, V. M. (1998). ‘Negotiating Identity and Status:  The Gladiators of Roman Nîmes’,
in J. Berry and R. Laurence (eds), Cultural Identity in the Roman Empire. London and
New York: Routledge, 179–​195.
Hope, V. M. (2000a). ‘Status and Identity in the Roman World’, in J. Huskinson (ed.),
Experiencing Rome:  Culture, Identity and Power in the Roman Empire. London and
New York: Routledge, 125–​152.
Hope, V. M. (2000b). ‘Inscription and Sculpture:  The Construction of Identity in the
Military Tombstones of Roman Mainz’, in G. Oliver (ed.), The Epigraphy of Death.
Liverpool: Liverpool University Press, 155–​186.
Hope, V. M. (2001). Constructing Identity: The Roman Funerary Monuments of Aquileia, Mainz
and Nimes. BAR International Series 960. Oxford: Archaeopress.
Hope, V. M. (2003). ‘Remembering Rome: Memory, Funerary Monuments and the Roman
Soldier’, in H. Williams (ed.), Archaeologies of Remembrance: Death and Memory in Past
Societies. New York: Kluwer/​Plenum, 113–​140.
Hope, V. M. (2007). ‘Age and the Roman Army: The Evidence of Tombstones’, in M. Harlow
and R. Laurence (eds), Age and Ageing in the Roman Empire. JRA Supplementary Series 65.
Portsmouth, RI: Journal of Roman Archaeology, 111–​130.
Hopkins, K. (1966). ‘On the Probable Age Structure of the Roman Population’, Population
Studies, 20: 245–​264.
Isserlin, R. M. J. (1998). ‘A Spirit of Improvement: Marble and the Culture of Roman Britain’,
in R. Laurence and J. Berry (eds), Cultural Identity in the Roman Empire. London and
New York: Routledge, 125–​178.
James, S. (1999). ‘The Community of Soldiers: A Major Identity and Centre of Power in the
Roman Empire’, in P. Barker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC 98: Proceedings
 Inscriptions and Identity    301

of the Theoretical Roman Archaeology Conference, Leicester 1998. Oxford:  Oxbow

Books, 14–​25.
Jones, B., and Mattingly D. (1990). An Atlas of Roman Britain. Oxford: Blackwells.
Kampen, N. (2006). ‘The Art of Soldiers on a Roman Frontier: Style and the Antonine Wall’, in
E. D’Ambra and G. Métraux (eds), The Art of Citizens, Soldiers and Freedmen in the Roman
World. BAR International Series 1526. Oxford: Archaeopress, 125–​134.
Keppie, L. (1998). Roman Inscribed and Sculptured Stones in the Hunterian Museum, University
of Glasgow. Britannia Monograph Series 13. London: Britannia.
Kleiner, D. E E. (1977). Roman Group Portraiture: The Funerary Reliefs of the Late Republic and
Early Empire. New York: Garland.
Knight, J. K. (2010). ‘An Inscription from Bavai and the Fifth Century Christian Epigraphy of
Britain’, Britannia, 41: 283–​292.
Kockel, V. (1993). Porträtreliefs Stadtrömischer Grabbauten. Mainz: Zabern.
Leach, S., Eckardt, H., Chenery, C., Müldner, G., and Lewis, M. (2010). ‘A Lady of
York: Migration, Ethnicity and Identity in Roman Britain’, Antiquity, 84: 131–​145.
MacMullen, R. (1982). ‘The Epigraphic Habit’, American Journal of Philology, 103: 234–​246.
MacMullen, R. (1984). ‘The Legion as a Society’, Historia, 33: 440–​456.
Mann, J. (1985). ‘Epigraphic Consciousness’, Journal of Roman Studies, 75: 204–​206
Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of
Roman Archaeology, 17: 5–​25.
Mattingly, D. (2006). Imperialism, Power and Identity:  Experiencing the Roman Empire.
Princeton: Princeton University Press.
Mattingly, D. (2008). ‘Urbanism, Epigraphy and Identity in the Towns of Britain under
Roman Rule’, in V. E. Hirschmann, A. Krieckhaus, and H. M. Schellenberg (eds),
A Roman Miscellenay:  Essays in Honour of R.  Birley on his Seventieth Birthday.
Gdańsk: Akanthina, 53–​7 1.
Meyer, E. (1990). ‘Explaining the Epigraphic Habit in the Roman Empire: The Evidence of
Epitaphs’, Journal of Roman Studies, 80: 74–​96.
Meyer, E. (2011). ‘Epigraphy and Communication’, in M. Peachin (ed.) The Oxford Handbook of
Social Relations in the Roman World. Oxford: Oxford University Press, 191–​226.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Mouritsen, H. (2005). ‘Freedmen and Decurions:  Epitaphs and Social History in Imperial
Italy’, Journal of Roman Studies, 95: 38–​63.
Mrozek, S. (1973). ‘A propos de la repartition chronologique des inscriptions Latines dans le
haut-​Empire’, Epigraphica, 35: 113–​118.
Mullen, A. (2012). ‘ Introduction: Multiple Languages, Multiple Identities’, in A. Mullen and
P. James (eds), Multilingualism in the Graeco-​ Roman Worlds. Cambridge:  Cambridge
University Press, 1–​35.
Noy, D. (2000). Foreigners at Rome: Citizens and Strangers. London: Duckworth.
Noy, D. (2010). ‘Epigraphic Evidence for Immigrants at Rome and in Roman Britain’, in H.
Eckardt (ed.), Roman Diasporas:  Archaeological Approaches to Mobility and Diversity
in the Roman Empire. JRA Supplementary Series 78. Portsmouth, RI:  Journal of Roman
Archaeology, 13–​26.
Parsons, D. N. (2011). ‘Sabrina in the Thorns: Place Names as Evidence for British and Latin in
Roman Britain’, Transactions of the Philological Society, 109/​2: 113–​137.
Pearce, J. (2004). ‘Archaeology, Writing Tablets and Literacy in Roman Britain’, Gallia, 61: 43–​51.
302

302   Valerie M. Hope

Pearce, J. (2011). ‘Marking the Dead: Tombs and Topography in the Roman Provinces’, in M.
Carroll and J. Rempel (eds), Living through the Dead: Burial and Commemoration in the
Classical World. Oxford: Oxbow Books, 134–​158.
Phang, S. (2001). The Marriage of Roman Soldiers (13 bc–​ad 235). Leiden: Brill.
Phillips, E. J. (1976). ‘A Workshop of Roman Sculptors at Carlisle’, Britannia, 7: 101–​108.
Raybould, M. (1999). A Study of the Inscribed Material from Roman Britain. BAR British Series
281. Oxford: Archaeopress.
Revell, L. (2007). ‘Architecture, Power and Politics: The Forum-​Basilica in Roman Britain’, in J.
Sofaer (ed.), Material Identities. Oxford: Blackwell, 127–​151.
Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge
University Press.
Richmond, I., and Wright, R. (1956). Catalogue of Roman Inscribed and Sculptured Stones in the
Grosvenor Museum Chester. Chester: Chester and North Wales Archaeological Society.
Russell, B. (2010). ‘Sarcophagi in Roman Britain’, Bollettino di Archeologia on line. Volume
Speciale:  XVII International Congress of Classical Archaeology, Roma 22–​26 September
2008  <http://​151.12.58.75/​archeologia/​bao_​document/​articoli/​2_​RUSSELL.pdf> (accessed
26 June 2012).
Saller, R. (1994). Patriarchy, Property and Death in the Roman Family. Cambridge: Cambridge
University Press.
Saller, R., and Shaw, B. (1984). ‘Tombstones and Roman Family Relations in the
Principate: Civilians, Soldiers and Slaves’, Journal of Roman Studies, 74: 124–​156.
Scheidel, W. (2012). ‘Epigraphy and Demography:  Birth, Marriage, Family and Death’, in J.
Davies and J. Wilkes (eds), Epigraphy and the Historical Sciences. Proceedings of the British
Academy, 177. Oxford: Oxford University Press, 101–​29.
Stewart, P. (2009). ‘Totenmahl Reliefs in the Northern Provinces: A Case-​Study in Imperial
Sculpture’, Journal of Roman Archaeology, 22: 253–​274.
Stewart, P. (2010). ‘Geographies of Provincialism in Roman Sculpture’, Journal of the
International Association of Research Institutes in the History of Art 0005 (27/​7/​2010) <http://​
www.riha-​journal.org/​articles/​2010/​stewart-​geographies-​of-​provincialism> (accessed 26
June 2012).
Strück, M. (2000). ‘High Status Burials in Roman Britain (First–​Third Centuries ad)—​
Potential of Interpretation’, in J. Pearce, M. Millet, and M. Strück (eds), Burial, Society and
Context in the Roman World. Oxford: Oxbow Books, 85–​96.
Tobler, G. W., and Adams R. (2007). Romano-​British Tombstones between the First and
Third Centuries:  Epigraphy, Gender and Familial Relations. BAR British Series 437.
Oxford: Archaeopress.
Tolia-​Kelly, D. P. (2010). ‘Narrating the Postcolonial Landscape:  Archaeologies of Race at
Hadrian’s Wall’, Transactions of the Institute of British Geographers, 36: 71–​88.
Tomlin, R. S. O. (2011). ‘Writing and Communication’, in L. Allason-​Jones (ed.), Artefacts in
Roman Britain: Their Purpose and Use. Cambridge: Cambridge University Press, 133–​152.
Vanderspoel, J. (2005). ‘Male Deities in Northern Roman Britain: Patterns and Perspectives’,
Religious Studies and Theology, 24/​1: 33–​59.
Woolf, G. (1996). ‘Monumental Writing and the Expansion of Roman Society in the Early
Roman Empire’, Journal of Roman Studies, 86: 22–​39.
Woolf, G. (1998). Becoming Roman:  The Origins of Provincial Civilization in Gaul.
Cambridge: Cambridge University Press.
 Chapter 15

Ideas of Ch i l dh o od
in Roman Bri ta i n
The Bioarchaeological and Material Evidence

Rebecca Gowland

Introduction

Once overlooked, ancient children are now subjects of overviews. But—​and

here we are less fortunate—​it is some 2,000 years too late to learn very much
about them.
(Golden 2011: 262)

Childhood in the past is a burgeoning field of study and has received a great deal of
attention from scholars of the ancient world, including Golden (2011: 262) quoted at the
head of this chapter. As a result of a focus on childhood within the discipline of history
(e.g. Ariès 1962; Stone 1977; Pollock 1983), it is now widely recognized that perceptions
of childhood do not subscribe to a universal reality, but instead are contingent upon his-
torical and cultural context. The seminal book by Ariès, Centuries of Childhood (1962),
led to a shift in the concept of the family from that of a fixed biological entity to a histori-
cally contingent construct (James et al. 1998).
Within archaeology, feminist and gender discourse, together with the influence of
studies of childhood within history, have acted as a stimulus to debates concerning chil-
dren in narratives of the past. Initial studies of childhood within archaeology were con-
cerned with children’s ‘invisibility’ in archaeological discourse. The neglect of children,
it was argued, stemmed from the same anthrocentric biases that had previously served
to marginalize women (Lillehammer 1989; Baker 1997):

children have been both absent/​invisible from the archaeological record, and invisi-
ble, unknowable at the conceptual level. Contemporary culturally constructed social
304

304   Rebecca Gowland

knowledge, embedded as it is in masculist ideologies, fits ‘children’ in the interpreta-

tive framework as incomplete humans, that is, not male/​masculine. (Baker 1997: 187)

Since this important initial research, studies of childhood have developed towards a
consideration of the culturally specific constructions of childhood within different time
periods and places (e.g. Sofaer Derevenski 1994; Moore and Scott 1997; Gowland 2001;
Halcrow and Tayles 2008). A useful way of conceptualizing childhood was described
within the discipline as follows:

Childhood provides an interpretive frame for contextualising the early years of

human life. Childhood, as distinct from biological immaturity, is neither a natural
or universal feature of human groups but appears a specific structural and cultural
component of many societies. (Prout and James 1997: 8)

Furthermore, studies of the past started to recognize that children are active agents
within society, rather than passive beings imprinted by the socializing forces of adults
(Prout and James 1997). The importance of children as a focus of study in their own right
is now acknowledged by numerous scholars and is reflected by the development of the
journal Childhood in the Past.
Studies have focused on childhood not as a separate or distinct entity, but a construct
that must be analysed and understood within the chronology of the life course as a
whole. Age is now conceptualized as a central aspect of social identity and a highly sig-
nificant structuring element within society. A life-​course perspective has been adopted
by most social scientists in their study of age identity. The life-​course approach differs
from previous approaches to age identity in that it considers the fluidity of identity
throughout the life of an individual from birth to death (see A. Moore, this volume).
This chapter provides an overview of these debates in relation to the archaeological evi-
dence for ideas of childhood in Roman Britain. In particular, this chapter will focus on
the bioarchaeological and funerary evidence. Before we discuss the various strands of
archaeological data, it is worth first outlining what classical sources had to say about
childhood in the Roman Empire in order to provide some context.

Childhood in the Roman

World: Classical Perspectives

Since the 1990s there has been a considerable amount of research on childhood and
the family in the Roman world, culminating in numerous books and edited volumes
(e.g. Dixon 1992; Rawson and Weaver 1997; Harlow and Laurence 2002; Rawson 2003,
2011; Dasen and Späth 2010). This research is almost exclusively derived from a histor-
ical or classical perspective, drawing upon literary sources, medical texts (especially
Soranus’ Gynaecology), as well as epigraphic and monumental evidence (e.g. Laurence
 Ideas of Childhood in Roman Britain    305

2000; Harlow and Laurence 2002). This work has been important for understanding
Roman perceptions of children. The aforementioned studies have demonstrated that
childhood was recognized as a distinctive stage of the life course in the Roman world,
and historical texts from this period allude to the charm of childish characteristics
(Dixon 1992). Iconographic evidence likewise depicts children learning and at play—​
activities that we would also regard as particularly childlike. The demarcation of this
developmental period into a series of distinct stages is also evident from historical
documentation. The term infantia, which literally means ‘not speaking’, was ascribed
to children until the age of 7 years. From then, terminology differentiated between the
sexes; puer or puella were employed, and males and females experienced divergent
life-​course trajectories in terms of social age transitions. Virginity is, for the first time,
attributed as a characteristic after the age of 7 years (Fraschetti 1997). Roman males
underwent a significant rite of passage at approximately 14–​16 years of age, during a
ceremony that took place in both public and private; they replaced their toga praetexta
with the toga virilis and removed their bulla (Eyben 1993: 6; Fraschetti 1997: 64). This
event signified a new social age for males, who were then considered ‘more respon-
sible’ individuals as adolescens, until approximately 25–​30 years of age (Weidemann
1989: 116).
Females have no similar rite of passage; the onset of menarche does not appear to
have been socially significant, and only upon marriage did they experience a change
in status (Fraschetti 1997:  63). Epigraphic and documentary evidence indicates that
high-​status females in Rome may have married as early as 12 years of age—​puberty
apparently not being a prerequisite for marriage (Hopkins 1965). However, in a re-​
examination of the epigraphic evidence from Rome and the wider empire, Shaw (1987)
convincingly demonstrated that, for the majority of Roman women, marriage tended
to occur from the late teens to early twenties and for males in the mid-​twenties. Shaw
(1987: 33) argues that those very few historical sources that discuss the age of marriage
for women tend to refer only to the ‘narrowest of elites’ and are thus largely irrelevant
for most of the population.
The great majority of historical sources pertaining to Roman childhood are biased
towards Rome and Italy. As a society’s life course is culturally constructed, it is subject
to temporal and spatial differences, which have been observed in studies of funerary
data from the Roman Empire (e.g. Gowland 2001; Revell 2005). Therefore, the relevance
of historical sources from Rome for Roman Britain may be regarded as questionable.
In the absence of such a rich corpus of historical data from Britain, we must turn to
the archaeological evidence. As highlighted above, one particularly fruitful source of
information is the funerary context. This context is unique, because it provides a cru-
cial link between individuals and material culture (Gowland and Knüsel 2006) and thus
one can infer aspects of age-​related social identity through patterns of deposition and
ritual treatment. When interpreting funerary evidence, one must be cautious not to be
overly simplistic; we are after all dealing with the treatment of the dead, not the living,
and glimpsing past identities through the distorting lens of ritual practice. Nevertheless,
the funerary context provides a key reservoir of data for understanding perceptions of
306

306   Rebecca Gowland

childhood in Roman Britain. It is not only the mode of burial that is of significance, but
also the skeletal remains. The bioarchaeological evidence has been an under-​exploited
resource for examining perceptions of childhood in Roman Britain, though research
over the last ten years has been highlighting its pivotal importance.

The Bioarchaeology
of Roman Childhood

The bioarchaeological analysis of children provides crucial direct evidence for their
well-​being in relation to their social and physical milieu. The potential for this biological
evidence to yield significant social information about the Roman family is immense and
yet it is under-​utilized. This is a consequence of sub-​and interdisciplinary boundaries
and a lack of communication across these—​a situation that is thankfully now changing
(see Redfern and Gowland 2012 for a discussion). The conceptualization of childhood as
a social construction may also have contributed to the marginalization of their physical
remains. Within such a schema, the body as a physical entity tends to be perceived as
irrelevant (Shilling 1993). However, while societies may construct their own perceptions
of the life course, one cannot overlook the corporeal aspects of childhood: the physical
and emotional changes that accompany growth and the attainment of bodily matura-
tion. An infant, for example, is helplessly dependent on adults for care and has basic
needs that must be met if he or she is to survive. As Franks (1991) writes, our bodies are
‘an obdurate fact’ and as such should be included in narratives of childhood in the past.
James and colleagues (1998: 51) also discuss the fact that ‘childhood is united by the uni-
versal biology of human physical development and cognitive potential but, in the same
moment, radically differentiated by the varied social contexts in which this growth can
be culturally enacted in the life course’. The interaction between the physical and the
social worlds in the forging of identities will be discussed later in the chapter. below.

Age and Sex
One of the key limitations of the skeletal analysis of non-​adults is the inability of osteo-
logical techniques to provide a reliable determination of biological sex. While numer-
ous studies of non-​adult remains have attempted to provide methodological sexing
criteria, unfortunately these have not performed sufficiently well during independent
tests. In the future it is possible that ancient DNA analysis will become more affordable
and accessible, in which case this problem may be circumvented. At present, however,
osteoarchaeologists rarely attempt to estimate the sex of non-​adults (<18 years). This
shortcoming provides some potential limitations with respect to the interpretation of
age and gender during this growth period.
 Ideas of Childhood in Roman Britain    307

In osteological analysis it is standard practice to translate the skeletal age of the indi-
vidual into a chronological age as a ‘universal’ (western) form of analysis. However, the
relationship between skeletal age and chronological age is problematic. Growth and
development are highly dependent on nutrition, environment, and genetics and may
thus vary considerably between populations. Social environment may also have pro-
found consequences for growth and maturation. For example, within modern con-
texts we observe the effect of socio-​economic environment on the age of menarche in
females, with higher social status often linked to a lower age of onset of menstruation
(Gowland 2006). A longitudinal study of modern British populations found that height
at age 7 years was a predictor of employment prospects in later life, because growth up
to this time was such a sensitive indicator of socio-​economic environment and psycho-​
social stressors (Blane 2006). This is relevant to studies of the Roman world, because
adult stature is used as an important indicator of well-​being in studies of the Roman
Empire (e.g. Kron 2005). Another example from the ancient world is highlighted by
Laurence (2000: 446), who discusses Galen’s view that puberty for Roman males began
at 14 years and ended at 25 years (somewhat later than the current western norm). If
real, this potentially has implications for the age estimation of Romano-​British skeletal
remains: delayed epiphyseal union (fusion of the bones after the cessation of growth) in
past populations may result in under-​ageing when using standards derived from mod-
ern populations. With regard to age-​at-​death, it is not accurate to consider skeletal age
in purely biological terms (Gowland 2006). During life, skeletons are comprised of liv-
ing tissue that responds to the social as well as physical environment in a dynamic way.
This potential variation in skeletal form should not necessarily be considered in purely
negative terms—​as a variable to be controlled for—​and instead has the potential to be
harnessed so that we might better understand the subtle ways in which the hard tissues
embody the Roman world.

Health and Care
Our ability to investigate the health and care of past children can be undertaken directly
on the skeletal remains of non-​adults or can be inferred from remnants of childhood
health stress retained in the adult skeleton (see Gowland and Redfern 2010). An exam-
ple of the latter is the presence of enamel defects in teeth. These relate to childhood peri-
ods of ‘health stress’ when the teeth are forming; because teeth do not remodel once
formed, this period of poor health is effectively ‘fossilized’ into adulthood. The study of
non-​adult skeletal remains has the ability to shed light not only on the care and health
of children, but also on broader social processes relating to adulthood and the family.
For example, Gowland and Redfern’s comparison (2010) of childhood stress indicators
from Roman London and Italy has shown how they may yield information concern-
ing living environment and population mobility. The growth period is a time during
which the skeleton is particularly susceptible to environmental onslaughts, and thus, as
Lewis (2007) has highlighted, children are sensitive ‘barometers’ of overall population
308

308   Rebecca Gowland

health. Research themes in relation to childhood in the Roman world include: demog-

raphy and infanticide, mobility, childcare and health, weaning and diet. It is not feasible
to review all of these here, and readers are referred to Redfern and Gowland (2012) for a
more thorough discussion. Here, though, I will outline a few bioarchaeological studies
that highlight the ways in which non-​adult skeletal remains and their treatment in death
may contribute to an understanding of childhood in Roman Britain.
A number of studies since 2000 have focused on the health and mortality of non-​
adult remains from Roman Britain and these have served to provide a growing body of
evidence concerning perceptions of children and childcare in the Roman world. A par-
ticular amount of attention has been focused on the health of children from the large
Romano-​British cemetery at Poundbury, Dorset (e.g. Molleson 1989, 1993; Redfern
2007; Lewis 2010; Redfern et  al. 2012). Poundbury is a unique cemetery in Roman
Britain in terms of its size and the large proportion of immature skeletal remains exca-
vated from the site, which include 364 ‘non-​adult’ individuals (<17 years of age), 75 of
whom were perinatal (dying around the time of birth) infants (Molleson 1989).
A distinctive feature of the infants at this site (aside from their being so many) was
that the bones of many show evidence of poor health related to dietary deficiency dis-
eases. Molleson (1989) has argued that the age of onset for this ill health was 3 months
of age, after which growth becomes stunted. She suggested that this was related to the
early introduction of weaning food and an exposure to dietary lead, possibly through
the use of pewter dishes. If these developmental problems were the consequence of the
introduction of weaning food, then weaning was occurring at an age prior to the advice
of contemporary medical writers from Rome (for example, Soranus), who counselled
against the introduction of foods prior to the second half of the first year (Prowse et al.
2008). This again brings into question the relevance of textual evidence from Rome for
childcare practices and perceptions in Roman Britain.
In a recent re-​evaluation of the Poundbury non-​adult skeletal remains, Mary Lewis
(2010) records that 31 per cent exhibited evidence of metabolic disease, and a number of
these showed very advanced forms of these diseases, which included rickets and scurvy
(vitamin D and C deficiency, respectively). Breast milk is a good source of vitamin C,
and the presence of scurvy in infants suggests complete cessation of breastfeeding,
unless of course the mother was also very deficient (Brickley and Ives 2008). Several bio-
archaeological studies of non-​adult remains from Roman Britain have focused on the
duration of breastfeeding using isotope analysis of nitrogen. Breastfed infants tend to
be enriched in the isotope 15Nitrogen when compared to their mother, and these values
then fall upon weaning (Jay 2009). Such studies are important because the age at which
food is introduced into a diet and breastfeeding ceases tends to be culturally subscribed
and has implications for health and mortality later in life. A study of the weaning age and
diet of the Poundbury infants and others from Iron Age and Roman Dorset was under-
taken by Redfern et al. (2012). They found that infant feeding practices changed from the
Iron Age to the Roman period, with the possible introduction of a special weaning diet
in the later period. These findings are consistent with other isotopic studies of breast-
feeding and weaning by Fuller et al. (2006) and Powell et al. (2014), which indicate that
 Ideas of Childhood in Roman Britain    309

the process of weaning was a gradual one and appears different from the isotope profile
obtained from the broadly contemporary site of Isola Sacra, near Rome. This therefore
indicates that, as one might expect, infant care did not subscribe to a standardized ideal
espoused by contemporary medical writers.
With regard to the presence of vitamin D deficiency at Poundbury, it should be noted
that approximately 90 per cent of vitamin D is synthesized in our own bodies on contact
with sunlight. Therefore, child-​rearing practices that involve swaddling and keeping the
children indoors are likely to lead to vitamin D deficiency—​particularly in Britain with
relatively few sunlight hours during winter months. In children, this leads to rickets,
a condition in which the bones are insufficiently mineralized. Infants with deficiency
diseases at Poundbury were recorded within the high-​status burials (within mausolea)
as well as the lower-​status graves (Lewis 2010). This reminds us that wealth does not
necessarily equate to health and certain high-​status child-​rearing practices (for exam-
ple, swaddling, keeping infants indoors) were likely to have been detrimental to health.
With respect to vitamin D deficiency in infancy, we must also consider the health status
of the mother. For example, high-​status cultural practices such as confinement indoors
during pregnancy are likely to lead to maternal vitamin D deficiency, which will affect
the developing foetus. At other sites, ‘high-​status’ burials of children have also been
excavated where the bones show signs of nutritional deficiencies. For example, the skel-
eton of a 7–​8-​year-​old girl excavated from the Eastern Cemetery of Roman London—​
buried in a lead coffin with grave goods—​exhibited signs of rickets.
Another piece of advice from Soranus was the withholding of colostrum. Colostrum
is the initial fluid produced by mothers, prior to the milk ‘coming in’ two to three days
after the birth of the infant. It is produced in very small quantities and is a yellowish,
thick liquid that contains important components for the newborn, including leucocytes
(white blood cells), antibodies, vitamin K, and protein. It has been suggested by Lewis
(2010) that the withholding of colostrum may have been a contributing factor towards
the deficiency diseases observed at Poundbury.
A final aspect of the skeletal pathologies of the children at Poundbury that warrants
special attention is the prevalence of what were initially identified as healed rib frac-
tures by Lewis (2010). Lewis (2012) has since reinterpreted these as thalassaemic rib
lesions. This palaeopathological information provides indirect evidence of mobility,
because thalassaemia is endemic to the Mediterranean but not Britain. Another recent
study of health in Roman Britain was undertaken by Jenny (2011) on the cemetery at
Butt Road, Colchester. A much smaller proportion of children were present at this site
compared to Poundbury. Skeletal preservation at the site was generally poor, and it is
possible that taphonomic factors are responsible for this under-​representation. Almost
half of the non-​adults exhibited the pathological condition cribra orbitalia, a non-​spe-
cific indicator of poor health linked to anaemia (Walker et al. 2009). This condition
appears as holes or perforations in the orbits of the skull. Jenny (2011: 166) found that
more children buried with grave goods had cribra orbitalia than those without and
concluded that grave goods are not reflective of socio-​economic status. Of course, the
implication in Jenny’s interpretation is that skeletal health is correlated with status,
310

310   Rebecca Gowland

which is not always the case, as illustrated by the Poundbury data. Similarly, at the
Romano-​British site of Baldock, Griffin and colleagues (2011) note that those individu-
als buried in graves with furnishings tended to have poorer health in childhood than
those without. They conclude from this that the inclination was for ‘social climbers to
feel the need to advertise their newfound status and identities through material display’
(Griffin et al. 2011: 546). However, the types of items classed as grave furnishings in this
study also included ‘animal bones’ and ‘rubbing stones’—​not the kind of extravagant
goods that one would associate with conspicuous displays of wealth. The relationship
between health and status is complex, and interpretations of the osteological evidence
in this regard is further complicated by the non-​specific nature of the lesions included
in the above studies and the paradoxical nature of lesion expression (that is, skeletons
with lesions may represent healthier individuals with stronger immune systems, while
those without lesions may have died before their skeletons were affected) (Wood et al.
1992).
The bioarchaeological data form a rich seam of evidence from which to make impor-
tant inferences concerning past perceptions and care of children in Roman Britain.
However, the palaeopathological data are complex to interpret and we must be careful
not to be overly simplistic in the correlations made between skeletal indicators of poor
health and cultural indicators of ‘wealth’. Techniques of analysis are developing rapidly,
and it is likely that skeletal data will play an increasingly prominent role in the explora-
tion of childhood in Roman Britain.

The Funerary Evidence

If you lose your parents you’re an orphan; if you lose your husband you’re
a widow and if you lose your wife you’re a widower … But there’s no word
for losing your child. It’s as though it’s so terrible they couldn’t even give it
a name.
(Sally Holland, whose only child died aged 14 years;
in Joanna Moorhead, Guardian, Saturday, 6 June 2009).

The funerary context has long been exploited as an important conduit for understand-
ing perceptions of childhood in Roman Britain, not least, because this material pro-
vides a direct connection between children of different ages and material evidence in
the form of ritual treatment. It thus allows archaeologists to integrate the bioarchaeo-
logical and funerary data as a means of assessing past perceptions of childhood. Studies
of the burial treatment of children from Roman Britain in relation to age-​at-​death have
demonstrated some interesting patterns in terms of spatial aspects of deposition, in
addition to grave-​good inclusions (for example, type and quantity) as well as the pro-
visioning of coffins. A review of the findings of some of these studies is presented here.
 Ideas of Childhood in Roman Britain    311

The Newborn: Between Person and Non-​Person

Infants were often the recipients of a distinct burial practice in Roman Britain as well
as elsewhere in the empire and, consequently, have attracted considerable discussion
in the archaeological literature. In most regions in Britain, infants have been excavated,
often in substantial numbers, from within and around settlements and villas rather than
formal cemetery sites (Scott 1991, 1992, 1999, 2001; Mays 1993, 2003; Strück 1993; Pearce
2001; Gowland and Chamberlain 2002; Moore 2009; Mays and Eyers 2011). This dif-
ferential treatment of infants in death has generally been interpreted in terms of either
ritual (though one un-​related to the funerary concerns of the infant) or disposal of the
unwanted child (e.g. Cocks 1921; Watts 1989; Mays 1993, 2003). Intrinsic to the latter
interpretation are preconceptions concerning a lack of emotional attachment to young
infants in response to either high infant mortality or the practice of infanticide. This
form of ‘demographic determinism’ has been disputed by a number of authors (Golden
1988). Interpretations of these burials solely in terms of the disposal of the body implies
a passivity that denies the agency of infants to effect those around them emotionally,
physically, and economically (Gowland et al. 2014).
Direct historical evidence for infanticide in Roman Britain does not exist, and histori-
cal evidence from Rome refers not to the direct killing of infants, but to the ‘putting-​out’
or abandonment of an infant (Boswell 1998). While these may lead to ultimately the
same outcome—​the death of the infant—​they are conceptually very different and have
different archaeological traces. Much of the historical evidence indicates that a large
proportion of these infants were subsequently taken in and raised by others (e.g. Grubbs
2010). Leaving a child potentially to be brought up by strangers is not the same as the
direct and violent act of killing. Crucially, in terms of the archaeological evidence, the
latter would lead to a body whereas the former would not (even were the child to die, the
body of an exposed infant is likely to be promptly dispersed by animals). The conflation
of infanticide and abandonment has been resoundingly critiqued within the historical
literature, but this does not appear to have filtered through to archaeological discussions
of Roman infancy (Gowland et al. 2014).
One of the most vociferous proponents of the infanticide theory has been Mays (1993,
2003; Mays and Eyers 2011), who has argued that the age distribution of infants recov-
ered from Romano-​British settlements exhibits a pronounced neonatal peak, incom-
patible with what one would expect from natural mortality. Gowland and Chamberlain
(2002) subsequently reassessed this evidence using a new methodology and concluded
that the ageing method used by Mays created biased results. When a different method
was employed, the age distribution obtained was compatible with that expected when
stillborn infants, as well as neonates, were accorded similar burial rites (that is, a much
broader range of ages at death). In other words, no osteological evidence was found to
substantiate interpretations of infanticide (Gowland and Chamberlain 2002). Mays
(2003) and Mays and Eyers (2011), however, continue to interpret the evidence in terms
312

312   Rebecca Gowland

of infanticide—​most recently in relation to the analysis of thirty-​three infants from

Yewden Villa, Hambledon (Mays and Eyers 2011).
It is important to note that the ages of the vast majority of infants buried within set-
tlements and villas range from approximately 24 gestational weeks to approximately
2 months of age, after which they tend to be accorded different burial rites (Gowland
2001; Moore 2009). Moore’s survey (2009) of this practice in Roman Britain revealed
that 76 per cent of the infants were perinates and neonates aged between birth and
1 month post-​partum. This phenomenon is, therefore, a repeated funerary ritual associ-
ated with a very specific age group. The vast majority of these infants were buried within
domestic contexts; beneath the floors of general domestic rooms or, when buried exter-
nally, they were close to the domestic building (Moore 2009).
The issue of when a foetus is considered a human being is much debated within mod-
ern western society. For example, technological advances relating to the imaging of the
foetus in utero have contributed to the shifting boundaries of personhood in contempo-
rary society (Gowland et al., 2014). The conferment of personhood is a fluid construct
and for many societies does not begin with birth. Historical evidence relating to Rome
indicates that an infant attained an individual social identity only on the day that it was
named (the lustratio)—​a ceremony that took place on the eighth day after birth for
females and the ninth for males (Weidemann 1989; Rawson 1991). Other historical evi-
dence indicates that infants were not perceived to have attained true personhood prior
to teething and, possibly, walking and talking (Watts 1989; Philpott 1991). The relevance
of historical evidence relating to Rome for attitudes in Roman Britain is open to consid-
erable debate, but, on the basis of such evidence, one could argue that the grouping of
stillbirths and infant deaths is to be expected: they were all considered non-​people and
of little importance. But, if this were so, why bury them within houses and settlements,
rather than simply dispose of them further from the dwelling space? The incorpora-
tion of the infant in death, firmly within the social sphere of the living, should not be
dismissed so readily. Perhaps the burial of infants within or close to the domestic sphere
was not because it was convenient for disposal (which can hardly have been so), but
because the household represented the social world of that child: their burial was con-
ducted within the small social arena of which they were a part (Gowland 2001). Moore
(2009: 48) likewise argues that the burials of these infants was ‘not the random disposal
of the unwanted or marginalised, but the result of careful choices and decisions relating
to concepts associated with the physical and spiritual worlds, the infant was inherently
ambiguous but was also, in certain senses, a being of power’.

Cemetery Evidence
When examining the burial treatment of children, it is of course important to consider
this evidence in relation to the graves and grave-​good assemblages included with indi-
viduals of all ages. By doing so, one can obtain a better sense of any age-​related patterns
in burial ritual spanning the entire life course, thus moving beyond the constraints of
 Ideas of Childhood in Roman Britain    313

the child/​adult dichotomy. Funerary studies of this nature are often limited to inhuma-
tion cemeteries owing to a dearth of full analyses currently undertaken on cremated
remains (though see Pearce 1999). Consequentially, for Roman Britain, there is a bias
towards later Roman cemeteries of the third and fourth centuries. A good example of
one such cemetery is the late Roman cemetery of Lankhills in Winchester. This cem-
etery was excavated originally in the 1960s (Clarke 1979), though another substantial
part of the site was excavated and published more recently (Booth et al. 2010). Gowland
(2001, 2002) conducted a reanalysis of the ages-​at-​death of the skeletons from the ear-
lier excavations and found distinctive age-​related patterns in burial treatment, which
will be summarized briefly. It was observed that grave-​good deposition was strongly
correlated with both age and sex. Children aged 4–​12 years and females in their early
twenties were buried with much greater quantities of grave goods than older females
and younger children. As well as an increase in the quantity of grave goods from 4 years
onwards, there was a shift towards the inclusion of those grave goods typically differen-
tiated by sex. The burial evidence points to a transition around the age threshold of 4–​
7 years, which coincides with the expression of a more strongly signified gender identity.
Likewise, at the Butt Road Cemetery, Colchester, Jenny (2011) found that only children
from the age of 7 onwards were buried with hairpins. Literary evidence pertaining to
Rome also indicates that the perceived identity of children underwent a transition after
the age of about 7 years, towards an identity that was more explicitly gendered (Harlow
and Laurence 2002).
While approximately equal quantities of jewellery were recovered from the graves
of adult females (96 items) and children (89 items) from the earlier excavations at the
Lankhills site, there were differences in terms of the deposition of these items within
the graves. Only 14 per cent of the items found with the females were worn, com-
pared to 39 per cent with the immature skeletons. The wearing of jewellery among
the children (4–​12 year olds) appeared to be governed partly by the type of item. For
example, necklaces were more likely to be worn, while finger rings were almost never
worn. Either it was either considered inappropriate for children of this age to be wear-
ing finger rings or it was functionally impossible if they were adult rings. If the latter,
then the finger rings had a symbolic role within the graves of these children (Gowland
2001, 2002). Of the unworn items of jewellery, those buried with the children were
more frequently buried near to the legs or feet, while those accompanying the adults
tended to be buried next to the head. This pattern is not observed at all Romano-​
British cemeteries. For example, at Butt Road, Colchester, the reverse is true. While
jewellery was much less frequent at this site, only 13 per cent of items buried with
children were worn (the majority having been placed in piles next to the head) com-
pared to 43 per cent buried with the adults. The important factor may not be whether
the items were worn or not, but rather that a distinction was maintained between the
adults and the young.
Another example from Lankhills concerns the burial of a belt set by the feet of a child
aged 4–​7 years. These belt sets were usually worn during burial, but were almost exclu-
sively buried with older males. At the late Roman cemetery of Butt Road, Colchester,
314

314   Rebecca Gowland

another child (aged 9–​11 years) was buried with two belt sets that had been placed beside
the head (Crummy et al. 1993). It is clear that, while the burial of these relatively rare
grave goods occurred with children, it was either not appropriate for them to be worn or
not practical owing to the object’s size. Either way, the symbolism of these grave goods
has altered because of the age of the deceased, and this has been reinforced through their
placement. It seems likely that these belt sets were associated with a particular status or
position of power that could not have been held in life by one so young, but perhaps they
would have been achieved or inherited had they lived longer. A parallel with the bur-
ial of late Roman belt sets can be drawn from O’Shea’s work (1995: 130) on the Mokrin
Bronze Age cemetery in Hungary. O’Shea also found that badges of office, when buried
with children, were not placed in the correct functional position in the grave. Instead,
they could be placed, for example, by the feet, to symbolize the fact that the role was not
actually held before death.
Literary sources from Rome refer to the display of the body of the deceased, especially
of the elite, prior to the funeral (Toynbee 1985). Whether the coffin was opened or closed
for view is not known; however, it has been suggested that at Lankhills and other sites in
Roman Winchester the body was displayed (Pearce 1999: 166). Recent chemical analyses
of burials from lead-​lined coffins in Roman Britain have suggested that the bodies of
some individuals were embalmed with natural plant exudates, presumably to preserve
the body (Brettell et al. 2014) (for further discussion of Romano-​British burial practice,
see Weekes, this volume). The recurrent distinction in the focus of deposition of grave
goods would have reinforced a visual difference in burial display that was linked to an
emphasis on different parts of the body that must be symbolic of a social identity related,
at least in part, to age. This also demonstrates how different meanings can be conferred
onto—​and imbued by—​the same items of material culture, depending not only on the
gender of an individual but also on his or her stage in the life course. When consider-
ing grave-​good assemblages, we are able not only to establish whether certain goods are
associated with either a masculine or a feminine gender, but also to observe the fluidity
of gender and status dynamics with age.
This analysis of the funerary evidence has been revealing in a number of ways. First,
perceptions of infancy differ from our own in terms of the grouping of stillbirths and
post-​neonatal deaths outside the cemetery context. In the modern world it has been
observed that ‘the social space of childhood is also a temporal phenomenon’ (James et al.
1998: 41). In other words, age identity is one means through which we manage social
space: there are inappropriate spaces for individuals of different ages to occupy. In the
Roman world we see that this is certainly true in death, with the burial of neonates fre-
quently occurring within the domestic sphere. Within the ‘normative’ cemeteries, there
are also age-​related patterns in burial practice. For example, a shift in identity occurs
around the 4–​7-​year age category extending to individuals of 8–​12 years. This shift is
associated with a greater expression of gendered identity, in death at least. The graves of
these individuals are among the richest for the entire cemetery, and this increase in bur-
ial wealth may also indicate a concurrent increase in the social status of children upon
reaching this age threshold, although direct correlations of this nature are problematic.
 Ideas of Childhood in Roman Britain    315

It is possible that the wealth of these graves may instead relate to the sense of loss accom-
panying the untimely death of a child.
In the contemporary western world we have constructed a separate and distinctive
material and social world of the child—​one that both reflects and reinforces our percep-
tions of their specialness and vulnerability. This material distinction is absent from the
cemetery evidence of Roman Britain and may indicate that the emphasis and percep-
tion of the lived reality of childhood were not based on difference, as they are today but,
instead, that children played a much more integrated role in the structuring and func-
tioning of Romano-​British society (Gowland 2001).

Future Directions

Developments within the field of human bioarchaeology are providing exciting oppor-
tunities to explore past perceptions of childhood and childcare as never before. Isotopic
studies of bones and teeth provide access to information on weaning practices, child-
hood diet, and mobility. Isotope samples taken from different tissues of the body pro-
vide information relating to different life-​course stages. For example, in an adult, ribs
samples may yield information for the couple of years prior to death, cortical bone
for the last decade or longer, and teeth to childhood. A technique recently developed
by Beaumont and colleagues (2013) has retrieved dietary isotope information on
nineteenth-​century Irish famine victims from multiple dentine increments within a
single tooth. This technique thus provides high-​resolution data for the months prior to
death and, in this instance, poignantly revealed the deteriorating health of the children.
Isotope information can, therefore, be used to construct chemical biographies of child-
hood experiences, even from adult remains. Such studies mean that skeletal remains
should no longer be considered as providing simply a snapshot of an individual at the
time of death but, instead, can yield a life-​course perspective (Robb 2002; Gowland and
Thompson 2013).
Likewise, developments in the palaeopathological analysis of growth and diagnosis of
nutritional deficiency diseases have the potential to provide complementary data relat-
ing to dietary and cultural practices. It is the integration of these different strands of data
that is crucial for providing a more informed interpretation of evidence for past con-
structions of childhood. Of course, these skeletal data are of limited meaning unless fully
contextualized and analysed with regard to the cultural evidence from specific grave
contexts, as well as a broader material understanding of Roman Britain and the empire
at large. It is important that bioarchaeologists work closely with Roman archaeologists,
including artefact specialists and experts in the economic history of the Roman Empire.
Finally, over more recent years bioarchaeologists are showing a burgeoning interest in
the integration of social theory into their studies of the hard tissues of the body (e.g.
Gowland and Knüsel 2006; Sofaer 2006; Knudson and Stojanowski 2009; Agarwal and
Glencross 2011; Gowland and Thompson 2013). The reconceptualization of the skeleton
316

316   Rebecca Gowland

as a social entity is helping to propel the discipline forward so that its significance for
debates concerning social identity generally and childhood specifically is fully realized.

Conclusions

Studies of childhood in Roman Britain are still comparatively scarce. While there is a
huge and growing corpus of historical research on children and the family in the ancient
world, this rarely touches on Britain, owing to a dearth of relevant sources. The archaeo-
logical evidence in this regard has so far been under-​utilized. Likewise, the bioarchaeo-
logical data—​the most direct form of evidence for Romano-​British childhood—​have
been neglected. Studies of the skeletal remains, both isotopic and palaeopathological,
have shown the richness and diversity of the information that can be retrieved. When
fully integrated with the archaeological evidence for settlement type, diet, and environ-
ment, these data can yield important insights into the lives of children in Roman Britain.
By adopting a life-​course approach to the funerary evidence—​that is, contextualizing
the study of the younger individuals within the entire lifespan—​it is possible to identify
age-​related transitions in gender and status. Some of these may coincide with histori-
cal evidence from Italy, but not always. A life-​course perspective enables archaeologists
more readily to identify those age thresholds that fall outside current or contemporane-
ous age paradigms, so that we might more readily arrive at an understanding of the per-
ceptions of childhood in Roman Britain.

Acknowledgements

Thank you to the editors for inviting me to contribute to this volume. Thanks also to
Rebecca Redfern for sharing her insights into Roman childhood over the last few years,
Tim Thompson for comments on an earlier draft, and Lindsay Powell for her assistance.
Any errors remain my own.

References
Agarwal, S. C., and Glencross, B. A. (2011) (eds). Social Bioarchaeology. Oxford:
Wiley-​Blackwell.
Ariès, P. (1962). Centuries of Childhood: A Social History of Family Life. London: Cape.
Baker M. (1997). ‘Invisibility as a Symptom of Gender Categories in Archaeology’, in J. Moore
and E. Scott (eds), Invisible People and Processes. London: Leicester University Press, 183–​191.
Beaumont, J., Gledhill, A., Lee-​Thorp, J., and Montgomery, J. (2013). ‘Childhood Diet: A Closer
Examination of the Evidence from Dental Tissues Using Stable Isotope Analysis of
Incremental Human Dentine’, Archaeometry, 55: 685–​693.
 Ideas of Childhood in Roman Britain    317

Booth, P., Simmonds, A., Boyle, A., Clough, S., Cool, H. E. M., and Poore, D. (2010) (eds). The
Late Roman Cemetery at Lankhills, Winchester: Excavations 2000–​2005. Oxford Archaeology
Monograph 10. Oxford: Oxford Archaeology.
Boswell, J. (1998). The Kindness of Strangers: The Abandonment of Children in Western Europe
from Late Antiquity to the Renaissance. Chicago: University of Chicago Press.
Blane, D. (2006). ‘The Life Course, the Social Gradients and Health’, in M. Marmot and R. G.
Wilkinson (eds), Social Determinants of Health. 2nd edn. Oxford: Oxford University Press.
Brettell, R. C., Stern, B., Reifarth, N., and Heron, C. (2014). ‘The “Semblance of Immortality”?
Resinous Materials and Mortuary Rites in Roman Britain’, Archaeometry, 56: 444–456.
Brickley, M., and Ives, R. (2008). The Bioarchaeology of Metabolic Bone Disease.
Amsterdam: Elsevier; London: Academic Press.
Clarke, G. (1979). The Roman Cemetery at Lankhills. Winchester Studies 3: Pre-​Roman and
Roman Winchester. Oxford: Clarendon Press.
Cocks, A. H. (1921). ‘A Romano-​ British Homestead in the Hambleden Valley, Bucks’,
Archaeologia, 71: 141–​198.
Crummy, N., Crummy, P., and Crossan, C. (1993). Excavations of Roman and Later Cemeteries,
Churches and Monastic Sites in Colchester, 1971–​88. Colchester Archaeological Reports
9. Colchester: Colchester Archaeological Trust.
Dasen, V., and Späth, T. (2010) (eds). Children, Memory and Family Culture in Roman Culture.
Oxford: Oxford University Press.
Dixon, S. (1992). The Roman Family. Baltimore: Johns Hopkins University Press.
Eyben, E. (1993). Restless Youth in Ancient Rome. London: Routledge.
Franks, A. W. (1991). ‘For a Sociology of the Body: An Analytical Review’, in M. Featherstone,
M. Hepworth, and B. Turner (eds), The Body, Social Process and Cultural Theory.
London: Sage, 36–​102.
Fraschetti, A. (1997). ‘Roman Youth’, in G. Levi and J.-​C. Schmitt (eds), A History of Young
People in the West. Volume 1. Cambridge, MA: Harvard University Press, 51–​82.
Fuller, B. T., Molleson, T. L., Harris, D. A., Gilmour, L. T., and Hedges, R. E. M. (2006). ‘Isotopic
Evidence for Breastfeeding and Possible Adult Dietary Differences from Late/​Sub Roman
Britain’, American Journal of Physical Anthropology, 129: 45–​54.
Golden, M. (1988). Did the Ancients Care when their Children Died?’, Greece & Rome,
35: 152–​163.
Golden, M. (2011). ‘Other People’s Children’, in B. Rawson (ed.), Companion to Families in the
Roman and Greek World. Oxford: Wiley-​Blackwell, 262–​275.
Gowland, R. L. (2001). ‘Playing Dead:  Implications of Mortuary Evidence for the Social
Construction of Childhood in Roman Britain’, in D. Davies, A. Gardner, and K. Lockyear
(eds), TRAC 2000:  Proceedings of the Tenth Annual Theoretical Roman Archaeology
Conference, London 2000. Oxford: Oxbow Books, 152–​168.
Gowland, R. L. (2002). ‘Age as an Aspect of Social Identity in Fourth-​to-​Sixth Century ad
England: The Archaeological Funerary Evidence’. Unpublished Ph.D. thesis, University of
Durham.
Gowland, R. L. (2006). ‘Ageing the Past: Examining Age Identity from Funerary Evidence’, in
R. Gowland and C. J. Knüsel (eds), Social Archaeology of Funerary Remains. Oxford: Oxbow
Books, 143–​154.
Gowland, R. L., and Chamberlain, A. T. (2002). ‘A Bayesian Approach to Ageing Perinatal
Skeletal Material from Archaeological Sites: Implications for the Evidence for Infanticide in
Roman Britain’, Journal of Archaeological Science, 29: 677–​85.
318

318   Rebecca Gowland

Gowland, R. L., and Knüsel, C. J. (2006). ‘Introduction’, in R. Gowland and C. J. Knüsel (eds),
Social Archaeology of Funerary Remains. Oxford: Oxbow Books, pp. ix–​xiv.
Gowland, R. L., and Redfern, R. C. (2010). ‘Childhood Health in the Roman World: Perspectives
from the Centre and Margin of the Empire’, Childhood in the Past, 3: 15–​42.
Gowland, R. L., and Thompson, T. J.  U. (2013). Human Identity and Identification.
Cambridge: Cambridge University Press.
Gowland, R. L., Chamberlain, A. T., and Redfern, R. C. (2014). ‘On the Brink of Being:
Re-​Evaluating Infanticide and Infant Burial in Roman Britain’, in P. M. Carroll and E.-​J.
Graham (eds), Infant Health and Death in Roman Italy and Beyond. Journal of Roman
Archaeology Supplementary Series 99. Ann Arbor: Journal of Roman Archaeology,
69–88.
Griffin, R., Pitts, M., Smith, R., and Brook, A. (2011). ‘Inequality at Late Roman Baldock,
UK: The Impact of Social Factors on Health and Diet’, Journal of Anthropological Research,
67: 533–​556.
Grubbs, J. E. (2010). ‘Hidden in Plain Sight: Expositi in the Community’, in V. Dasen and T.
Späth (eds), Children, Memory and Family Identity in Roman Culture. Oxford:  Oxford
University Press, 293–​310.
Halcrow S., and Tayles, N. (2008). ‘The Bioarchaeological Investigation of Childhood
and Social Age:  Problems and Prospects’, Journal of Archaeological Method and Theory,
15: 190–​215.
Harlow, M., and Laurence, R. (2002). Growing Up and Growing Old in Ancient Rome.
London: Routledge.
Hopkins, K. (1965). ‘The Age of Roman Girls at Marriage’, Population Studies, 18: 309–​327.
James, A., Jenks, C., and Prout, A. (1998). Theorizing Childhood. Cambridge: Polity Press.
Jay, M. (2009). ‘Breastfeeding and Weaning Behaviour in Archaeological Populations:
Evidence from the Isotopic Analysis of Skeletal Materials’, Childhood in the Past, 2: 163–​179.
Jenny, L. L. (2011). ‘A Bioarchaeological Study of Local Roman Identity: Skeletal Stress and
Mortuary Treatment in the Butt Road Cemetery’. Unpublished Ph.D. dissertation, Michigan
State University.
Knudson, K. J., and Stojanowski, C. M. (2009) (eds). Bioarchaeology and Identity in the
Americas. Gainesville, FL: University Press of Florida.
Kron, G. (2005). ‘Anthropometry, Physical Anthropology and the Reconstruction of Ancient
Health, Nutrition and Living Standards’, Historia, 54: 68–​83.
Laurence, R. (2000). ‘Metaphors, Monuments and Texts: The Life Course in Roman Culture’,
World Archaeology, 3: 442–​455.
Lewis, M. E. (2007). The Bioarchaeology of Children. Cambridge: Cambridge University Press.
Lewis, M. E. (2010). ‘Life and Death in a civitas Capital: Metabolic Disease and Trauma in the
Children from Late Roman Dorchester, Dorset’, American Journal of Physical Anthropology,
142: 405–​416.
Lewis, M. E. (2012). ‘Thalassaemia: its Diagnosis and Interpretation in Past Skeletal
Populations’, International Journal of Osteoarchaeology, 22: 685–​693.
Lillehammer, G. (1989). ‘A Child Is Born: The Child’s World in an Archaeological Perspective’,
Norwegian Archaeological Review, 22: 89–​105.
Mays, S. (1993). ‘Infanticide in Roman Britain’, Antiquity, 67: 883–​888.
Mays, S. (2003). ‘Comment on “A Bayesian Approach to Ageing Perinatal Skeletal Material
from Archaeological Sites: Implications for the Evidence for Infanticide in Roman Britain”
by R. L. Gowland and A. T. Chamberlain’, Journal of Archaeological Science, 30: 1695–​1700.
 Ideas of Childhood in Roman Britain    319

Mays, S., and Eyers, J. (2011). ‘Perinatal Infant Death at the Roman Villa Site at Hambleden,
Buckinghamshire, England’, Journal of Archaeological Science, 38: 1931–​1938.
Molleson, T. I. (1989). ‘Social Implications of Mortality Patterns of Juveniles from Poundbury
Camp, Romano-​British Cemetery’, Anthropologischer Anzeiger, 47: 27–​38.
Molleson, T. I. (1993). ‘The Human Remains’, in D. E. Farwell and T. I. Molleson (eds),
Poundbury, Volume 2: The Cemeteries. Dorset Natural History and Archaeological Society
Monograph Series 11. Dorset: Dorset Natural History and Archaeological Society, 142–​213.
Moore, A. (2009). ‘Hearth and Home: The Burial of Infants within Romano-​British Domestic
Contexts’, Childhood in the Past, 2: 33–​54.
Moore, J., and Scott, E. (1997) (eds). Invisible People and Processes. London: Leicester University Press.
O’Shea, J. M. (1995). ‘Mortuary Customs in the Bronze Age of South-​ Eastern
Hungary: Diachronic and Synchronic Perspectives’, in A. L. Beck (ed.), Regional Approaches
to Mortuary Analysis. New York: Plenum, 126–​146.
Pearce, J. (1999). ‘The Dispersed Dead: Preliminary Observations on Burial and Settlement
Space in Rural Roman Britain’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC
98:  Proceedings of the Eighth Theoretical Roman Archaeology Conference, Leicester 1998.
Oxford: Oxbow Books, 151–​162.
Pearce, J. (2001). ‘Infants, Cemeteries and Communities in the Roman Provinces’, in D. Davies,
A. Gardner, and K. Lockyear (eds), TRAC 2000: Proceedings of the Tenth Annual Theoretical
Roman Archaeology Conference, London 2000. Oxford: Oxbow Books, 125–​142.
Philpott, R. (1991). Burial Practices in Roman Britain: A Survey of Grave Treatment and
Furnishing ad 43-​410. BAR British Series 219. Oxford: Archaeopress.
Pollock, L. A. (1983). Forgotten Children. Cambridge: Cambridge University Press.
Powell, L. A., Redfern, R. C., and Millard, A. R. (2014). ‘Infant Feeding Practices in Roman
London: The Isotopic Evidence’, in P. M. Carroll and E.-​J. Graham (eds), Infant Health and
Death in Roman Italy and Beyond. Journal of Roman Archaeology Supplementary Series, 96.
Ann Arbor: Journal of Roman Archaeology, 89–110.
Prout, A., and James, A. (1997). ‘A New Paradigm for the Sociology of Childhood? Provenance,
Promise and Problems’, in A. James and A. Prout (eds), Constructing and Reconstructing
Childhood. Basingstoke: Falmer Press, 7–​32.
Prowse, T. L., Schwarcz, H. P., Garnsey, P., Knyf, M., Macchiarelli, R., and Bondioli, L. (2008).
‘Isotopic and Dental Evidence for Infant and Young Child Feeding Practices in an Imperial
Roman Skeletal Sample’, American Journal of Physical Anthropology, 133: 294–​308.
Rawson, B. (1991). ‘Adult–​Child Relationships in Roman Society’, in B. Rawson (ed.), Marriage,
Divorce and Children in Ancient Rome. Oxford: Clarendon Press, 7–​30.
Rawson, B. (2003). Children and Childhood in Roman Italy. Oxford: Oxford University Press.
Rawson, B. (2011) (ed.). Companion to Families in the Greek and Roman Worlds. Oxford:
Oxford University Press.
Rawson, B., and Weaver, P. (1997) (eds). The Family in Roman Italy. Oxford: Oxford University
Press.
Redfern, R. C. (2007). ‘The Influence of Culture upon Childhood: An Osteological Study of
Iron Age and Romano-​British Dorset, England’, in M. Harlow and R. Laurence (eds), Age
and Ageing in the Roman Empire: Approaches to the Roman Life Course. Journal of Roman
Archaeology Supplementary Series 64. Ann Arbor: Journal of Roman Archaeology,
171–​90.
Redfern, R. C., and Gowland, R. L. (2012). ‘A Bioarchaeological Perspective on the Pre-​Adult
Stages of the Life Course: Implications for the Care and Health of Children in the Roman
320

320   Rebecca Gowland

Empire’, in M. Harlow and L. Larsson Loven (eds), Families in the Roman and Late Antique
World. London: Continuum Press, 111–​140.
Redfern, R. C., Millard, A. R., and Hamlin, C. (2012). ‘A Regional Investigation of Subadult
Dietary Patterns and Health in Late Iron Age and Roman Dorset, England’, Journal of
Archaeological Science, 39: 1249–​1259.
Revell, L. (2005). ‘The Roman Life Course: A View from Inscriptions’, European Journal of
Archaeology, 8: 43–​63.
Robb, J. (2002). ‘Time and Biography: Osteobiography of the Italian Neolithic Lifespan’, in Y.
Hamilakis, M. Pluciennik, and S. Tarlow (eds), Thinking through the Body: Archaeologies of
Corporeality. New York: Kluwer Academic/​Plenum Publishers, 153–​172.
Scott, E. (1991). ‘Animal and Infant Burials in Romano-​ British Villas:  A  Revitalization
Movement’, in P. Garwood, D. Jennings, R. Skeates, and J. Toms (eds), Sacred and
Profane: Proceedings of a Conference on Archaeology, Ritual and Religion. Oxford University
Committee for Archaeology Monograph 32. Oxford:  Oxford University Committee for
Archaeology, 115–​121.
Scott, E. (1992). ‘Images and Contexts of Infants and Infant Burials: Some Thoughts on Cross
Cultural Evidence’, Archaeological Review from Cambridge, 11: 77–​92.
Scott, E. (1999). The Archaeology of Infancy. BAR International Series 819. Oxford;
Archaeopress.
Scott, E. (2001). ‘Unpicking a Myth:  The Infanticide of Female and Disabled Infants in
Antiquity’, in D. Davies, A. Gardner, and K. Lockyear (eds), TRAC 2000: Proceedings of the
Tenth Annual Theoretical Roman Archaeology Conference, London 2000. Oxford:  Oxbow
Books, 143–​151.
Shaw, B. D. (1987). ‘The Age of Roman Girls at Marriage: Some Reconsiderations’, Journal of
Roman Studies, 77: 30–​46.
Shilling, C. (1993). Body and Society. Cambridge: Cambridge University Press.
Sofaer, J. R. (2006). The Body as Material Culture: A Theoretical Osteoarchaeology. Cambridge:
Cambridge University Press.
Sofaer Derevenski, J. R. (1994). ‘Where are the Children? Accessing Children in the Past’,
Archaeological Review from Cambridge, 13: 7–​33.
Stone, L. (1977). The Family, Sex and Marriage in England 1500–​1800. London: Weidenfeld &
Nicholson.
Toynbee, J. M. C. (1985). Death and Burial in the Roman World. London: Thames and Hudson.
Walker, P. L., Bathurst, R. R., Richman, R., Gjerdrum, T., and Andrushko, V. A. (2009). ‘The
Causes of Porotic Hyperostosis and Cribra Orbitalia: A Reappraisal of the Iron Deficiency-​
Anemia Hypothesis’, American Journal of Physical Anthropology, 139: 109–​125.
Watts, D. J. (1989). ‘Infant Burials and Romano-​British Christianity’, Archaeological Journal,
46: 372–​383.
Weidemann, T. (1989). Adults and Children in the Roman Empire. London: Routledge.
Wood, J. W., Milner, G. R., Harpending, H. C., and Weiss, K. M. (1992). ‘The Osteological
Paradox: Problems of Inferring Prehistoric Health from Skeletal Samples’, Current
Anthropology, 33: 343–​370.
 Chapter 16

The Life C ou rse

Alison Moore

Introduction

Until relatively recently, age has been an under-​examined constituent of social identity
and social organization in past societies: either consciously or unconsciously, the past
was populated by (predominantly male) adults (Revell 2000; 2010: 2; Lucy 2007: 43).
Although the development of gender studies in archaeology during the 1980s and 1990s
(e.g. Conkey and Spector 1984; Scott 1995, 1997; Sørensen 2000: 16–​20) and a growing
interest in children and childhood (e.g. Lillehammer 1989; Sofaer Derevenski 1994;
Crawford 1999, 2000; Kamp 2001; see also papers in Moore and Scott 1997) have bought
the study of age into greater prominence, much work still needs to be done on other age
categories, such as the elderly (Welinder 2001; Appleby 2010; Moore 2010), and the role
(and variability) of age as an organizational tool. Age, alongside gender, was a funda-
mental aspect of identity in the Roman world (see Harlow and Laurence 2008 for a dis-
cussion of Roman age systems), and the ‘life course’ (the study of the social construction
of age categories over the entirety of the human lifespan) is a useful framework through
which to explore how age was understood and constructed in the Roman provincial
contexts. Whilst life-​course studies have been used to good effect in studies from prehis-
toric and historic eras (e.g. Sofaer Derevenski 1997, 2000; Stoodley 1999, 2000; Meskell
2000; Gowland 2006; Gilchrist 2012), understanding of age construction in Roman con-
texts remains, to a large extent, dominated by a reliance on elite ‘Rome-​centric’ textual
sources (e.g. Finley 1989; Fraschetti 1997; Harlow and Laurence 2002; Cokayne 2003;
Parkin 2003), studies of the construction of the Roman family unit (e.g. Wiedemann
1989; Bradley 1991; Dixon 1992), and demographic studies based on epigraphy from
Rome (e.g. Hopkins 1966; Saller and Shaw 1984; Saller 1987; Shaw 1987; Scheidel 2001).
However, as age is a variable and how it is constructed can vary considerably between
societies, over time, and over geographical distance (Hareven 1978:  1–​2; Gowland
2006: 143) it is unlikely that how age was constructed in Rome would be relevant in
provincial contexts. For example, Revell’s study (2005) of the life course as recorded
322

322   Alison Moore

on epitaphs from regiones in Italy and the provinces of Africa, Spain, and Britain
showed a wide degree of variability in ages marked, influenced by both pre-​Roman and
Romanized customs (see also Revell 2010: 4). Similarly, Gowland’s work (2002) on the
construction of age at the late Romano-​British cemeteries at Lankhills, Winchester,
and Butt Road, Colchester, has shown variability in how age was marked within urban
contexts, while Moore (2009a) has recorded variability in age construction between
Romano-​British urban and rural contexts, and between eastern and western regions of
Roman Britain. With its long history of Roman excavations and large number of cem-
etery contexts, Roman Britain should be an ideal arena in which to study age and age
construction on a provincial and inter-​provincial level, and this chapter summarizes
the work undertaken so far on age construction in Roman Britain and suggests possible
avenues of enquiry for the future.

The Life Course in Theory and Practice

Alongside gender, ethnicity, status, and religious affiliation, age is a fundamental part
of an individual’s personal and social identity. As an aspect of identity, age can also be
socially constructed: that is, it is utilized as an organizational category by societies to
form one of the ‘building blocks’ that supports how a society is organized and func-
tions (Gilchrist 2000; Gowland 2006; Lucy 2007). As individuals (or age cohorts) grow,
develop, and age, they enter new life phases, and these ‘age transitions’ are often marked
by rites and rituals (Sofaer 2011; Gilchrist 2012: 9). By taking the human lifespan in its
entirety, the life course enables the role of age in social organization to be understood, by
tracing when ‘age transitions’ take place across the linear trajectory from birth to death
and comparing and contrasting how each age category is constructed and displayed.
However, age is more complex than just a number and can be defined in three ways:
biological (the age of the body), chronological (amount of time passed since birth), and
social (age-​related behaviour and social expectations associated with a given age stage)
(Gowland 2006: 143; Sofaer 2011: 286–​287). The different concepts of age are inter-
twined, and ‘social’ age is not primarily biological or chronological but encapsulates the
fulfilment of social roles and obligations (Gowland 2006: 143). It is only by attaining
a combination of a biological, chronological, or social age that a life-​course transition
takes place, often marked by ‘rites of passage’, which in themselves are socially varia-
ble (Gilchrist 2000: 235). Working in tandem with these three forms of age is gender.
How and when an individual enters a new life stage is dependent upon gendered socio-​
cultural norms, which are, in turn, influenced by the social status, ethnicity, and reli-
gious affiliation of the individual (Ginn and Arber 1995: 2; Sofaer Derevenski 1997: 876;
Gilchrist 1999).
Within archaeology, a life-​course approach explores the timing and expressions of
‘age transitions’ as represented in the mortuary record. A society’s burial practice often
encapsulates the social norms and expectations relating to identity, and it is often
 The Life Course   323

through the quantity and type of material culture provided with the burial—​which can
have both a functional and a symbolic meaning—​that concepts of identity are created,
expressed, and reinforced (Sofaer Derevenski 1994:  8; 1997:  876; Gilchrist 2000:  325;
Sørensen 2000). These concepts of identity are created not by the individual but by the
mourners, and, as individuals grow, mature, and age, the nature of the primary mourn-
ers changes, from parents in childhood, spouse in adulthood, and descendants in older
age (Parkin 1992; Gowland 2006: 152). As such, transitions within the composition of
grave-​good assemblages and within other aspects of the mortuary data can signal when
the social group considers that a new age stage has been attained. It is through trac-
ing these age-​and gender-​related changes across the lifespan that socio-​cultural atti-
tudes towards different age cohorts can be identified and explored on a local, regional,
or societal level. Furthermore, as well as representing cultural and gendered attitudes
towards age across the lifespan, a life-​course framework can also be utilized to iden-
tify which age stages or cohorts were considered the most socially worthy or significant
(Revell 2005: 44). As burial practice can reflect social hierarchies extant in life, socially
significant age cohorts would be emphasized through a high provision of grave goods,
the inclusion of a specific burial assemblage, and/​or a distinctive burial rite. In relation
to the most prominent age cohorts, the remaining age cohorts would show ‘degrees of
marginality’, reflecting their social position in relation to the pre-​eminent socio-​cultural
ideal. Thus a life-​course approach can provide a framework in which to understand
shifting concepts of social organization within a society over time or in relation to the
impact of external factors.
However, the reality of life-​course studies is not as clear-​cut as the theory would sug-
gest. Problems exist both with the data and with the various approaches employed by
archaeologists in ageing and conceptualizing age stages. These issues relate to the life-​
course approach in general and to the Romano-​British period in particular. One major
issue relates to correctly assigning age to skeletal material. For example, prior to puberty
and the development of the sexually diamorphic adult skeleton, it is problematic cor-
rectly to assign biological sex to individuals (Mays and Cox 2000: 123). This issue pre-
cludes a detailed understanding of when a gendered identity may have impacted on
the lives of children in the past. Conversely, with adult skeletal material the issue lies
with assigning a correct biological age using modern osteological techniques, leading
to a tendency to underestimate the age of adult skeletons, particularly in regard to older
and elderly adults (e.g. Molleson and Cox 1993; Aykroyd et al. 1999; Cox 2000: 62–6​3;
Appleby 2010: 148). In relation to life-​course study, the underestimation of adult age
means that identifying old age and understanding its impact on an individual and soci-
etal level are problematic (Welinder 2001; Appleby 2010: 146–​147; Moore 2010: 107).
A further issue in relation to the life course comes from a general lack of consensus
among archaeologists as to what ages constitute biological age cohorts and the terminol-
ogy used to describe the different age cohorts (Scheuer and Black 2000: 10; Gowland
2001: 153). For example, a brief trawl of Romano-​British cemetery reports shows a wide
range of ages coming under the category of ‘child’: 0–​15 years at Poundbury (Woodward
1993: 222), 2–​12 years at Folly Lane (Mays and Steele 1999: 308, table 39), and 5–​12 years
324

324   Alison Moore

at Brougham (McKinley 2004: 288, table 6.1). Other reports subdivide childhood into
young child and older child but show little consensus as to the ages being described—​for
example, 4–​7 years and 8–​12 years at Lankhills (Booth et al. 2010: 340, table 5.2), and
2–​5 years and 5–​12 years at London Road, Gloucester (Marquez-​Grant and Loe 2008:
30, table 3.2). Similarly, terms such as child, juvenile, sub-​adult, immature, and teenager
have all been employed in Romano-​British archaeological reports to describe individu-
als in pre-​adulthood—​terms that, as Gowland (2001: 153) argues, are culturally loaded
and carry western assumptions of behaviour and identity not relevant to the past. This
vagueness of data and terminology is further compounded by the nature of archaeologi-
cal cemetery reports in Britain, where the strands of burial evidence are often divided
between different specialists and reports (Gowland 2004: 135–​136). This division means
that taking a ‘holistic’ approach such as the life course to the burial evidence is com-
plicated by having to re-​contextualize the skeletal material with the appropriate grave
goods and other data. While this has been less of a problem in recent archaeological
cemetery reports, which often include a full catalogue of burials, it does preclude the
reliable use of older published material. This is a problem of particular relevance to life-​
course study for Roman Britain, where a large number of cemeteries were subject to
antiquarian excavation where the emphasis was on establishing pottery typologies and
dating.
A further issue that impacts on the use of life-​course analysis on Romano-​British
material is the use of cremation as the predominant burial rite in the late Iron Age and
early Romano-​British period. The difficulties inherent in the osteological ageing and
sexing of cremated skeletal remains (see McKinley 2000) means that a lot of early buri-
als are not usable in a life-​course analysis and the majority of life-​course analysis is based
primarily upon late, mainly fourth-​century, cemeteries. This precludes a clear under-
standing of if, and how, age construction may have been impacted on by Romanization.
Allied to a reliance on late Roman burial material is the decline in grave-​good use in the
later Roman period. This means that we are looking only at those segments of the popu-
lation who still included grave goods as part of the burial rite and, therefore, how social
age was constructed may not have been relevant to all. However, the issues outlined
can be addressed with an awareness of the limitations of the data, a general consensus
regarding age terminology, and a greater attention paid to the importance of age identity
within archaeological reporting in the future.

The Romano-​British Life Course

So, bearing these caveats in mind, what form does the Romano-​British life course
take? Overall, the civilian Romano-​British life course appears to consist of four main
age stages: infancy/​early childhood (0–​3 years), older childhood/​adolescence (4–​c. 17
years), young/​prime adulthood (c. 18–​39 years), and mature/​older adulthood (c. 40+
years) (Gowland 2002; Revell 2005; Moore 2009a), although variation in the timing
 The Life Course   325

and representation of these age stages is visible, dependent upon regional and localized
modes of burial.

Newborns and Infants

The first visible stage is that of infancy (0–​3 years) (see also Gowland, this volume).
In general, this stage is characterized by either a few grave goods or the inclusion of
a limited range of, predominantly, non-​gendered items. For example, at Colchester
(Crummy et al. 1993), pottery and shoes were the most frequent type of grave goods with
this age category, coins predominated at Poundbury (Farwell and Molleson 1993), while
at Lankhills, Winchester (Booth et al. 2010), pottery and coins were common (Gowland
2001: 159–​160). However, there are exceptions to this pattern, with some individuals
who died during infancy being buried with a large quantity and wide variety of grave
goods; for example, G503 at Butt Road, Colchester, an infant aged less than 1 year, was
provided with necklaces, armlets, and an iron ring (Crummy et al. 1993). Such variations
underline that factors such as wealth and status, as well as parental grief and affection,
need to be taken into account when looking at age identity in the burial record for this
age group. Another common feature of burial for infants, in particular neonates (aged
0–​1 month), is interment away from formalized cemetery contexts, particularly prior
to the ‘managed’ urban cemeteries of the fourth century ad (e.g. Watts 1989). Burials
of neonates and infants are commonly found within domestic or agricultural contexts
(Scott 1991), and evidence has been found for ‘zoning’ of infant burials within some cem-
eteries, sometimes associated with features such as boundary ditches (e.g. Gray 1930;
Blockley 1985; Scott 1999; Booth and Evans 2001), interpreted as marking the beginning
of use, or demarcating areas of different character within the cemeteries (Pearce 2001:
136). Age-​related ‘zoning’ of burials also occurred in some rural settlement contexts in
southern Britain. For example, Pearce (1999) identified a concentric pattern of burials,
with infants buried closer to the interior of the settlement space and older children and
adults marking site peripheries. Analysis of the age ranges of infants within agricultural
and domestic contexts indicates that infancy can be subdivided further, with neonates
(0–​1 month) and very young infants (1–​6 months) being buried in or adjacent to settle-
ment structures, often associated with features such as hypocausts, bathhouse furnaces,
and hearths or adjacent to ‘protective’ features such as walls and corners (Moore 2009b).
The exclusion of infants, particularly neonates, from formal cemetery contexts and
the appearance of this age group within domestic spaces has led to a variety of inter-
pretations, including infanticide (Mays 1993, 2000) and the disposal of unwanted
disabled infants (Molleson 1999). However, the burial of very young infants on settle-
ment sites has its roots deep in prehistory, both in Britain (e.g. Lally 2008; Lally and
Ardren 2008; McLaren 2011) and on the continent (e.g. Becker 1994, 1997, 2011; Karl
and Löcker 2011). The differentiation in treatment for this age group may correspond to
the concept of ‘personhood’ (the acquisition of human characteristics). The attainment
326

326   Alison Moore

of ‘personhood’ was (and remains) a fluid process and one not always immediately
acquired at birth; rather, when a newborn achieved personhood was defined by a series
of legal, religious, and cultural mores, bound about by ritual interventions and rites of
passage (Finlay 2013: 207; for rites of passage associated with newborns in the Roman
world, see, e.g., Hänninen 2005; Dasen 2009; Graham 2013). Prior to attaining person-
hood, newborns and the very young may have been conceptualized as ‘other’—​that is,
transitional beings not yet fully separated from the spiritual world. As ‘spiritual’ beings,
newborns and very young infants may have been buried within agricultural and domes-
tic buildings and features to act as a possible conduit to promote regeneration, both
agricultural and human (Moore 2009a: 184–​189; 2009b: 45; Carroll 2012; Finlay 2013:
210). This concept of this age group as a conduit to the spiritual is further borne out
by the inclusion of newborns and very young infants within ritual contexts—​for exam-
ple, Temple IV at Springhead, Kent (Penn 1960), Uley, Gloucestershire (Woodward and
Leach 1993), and Walcot, Bristol (Frere 1992: 296).

Childhood and Adolescence

The second life-​course stage is that of childhood/​adolescence (c. 4–​16/​17 years) and,
again, this can be further subdivided between children (c. 4–​12 years) and adolescence
(c. 13–​17 years) (see also Gowland, this volume, for further discussion on childhood in
Roman Britain). Revell’s study (2005: 58) of age from epitaphs in Roman Britain records
a relatively consistent percentage of epitaphs for both males and females in childhood
and adolescence, with slightly more males being commemorated at 0–​10 years, while
female commemoration was slightly higher at 11–​20 years. In burial, the second life-​
course stage is characterized by a greater formality in burial treatment with inclusion
within formal cemetery contexts and an increase in the number, and variety, of grave
goods provided (Figure 16.1). In her study of age identity based on the earlier excava-
tion of the cemetery at Lankhills, Gowland (2001: 160; 2002; see also Cool 2010: 31, table
3.4) identified a peak in grave-​good provision among older children (8–​12 years), a pat-
tern also seen at Colchester (26 per cent of older children given grave goods) and at
Poundbury (17.4 per cent of older children). A feature of the burials of older children
and adolescents at cemeteries such as Lankhills, Poundbury, and Colchester is the intro-
duction of/​or increase in the number of gendered items, in particular jewellery items,
usually bracelets. Approximately half of the older children (8–​12 years) and adolescents
(13–​17 years) at Lankhills were provisioned with bracelets (Gowland 2001: 160; 2002;
Cool 2010: 31, table 3.4), and here the bracelets were predominantly placed within the
grave for children and adolescents, rather than worn (Booth et al. 2010: 517). Similarly at
Colchester, bracelets were the predominant form of jewellery, again unworn (Gowland
2001: 162), and formed 78.3 per cent (36/​46) of all the jewellery items placed with child/​
adolescent burials (Crummy et al. 1993; Moore 2009a), while 64 per cent (16/​25) of jewel-
lery items in the 4–​17-​year age category at Poundbury (Woodward 1993) were bracelets.
 The Life Course   327

100

90

80
% of burials with grave goods

70

60

50

40

30

20

10

0
0–3 years 4–12 years 13–17 years 18–39 years 40+ years
Age categories
Colchester Poundbury Lankhills

Figure 16.1  Percentage of burials with grave goods with each age category at Colchester,
Poundbury, and Lankhills.
Source: © Alison Moore.

For Lankhills, Cool (2010: 31–​2) suggests that, as a percentage of the burials for the
child/​adolescent age category would have been males, the inclusion of so many brace-
lets, argued as being a predominantly female accoutrement, may indicate that this age
stage was, in particular, an important marker in the female life course. Unfortunately,
the difficulties inherent in correctly assigning biological sex to sub-​adult skeletal material
at present do not allow a definitive answer to the question of when a gendered identity
was introduced. However, the presence of bracelets and other items of jewellery, also a
feature associated with the adult female from across provincial contexts (e.g. Wyke 1994;
Montserrat 1996), is suggestive. At present, in the context of Roman Britain, it may be
easier to argue that these items represent cultus (appearance and care for the body) (see
Hill 1997), as the body was often employed as a vehicle to display aspects of social identity
in the Roman provinces (e.g. Swift 2000; Bartman 2001; Davies 2005; Harlow 2005).
The inclusion of the child/​adolescent age category within formalized urban cemetery
contexts, the increase and variety of grave goods, and, in some contexts at least, the
introduction of gendered items signal a new stage of the life course: that of the ‘potential’
adult. As potential adults, those in this age category represent the future well-​being of
the social group and, as such, are perceived to have a high level of social worth, second
only to prime adults (see, e.g., MacDonald 2001). In the context of the Roman world,
the focus on this age group on epitaphs has been interpreted as representing the sense of
loss, by an early death (mors immatura), felt by kin of this future potential (e.g. Martin-​
Kilcher 2000; Laes 2007; Sigismund-​Nielson 2007), and it is likely that the distinct focus
328

328   Alison Moore

on the child/​adolescent in some Romano-​British cemeteries is a provincial expression

of the same sense of loss in a society that did not embrace the epigraphic habit. Similarly,
if the provision of jewellery items such as bracelets with child/​adolescent burials can be
equated with the female, as Cool (2010: 31–3​2) suggests, then it can be argued that these
items may symbolize their dowries, thereby representing the loss of the potential fertil-
ity of the familial group and wider social community (Oliver 2002; Puttock 2002: 41–​55).

Young/​P rime Adulthood

The third stage of the Romano-​British life course equates with young/​prime adulthood
(18–​39 years) and it is among this age category that a gendered identity is more reliable
to identify. Revell (2005: 57–​58) records a peak in the age distribution of epitaphs among
females in Roman Britain, with approximately 80 per cent of tombstones dedicated to
women up to the age of 40 years, with clustering around the 25–​35 year age range and
the peak age of commemoration at 30 and 35 years. Among the adult age categories at
Lankhills, Gowland (2001: 160; 2002) noted that over 50 per cent of the total quantity of
grave goods were placed with females in their late teens and early twenties and, among
adults, the items of personal adornment were made from more exclusive materials than
those placed with sub-​adults. Within some other urban cemetery contexts a similar
peak in grave-​good provision can be seen in comparison to males of an equivalent age,
and items of material culture related to appearance, dress, or personal ornamentation
were also prevalent in young/​prime-​adult female burials (Table 16.1). For example, at
Colchester (Crummy et al. 1993) 25.4 per cent (15/​59) of burials of young/​prime-​adult
females were provided with grave goods, of which 71.7 per cent (38/​53) were items of
material culture related to appearance. In contrast, only 15 per cent (10/​66) of males of a
similar age were provided with grave goods, predominantly pottery and hobnail shoes.

Table 16.1 Distribution of grave goods with young/​prime adults and percentage

of items related to personal appearance with female burials at four
major urban cemeteries
% of the amount of
material culture
% of male burials with grave % of female burials with related to appearance
Cemetery goods grave goods in female burials

Colchester 15% (10/​66) 25.4% (15/​59) 71.7% (38/​53)

Cirencester 15.2% (7/​46) 21.7% (15/​69) 33% (4/​12)
Poundbury 8.7% (13/​150) 11.8% (23/​192) 48% (24/​50)
York 11.3% (15/​132) 28.1% (9/​32) 36% (4/​11)

Source: © Alison Moore.
 The Life Course   329

The peak in age distribution on epitaphs and the focus on the young/​prime adult
female seen in some of the urban cemeteries suggest that this was another important
stage in the female life course in Roman Britain, one that began in adolescence and
continued through into adulthood. However, rather than representing possible dow-
ries or potential fertility as with older children and adolescents, it may be that the
focus represented a fulfilment of these goals. The primary social roles of the female in
Roman Britain, as in the rest of the empire, were that of wife and mother (Revell 2005:
58; Harlow 2007: 197). Although at what age women in Roman Britain married is open
to debate, it is probable that most women would have been married and bearing chil-
dren by their late 20s (Allason-​Jones 2004: 380–​381; see Shaw 1987 for a discussion of
the probable ages of female marriage during the Roman period). As such, the years of a
female’s young/​prime adulthood and identity would have been defined by her gendered
social role: that of a biologically mature (fertile) adult (Cool 2010).
In contrast to the greater definition in the young/​prime adult female burial assem-
blage, young/​prime adult male age identity appears more complex and ambiguous.
There was no easily recognizable ‘male’ burial assemblage, and, in general, adult males
were buried with a variety of material culture in urban cemeteries; for example, coins
predominated among young/​prime adult males at Cirencester, and hobnail shoes and
pottery at Lankhills and Poundbury. A similar pattern of complexity can be seen in the
age distribution of epitaphs on civilian male tombstones. Here, the pattern of male com-
memoration rises and peaks at 31–​40 years, rather than between 21–​30 years, as seen
in other western provinces, where magistracies and a political career were interpreted
as being important markers in the male life course (Revell 2005: 58). Rather, in Roman
Britain other factors may have defined the male life course. For example, life events such
as the death of the father, marriage, or the birth of the first child could have been impor-
tant transitional markers among young/​prime males. Other factors may also have come
into play. Like the majority of societies before the modern era, Roman Britain was pre-
dominantly patriarchal. Unlike young/​prime adult females who were socially defined
by their biology, domestic roles, and familial relationships (Parkin 2003: 246), the social
identity of the male was more ‘externalized’ and may have been influenced by factors
such as ability to work, status among peers, or position within the community (Moore
2009a: 173), and further studies on masculine identity (see Revell 1999, 2010) are needed
in order to help pinpoint life-​course transitions among non-​elite males in provincial
contexts.

Mature/​Older Adulthood

The fourth visible stage in the life course was that of mature/​older adulthood (40/​50+
years). The general characteristic for this stage of the life course was a greater degree of
ambiguity, in contrast to young/​prime adulthood, in the levels and types of grave goods
provided, but the mature/​older adult life-​course stage remained embedded within the
330

330   Alison Moore

general pattern of adult burials (Moore 2010: 112). Overall, there was a tendency towards
fewer grave goods with advancing age, but there was no defined grave-​good ‘assemblage’
related to mature/​older adults, with jewellery, pottery, coins, hobnail shoes, and ani-
mal bone all featuring within this age category (Moore 2009a, 2010). At certain cem-
eteries gender remained a visible aspect of the mature/​older adult burial pattern. For
example, Gowland (2001: 160–​162) noted a slight emphasis in grave-​good provision for
older males at Lankhills, and a similar pattern can be seen at Poundbury (Woodward
1993), where slightly more males (15.6 per cent or 33/​211 burials) than females (11.7 per
cent or 25/​214 burials) were provided with grave goods. Similarly, Revell’s study (2005:
58) of Romano-​British patterns of commemoration also recorded a slight emphasis
on the older civilian male from 50 years onwards. This focus on the older male within
the epigraphic record and at the cemeteries at Lankhills and Poundbury is often seen
in traditional societies, as he represents the oldest male of the family and the focus of
life experience and learnt knowledge, particularly within tightly knit communities
(Moore 2010: 112). However, the pattern for mature/​older adult burial is more complex
elsewhere, with other sites recording an increase in grave-​good provision among the
mature/​older adults (Table 16.2) and a continued emphasis on the mature/​older female
in levels of grave-​good provision.
Whilst this emphasis on the older female at certain urban cemeteries continues the
focus on the female seen with young/​prime adults, in the context of older age it may
represent changing social and familial relationships. If women were defined primarily
by their gender roles and their relationships to family members, then changes in these
relationships may well have impacted significantly on how older women were perceived
(Moore 2010: 113). For example, on the death of a husband, an older widow may have had
more limited economic power, while generational change—​through the establishment
of new households on the marriage of adult children or the arrival of grandchildren—​
may have removed her from the centre of family life to a more peripheral position
(Harlow 2007:  200; Hin 2007:  4). However, such transitions would not necessarily
have been entirely negative. Such factors could have freed up the older female from the

Table 16.2 Percentage of grave goods with mature/​older male and female burials

at five urban cemeteries
Cemetery % of female burials with grave goods % of male burials with grave goods

Colchester 6.4% (7/​109) 6.1% (4/​66)

Cirencester 33% (14/​42) 26.9% (38/​141)
Eastern London 54.1% (33/​61) 46.1% (18/​39)
Gloucester 35.7% (5/​14) 24% (6/​25)
York 21.5% (4/​19) 18.5% (15/​81)

Source: © Alison Moore.
 The Life Course   331

familial responsibility of raising children while providing her with a new social role as
a grandmother (see also Nielson 1989 on women’s social roles). Therefore, the older
female continued to maintain a social role, which, in certain contexts, was reflected in
the burial record (see Moore 2010 for further discussion).
The ambiguity within the mature/​older adulthood category, both in terms of an unde-
fined burial assemblage and in levels of grave goods between males and females, sug-
gests that, unlike in modern western societies where an established age of retirement
represents the onset of older age, chronological age was not significant (see Revell 2005:
59–​60 for a discussion on ‘age-​rounding’). Rather, archaeologically ‘hidden’ factors
may have been at work, ones that were dependent upon the physicality of older age and
attendant familial, economic, and social circumstances of the individual. While there
are many possible interpretations as to when a person was considered ‘old’ in Roman
Britain, it may be that the ambiguity seen with mature/​older adults in certain contexts
may also be representative of both the changing identity of mourners and the concep-
tual position of older adults within the life course. In many traditional societies, the least
influential in society—​such as the young and old—​were often accorded a more vari-
able or, conversely, a highly conservative burial rite in opposition to the prevailing adult
norm (Crawford 1999: 79–​80; Moore 2009a: 179–1​80). If this argument can be applied
to Roman Britain, the burial of older members of society may reflect more closely ideas
and modes of burial relating to localized concepts of identity and, therefore, hold a key
to help us recognize regional expressions of age patterning in the burial record.

Local and Regional Identities in 

the Life Course

However, the broad four-​stage Romano-​British life course seen at the civilian cemeteries
of Lankhills, Poundbury, and Colchester, and outlined here, was not consistent across
the province but was influenced by regional and local habits of burial and by different
social groups within the province. Even within these cemeteries, how age was expressed
was subject to localized habits. For example, Gowland (2001: 162) identified that brace-
lets tended to be worn in adult burials at Colchester but were placed unworn in child/​
adolescent burials. In contrast, at Lankhills, there was no obvious link between age and
whether bracelets were worn or unworn (Cool 2010: 31, table 3.3). Rather, age differen-
tiation was marked by the large number of bracelets with older children and adolescents
(Cool 2010: 31, table 3.4). Similarly, other forms of material culture—​for example, dif-
ferent pottery forms and types of pottery fabrics—​may have been selected as being ‘age
appropriate,’ dependent upon the availability of regional and local types.
Analysis of age patterning at other urban cemeteries in Roman Britain shows an
emphasis on different life stages over the life course, indicating the possibility of mul-
tiple expressions of age identity in the province. At London, Gloucester, and York, the
332

332   Alison Moore

focus on grave-​good provision was firmly placed on the adult age categories (Figure
16.2), in contrast to the emphasis on the child/​adolescent category seen at cemeteries
such as Lankhills and Poundbury. What may be significant with the expression of the
life course in these cemeteries is the development of their urban milieux as a new type
of urban community (Pearce 1997: 179). London was a major port and trading centre
(Perring 1991; Porter 1995) bringing in people from across the empire, the colonia of
Gloucester retained a dual aspect of Romanized nucleus and suburban area, and, possi-
bly, a dual identity (see Hurst 1999, 2005, for a full discussion), and York was a legionary
base (Addyman 1976). As such, these were contexts in which peoples from various back-
grounds and cultural norms, divorced from their original familial groups and social net-
works, eventually formed and expressed a new concept of age identity, one based around
the adult as the nucleus of the familial structure and tradition (Moore 2009a: 180–​182).
A similar process may be seen in Revell’s study (2005: 58) of age identity from Romano-​
British military tombstones. Here the peak age for male commemoration was between
30 and 50 years, the prime years of military service, rather than after retirement. Thus,
how the life course of the male, regardless of ethnicity, place of origin, or cultural norms,
was defined was primarily by his adopted identity as a soldier.
That multiple life courses were in action in Roman Britain can also be seen on a
regional level. For example, in southern Britain a comparison of late Roman burials
from sites from rural contexts in the west of province with those from the eastern region
shows a slightly different life course (Figure 16.3). Not only were there visible differences

45

40

35
% of total burials with grave goods

30

25

20

15

10

0
0–3 years 4–12 years 13–17 years 18–39 years 40+ years
Age categories

East London Gloucester York

Figure 16.2  Percentage of total burials with grave goods at East London, Gloucester, and York.
Source: © Alison Moore.
 The Life Course   333

60

50
% of total burials with grave goods

40

30

20

10

0
0–3 years 4–12 years 13–17 years 18–39 years 40+ years
Age categories

Western rural Eastern rural

Figure 16.3  Percentage of aged burials with grave goods at rural sites in western and eastern
regions of southern Roman Britain.
Source: © Alison Moore.

in the level of material culture between regions, with all age groups in the eastern area
being provided with more grave goods, but in the western region, broadly representing
the area covered by the modern counties of Gloucestershire, Somerset, and Wiltshire,
the focus in burial is with 13–​17 years, similar to Lankhills and Poundbury. In contrast,
in the eastern region (broadly the area encompassing the modern counties of Essex,
Hertfordshire, Cambridgeshire, and Suffolk) the age category highlighted is child-
hood (4–​12 years), a pattern also reflected at the urban site of Colchester (Figure 16.1).
This patterning may indicate that regional habits of burial and concepts of age iden-
tity retained a strong influence, and further studies need to be undertaken to build up a
more detailed picture of the regionality of the Romano-​British life course.

Conclusions

How age identity, with its attendant age expectations and socio-​cultural norms, shaped
and influenced the lives of people in Roman Britain holds a great deal of potential to
inform us about social structure and organization on the periphery of empire. Although
problems remain, such as osteological issues with assigning correct age and a lack of
consensus among archaeologists regarding age categories, adopting a life-​ course
334

334   Alison Moore

approach to the burial evidence provides a methodological framework by which to

explore the complexities and fluidity of social identity, allowing us to move away from
a reliance on the textual evidence of ‘Rome-​centric’ concepts of age, towards a more
nuanced understanding of provincial life. The studies undertaken so far and outlined
above have indicated that a broad four-​stage life course can be traced at a variety of dif-
ferent sites, consisting of infancy, childhood/​adolescence, young/​prime adulthood, and
older adulthood, and the transitions between each category would have been initiated
by a complex mix of physiological and social factors. Similarly, social structures can be
outlined, with infants and mature/​older adults appearing to hold a conceptually mar-
ginal position, while the treatment in burial of older children and adolescents may be
representative of the sense of loss of the future potential of this age cohort as adults.
However, familial and social organization was centred on the young/​prime adult, rep-
resented primarily by the young/​prime adult female, and the focus on the young/​prime
adult may be associated with a perceived break in the flow of the natural life course. As
an age cohort, the young/​prime adults encapsulated familial continuity through parent-
hood, while on a social level they represent social stability and economic prosperity.
A life-​course framework is an ideal vehicle through which to examine aspects of social
identity in the Roman provinces, as there is a wealth of age-​related evidence within burial
practice and surviving epigraphic sources that needs to be integrated fully in order to
understand the complexities of the provincial life course. In the context of Roman Britain,
the four-​stage life course is visible at sites from across the province, but which age cohorts
are highlighted is dependent upon the context of the site in question, whether urban or
rural, and between geographical areas. This suggests that social identity was primarily
expressed on a local or regional level. However, age identity and the life course are still
a relatively recent area of academic enquiry, particularly in the context of Roman stud-
ies, and further work is needed to provide a more nuanced picture of regional variation,
both in Britain and in the other Roman provinces. In order to facilitate this, a consensus
with regards to ageing and age categories needs to be reached among archaeologists and
a life-​course framework needs to be integrated into future cemetery excavations reports
to build up a strong data set. This will enable questions of regionalism in Roman Britain
and across the empire to be addressed, which, allied with other forms of evidence, will
advance our understanding of the level of continuation of pre-​Roman traditions, the
adoption of Romanized customs, and the expression of social identity.

References
Addyman, P. V. (1976). The Legionary Fortress. Archaeology of York 3. London: Council for
British Archaeology.
Allason-​Jones, L. (2004). ‘The Family in Roman Britain’, in M. Todd (ed.), A Companion to
Roman Britain. Oxford: Blackwell Publishing, 273–​287.
Appleby, J. E. P. (2010). ‘Why we Need an Archaeology of Old Age and a Suggested Approach’,
Norwegian Archaeological Review, 43/​2: 145–​168.
 The Life Course   335

Aykroyd, R., Lucy, D., Pollard, M. A., and Roberts, C. A. (1999). ‘Nasty, Brutish, but not
Necessarily Short: A Reconsideration of the Statistical Methods used to Calculate Age at Death
from Adult Human Skeletal and Dental Age Indicators’, American Antiquity, 64/​1: 55–​70.
Bartman, E. (2001). ‘Hair and the Artifice of Roman Female Adornment’, American Journal of
Archaeology, 105/​1: 1–​25.
Becker, M. J. (1994). ‘Suggrundaria and Roman Villas: A Note on an Infant’s Burial at Satricum,
Lazio, Italy’, Old World Archaeology Newsletter, 17/​2: 7–​10.
Becker, M. J. (1997). ‘Perinatal and Other Burial Patterns in Lucania: Evidence from Collina
San Salvatore at Timmari, Italy, during the 7th–​6th centuries bc’, Old World Archaeology
Newsletter, 20/​3: 1–​12.
Becker, M. J. (2011). ‘Etruscan Infants: Children’s Cemeteries at Tarquinia, Italy, as Indicators of
an Age of Transition’, in M. Lally and A. Moore (eds), (Re)Thinking the Little Ancestor: New
Perspectives on the Archaeology of Infancy and Childhood. BAR International Series 2271.
Oxford: Archaeopress, 24–​36.
Blockley, K. (1985). Marshfield Ironmonger’s Piece Excavations 1982–​ 83:  An Iron-​ Age
and Romano-​ British Settlement in the South Cotswolds. BAR British Series 141.
Oxford: Archaeopress.
Booth, P., and Evans, J. (2001). Roman Alcester:  Northern Extramural Area, 1969–​ 1988
Excavations. CBA Research Report 127/​Roman Alcester Series Volume 3. London: Council
for British Archaeology.
Booth, P., Simmonds, A., Boyle, A., Clough, S., Cool, H. E. M., and Poore, D. (2010). The Late
Roman Cemetery at Lankhills, Winchester: Excavations 2000–​2005. Oxford Archaeology
Monograph 10. Oxford: Oxford Archaeology.
Bradley, K. R. (1991). Discovering the Roman Family. Oxford: Oxford University Press.
Carroll, M. (2012). ‘“No Part in Earthly Things”: The Death, Burial and Commemoration of
Newborn Children and Infants in Roman Italy’, in M. Harlow and L. Larrson Lovén (eds),
Families in the Roman and Late Antique World. London: Continuum, 41–​63.
Cokayne, K. 2003. Experiencing Old Age in Ancient Rome. London: Routledge.
Conkey, M. W., and Spector, J. (1984). ‘Archaeology and the Study of Gender’, in M. B. Schiffer
(ed.), Advances in Archaeological Method and Theory 7. New York: Academic Press, 1–​38.
Cool, H. E. M. (2010). ‘Finding the Foreigners’, in H. Eckardt (ed.), Roman Diasporas:
Archaeological Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman
Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 27–​44.
Cox, M. (2000). ‘Ageing adults from the skeleton’, in M. Cox and S. Mays (eds) Human
Osteology in Archaeology and Forensic Science. London: Greenwich Medical Media Ltd,
61–​81.
Crawford, S. (1999). Children in Anglo-​Saxon England. Stroud: Sutton Publishing.
Crawford, S. (2000). ‘Children, Grave Goods and Social Status in Early Anglo-​Saxon England’,
in J. Sofaer Derevenski (ed.), Children and Material Culture. London: Routledge, 169–​179.
Crummy, N., Crummy, P., and Crossan, C. (1993) (eds). Excavation of Roman and Later
Cemeteries, Churches and Monastic Sites in Colchester, 1978–​1988. Colchester Archaeological
Report 9. Colchester: Colchester Archaeological Trust.
Dasen, V. (2009). ‘Roman Birth Rites of Passage Revisited’, Journal of Roman Archaeology,
22: 199–​214.
Davies, G. (2005). ‘What Made the Roman Toga Virilise?’, in L. Cleland, M. Harlow, and L.
Llewellyn-​Jones (eds), The Clothed Body in the Ancient World. Oxford:  Oxbow Books,
121–​131.
336

336   Alison Moore

Dixon, S. (1992). The Roman Family. Baltimore: Johns Hopkins University Press.

Farwell, D. E., and Molleson, T. I. (1993) (eds). Poundbury, Volume II: The Cemeteries. Dorset
Natural History and Archaeological Society Monograph Series 11. Dorchester: Dorset
Natural History and Archaeological Society.
Finlay, N. (2013). ‘Archaeologies of the Beginning of Life’, World Archaeology, 45/​2: 207–​214.
Finley, M. I. (1989). ‘Introduction: The Elderly in Classical Antiquity’, in T. M. Falkner and J.
de Luce (eds), Old Age in Greek and Latin Literature. Albany, NY: University of New York
Press, 1–​20.
Fraschetti, A. (1997). ‘Roman Youth’, in G. Levi and J. C. Schmitt (eds), A History of Young People
in the West. Volume One: Ancient and Medieval Rites of Passage. Cambridge, MA: Belknap
Press, 51–​82.
Frere, S. S. (1992). ‘Roman Britain in 1991’, Britannia, 22: 296.
Gilchrist, R. (1999). Gender and Archaeology: Contesting the Past. London: Routledge.
Gilchrist, R. (2000). ‘Archaeological Biographies: Realizing Human Lifecycles, -​Courses and
-Histories’, World Archaeology, 31/​3, Human Lifecycles, 325–​328.
Gilchrist, R. (2012). Medieval Life: Archaeology and the Life Course. Woodbridge: Boydell Press.
Ginn, J., and Arber, S. (1995). ‘“Only Connect”: Gender Relations and Ageing’, in S. Arber and
J. Ginn (eds), Connecting Gender and Ageing: A Sociological Approach. Buckingham: Open
University Press, 1–​14.
Gowland, R. (2001). ‘“Playing Dead”: Implications of Mortuary Evidence for the Social
Construction of Childhood in Roman Britain’, in G. Davies, A. Gardiner, and K. Lockyear
(eds), TRAC 2000. Proceedings of the Tenth Theoretical Roman Archaeology Conference,
London 2000. Oxford: Oxbow Books, 152–​168.
Gowland, R. (2002). ‘Examining Age as an Aspect of Social Identity in Fourth to Sixth Century
England:  The Analysis of Mortuary Evidence’. Unpublished Ph.D.  thesis, University of
Durham.
Gowland, R. (2006). ‘Ageing the Past: Examining Age Identity from Funerary Evidence’, in
R. Gowland and C. Knüsel (eds), Social Archaeology of Funerary Remains. Oxford: Oxbow
Books, 141–​154.
Graham, E.-​J. (2013). ‘The Making of Infants in Hellenistic and Early Roman Italy: A Votive
Perspective’, World Archaeology, 45/​2: 215–​231.
Gray, H. St G. (1930). ‘Excavations at Kingsdown Camp, Mells, Somerset, 1927–​9’, Archaeologia,
80: 59–​98.
Hänninen, M.-​L. (2005). ‘From Womb to Family: Rituals and Social Conventions Connected
to Roman Birth’, in K. Mustakallio, J. Hanska, H.-​L. Saino, and V. Vuolanto (eds), Hoping for
Continuity: Childhood, Education and Death in Antiquity and the Middle Ages. Rome: Acta
Instituti Romani Finlandiae 33, 49–​59.
Hareven, T. K. (1978). ‘Introduction: The Historical Study of the Life Course’, in T. K. Hareven
(ed.), Transitions: The Family and the Life Course in Historical Perspective. London: Academic
Press, 1–​16.
Harlow, M. (2005). ‘Dress in the Historia Augusta: The Role of Dress in Historical Narrative’, in
L. Cleland, M. Harlow, and L. Llewellyn-​Jones (eds), The Clothed Body in the Ancient World.
Oxford: Oxbow Books, 113–​120.
Harlow, M. (2007). ‘Blurred Visions: Male Perceptions of the Female Life Course—​the Case
of Aemilia Pudentilla’, in M. Harlow and R. Laurence (eds), Age and Ageing in the Roman
Empire. Journal of Roman Archaeology Supplementary Series 65. Portsmouth, RI: Journal
of Roman Archaeology, 195–​207.
 The Life Course   337

Harlow, M., and Laurence, R. (2002). Growing Up and Growing Old in Ancient Rome: A Life
Course Approach. London: Routledge.
Harlow, M., and Laurence, R. (2008). ‘The Representation of Age:  Towards a Life Course
Approach’, in P. P. A. Funari, R. S. Garrafuni, and B. Lelalien (eds), New Perspectives on the
Ancient World: Modern Perceptions, Ancient Representations. BAR International Series 1782.
Oxford: Archaeopress, 205–​212.
Hill, J. D. (1997). ‘The End of One Kind of Body and the Beginning of Another Kind of Body?
Toilet Instruments and “Romanization” in Southern England during the First Century ad’,
in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies: New Approaches to the
British Iron Age. Oxbow Monograph 71. Oxford: Oxbow Books, 97–​107.
Hin, S. (2007). Family Matters: Economy, Culture and Biology. Fertility and its Constraints in
Roman Italy. Princeton: Stanford Working Papers in Classics.
Hopkins, K. (1966). ‘On the Probable Age Structure of the Roman Population’, Population
Studies, 20/​2: 245–​264.
Hurst, H. (1999). ‘Topography and Identity in Glevum colonia’, in H. Hurst (ed.), The Coloniae
of Roman Britain: New Studies and a Review. Papers of the Conference held at Gloucester 5–​6
July, 1997. Journal of Roman Archaeology Supplementary Series 36. Portsmouth, RI: Journal
of Roman Archaeology, 153–​160.
Hurst, H. (2005). ‘Roman Cirencester and Gloucester Compared’, Oxford Journal of
Archaeology, 24/​3: 293–​305.
Karl, R., and Löcker, K. (2011). ‘Thrown out with the Bathwater or Properly Buried? Neonate and
Infant Skeletons in a Settlement Context on the Durrnberg bei Hallein, Austria’, in M. Lally
and A. Moore (eds), (Re)Thinking the Little Ancestor: New Perspectives on the Archaeology of
Infancy and Childhood. BAR International Series 2271. Oxford: Archaeopress, 37–​46.
Kamp, K. A. (2001). ‘Where Have All the Children Gone? The Archaeology of Childhood’,
Journal of Archaeological Method and Theory, 8/​1: 1–​34.
Laes, C. (2007). ‘Inscriptions from Rome and the History of Childhood’, in M. Harlow and
R. Laurence (eds), Age and Ageing in the Roman Empire. Journal of Roman Archaeology
Supplementary Series 65. Portsmouth, RI: Journal of Roman Archaeology, 25–​36.
Lally, M. (2008). ‘Bodies of Difference in Iron Age Southern England’, in O. Davies, N. Sharples,
and K. Waddington (eds), Changing Perspectives on the First Millennium bc. Oxford: Oxbow
Books, 119–​138.
Lally, M., and Ardren, T. (2008). ‘Little Artefacts: Rethinking the Constitution of the
Archaeological Infant’, Childhood in the Past, 1: 62–​77.
Lillehammer, G. (1989). ‘A Child is Born: The Child’s World in an Archaeological Perspective’,
Norwegian Archaeological Review, 22/​2: 89–​105.
Lucy, S. (2007). ‘The Archaeology of Age’, in M. Diaz-​Andreu, S. Lucy, S. Babić, and D. N.
Edwards (eds), The Archaeology of Identity: Approaches to Gender, Age, Status, Ethnicity and
Religion. London: Routledge, 43–​66.
MacDonald, D. H. (2001). ‘Grief and Burial in the American Southwest:  The Role of
Evolutionary Theory in the Interpretation of Mortuary Remains’, American Antiquity, 66/​
4: 704–​7 14.
McKinley, J. (2000). ‘The Analysis of Cremated Bone’, in M. Cox and S. Mays (eds), Human
Osteology in Archaeology and Forensic Science. London: Greenwich Medical Media, 403–​421.
McKinley, J. (2004). ‘The Human Remains’, in H. E. Cool, The Roman Cemetery at Brougham,
Cumbria: Excavations 1966–​67. Britannia Monograph Series 21. London: English Heritage,
283–​309.
338

338   Alison Moore

McLaren, D. (2011). ‘Where have All the Flowers Gone? Bronze Age Children’s Burials in
South-​East England: Initial Thoughts’, in M. Lally and A. Moore (eds), (Re)Thinking the Little
Ancestor: New Perspectives on the Archaeology of Infancy and Childhood. BAR International
Series 2271. Oxford: Archaeopress, 85–​99.
Marquez-​Grant, N., and Loe, L. (2008). ‘The Human Remains’, in A. Simmonds, N. Marquez-​
Grant, and L. Loe (eds), Life and Death in a Roman City: Excavation of a Roman Cemetery
with a Mass Grave at 120–​122 London Road, Gloucester. Oxford Archaeology Monograph 6.
Oxford: Oxford Archaeology, 29–​32.
Martin-​Kilcher, S. (2000). ‘Mors immatura in the Roman World: A Mirror of Society and
Tradition’, in J. Pearce, M. Millett, and M. Strück (eds), Burial, Society and Context in the
Roman World. Oxford: Oxbow Books, 63–​77.
Mays, S. (1993). ‘Infanticide in Roman Britain’, Antiquity, 67: 883–​888.
Mays, S. (2000). ‘The Archaeology and History of Infanticide, and its Occurrence in
Earlier British Populations’, in J. Sofaer Derevenski (ed.), Children and Material Culture.
London: Routledge, 180–​190.
Mays, S., and Cox, M. (2000). ‘Sex Determination in Skeletal Remains’, in M. Cox and S. Mays
(eds), Human Osteology in Archaeology and Forensic Science. London: Greenwich Medical
Media, 117–​130.
Mays, S. A., and Steele, J. (1999). ‘The Human Bone’, in R. Niblett, The Excavation of a
Ceremonial Site at Folly Lane, Verulamium. Britannia Monograph Series 14. London: Society
for the Promotion of Roman Studies, 299–​324.
Meskell, L. (2000). ‘Cycles of Life and Death:  Narrative Homology and Archaeological
Realities’, World Archaeology, 31/​3: 324–​441.
Molleson, T. I. (1999). ‘Archaeological Evidence for Attitudes to Disability in the Past’,
Archaeological Review from Cambridge, 15: 69–​72.
Molleson, T. I., and Cox, M. (1993). The Spitalfields Project. Volume 2: The Anthropology. CBA
Research Report 86. York: Council for British Archaeology.
Montserrat, D. (1996). Sex and Society in Greco-​ Roman Egypt. London:  Kegan Paul
International.
Moore, A. (2009a). ‘Young and Old in Roman Britain:  Aspects of Age Identity and Life-​
Course Transitions in Regional Burial Practice’. Unpublished Ph.D.  thesis, University of
Southampton.
Moore, A. (2009b). ‘Hearth and Home: The Burial of Infants within Romano-​British Domestic
Contexts’, Childhood in the Past, 2: 33–​54.
Moore, A. (2010). ‘Age and Identity in Funerary Contexts: The Elderly in Southern Roman
Britain’, in A. Moore, G. Taylor, E. Harris, P. Girdwood, and L. Shipley (eds), TRAC
2009:  Proceedings of the 19th Theoretical Roman Archaeology Conference, Michigan and
Southampton 2009. Oxford: Oxbow Books, 105–​119.
Moore, J., and Scott, E. (1997) (eds). Invisible People and Processes:  Writing Gender and
Childhood into European Archaeology. Leicester: Leicester University Press.
Nielson, M. (1989). ‘Women and Family in a Changing Society: A Quantitative Approach to
Late Etruscan Burials’, Analecta Romana Instituti Danici, 17–​18, 53–​89.
Oliver, A. (2002). ‘Jewellery for the Unmarried’, in D. Kleiner and S. Matheson (eds), I Claudia
II: Women in Roman Art and Society. Austin, TX: University of Texas Press, 115–​124.
Parkin, T. G. (1992). Demography and Roman Society. Baltimore:  Johns Hopkins
University Press.
 The Life Course   339

Parkin, T. G. (2003). Old Age in the Roman World: A Cultural and Social History. Baltimore and
London: Johns Hopkins University Press.
Pearce, J. (1997). ‘Death and Time: The Structure of Late Iron Age Mortuary Ritual’, in A. Gwilt
and C. Haselgrove (eds), Reconstructing Iron Age Societies: New Approaches to the British
Iron Age. Oxford: Oxbow Books, 174–​180.
Pearce, J. (1999). ‘The Dispersed Dead: Preliminary Observations on Burial and Settlement
Space in Rural Roman Britain’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC
98:  Proceedings of the Eighth Theoretical Roman Archaeology Conference, Leicester 1998.
Oxford: Oxbow Books, 151–​162.
Pearce, J. (2001). ‘Infants, Cemeteries and Communities in the Roman Provinces’, in G. Davies,
A. Gardner, and K. Lockyer (eds), TRAC 2000: Proceedings of the Tenth Theoretical Roman
Archaeology Conference, London 2000. Oxford: Oxbow Books, 125–​142.
Penn, W. S. (1960). ‘Springhead Temples III and IV’, Archaeologia Cantiana, 74: 113–​140.
Perring, D. (1991). Roman London. London: Seaby.
Porter, R. (1995). London: A Social History. Cambridge, MA: Harvard University Press.
Puttock, S. (2002). Ritual Significance of Personal Ornament in Roman Britain. BAR British
Series 327. Oxford: Archaeopress.
Revell, L. (1999). ‘Constructing Romanitas: Roman Public Architecture and the Archaeology
of Practice’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC 98: Proceedings of the
Eighth Annual Theoretical Roman Archaeology Conference, Leicester 1998. Oxford: Oxbow
Books, 52–​58.
Revell, L. (2000). ‘The Creation of Multiple Identities in Roman Italica’, in G. Fincham, G.
Harrison, R. Holland, and L. Revell (eds), TRAC 99:  Proceedings of the Ninth Annual
Theoretical Roman Archaeology Conference, Durham 1999. Oxford: Oxbow Books, 1–​9.
Revell, L. (2005). ‘The Roman Life Course: A View from the Inscriptions’, European Journal of
Archaeology, 8/​1: 43–​63.
Revell, L. (2010). Romanization:  A  Feminist Critique’, in A. Moore, G. Taylor, E. Harris, P.
Girdwood, and L. Shipley (eds), TRAC 2009: Proceedings of the Nineteenth Annual Theoretical
Roman Archaeology Conference, Michigan and Southampton 2009. Oxford:  Oxbow
Books, 1–​10.
Saller, R. P. (1987). ‘Men’s Age at Marriage and its Consequences in the Roman Family’, Classical
Philology, 82/​1: 21–​34.
Saller, R. P., and Shaw, B. D. (1984). ‘Tombstones and Roman Family Relations in the
Principiate: Civilians, Soldiers and Slaves’, Journal of Roman Studies, 74: 124–​156.
Scheidel, W. (2001). ‘Roman Age Structure: Evidence and Models’, Journal of Roman Studies,
91: 1–​26.
Scheuer, L., and Black, S. (2000). ‘Development and Ageing of the Juvenile Skeleton’, in M.
Cox and S. Mays (eds), Human Osteology in Archaeology and Forensic Science. London:
Greenwich Medical Media, 9–​22.
Scott, E. (1991). ‘Animal and Infant Burials in Romano-​ British Villas: A Revitalization
Movement’, in P. Garwood, D. Jennings, R. Skeates, and J. Toms (eds), Sacred and Profane:
Proceedings of a Conference on Archaeology, Ritual and Religion, Oxford 1989. Oxford
Committee for Archaeology Monograph 32. Oxford: Oxbow Books, 115–​121.
Scott, E. (1995). ‘Women and Gender Relations in the Roman Empire’, in P. Rush (ed.),
Theoretical Roman Archaeology:  Second Conference Proceedings. Aldershot:  Avebury,
174–​189.
340

340   Alison Moore

Scott, E. (1997). ‘Introduction:  On the Incompleteness of Archaeological Narratives’, in J.

Moore and E. Scott (eds), Invisible People and Processes: Writing Gender and Childhood in
European Archaeology. Leicester: Leicester University Press, 1–​12.
Scott, E. (1999). The Archaeology of Infancy and Infant Death. Oxford: BAR International Series
819. Oxford: Archaeopress.
Shaw, B. D. (1987). ‘The Age of Roman Girls at Marriage: Some Reconsiderations’, Journal of
Roman Studies, 77: 30–​46.
Sigismund-​Nielson, H. (2007). ‘Children for Profit and Pleasure’, in M. Harlow and R. Laurence
(eds), Age and Ageing in the Roman Empire. Journal of Roman Archaeology Supplementary
Series 65. Portsmouth, RI: Journal of Roman Archaeology, 37–​54.
Sofaer, J. (2011). ‘Towards a Social Bioarchaeology of Age’, in S. C. Agarwal and B. A. Glencross
(eds) Social Bioarchaeology. Chichester: Blackwell Publishing, 285–​311.
Sofaer Derevenski, J. (1994). ‘Where are the Children? Accessing Children in the Past’,
Archaeological Review from Cambridge, 13/​2: 7–​20.
Sofaer Derevenski, J. (1997). ‘Age and Gender at the Site of Tiszapolgar-​Basantanya, Hungary’,
Antiquity, 71/​274: 875–​888.
Sofaer Derevenski, J. (2000). ‘Rings of Life: The Role of Early Metalwork in Mediating the
Gendered Life Course’, World Archaeology, 31/​3, Human Lifecycles, 389–​406.
Sørensen, M.-​L. (2000). Gender Archaeology. Cambridge: Polity Press.
Stoodley, N. (1999). The Spindle and the Spear:  A  Critical Enquiry into the Construction
and Meaning of Gender in the Early Anglo-​Saxon Burial Rite. BAR British Series 288.
Oxford: Archaeopress.
Stoodley, N. (2000). ‘From the Cradle to the Grave: Age Organisation and the Early Anglo-​
Saxon Burial Rite’, World Archaeology, 31/​3, Human Lifecycles, 456–​472.
Swift, E. (2000). Regionality in Dress Accessories in the Late Roman West. Instrumentum
Monograph 11. Montagnac: Instrumentum.
Watts, D. J. (1989). ‘Infant Burials and Romano-​British Christianity’, Archaeological Journal,
46: 372–​383.
Welinder, S. (2001). ‘The Archaeology of Old Age’, Current Swedish Archaeology, 9: 163–​178.
Wiedemann, T. (1989). Adults and Children in the Roman Empire. London: Routledge.
Woodward, A. B. (1993). ‘Discussion’, in D. E. Farwell and T. I. Molleson (eds), Poundbury,
Volume II: The Cemeteries. Dorset Natural History and Archaeological Society Monograph
Series 11. Dorchester: Dorset Natural History and Archaeological Society, 215–​239.
Woodward, A., and Leach, P. (1993). The Uley Shrines:  Excavation of a Ritual Complex on
West Hill, Uley, Gloucestershire, 1977–​ 9. English Heritage Archaeological Report 17.
London: British Museum Press.
Wyke, M. (1994). ‘Woman in the Mirror: The Rhetoric of Adornment in the Roman World’, in
L. J. Archer, S. Fischler, and M. Wyke (eds), Women in Ancient Societies: ‘An Illusion of the
Night.’ London: Macmillan, 134–​151.
 Chapter 17

Statu s and Bu ria l

John Pearce

Introduction

A few years after the invasion initiated by the emperor Claudius, a mortuary complex
was created on the hillside across the river from the nascent city at Verulamium on a
scale impressive by any standard in the ancient world (see also Wallace, this volume).
A ditch up to 3 metres deep and 6 metres wide demarcated a 2-​hectare enclosure in
which, after the likely display of a corpse in a burial chamber at its centre, a spectacular
cremation took place. Its residues, buried close to the site of the chamber, included
burnt and broken fragments of many amphorae and ceramic finewares from Britain
and Gaul, molten silver and bronze, including fragments of a carnyx, horse harness,
and vehicle fittings, chain mail, ivory-​inlaid furniture, and animal bones including
claws from a bear skin. In contrast, a mass burial pit from a cemetery north-​east of
Gloucester, dug a century or so later, contained the remains of more than ninety indi-
viduals, dumped without apparent ceremony and with scarcely a single object among
them (Niblett 1999; Simmonds et al. 2008). If we leave aside specific questions these
discoveries raise—​the identity of the individual cremated at Folly Lane, for example, or
the possibility of an epidemic at Gloucester—​the contrast starkly evokes the inequali-
ties of life in the province. Yet burial has rarely been used to inform the study of social
hierarchy in Roman Britain. Outside the garrison communities, epitaphs were not in
common use, depriving us of a source of information crucial for understanding social
structures in other provinces. As objects that signify a particular status or office were
also rarely buried with the dead, funerary evidence has added little to the typical model
of Romano-​British social structure, with its narrow apex of elites, and the mass of the
population, including free farmers, urban artisans, and a larger group of lower-​status
individuals, including dependent labourers and slaves (Mattingly 2006; cf. Roymans
and Derks 2011:  27, fig.  9). The character of grave furnishing, typically comprising
small numbers of artefacts in common circulation, also contrasts with the wealth dis-
played in public or domestic architecture. The impression that social status is not often
342

342   John Pearce

strongly articulated in death rites is epitomized in an observation related to the fourth-​

century ad cemetery at Lankhills, Winchester: ‘in none of the burials was the practice
of burying grave goods with the dead used as an opportunity to display the wealth of
the deceased or their mourners. The majority of the grave goods… were of little finan-
cial value’ (Booth et al. 2010: 517).
Scholarship on Roman cemeteries has traditionally paid more heed to the cultural
or religious identities of the dead than to status or rank, burials being central to debates
over pagan or Christian affiliation or population origins and cultural identity (Morris
1992). Recent work has accommodated these long-​standing interests within a theoretical
framework that characterizes burial rites as performative—​that is, where the religious
or cultural identities of the deceased (perhaps) and, more importantly, of the mourners
are asserted and thus reproduced through their enactment (Petts 2003; Eckardt 2010;
see also Petts and Eckardt, this volume). After acknowledging key methodological and
theoretical caveats, this chapter explores status from the same perspective, and consid-
ers how it may have been marked out in a funerary setting. It would be curious were
the representation of differences of rank, wealth, and so on not to be a central element
of funerary behaviour in a world otherwise ‘fundamentally characterized by a strik-
ing predilection for establishing acute social hierarchies’ (Peachin 2011: 22). Study of
Roman funerary practice in continental Europe also offers a useful shorthand category,
the sépulture privilegiée, for Roman period burials marked out by the scale, complexity,
symbolic richness, and expense in ritual process and tomb markers (Crowley 2011: 195).
After considering approaches to status, the following pages explore distinctions of this
type among the Romano-​British dead. Except where they give insights into rituals,
human skeletal remains will not be considered here, although osteological and isotopic
analyses show considerable promise for mapping inequality (e.g. Redfern and DeWitte
2011; Pitts and Griffin 2012; Müldner 2013; see also Gowland, this volume). The focus of
this chapter lies therefore on the representation of difference through ritual, rather than
difference in lifelong experience as laid down in osteological and biochemical markers.

Approaches to Status

Despite reservations of the type outlined above, burial evidence has been intermittently
considered in relation to status hierarchies. The province’s elites, principally the landed
aristocrats who form the mainstay of the curial class, are sometimes argued to be distin-
guished by a special treatment of the corpse, abundance of grave goods, or substantial
monuments (e.g. Philpott 1991: 229–​232; Strück 2000: 85–​87; Mattingly 2006: 377–​378).
Variability in ritual, as expressed through numbers or types of grave goods or differing
containers for the dead, has also been mapped in order to derive insights into wider
social structure. Through multivariate analysis, for example, Jones identified groupings
of graves within burial samples from communities from Britain and beyond. These he
initially related to groups of differing social status, although in later work he gave greater
 Status and Burial   343

weight to cultural differences (Jones 1983, 1987, 1991). Rural samples have been shown by
numbers of grave goods and more complex statistical methods to be more unequal than
urban in their furnishing or other characteristics such as burial containers (e.g. Millett
1987; Pitts and Griffin 2012). Studies of single cemeteries also suggest complex varia-
bility within burying communities. At the King Harry Lane cemetery at St Albans, for
example, Millett (1993) found that in the decades before and after the Claudian conquest
the average quantity of grave goods per burial declined over time, especially among
interments outside the ditch-​demarcated enclosures at the cemetery’s centre. This he
interpreted as indicating the disruption of social networks consequent on urbanization,
less-​well-​furnished burials being those where fewer people participated in the funeral.
In the third-​century cremation cemetery of the garrison community at Brougham, the
number and type of objects placed on pyre and grave showed substantial differentiation,
which also related to the age and sex of the deceased (Cool 2004). At different scales and
across different settings, urban, rural, military, these studies demonstrate significant
variability in the material investment in the burial ceremony, indicating more complex
status gradations than an elite: non-​elite dichotomy. It is, however, essential to discuss
the key methodological and theoretical caveats that qualify the conclusions that may be
drawn from studies of this kind.
The relationship of excavation to modern economic development, especially in
recent decades, means a bias in our knowledge of burial practice by region and site
type. South-​east and urban Britain predominate in the excavated data in general and
in the samples used for the particular analyses referenced in the previous paragraph,
but no single region furnishes burial data distributed extensively across different settle-
ment types. This compromises study of a society where social hierarchies are unlikely
to be fully represented in the cemeteries of a single community. In general, late Roman
cemeteries comprise the majority of published examples (Pearce 2013:  23–​26). What
survives of burial is also a fraction of its original material residues, reduced by natural
and human agents from the Roman period onwards. With rare exceptions, decomposi-
tion has meant the loss of burial containers, objects, and markers in organic materi-
als. Indignities reported for individual burials, such as the reuse of a lead coffin liner to
repair a gas main in Sittingbourne or the setting of a gin trap within an emptied stone
coffin at Dartford, illuminate an otherwise silent destruction through urban expansion,
agriculture, and so on (Payne 1893; Herbert 2011). Early excavations often treated cem-
eteries as quarries for complete objects, leaving little more than the disiecta membra of
assemblages—​for example, from nineteenth-​and early twentieth-​century excavations
at Colchester (May 1930). Pessimism should not be overstated, since the sample of well-​
studied and published data is one of the largest for any Roman province, but one par-
ticular casualty of post-​depositional attrition is the evidence from cemetery surfaces,
pyre sites, or debris from funerary meals, for example. A focus on what survives of inter-
ment may not give a representative impression of funerary rituals; Roman period prac-
tice leads us reasonably to anticipate an extended sequence of rituals preceding and/​
or following this act in Britain too, where the resources of participants allowed (Scheid
2005: 161–​188; see also Weekes, this volume).
344

344   John Pearce

The nature of burial practice, in particular the diversity of tradition before the fourth
century ad, also hinders extensive quantified comparisons of the types favoured in
burial analysis (cf. Härke 2014). Indeed, archaeologically visible burials, whether cre-
mation or inhumation, may not be fully representative of practice in the province. In
much of southern and central rural Roman Britain, few burials of pre-​fourth century ad
are documented, despite good conditions for preservation of human bone and exten-
sive large-​scale excavations, suggesting continuity of an archaeologically invisible bur-
ial form from the Iron Age for much of the population (Pearce 2013: 24–​26). Among
excavated burials, the mosaic of practices inherited from the pre-​conquest period was
complemented by rituals introduced after conquest, especially around towns and gar-
risons. Below the broad-​brush variation—​for example, the more generous furnishing of
graves in south-​east Britain or greater frequency of epitaphs in garrison communities—​
more localized differences in burial traditions are also emerging. More than a decade
ago the diversity of rites, containers, and objects mapped from the cemeteries of Roman
London, for example, frustrated classification of excavated burials in relation to status
hierarchies (Barber and Hall 2000: 116–​118); since then those rituals so far documented
have grown ever more diverse, now including apparent use of the Walbrook and its trib-
utary streams as agents of secondary burial for bodies buried on their banks (Merrifield
and Hall 2008: 126–​127). In Roman Winchester’s final decades the evidence from dif-
ferent cemeteries shows a striking aversion to placing grave goods in burials away from
Lankhills, despite an otherwise close similarity of rituals (Booth et al. 2010; Ottaway et
al. 2012).
Even if such problems are allowed for, work in funerary archaeology from a post-​
processual viewpoint makes clear that quantitative or qualitative differences in burial
cannot be directly mapped onto deceased individuals’ relative positions within status
hierarchies, legally defined or otherwise, which organize the living. Such an approach
neglects the contextual meaning of the objects and treatments applied to the dead and
the capacity for manipulation of burial rites, especially where death requires politi-
cal roles or control of major resources to be renegotiated (Parker Pearson 1999; Pearce
2000). The mortuary complexes created around the time of the Claudian conquest
such as Folly Lane (Verulamium) and Stanway (Colchester), associated perhaps with
the manœuvres of dynasties glimpsed in Iron Age coin legends as they accommodated
themselves to Roman power, or the funeral of Septimius Severus at York, reported by
Herodian as a feigned display of concord with his brother on the part of Caracalla, pro-
vide potential examples from the province of death rites put to just such ends (Creighton
2006; Herodian 3.15). Freer from the constraints dependent on political and legal status
that affected public spaces, burial also lent an opportunity for the self-​representation by
individuals and groups whose participation in other arenas was affected by their mar-
ginal status, especially freedmen (Mouritsen 2011: 281–​289). The latter are rarely docu-
mented in Britain, but the massive stelae set up here and in other northern provinces
by auxiliary cavalrymen, to commemorate their comrades, illustrate a similar phenom-
enon. The portraits and texts celebrated feats of arms and horsemanship of individuals
who, despite their service and relative wealth, did not possess, until discharge, rights
 Status and Burial   345

of citizenship enjoyed by their legionary counterparts. Notwithstanding the martial

identity expressed by motifs drawn from Roman triumphal art—​for example, the subju-
gated or exterminated barbarian, naked and shaggy, embodying the antithesis to Greco-​
Roman culture—​these soldiers themselves had been recruited among societies only
recently made part of the Roman Empire (Hope 1997; Coulston 2007; see also Hope, this
volume).
From such a perspective, death rites and monuments provide representations of the
identity of the deceased, conditioned by the relationship to them of the actors respon-
sible for burial and commemoration. Such representations are situated within the
continuous construction of social identities in the routines of domestic and commu-
nal life (Revell 2009; Gardner 2013). In its assembly of a wider community, depending
on circumstances, the funeral provided a context for the provision of hospitality in
rituals that are focused on the structured consumption of food and drink. The treat-
ment of the corpse is often the visible medium by which participants witness the
transformations effected at the funeral, integrating the mourners again with the liv-
ing and the deceased with the dead and the ancestors, again in a Roman setting and
beyond (Scheid 2005: 161–​188; Fowler 2013). Study of Bronze and Iron Age burials
has emphasized the powerful impression made on participants by the presentation
of the corpse associated with prestigious objects, for example, related to feasting or
warfare, enhanced by the special resonance such objects possessed owing to their his-
tories of creation, exchange, and use within the group in mourning (Barrett 1993: 115–​
120; Treherne 1995; Gosden and Garrow 2012: 196). The archaeological evidence from
Roman Britain obliges a focus on these ‘snapshots’ in the acts that preceded burial, the
containers for human remains, and the objects placed with the dead. The contextual
data for the Roman period also inform both an appreciation of objects and substances
used in funerary rituals as a material resource and their symbolic resonance. Only the
tomb marker manifested a more enduringly visible sign of commemoration. While
epitaphs and figural representations are rare, especially away from garrisons, other
types of monument allow some exploration of these more lasting representations
through examination of form and setting.

A Death less Ordinary

Pre-​Interment Rituals for Cremation and Inhumation

Our inability to observe the full sequence of ritual is often used to qualify the conclu-
sions to be drawn concerning social status from burial. While textual descriptions of
funerals are generally lacking, archaeological evidence gives some insights into prac-
tices preceding interment. This is more commonly the case for cremation, where osteo-
logical study of cremated human bone, combined with attention to burnt and broken
residues from the pyre, especially of animal and plant remains, has transformed the
346

346   John Pearce

study of burial ritual. Cremation itself is a resource-​intensive ceremony, requiring very

substantial amounts of fuel and tending of the pyre for the conversion of the corpse to
the white calcined bone seen in most Romano-​British cremations (McKinley 2013). In
most examples, scraps of animal bone, molten glass, or metal, and occasional sherds or
nails, are all that remain in the small quantities of clean cremated bone formally buried
(e.g. Evans and Maynard 1997; Barber and Bowsher 2000: 60–​76; Pooley et al. 2011: 153–​
158, 1128–​1131). While the Folly Lane cremation described in the opening paragraph
remains unparalleled in Britain, for the following centuries other pyre assemblages are
documented that also stand out against this background.
The fragments of bone inlay documented in a conquest period burial west of the
Colchester colonia represent one of the few other instances of an elaborately decorated
bier being burnt on the pyre (Eckardt 1999). Dating a half century or so later, a probable
bustum sited close to Watling Street, where it entered Southwark, contained rich plant
material, including fig, date, and almond and stone pine cone as well as abundant cereal
remains and burnt chicken (Mackinder 2000: 11–​12). In the remnant of a third-​century
cremation burial associated with a tower tomb west of the Balkerne gate at Colchester,
ceramics, molten glass, copper alloy, and a bone pyxis were complemented by an exten-
sive animal bone assemblage, including a falcon and several other birds, as well as
domesticates (Brooks 2006). In a contemporary rural example from Holborough, Kent,
deposits by the burial included sherds from at least five amphorae, broken and burnt
tablewares and glass, animal bone, and a folding iron chair on which the deceased had,
perhaps, been cremated (Jessup 1954). Pyre residues documented in the contemporary
burials and related deposits of the garrison community at Brougham, Cumbria, also
included abundant animal bone (in at least one instance remnants from an entire horse
carcass), ceramic, glass and metal vessels, personal ornament in silver, copper alloy,
and glass, bone veneer from caskets, as well as military equipment (Cool 2004). As it is
rarely possible to distinguish the precise role of such objects in the ceremony, whether
placed with the dead or used in the ceremony and then destroyed on the pyre, this mate-
rial offers an aggregate view of resources and objects displayed and consumed prior to
interment. Little documented as yet in Britain are the assemblages related to larger-​scale
funerary feasting recorded elsewhere is northern Europe, exemplified by the contents
of a mid-​first-​century ad fosse à offrandes from a cemetery at Mécleuves, south of Metz,
containing sherds from at least fifty-​three amphorae and 8 kilos of burnt animal bone
from many pigs and other animals, including hare and fish (Ancel 2012: 69–​70).
Similar illumination for inhumation is rarer, but recent and older finds where organic
material has been preserved allow some insights into the preparation and likely display
of the corpse. Some discoveries of this type reported by antiquarians tantalize with their
reference to exceptional preservation (and lost analytical opportunities), like the follow-
ing account of the 1797 opening of a stone coffin at Dartford:

When the coffin was first opened the hair appeared a light brown colour, appar-
ently clubbed on the crown of the head, and fastened with a brooch or bandeau of
pearls; but in a few moments the whole fell to dust… The body had been swathed in
 Status and Burial   347

linen, some of which was visible and covered with cement. A coating of gum strongly
adhered to the larger bones, which retained an aromatic smell… (J. Dunkin, History
and Antiquities of Dartford (1844), 91 ff.; quoted by Haverfield et al. 1932: 89)

The ‘cement’ described here makes this a likely example of a ‘plaster burial’, a widespread
but mostly urban phenomenon where white minerals of varied types (chalk in London,
elsewhere also lime and gypsum) were deposited within the coffin, sometimes inter-
leaved with the wrapping of the corpse to encase it (Philpott 1991: 438–​439; Sparey Green
2003). In one instance from Colchester this ritual demonstrably took place in the cem-
etery itself (Crummy et al. 1993: 36). Other aspects of pre-​interment preparation have
also been illuminated where anaerobic conditions pertain, revealing behaviour of which
little or no trace would otherwise survive. Scraps of silk and gold thread are attested
among textiles in late Roman burials in London, Colchester, and elsewhere, while at
Dorchester a child in a lead coffin liner was dressed in a woollen tunic with clavi (verti-
cal stripes) dyed with Tyrian purple (Swain and Roberts 2001; Davies et al. 2002: 133–​135,
158–​159; Booth et al. 2010: 517–​518). Two pairs of shoes in a stone coffin at Boscombe
Down, one of laced calfskin and the other cork-​soled, perhaps lined with deerskin,
contrast with the robust footwear represented by hobnails in other burials (Pearce et
al. 2008: 134–​135). Plaster burials at York and Dorchester preserved elaborate coiffures
(RCHME 1962: 79; Green et al. 1981; Farwell and Molleson 1993: 205–​206). Occasional
finds such as bay leaves indicate the use of plant material in ritual (Swain and Roberts
2001). Molecular analysis has shown the use of mastic, a resin of east Mediterranean or
Levantine origin, in a second-​century child burial from Cambridgeshire, and of similar
materials in other cases (Taylor 1993: 207–​208; Brettell et al. 2013; Brettell, pers. comm.).
The description of the Dartford discovery, and others like it, may not then be as fanciful
as they might seem. Whether through elaborate dress or treatment that embellished or
maintained the integrity of the corpse during extended funerary rituals, the application
of commodities from the furthest reaches of the empire and beyond powerfully illus-
trates the resources lavished on materials consumed in some ceremonies.

Containing the Dead
Among both cremations and inhumations, but especially the latter, some burials were
distinguished by containers exceptional for their materials, elaboration, or dimensions.
With occasional exceptions, traces of organic containers are usually lost—​for example,
the textile or leather bags or wrappings whose existence is indirectly attested for cre-
mated bone with some rich grave-​good assemblages, such as at Tollgate, Kent.
In contrast to the common use of jars in local ceramic fabrics, cremated bone was
occasionally placed in stone and lead containers, glass urns, wooden caskets, and chests,
usually made specifically for this funerary purpose and likely to have incurred signifi-
cantly greater cost for their materials or making (Philpott 1991). Two examples illustrate
the varied types: a cluster of first-​century ad burials from Paternoster Square, London,
348

348   John Pearce

including four lead ossuaria, glass, and, otherwise unique in Britain, a possible por-
phyry urn; and an early third-​century burial found near Springhead, where cremated
bone was placed in two omega-​handled glass jars within a child-​sized limestone coffin
(Shepherd 1988: 28–​29; Walker 1990: 57).
Most inhumations were interred in timber coffins, inferred from nails, stains, or skel-
etal taphonomy; others were buried directly in the earth. Decay means that the massive
character of a small number of wooden coffins (mainly oak) must be inferred by metal
coffin fittings, especially nail size (Booth et al. 2010: 322–​331). Stone coffins and lead cof-
fin liners are more readily appreciable as tokens of display. From the later second cen-
tury onwards, these are widely attested in and around cities, especially York, Dorchester,
Bath, and London (Toller 1977; Taylor 1993). Recent excavations demonstrate the rarity
of their ancient use; in a half century of fieldwork in London, for example, only three
lead liners and two complete stone coffins have been excavated, as well as occasional
fragments (Barber and Bowsher 2000: 95–​96; Thomas 2004: 18–​29). Even at Poundbury,
much closer to stone quarries, only four burials of more than 1,000 excavated in the
principal excavation were in stone coffins (Farwell and Molleson 1993: 63–​65). Labour
estimates for quarrying, shaping, and transport alone suggest that cost put stone cof-
fins beyond the reach of all but the richest; similar constraints will have applied to lead
(Russell 2010). The expense of such containers probably accounts for their reuse. The
considerable effort expended at the end of the fourth century ad to disinter stone coffins
from mausolea at Lullingstone and Poundbury illustrates the labour of preparation and
transport from the quarry, which the robbers sought to avoid (Farwell and Molleson
1993: 51–​52; Meates 1987: 127–​132).
Although stone coffins are mostly plain, roughly one-​quarter of lead liners carried
exterior decoration, most commonly pecten shell motifs on their lids within panels
defined by bead and reel interlace, as well as lions and other Dionysiac figures (Toller
1977). Toynbee’s finding (1964:  347–​352) of allusions in these motifs to an afterlife
free from earthly toil in the Blessed Isles has been influential, but these symbols also
enhanced domestic space and portable objects and may have a had a wider range of
resonance for the living. The pecten’s association with the birth of Venus allows it, for
example, to stand pars pro toto as an allusion to the embellishment of the person in toi-
lette rituals and perhaps to the preparation of the corpse within (Swift 2009: 150–​151).
Decoration of this type was primarily visible in the period between the sealing of the
liner and the closure of the timber or stone coffin within which it was encased. Whatever
the impression made by other containers similarly distinguished by size, materials, or
decoration, the occasion of burial provided the principal (and ephemeral) context in
which it was achieved. The Lullingstone and Poundbury tombs illustrate their likely
common placing beyond access by the living below the floors of structures (Meates
1987: 127–​132; Farwell and Molleson 1993: 51–​52). The epitaphs on a small number of
stone coffins from York and London may suggest anticipation of future accessibility
and audience, but their discovery divorced from their primary context makes this little
more than guesswork; in the burial vault at the Mount (York), size dictates that the cof-
fin must have been placed in the tomb during its construction and the body later passed
 Status and Burial   349

in (RCHME 1962: 95–​96; Russell 2010). The cella of the temple-​like tomb at Welwyn
Hall may have displayed an Attic marble sarcophagus, otherwise extraordinarily rare in
Britain, but its discovery as shattered and scattered fragments makes this impossible to
confirm (Rook et al. 1984).

Grave Goods
Objects interred with the dead have received the lion’s share of attention on Romano-​
British burials (Philpott 1991; Cool 2011). Their properties may not always have
impressed archaeologists, but we should beware of overemphasizing their poverty.
The abundance, archaeologically speaking, of typical grave objects such as ceramics,
coins, or copper-​alloy ornaments, should not be confused necessarily with easy avail-
ability in antiquity. Careful inventorization of portable property documented in the
Vindolanda tablets, or the ritual labour invested in recovery of garments and small
sums of money in curses at Bath and elsewhere, for example, speak against underesti-
mating the economic significance of mundane objects buried with the dead (Bowman
2003; Tomlin 2003). The comparison of assemblages between contexts also assists in
assessing the availability of objects chosen to accompany the dead. The high frequency
of samian ware among accessory vessels in rural burials, for example, contrasts with
its low representation in ceramic assemblages and suggests its preferential selection
(Willis 2011).
If the placing even of single objects or joints of meat beyond use in burial is a non-​triv-
ial sacrifice, then large numbers of objects in single graves, as occurs especially in south-​
east Britain in the early Roman period, carry even greater weight as indices of resources
consumed in mortuary rites (Pearce 2013: 130–​140). Two examples come from Kent, one
from Tollgate (site D), near the Springhead small town, and another from Bayford were
discovered during brickearth extraction near Milton, a roadside settlement on the river
Swale in the late nineteenth century. The two graves reveal first-​and second-​century ad
assemblages typical of those at the apex of grave furnishing (Payne 1893: 44–​54; Allen et
al. 2012: 354–​386). Both contain extensive ceramic assemblages, especially of table wares,
the former mainly Gallo-​Belgic wares and local copies, the latter terra sigillata, as well
as bronze vessels, complemented at Bayford by glass jugs and bottles (Figure 17.1). The
Bayford assemblage also included an iron lampstand and bronze lamp, as well as bath-
ing gear including strigils and a bronze balsamarium. The presence of glass and bronze
vessels for consumption of liquids and hand-​washing typifies the diversity of materials
in such grave groups. A similar diversity is met in the repertoire of personal ornament,
especially among objects more frequently buried in the later Roman period such as hair-
pins and bracelets; as examples from London, Winchester, and York show, jet, silver,
and ivory examples replaced or complemented those in commoner materials such as
bone, copper alloy, iron, and shale (e.g. Clarke 1979: 312–​313; Barber and Bowsher 2000:
117–​120; Leach et al. 2010). Gold ear and finger rings are occasionally documented, but
the bracelets, rings, and, in one case, a necklace buried with children at Southfleet and
350

350   John Pearce

Figure 17.1  Bronze and ceramic vessels in grave 6260, the cremation burial of an adult (ad
50–​80), from a rural settlement at Tollgate, Springhead. The bronze jug, pan, and mixing bowl as
well as two ceramic flagons were placed beneath a table. On this were piled terra nigra and terra
rubra platters, beakers, cups, and local copies. The skull and forelimb of a pig were found close by,
as was the cremated bone. One or more boards with glass gaming pieces and bone dice lay closer
to the grave’s centre.
Source: courtesy of Oxford Archaeology.

Borden in Kent are exceptional for burials in Britain (Payne 1893: 54–​58; Jessup 1959:
29–​30; Philpott 1991: 152).
The perception of the burial furniture also draws on symbolic associations of these
objects beyond rarity and (likely) expense. In funerary portraiture objects are argued
to be props that, by association with particular occasions or behaviours, manifest the
virtues claimed for the deceased, often metonymically. For example pugillaria (blocks
of writing tablets) allude to the privileges of the military scribe, or wool-​working tools
to the matron’s domestic accomplishments (Hope 1997; Carroll 2013). Objects in graves
may be examined from the same perspective. Some may relate to the specific role or
office held by the deceased in life—​for example, therapeutic and/​or religious exper-
tise at Stanway, Colchester, where the rich assemblage of one conquest-​period burial
contained medical instruments, or to military or bureaucratic service in those few
fourth-​century burials that contain crossbow brooches or elaborate belt sets (Jackson
2007: 236–​251; Cool 2010: 278–​284; Esmonde Cleary 2013: 84–​85). The allusive power
of objects is, however, normally less specific. In the grave assemblages of the early
Roman period, for example, such allusions are overwhelmingly to the dinner party and
associated social routines. These may be emblematic of cultivated hospitality, urbane
leisure, and the inequalities characteristic of Roman commensalism marked by differ-
entiated seating, serving, food and drink, manner of consumption, and (sometimes) the
 Status and Burial   351

impossibility of reciprocation (Dunbabin and Slater 2011). Burials like those at Tollgate
and Bayford evoke the complexities of the Roman table with vessels for ablutions and
consumption of food and drink in multiple courses using vessels of diverse sizes, mate-
rials, and shapes. Gaming and bathing are called to mind by the presence of objects such
as gaming counters and strigils (Pearce 2013: 136–​140). Hospitality endures as a motif
through the continued deposition of vessels in third-​and fourth-​century graves, espe-
cially of glass in otherwise rare forms (e.g. Price and Jackson 2010). The main emphasis
in late Roman burials, however, lies on personal appearance, manifested in jewellery or
objects related to the enhancement of appearance, extending the parading of difference
among the living through dress and public performance of complex toilette routines
into the sphere of the dead. The toilet knife with an ivory handle in leopard form and oil
flask buried with an adolescent (BL15) in the fourth-​century ad cemetery at Lant Street,
Southwark, allude to routines of bathing, grooming, and dress that could themselves
serve as a spectacle in their own right (Swift 2009; Ridgeway et al. 2013: 79). The toilet
knife was associated in the grave with the remnants of a bone-​inlaid small casket, also
found in burials of other women and children with similar object repertoires; in these
and as figured on funerary monuments, such as that of Regina at South Shields, caskets
served as signifiers of affluence (Swift 2009: 150–​151; Carroll 2013). Whatever their spe-
cific affiliation to office-​holding, the elaborate belt sets and crossbow brooches are the
surviving elements of the equivalent male burials (Esmonde Cleary 2013: 84–​87).
The bronze vessels referred to from Tollgate and Bayford provide an illustration of the
capacity of decoration on grave objects to accentuate this allusive capacity beyond sim-
ple embellishment. Their motifs are often self-​referential, especially to the symposium
and associated behaviour. The Bayford balsamarium, for example, bears the faces of
three African boys modelled in high relief, likely to be read as images of pueri delicati—​
that is, young slaves to attend on the symposium and bathing that preceded it (Tanner
2010: 34). Bacchic figures likewise evoke the symposium and its conviviality as the fruit
of labour and luck (Figure 17.2). The madness of Ajax represented on a Bayford jug is a
rare representation in bronze of mythological motifs that more commonly exemplify
paideia in other dining-​related media—​for example, precious metal vessels or mosaics
(Toynbee 1964: 324).
As well as generic allusions to prosperity, cultivation, and so on, it can also be sug-
gested that objects used at funerals could prompt the recall of previous gatherings,
funerary or not, where their ‘biography’—​for example, the circumstances of their acqui-
sition, exchange, or use—​evoked the specific history of the participants (Gosden and
Garrow 2012). Howard Williams (2004), for example, has proposed a mnemonic power
for the ceramics that predominate in Romano-​British grave assemblages. He argues that
their presence at the burial ceremony prompted recall of previous occasions where they
manifested the hospitality extended by the deceased or the funeral’s organizers. The
capacity of commensalism for exhibiting inequalities has already been noted.
The presence among grave goods of objects that are rare, are of significant age, or
carry some other highly visible manifestation of their history is also a recurring phe-
nomenon and perhaps indicative of a special potency to trigger memory; for example,
352

352   John Pearce

Figure  17.2  A bronze drinking vessel found at Llantilio Pertholey near Abergavenny with a
handle in the form of a snarling leopard with inlaid silver and amber (PAS NMGW-​9A9D16).
Fieldwork subsequent to its discovery by metal-​detecting revealed the original funerary context.
Source: courtesy of National Museums and Galleries of Wales.

inscriptions on some objects reference the circumstances in which they were made or
exchanged (e.g. Biddulph 2006: 358; Wallace 2006). A crossbow brooch from Lankhills
carries a typical exhortation to felicity or long life (utere felix and v{b}ene vivas)—​its
worn state suggesting a lengthy use after its initial presentation (Cool 2010: 279–​282).
On a very large beaker of late-​second-​century date from Lexden, Colchester, a text cut
after firing and relief images on the vessel’s exterior record arena encounters by two
combatant pairs (Toynbee 1964: 341–​343; RIB II.8 2503, 119). In these cases an intrinsic
value of the object, should it be calculable, seems less significant than the histories and
relationships they embodied—​promotion perhaps through military or administrative
posts in the former instance, shared partisanship at spectacles and associated convivial-
ity in the latter (Fagan 2011; Esmonde Cleary 2013: 84–​85). The final use in a burial cer-
emony evoked those previous occasions where such relationships were made tangible in
shared ceremonies (cf. Ekengren 2013).

Marking the Dead
The final aspect to consider is that of the place of burial and its marker. By largely
unknown mechanisms—​affiliation to a collegium or provision of a space by family,
friends, and heirs, patron or owner—​most individuals were interred singly in separate
 Status and Burial   353

burials. The mass burial from London Road, Gloucester, noted in the opening para-
graph remains exceptional, but its occupants were at least interred within a formal cem-
etery setting; in exceptional cases the context and condition of skeletal remains suggest
behaviour intended to exclude individuals from formal burial, denying the opportunity
for later commemoration and marking a position outside civic structures. The young
men whose crania and other skeletal elements were deposited in pits by the Walbrook
in London are one such example, their pathology revealing lives as well as deaths of vio-
lence, perhaps as arena performers, criminals, or military enemies (Redfern and Bonney
2014). To protect the grave’s integrity, to enable commemoration, and sometimes to
compete for the attention of passers-​by, most burials were probably marked on the sur-
face; postholes detected at Great Dunmow, Essex, or small mounds over cremation bur-
ials at Petty Knowes, Northumberland, exemplify likely frequent forms (Charlton and
Mitcheson 1984; Brooks and Wightman 2011: 5). More significant in their demands on
space, materials, and skills are the markers in permanent media documented in diverse
contexts.
Among military communities, epitaphs and funerary portraits, especially in stele
form, are more common than in any other context, sometimes commemorating indi-
viduals of modest status who through text and image make prominent differences of
ranks and special(ist) roles or familiarity with respectable and leisured domesticity
(Hope 1997; Coulston 2007; see also Hope, this volume). Evidence from their founda-
tions detected in prospection and excavation suggests the recurring presence of much
larger built tombs (e.g. Fitzpatrick 2004: 28–​33; Pollock 2006: 44–​45). These provide a
likely context for the spolia identified in Roman or later reuse, although the dearth of
inscriptions from larger tombs complicates the identification of their specific social con-
text (e.g. Blagg 1998: 251–​252; Bidwell 2010).
Inscribed markers, especially stelae, and funerary sculpture occur in modest num-
bers in cities, disproportionately commemorating soldiers, bureaucrats, and individu-
als of extra-​provincial origin (Mattingly 2008; Stewart 2010; Pearce 2013). Surviving
stelae, again the most frequently occurring form, show considerable diversity in size,
varying from a few tens of centimetres to more than 2 metres in height (e.g. RIB 9 and
17 from London). Ground plans again point to more complex monuments, although
reduction to robbed-​out foundations typically obstructs reconstruction of superstruc-
tures or even definitive identification of funerary purpose (e.g. Lakin et al. 2002: 10–​11,
26). Foundations from masonry-​built mausolea constructed over two centuries hard by
Watling Street in Southwark provide the clearest instance of a monumentalized urban
Gräberstrasse (street of tombs) (Mackinder 2000). Fragments of superstructure occa-
sionally survive, often where reused as spolia, for example, the blocks of the tomb-​altar
of the first century ad procurator Classicianus in the Trinity Square bastion of London’s
wall (Grasby and Tomlin 2002). The letters cut for this same inscription, among the larg-
est and most elegant from the province and comparable with metropolitan work, or the
carving of London’s best-​preserved funerary portrait, another bastion find, a soldier
in tunic and cloak with sword belt, scroll, and stylus tablets, reveal the effort expended
beyond scale alone to engage passers-​by (Bishop 1983; Grasby and Tomlin 2002: 71;
354

354   John Pearce

see also Hope, this volume). The use of limestones from diverse sources in Britain and
France as revealed by petrographic analysis (Hayward 2009) as well as Purbeck marble
(e.g. RIB 3131; Tomlin 2008: 370–​371, no. 3) and imported white marbles (e.g. RIB 3009,
3012, 3026) are to be attributed to the same motivation. Painted plaster enhanced tomb
exteriors and interiors, though, save for exceptions at Poundbury, where mausoleum
R8’s walls depict a procession and an architectural setting as well as a chi-​rho mono-
gram, what it shows cannot be reconstructed (Sparey-​Green 1993: 135–​139).
The countryside east of London on the road to Canterbury illustrates the larger funer-
ary monuments characteristic of rural settings, especially south-​east Britain. Masonry
monuments built from the first to late third centuries ad, some perhaps tower tombs like
those of Gallia Belgica, others resembling the Romano-​Celtic temple with central tower
and ambulatory, punctuate this route. At Southfleet an outer precinct wall enclosed an
area of exceptional size, c. 120 metres long on one side, around the central third-​century
enclosure containing burials with stone coffins and precious metal jewellery (Jessup
1959; Meates 1987; Andrews et al. 2011: 132–​134). The variety of form and scale shown by
barrows is illustrated by the contrast between Keston, c. 9 metres in diameter, revetted
and buttressed with a red painted exterior (and white marble inscription) and the near-​
contemporary ‘Holborough Knob’, an earthen mound c. 30 metres in diameter and 5.5
metres high when excavated (Jessup 1954; Philp et al. 1999; Hassall and Tomlin 2006:
467). The form owes its popularity to its resonance in local and Roman traditions, being
both the burial monument of the northern European landscape and a preferred tomb
type in Roman Italy (Morris 1992: 51; Schwarz 2002).
The location of many such monuments within a few kilometres of towns suggests that
they serve as the tombs of the curial class, who, with the exception of York, are other-
wise near invisible among epitaphs (Pearce 2013: 125–​127). Their immediate context is
not generally well established, though examples such as Keston or Lullingstone reveal
the contribution of tombs to the general monumentality of villas. Intuitively, the scale
of some monuments suggests an intention to contrive widespread visibility by land or
water, but, in contrast to prehistoric monuments, the landscape setting of Roman tombs
has received limited attention. Viewshed analysis of the Bartlow Hills barrows, claimed
as among the largest in Roman northern Europe (i.e. an examination using Geographical
Information Systems of the vantage points from which they could be seen), unexpectedly
revealed limited visibility, a spectacle reserved for those who followed the byways of a
region on the margins of three civitas territories (Eckardt et al. 2009).

Conclusion

Hopes that excavated burials will provide a map of Romano-​British social hierarchies
to compensate for the limitations of the texts will be ever frustrated. The irrevocable
loss of the legally defined status of individuals encountered in Romano-​British graves is
not, however, a reason for despair. While the surviving evidence is imperfect, its study
 Status and Burial   355

nonetheless gives key insights into the dynamics of status relationships. The ‘intense
[Roman] appetite for ranking people socially, for constantly establishing finely tuned
hierarchies’ identified by Peachin (2011: 27) clearly manifests itself in funerary prac-
tice. For even commonplace burials, resource requirements were substantial, including
matériel and labour for rituals and a burial site, especially so on busy urban margins.
In the countryside, formal burial itself may have marked a status distinction, making it
implausible to identify the small groups of ‘modest’ burials frequently excavated on set-
tlement margins as those of slaves or the rural poor (Pearce 2013: 26). Among archaeo-
logically visible burials, differentiation in death spans the period under scrutiny. For
some excavated cemeteries statistical analyses have identified groupings of burials
based on varying combinations and quantities of grave goods; these have scope for
further application (Cool and Baxter 2005). In this chapter a qualitative evaluation has
revealed similar variability between burials; in preparation of the body, interment and
tomb marker, archaeological evidence reveals both micro-​variation within a given cem-
etery or region and a small group of tombs recurrently differentiated to a much greater
degree, especially through furnishing with objects that are exceptional in terms of their
material, in either its quantity or its rarity, and craftsmanship. Among such burials the
nature of display changes over time; inscriptions are little used as markers after the mid-​
third century ad (Hope, this volume), and during the same century large grave-​good
assemblages give way to massive containers, housing sometimes elaborately prepared
and dressed corpses. The homogeneity of late Roman burials should not be overempha-
sized; it is difficult to identify the cyclical character to funerary display from ostentation
to simplicity documented in other settings (cf. Parker Pearson 1999: 87).
Contextual information allows some assessment of the social and symbolic connota-
tions of burial treatment, especially for the sépultures privilegiées at the time of interment.
Objects displayed with the dead embodied the savoir faire in presentation of the person,
consumption of food and drink, leisure, and so on that was central to presenting status
difference (cf. Ekengren 2013). These ‘snapshots’ of the dead encapsulated, fleetingly, the
essence of the cultivated citizen. Less easily captured, but equally significant perhaps for
funerary display, was the mnemonic potency of objects, their capacity to bring to mourn-
ers’ minds prior occasions, funerary or otherwise, central to the reproduction of rela-
tionships within the burying group, household, collegium, clientage, or citizen body, with
their inherent status asymmetries. The spatial settings of some monuments or the fresco
fragments from Poundbury hint at the scale of assemblies for funerary sacrifice and feast-
ing, although for this aspect of ritual we depend mainly on evidence from other prov-
inces (Woolf 1998: 166–​168; Dunbabin and Slater 2011; Ancel 2012: 216–​228).
Arguments for status display risk being instrumentalist, reducing the emotional and
ritual complexity of funerals to social stratagems. If pressed too far, they risk suggest-
ing a misleading homogeneity of funerary practice in the province and underplay the
conditioning of ritual by other aspects of the identity of the deceased or mourners, espe-
cially age and gender (see also Moore, Gowland, and Cool, this volume). A focus on the
funeral evidence in isolation also risks exaggerating its importance. Even those monu-
ments contrived to perpetuate the memory of the deceased enjoyed a limited lifespan,
356

356   John Pearce

especially those whose appeal as spolia was irresistible after the energy to commemo-
rate dissipated. The placing of a scarcely legible inscribed marble over the face of a late
Roman inhumation from Canterbury hints at a residual sense of the stone’s purpose but
otherwise illustrates the reuse within the Roman period itself of many memorials (Helm
and Rady 2010: 27–​28). Yet, although the tomb once closed ‘could not have pleased them
[the living] aesthetically, instructed or comforted them spiritually, or gratified any taste
for display’ (Toynbee 1964: 352–​353), its significance as a setting for reproducing sta-
tus relationships should not be underestimated. The diversity of deposit types related
to the funerary process, the abundant environmental and artefactual assemblages,
the evidence of monuments, including epitaphs and portraits, and abundant contex-
tual information should make Roman data the envy of prehistorians (Parker Pearson
1999:  32–​44). While it may lack the historiography of other eras—​for example, the
Anglo-​Saxon period (Härke 2014)—​burial archaeology from Roman Britain possesses
rich analytical possibilities for exploring how fundamental social relationships were
expressed and negotiated. These relationships include the inequalities likely to have
been so marked a characteristic of life and death in the province.

Acknowledgements

I would like to express my gratitude to the editors for their comments and their patience,
Sally Worrell for reading the chapter in draft, and Paul Booth (Oxford Archaeology) and
Evan Chapman (National Museums and Galleries of Wales) for assistance with images.

Abbreviation
RIB R. G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume
I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Primary Source
Herodian (1969). Roman History, trans. C. R. Whittaker. Loeb Classical Library. 2  vols.
Cambridge, MA: Harvard University Press.

Secondary Sources
Allen, T., Donnelly, M., Hardy, A., Hayden, C., and Powell, K. (2012). A Road through the
Past:  Archaeological Discoveries on the A2 Pepperhill to Cobham Road-​Scheme in Kent.
Oxford: Oxford Archaeology.
 Status and Burial   357

Ancel, M.-​J. (2012). Pratiques et espaces funéraires: La Crémation dans les campagnes romaines
de la Gaule Belgique. Éditions Mergoil: Montagnac.
Andrews, P., Biddulph, E., Hardy, A., and Brown, R. (2011). Settling the Ebbsfleet Valley. High
Speed 1 Excavations at Springhead and Northfleet, Kent. The Late Iron Age, Roman, Saxon and
Medieval Landscape, i. The Sites. Salisbury: Oxford and Wessex Archaeology.
Barber, B., and Bowsher, D. (2000). The Eastern Cemetery of Roman London. Excavations 1983–​
1990. London: MoLAS.
Barber, B., and Hall, J. (2000). ‘Digging up the People of Roman London’, in I. Haynes,
H. Sheldon, and L. Hannigan (eds), London Underground:  The Archaeology of a City.
Oxford: Oxbow Books, 102–​120.
Barrett, J. C. (1993). Fragments from Antiquity: An Archaeology of Social Life in Britain 2900–​
1200 BC. Oxford: Blackwell.
Biddulph, E. (2006). ‘Graffiti on Funerary Pottery’, Britannia, 37: 355–​359.
Bidwell, P. (2010). ‘A Survey of the Anglo-​Saxon Crypt at Hexham and its Reused Roman
Stonework’, Archaeologia Aeliana, 39: 53–​145.
Bishop, M. (1983). ‘The Camomile Street Soldier Reconsidered’, Transactions of the London and
Middlesex Archaeological Society, 34: 31–​48.
Blagg, T. F.  C. (1998). ‘The Sculpture’, in H. E.  M. Cool and C. Philo (eds), Roman Castleford
Excavations 1974–​85, i. The Small Finds. Wakefield: West Yorkshire Archaeology Service, 247–​253.
Booth, P., Boyle, A., Clough, S., and Simmonds, A. (2010). The Late Roman Cemetery at
Lankhills, Winchester. Excavations 2000–​2005. Oxford: Oxford Archaeology.
Bowman, A. (2003). Life and Letters on the Roman Frontier. London: British Museum.
Brettell, R., Stern, B., Reifarth, N., and Heron, C. (2013). ‘The “Semblance of Immortality”?
Resinous Materials and Mortuary Rites in Roman Britain’, Archaeometry Early View <doi:
10.1111/​arcm.12027>
Brooks, H. (2006). ‘A Roman Temple-​Tomb at Colchester Royal Grammar School, 6 Lexden
Road, Colchester, Essex. August–​September 2005’. Colchester Archaeological Trust Report
345 <http://​cat.essex.ac.uk/​reports/​CAT-​report-​0345.pdf> (accessed 10 February 2012).
Brooks, H., and Wightman, A. (2011). ‘Report on an Archaeological Evaluation, Excavation
and Watching Brief at the Salerooms Site, Chequers Lane, Great Dunmow, Essex’. Colchester
Archaeological Trust Report 535  <http://​cat.essex.ac.uk/​reports/​CAT-​report-​0535.pdf>
(accessed 10 February 2012).
Carroll, M. (2013). ‘The “Insignia of Women”:  Dress, Gender and Identity on the Roman
Funerary Monument of Regina from Arbeia’, Archaeological Journal, 169: 281–​311.
Charlton, B., and Mitcheson, M. (1984). ‘The Roman Cemetery at Petty Knowes, Rochester,
Northumberland’, Archaeologia Aeliana, 12: 1–​31.
Clarke, G. (1979). Pre-​Roman and Roman Winchester. Part 2: The Roman Cemetery at Lankhills.
Oxford: Oxford University.
Cool, H. E.  M. (2004). The Roman Cemetery at Brougham, Cumbria:  Excavations 1966–​67.
London: Society for the Promotion of Roman Studies.
Cool, H. E. M. (2010). ‘Objects of Glass, Shale, Bone and Lead’, in P. Booth, A. Boyle, S. Clough,
and A. Simmonds (eds), The Late Roman Cemetery at Lankhills, Winchester: Excavations
2000–​2005. Oxford: Oxford Archaeology, 267–​309.
Cool, H. E. M. (2011). ‘Funerary Contexts’, in L. Allason-​Jones (ed.), Artefacts from Roman
Britain: Their Purpose and Use. Cambridge: Cambridge University Press, 293–​313.
Cool, H. E. M., and Baxter, M. J. (2005). ‘Cemeteries and Significance Tests’, Journal of Roman
Archaeology, 18: 397–​404.
358

358   John Pearce

Coulston, J. (2007). ‘Art, Culture and Service: The Depiction of Roman Soldiers on Funerary
Monuments of the Third Century ad’, in L. de Blois and E. Lo Cascio (eds), Impact of Empire
VI: The Impact of the Roman Army (200 BC–​AD 476). Leiden: Brill, 529–​561.
Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge.
Crowley, L. (2011). ‘The Role of Mortuary Ritual in the Construction of Social Boundaries
by Privileged Social Groups within Villa Landscapes’, in N. Roymans and T. Derks (eds),
Villa Landscapes in the Roman North: Economy, Culture, Lifestyles. Amsterdam: Amsterdam
University Press, 195–​210.
Crummy N., Crummy P., and Crossan C. (1993). Excavations of Roman and Later Cemeteries,
Churches and Monastic Sites in Colchester. Colchester: Colchester Archaeological Trust.
Davies, S. M., Bellamy, P. S., Heaton, M. J., and Woodward, P. J. (2002). Excavations at Alington
Avenue, Fordington, Dorchester, Dorset 1984–​7. Dorchester:  Dorset Natural History and
Archaeological Society.
Dunbabin, K. M. D., and Slater, W. J. (2011). ‘Roman Dining’, in M. Peachin (ed.), The Oxford
Handbook of Social Relations in the Roman World. Oxford: Oxford University Press, 428–​436.
Eckardt, H. (1999). ‘The Colchester “Child’s Grave”’, Britannia, 30: 57–​90.
Eckardt, H. (2010). ‘A Long Way from Home’, in H. Eckardt (ed.), Roman Diasporas:
Archaeological Approaches to Mobility and Diversity in the Roman Empire. JRA
Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 99–​130.
Eckardt H., Brewer, P., Hay, S., and Poppy, S. (2009). ‘Roman Barrows and their Landscape
Context: A GIS Case Study at Bartlow, Cambridgeshire’, Britannia, 40: 65–​98.
Ekengren, F. (2013). ‘Contextualizing Grave Goods:  Theoretical Perspectives and
Methodological Implications’, in S. Tarlow and L. Nilsson Stutz (eds), The Oxford Handbook
of the Archaeology of Death and Burial. Oxford: Oxford University Press, 173–​192.
Esmonde Cleary, S. (2013). The Roman West ad 200–​ 500:  An Archaeological Study.
Cambridge: Cambridge University Press.
Evans, E. M., and Maynard, D. J. (1997). ‘Caerleon Lodge Hill Cemetery: The Abbeyfield Site
1992’, Britannia, 28: 169–​244.
Fagan, G. (2011). The Lure of the Arena: Social Psychology and the Crowd at the Roman Games.
Cambridge: Cambridge University Press.
Farwell, D., and Molleson, T. (1993). Excavations at Poundbury, Dorchester, Dorset, 1966–​1980,
ii. The Cemeteries. Dorchester: Dorset Natural History and Archaeological Society.
Fitzpatrick, A. P. (2004). ‘The Masonry Monuments’, in H. E. M. Cool (ed.), The Roman
Cemetery at Brougham, Cumbria: Excavations 1966–​67. London: Society for the Promotion
of Roman Studies, 28–​33.
Fowler, C. (2013). ‘Identities in Transformation: Identities, Funerary Rites and the Mortuary
Process’, in S. Tarlow and L. Nilsson Stutz (eds), The Oxford Handbook of the Archaeology of
Death and Burial. Oxford: Oxford University Press, 511–​526.
Gardner, A. (2013). ‘Thinking about Roman Imperialism: Postcolonialism, Globalization and
Beyond?’, Britannia, 44: 1–​25.
Gosden, C., and Garrow, D. (2012). Technologies of Enchantment? Exploring Celtic Art 400 bc to
ad 100. Oxford: Oxford University Press.
Grasby, R. D., and Tomlin, R. S.  O. (2002). ‘The Sepulchral Monument of the Procurator
C. Julius Classicianus’, Britannia, 33: 43–​76.
Green, C., Paterson, M., and Biek, L. (1981). ‘The Romano-​British Cemetery at the Site of the
Crown Buildings, Dorchester’, Proceedings of the Dorset Natural History and Archaeological
Society, 103: 67–​100.
 Status and Burial   359

Härke, H. (2014). ‘Grave Goods in Early Medieval Burials: Messages and Meanings’, Mortality,
19/​1: 1–​21 <doi: 10.1080/​13576275.2013.870544>
Hassall, M., and Tomlin, R. S. O. (2006). ‘Roman Britain in 2005: Inscriptions’, Britannia, 37:
467–​488.
Haverfield, F., Taylor, M. V., and Wheeler, R. E. M. (1932). ‘The Romano-​British Period’, in The
Victoria County History of the Counties of England, iii. Kent. London: Constable, 1–​176.
Hayward, K. M. J. (2009). Roman Quarrying and Stone Supply on the Periphery. Southern
England: A Geological Study of First-​ Century Funerary Monuments and Monumental
Architecture. BAR British Series 500. Oxford: Archaeopress.
Helm, R., and Rady, J. (2010). Excavations at Market Way, St Stephen’s, Canterbury.
Canterbury: Canterbury Archaeological Trust.
Herbert, A. (2011). ‘Excavations and Discoveries at the Romano-​British Cemetery, East Hill,
Dartford’, Archaeologia Cantiana, 131: 85–​110.
Hope, V. (1997). ‘Words and Pictures:  The Interpretation of Romano-​British Tombstones’,
Britannia, 28: 245–​258.
Jackson, R. (2007). ‘The Surgical Instruments’, in P. Crummy, S. Benfield, N. Crummy, V. Rigby,
and D. Shimmin (eds), Stanway: An Élite Burial Site at Camulodunum. London: Society for
the Promotion of Roman Studies, 236–​252.
Jessup, R. (1954). ‘Excavation of a Roman Barrow at Holborough, Snodland’, Archaeologia
Cantiana, 68: 1–​61.
Jessup, R. (1959). ‘Barrows and Walled Cemeteries in Roman Britain’, Journal of the British
Archaeological Association, 22: 1–​32.
Jones, R. F. J. (1983). ‘Cemeteries and Burial Practice in the Western Provinces of the Roman
Empire’. Unpublished Ph.D. Thesis, London University.
Jones, R. F. J. (1987). ‘Burial Customs of Rome and the Provinces’, in J. Wacher (ed.), The Roman
World. London: Routledge, 812–​838.
Jones, R. F. J. (1991). ‘Cultural Change in Roman Britain’, in R. F. J. Jones (ed.), Britain in the
Roman Period: Recent Trends. Sheffield: J. R. Collis Publications, 115–​120.
Lakin, D., Seeley, F., and Rielly, K. (2002). The Roman Tower at Shadwell, London:  A  Re-​
Appraisal. London: MoLAS.
Leach S., Eckardt H., Chenery C., Müldner, G., and Lewis, M. (2010). ‘A Lady of York: Migration,
Ethnicity and Identity in Roman Britain’, Antiquity, 84: 131–​145.
Mackinder, A. (2000). A Romano-​British Cemetery on Watling Street. London: MoLAS.
Mattingly, D. J. (2006). An Imperial Possession:  Britain in the Roman Empire.
London: Allen Lane.
Mattingly, D. J. (2008). ‘Urbanism, Epigraphy and Identity in the Towns of Britain under Roman
Rule’, in H. M. Schellenberg, V. Hirschmann, and A. Krieckhaus (eds), A Roman Miscellany: Essays
in Honour of Anthony R. Birley on his 70th Birthday. Gdansk: Gdansk University, 53–​71.
May, T. (1930). Catalogue of the Roman Pottery in Colchester and Essex Museum.
Cambridge: Colchester Corporation.
McKinley, J. (2013). ‘Cremation:  Excavation, Analysis and Interpretation of Material from
Cremation-​Related Contexts’, in S. Tarlow and L. Nilsson Stutz (eds), The Oxford Handbook
of the Archaeology of Death and Burial. Oxford: Oxford University Press, 147–​171.
Meates, G. W. (1987). The Roman Villa at Lullingstone, Kent. Volume 2. Maidstone:  Kent
Archaeological Society.
Merrifield, R., and Hall, J. (2008). ‘In its Depths what Treasures? The Nature of the Walbrook
Stream Valley and the Roman Metalwork Found Therein’, in J. Clark, J. Cotton, J. Hall,
360

360   John Pearce

R. Sherris, and H. Swain (eds), Londinium and Beyond:  Essays on Roman London and
its Hinterland for Harvey Sheldon. CBA Research Report 156. York:  Council for British
Archaeology, 121–​127.
Millett, M. (1987). ‘An Early Roman Burial Tradition in Central Southern England’, Oxford
Journal of Archaeology, 6: 63–​68.
Millett, M. (1993). ‘A Cemetery in Transition:  King Harry Lane Reconsidered’, in M.
Strück (ed.), Römerzeitliche Gräber als Quellen zur Religion, Bevölkerungsstruktur und
Sozialgeschichte. Mainz: Johannes Gutenberg Institut für Vor-​und Frühgeschichte, 255–​282.
Morris, I. (1992). Death Ritual and Social Structure in Classical Antiquity.
Cambridge: Cambridge University Press.
Mouritsen, H. (2011). The Freedman in the Roman World. Cambridge:  Cambridge
University Press.
Müldner, G. (2013). ‘Stable Isotopes and Diet: Their Contribution to Romano-​British Research’,
Antiquity, 87: 137–​149.
Niblett R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium.
London: Society for the Promotion of Roman Studies.
Ottaway, P. J., Qualmann, K., Rees, H., and Scobie, G. (2012). The Roman Cemeteries and
Suburbs of Winchester Excavations 1971–​86. Winchester: Winchester Museums Service.
Parker Pearson, M. (1999). The Archaeology of Death and Burial. Sutton: Stroud.
Payne, G. (1893). Collectanea Cantiana, or, Archaeological Researches in the Neighbourhood of
Sittingbourne and Other Parts of Kent. London: Mitchell and Hughes.
Peachin, M. (2011). ‘Introduction’, in M. Peachin (ed.), The Oxford Handbook of Social Relations
in the Roman World. Oxford: Oxford University Press, 3–​36.
Pearce, J. (2000). ‘Burial, Society and Context in the Provincial Roman World’, in J. Pearce, M.
Millett, and M. Strück (eds), Burial, Society and Context in the Roman World. Oxford: Oxbow
Books, 1–​12.
Pearce, J. (2013). Contextual Archaeology of Burial Practice: Case Studies from Roman Britain.
BAR British Series 588. Oxford: Archaeopress.
Pearce, J., Booth, P., and Niblett, R. (2008). ‘Les Nécropoles romaines en Grande
Bretagne: Nouvelles découvertes’, Dossiers d’archéologie, 330: 28–​35.
Petts D. (2003). Christianity in Roman Britain. Stroud: Sutton.
Philp, B., Parfitt, K., Willson, J., and Williams, W. (1999). The Roman Villa Site at Keston,
Kent:  Second Report (Excavations 1967 and 1978–​ 1990). Dover:  Kent Archaeological
Rescue Unit.
Philpott, R. (1991). Burial Practice in Roman Britain: A Survey of Grave Treatment and
Furnishing ad 43–​410. BAR British Series 219. Oxford: Archaeopress.
Pitts, M., and Griffin, R. (2012). ‘Exploring Health and Social Well-​Being in Late Roman
Britain: An Inter-​Cemetery Approach’, American Journal of Archaeology, 116: 253–​276.
Pollock, K. J. (2006). The Evolution and Role of Burial Practice in Roman Wales.
Oxford: Hadrian Books.
Pooley, L., Crummy, P., Shimmin, D., Brooks, H., Holloway, B., and Masefield, R. (2011).
‘Archaeological Investigations on the “Alienated Land”, Colchester Garrison’. Colchester
Archaeological Report 412  <http://​cat.essex.ac.uk/​summaries/​CAT-​0412.html> (accessed
10 June 2013).
Price, J., and Jackson, C. (2010). ‘Assessment. Roman Glass Vessels from Burial SGR 73’ <http://​
www.lparchaeology.com/​prescot/​about/​assessment-​g lass-​vessels-​f rom-​burial-​sgr73>
(accessed 10 January 2013).
 Status and Burial   361

RCHME (1962). Royal Commission on Historical Monuments, Eburacum:  Roman York.

London: HMSO.
Redfern, R., and Bonney, H. (2014). ‘Headhunting and Amphitheatre Combat in Roman
London, England: New Evidence from the Walbrook Valley’, Journal of Archaeological
Science <doi: 10.1016/​j.jas.2013.12.013>
Redfern, R., and DeWitte, S. (2011). ‘Status and Health in Roman Dorset: The Effect of Status on
Risk of Mortality in Post-​Conquest Populations’, American Journal of Physical Anthropology,
146: 197–​208.
Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge
University Press.
Ridgeway, V., Leary, K., and Sudds, B. (2013). Roman Burials in Southwark. Excavations at 52–​56
Lant Street and 56 Southwark Bridge Road, London SE1. London: Pre-​Construct Archaeology.
Rook, A. G., Walker, S., and Denston, C. B. (1984). ‘A Roman Mausoleum and Associated
Marble Sarcophagus and Burials from Welwyn, Herts’, Britannia, 15: 143–​162.
Roymans, N., and Derks, T. (2011). ‘Studying Roman Villa Landscapes in the 21st Century:
A Multi-​Dimensional Approach’, in N. Roymans and T. Derks (eds), Villa Landscapes in
the Roman North: Economy, Culture, Lifestyles. Amsterdam: Amsterdam University Press,
1–​44.
Russell, B. (2010). ‘Sarcophagi in Roman Britain’, Bollettino di Archeologia on line. Volume
Speciale:  XVII International Congress of Classical Archaeology, Roma 22-​26 Sept. 2008
<http://​151.12.58.75/​archeologia/​bao_​document/​articoli/​2_​RUSSELL.  pdf> (accessed 22
October 2013).
Scheid, J. (2005). Quand faire c’est croire: Les Rites sacrificiels des Romains. Paris: Aubier.
Schwarz, M. (2002). Tumulat Italia tellus: Gestaltung, Chronologie und Bedeutung der römis-
chen Rundgräber in Italien. Rahden: Leidorf.
Shepherd, J. D. (1988). ‘The Roman Occupation in the Area of Paternoster Square, City of
London’, Transactions of the London and Middlesex Archaeological Society, 39: 1–​31.
Simmonds, A., Márquez-​Grant, N., and Loe, L. (2008). Life and Death in a Roman City.
Excavation of a Roman Cemetery with a Mass Grave at 120–​122 London Road, Gloucester.
Oxford: Oxford Archaeology.
Sparey-​ Green, C. (1993). ‘The Mausolea Painted Plaster’, in D. Farwell and Molleson
(eds), Excavations at Poundbury, Dorchester, Dorset, 1966–​ 1980, ii. The Cemeteries.
Dorchester: Dorset Natural History and Archaeological Society, 135–​140.
Sparey-​Green, C. (2003). ‘Where are the Christians? Late Roman Cemeteries in Britain’, in M.
Carver (ed.), The Cross Goes North: Processes of Conversion in Northern Europe, ad 300–​1300.
Woodbridge: Boydell and Brewer, 93–​107.
Stewart, P. (2010). ‘Geographies of Provincialism in Roman Sculpture’, Journal of the
International Association of Research Institutes in the History of Art 5 <http://​www.riha-​jour-
nal.org/​articles/​2010/​stewart-​geographies-​of-​provincialism> (accessed June 2013).
Strück, M. (2000). ‘High Status Burials in Roman Britain (1st to 3rd Century ad)—​Potential of
Interpretation’, in J. Pearce, M. Millett, and M. Strück (eds) Burial, Society and Context in the
Roman World. Oxford: Oxbow Books, 85–​96.
Swain, H., and Roberts, M. (2001). The Spitalfields Roman. 2nd edn. London: MoLAS.
Swift, E. (2009). Style and Function in Roman Decoration. Farnham: Ashgate.
Tanner, J. (2010). ‘Race and Representation in Ancient Art:  Black Athena and After’, in D.
Bindman and H. Gates (eds), The Image of the Black in Western Art, i. From the Pharaohs to
the Fall of the Roman Empire. Cambridge, MA: Harvard University Press, 1–​39.
362

362   John Pearce

Taylor, A. F. (1993). ‘A Roman Lead Coffin with Pipeclay Figurines from Arrington,
Cambridgeshire’, Britannia, 24: 191–​227.
Thomas, C. (2004). Life and Death in London’s East End:  2000 Years at Spitalfields.
London: MoLAS.
Toller, H. (1977). Roman Lead Coffins and Ossuaria. BAR British Series 38. Oxford: British
Archaeological Reports.
Tomlin, R. S. O. (2003). ‘Writing to the Gods in Britain’, in A. E. Cooley (ed.), Becoming
Roman, Writing Latin? JAR Supplementary Series 48. Portsmouth, RI: Journal of Roman
Archaeology, 165–​180.
Tomlin, R. S. O. (2008). ‘Roman Britain in 2007: III Inscriptions’, Britannia, 39: 369–​389.
Toynbee, J. (1964). Art in Britain under the Romans. Oxford: Clarendon Press.
Treherne, P. (1995). ‘The Warrior’s Beauty: The Masculine Body and Self Identity in Bronze Age
Europe’, Journal of European Archaeology, 3/​1: 105–​144.
Walker, S. (1990). Catalogue of Roman Sarcophagi in the British Museum. London:  British
Museum.
Wallace, C. (2006). ‘Long-​Lived Samian?’ Britannia, 37: 259–​272.
Williams, H. (2004). ‘Potted Histories: Cremation, Ceramics and Social Memory in Early
Roman Britain’, Oxford Journal of Archaeology, 23/​4: 417–​427.
Willis, S. (2011). ‘Samian Ware and Society in Britain and Beyond’, Britannia, 42: 167–​242.
Woolf, G. (1998). Becoming Roman in Gaul. Cambridge: Cambridge University Press.
 Chapter 18

Gender in Rom a n Bri ta i n

Melanie Sherratt and Alison Moore

Introduction

This chapter will explore how the concept and theory of gender have been treated within
studies of Roman-​period Britain. Gender theory is argued to have been introduced to
archaeology through the paper authored by Conkey and Spector in 1984, yet it took
nearly a decade for the concept to penetrate into discourse concerning Roman Britain.
The introduction of gender into Roman archaeology was championed in the early years
through the Theoretical Roman Archaeology Conferences (Allason-​Jones 1995; Scott
1993, 1995; van Driel-​Murray 1995), and in the years since, while some researchers have
continued the study of gender using a diverse range of approaches and material evi-
dence, gender still remains under-​theorized and, to a certain extent, sidelined within
mainstream Romano-​British archaeology (Revell 2010). This chapter discusses how
the study of gender developed in relation to Roman Britain, exploring how gender is
approached in present scholarship and seeks to identify new approaches to the subject
in order to integrate the study of gender into mainstream discourse.

What is Gender?

One reason why gender has not been embraced in mainstream interpretations is due, in
part, to conflicting definitions, and this has created confusion as to what gender is and
how it can be applied to archaeological interpretations. As a basic working definition,
Díaz-​Andreu (2005: 14) has argued that gender can be defined ‘as an individual’s self-​
identification and the identification by others to a specific gender category on grounds
of their culturally perceived sexual difference’. In other words, gender is a social con-
struct, created through biological and sociocultural interpretations of sexual difference
(Lesick 1997: 34; Gilchrist 1999: 1). As a social construct, concepts of gender intersect
364

364    Melanie Sherratt and Alison Moore

with many other aspects of social identity, such as status, age, and ethnicity (Scott 1995,
1997; Gilchrist 1999; Baker 2003: 142; Morelli 2009: 4), making gender—​as a topic—​
problematic to explore in isolation, without bringing these other aspects of identity
into play. Gender is not a fixed male–​female dichotomy based only upon biological sex;
rather it is a fluid state that can change as social identities develop. For example, age can
break down gender homogeneities, as in many societies gender is differentiated through
differing roles in the reproductive process (Lesick 1997: 35; see also A. Moore, this vol-
ume, for discussion of the Romano-​British life course). Those who fall outside this cat-
egory, such as prepubescent children or post-​menopausal women, may not have been
accorded a defined gendered identity or may have different gender roles completely.
Similarly, oppositional gendered identities may have been adopted at different stages
of the life cycle depending on social status—​for example, women of a particular social
status, not necessarily elite, who take on financial or employment responsibilities usu-
ally associated with males. The degree of fluidity between gender categories over time,
within society, and between cultures means that gender should be examined in terms of
constant change, rather than being a fixed aspect of identity.
Gender studies are also a search for processes—​that is, the cultural processes that
defined and differentiated between the genders. As a social category, gender can be con-
structed, interpreted, and reinforced in two ways: what Allison (2015: 107) describes as
‘being’ gender and ‘doing’ gender. ‘Being’ gender refers to the acceptance and mainte-
nance of the external, socially acceptable appearance of an individual of a particular
gender. For example, in modern-​day western societies little girls are often dressed in
pink clothing and little boys in blue; this process of gendering often begins at birth and
is accepted and maintained by individuals into the next generation, when their children,
in turn, are dressed in pink or blue. In terms of archaeology, ‘being’ gender is usually
associated with items and objects that relate to the maintenance of personal appear-
ance or personal hygiene, such as dress accessories and unguent bottles. ‘Doing’ gender
refers to occupations and behaviours socially and culturally associated with a gender.
To use again the modern analogy, this refers to little girls being given, and playing with,
dolls to encourage care-​giving ‘feminine’ behaviours while little boys may have Lego,
to develop their analytical ‘masculine’ skills. Within the archaeological record, items
such as needles, spindles, swords, and tools have all been used to identify gender-​related
behaviours.
Often, the term ‘gender’ is misused by being focused exclusively on a discussion of
the male–​female dichotomy, rather than being used as a theoretical term to address a
wide variety of social expression that crosses biological sex categories. In some cases
biological sex may not have determined gender. Anthropological and archaeologi-
cal research demonstrates that varied gender categories are present in societies (Lesick
1997: 34; Gilchrist 1999) but the heterosexual bias inherent in the male–​female dichot-
omy limits discussion on the expression of and attitudes to homosexuality in the past
(Morelli 2009: 4). Matthews (1994) has attempted to bring discussions of homosexu-
ality into Romano-​British studies by questioning and contrasting modern concepts of
homosexuality with Roman attitudes. For example, he explores material culture that
 Gender in Roman Britain    365

may represent a homosexual subculture and suggests that ‘female’ items buried in male
burials may hint at the presence of homosexual males adopting female items to express
their gender. However, Matthews (1994: 129) acknowledges the pitfalls of such associa-
tions, as grave goods within a grave may not have been the possessions of the individ-
ual buried but tokens from the mourners. The unspoken heterosexual bias means that
there is little critical exploration of so-​called gendered items in the material record: why
have certain artefacts been labelled as ‘male’ or ‘female’? What purpose did they serve
in ‘gender’ expression? What can it mean when the artefact is found within the context
of the ‘wrong’ biological sex? It is only by thinking critically around these type of ques-
tions and exploring gender within the context of other expressions of social identity that
archaeologists can move beyond the present state of knowledge and integrate gender
studies into the mainstream of archaeological discourse. It is important to do so, as the
past can be used as a political tool (see chapters in Little and Shackel 2007); we need to
challenge gender assumptions based on preconceptions of the past, as these can be used
to reinforce ideas in the present that certain roles attributed to gender groups are ‘natu-
ral’ (Scott 1995: 175).

Development of Romano-​British
Gender Studies

In general, Roman archaeology has been rather slow to engage in, and develop
approaches to, the study of gender. In part, this is due to an unspoken assumption of
the ‘maleness’ of Roman (particularly military) archaeological remains and the long-​
standing division between ‘classical’ studies, with its focus on art history and literary
evidence, and ‘provincial’ archaeology, where theoretical approaches have been influ-
enced by the archaeology of other time periods (Allison 2015: 106). However, as Hill
(2001: 15) argues, there are remarkable quality and wealth of Roman archaeological
data through which to address gender issues, which exceed those of prehistory and,
in certain cases, the later medieval period. The engagement with gender meant that it
was only really in the late 1980s and 1990s that gender was introduced into Romano-​
British studies, primarily through the work of Lindsay Allason-​Jones (1989, 1995, 1999,
2012), Carol van Driel Murray (1994, 1995, 1997), and Eleanor Scott (1993, 1995, 1997)
(see also Revell 2009). Much of the initial work focused on the identification of women
within the province and investigations of their lives. Using epigraphic sources and
archaeological evidence, Allason-​Jones and van Driel Murray explored the presence
of women within the militarized zone of northern Britain, identifying that, rather than
being ‘male-​only’ bastions, military forts were also home to women and children. For
example, van Driel Murray (1995) used the measurements of leather shoes preserved
in the waterlogged conditions at sites such as Vindolanda to identify the presence of
women and children within the fort, indicating that they were not just confined to the
366

366    Melanie Sherratt and Alison Moore

periphery of the vicus. Similarly, Allason-​Jones (1999) argued that the size of the cen-
turions’ rooms at military sites, usually considered too small to accommodate a family,
were no smaller than excavated urban houses at Verulamium or Colchester, and she
utilized epigraphic evidence from funerary monuments, altars, and surviving letters to
explore their lives.
In contrast, Eleanor Scott (1993, 1995, 1997) championed a more theoretical approach
to gender archaeology in Roman Britain, challenging the androcentric interpreta-
tions, with their roots in the male-​world view of the nineteenth century and its focus
on the political and imperial systems of the Roman Empire (Revell 2010: 1), and lack of
engagement with feminist social theory (Scott 1995: 174). Scott (1995; 1997: 2) argued that
women were overlooked in three ways, by exclusion, pseudo-​inclusion, and alienation.
Women were excluded via research that focused on male-​dominated areas of society
(for example, politics, administration, or the military); as women were not deemed to
have any roles in these spheres in the Roman period, their agency was omitted from
discussion (Scott 1995: 178; Hill 2001; Revell 2010: 1). Within research that did touch
upon women, they were often marginalized or treated as ‘special cases’ or ‘exceptions’
(pseudo-​inclusion), and imaginative stories were sometimes put forward, in the absence
of factual evidence, to explain certain finds or events (Scott 1995, 1997). For example, in
the case of the infant burials at Hambledon villa excavated in the 1920s, the excavator
explained the infants as being the unwanted female offspring of slaves, despite the facts
that the infants were unsexed (Cocks 1921). Scott (2001) argues that the interpretation,
underpinned by the unspoken social mores of the time of excavation, is of male power
over women, with the outcome that, if males were socially valued, females were disre-
garded. The final category, that of alienation, meant that the perspective on, and choice
of, the parts of women’s lives deemed significant were selected and discussed through
the prism of male-​centred methodologies and values (Scott 1995: 179; 1997: 3). Women
were discussed primarily in relation to men—​as mothers, wives, daughters—​and given
a token paragraph or chapter, usually in relation to topics such as family, jewellery, or
fashion, but this discussion never included, or the women were never assumed to have
had, agency in the wider Roman world.
Since the pioneering work of archaeologists such as Allason-​Jones, van Driel
Murray, and Eleanor Scott, gender—​and specifically the roles of women—​has been
discussed more frequently in Romano-​British contexts, particularly through plat-
forms such as the Theoretical Roman Archaeology Conference (Baker 2003). Recent
research has also focused on identifying and categorizing burial assemblages with
female or male attributes within the context of securely osteologically sexed burials
(e.g. Cool 2004; Pearce 2011; for further discussion, see section ‘Burial Evidence’).
However, gender as a separate theoretical subject has waned in interest, even though
studies of other areas of social identity such as age and ethnicity are flourishing (see
chapters by Gowland, A. Moore, Ivleva, and Eckardt and Müldner, this volume).
This lack of interest has been commented on. Allison (2015: 116) suggests that the
lack of secure evidence for gendered behaviour means that archaeologists shy away
from making characterizations in relation to gender practices because of a risk of
 Gender in Roman Britain    367

stereotyping. Both Baker (2003) and Revell (2009, 2010) argue that the still unac-
knowledged androcentricism is leading to gender studies becoming a marginalized
subject unintegrated into mainstream discussion, and that it risks becoming a byword
for the study of women by women. This ‘ghettoization’ has led gender to be a topic that
male archaeologists feel they cannot engage with (Baker 2003: 142). This latter point is
particularly pertinent as there is also little discussion of male identity and the material
expression of masculinity in Roman archaeology (Matthews 1994; Baker 2003: 142;
Revell 2010; for a discussion of masculinity in archaeology, see also Knapp 1998). As
is known, there was mobility within the Roman Empire (see Ivleva, and Eckardt and
Müldner, this volume), and research is needed into how men from different ethnic
and cultural systems negotiated their identity within, or expressed their resistance to,
the wider homogeny of the Roman world.

Identifying Gender in Roman Britain

A core idea within gender theory, and theories of identity more widely, is that gen-
der is performative (Butler 1999; Sørensen 2000). It is not fixed through biology, or
stable through the lifetime, but rather is an ongoing project maintained through the
processes and interactions of daily living. The repeated actions and encounters of living
a life involve interaction with others, and within these interactions we communicate
verbally and non-​verbally information about ourselves and our identity. In particular,
we mark our affinity with particular social groups and our differences with others. We
draw particular items of material culture into this communication, and it is through
such mundane performances that material takes on an association with particular gen-
ders. The project for the archaeologist investigating gender in Roman Britain is taking
the material record of the past and identifying the processes of differentiation, the dif-
fering identities that might be created, and, consequently, the processes of gender in
the past.
This is in itself a problematic task, as we are required to understand the meanings
given to materials in the past. It has proved too easy to project backwards modern
meanings of particular items. We might assume that activities such as sewing are asso-
ciated with women, and so needles are bound up with ideas of femaleness. However,
they might also have formed part of a soldier’s standard equipment, for mending when
on campaign, thus rendering attempts to determine gendered space within forts and
fortresses more complex. Similar issues apply to jewellery, where not all items are gen-
dered, or some, such as earrings, are worn by men in only certain areas of the empire
(Allason-​Jones 1995). Each form of material evidence comes with its own problems, but
how might we identify gender difference from the archaeology evidence? In the rest of
this chapter, we will examine four theatres of activity where gender definition and dif-
ferentiation may have occurred: the burial evidence, dress and adornment, personal
appearance, economic activity, and family roles.
368

368    Melanie Sherratt and Alison Moore

Burial Evidence
One key area of evidence is within the burial record, where funerary monuments
and grave goods can provide clues to appearance, familial relationships, and the pos-
sibility of sexed anthropological remains suggesting the ability to ‘read off ’ gender.
However, although Roman Britain has a wealth of funerary data, there are issues in
‘reading’ a gender straight from any aspect of the burial record. On a general level,
we are hampered by the type of burial ritual itself. In the early Roman period, the
tendency was towards cremation, which causes problems in accurately osteologically
sexing the bone material (see McKinley 2000), with the deposition with multiple
forms of ceramics, the selection of which also shows little in the way of gender dif-
ferentiation (Pearce 2011: 238). Conversely, in the later Roman period, when inhuma-
tion became the more dominant burial form (thus making the osteological sexing
of skeletal material straightforward), the pattern—​particularly in the later Roman
urban cemeteries—​was towards east–​west, coffined burials with few grave goods for
both sexes, a pattern also borne out in the burial rituals (Pearce 2011: 238; see also
Weekes, this volume, on burial rituals and Pearce, this volume, on the place of status
in burial). Another, still unresolved, issue is an apparent discrepancy in the num-
bers of male and female burials, most particularly in urban contexts (e.g. Davidson
2000; Crowe 2001). Davidson (2000), utilizing data from excavations published up
to the mid-​1990s, shows that 62 per cent of adult burials were identified as male and
38 per cent as female, as against the expected adult ratio of 1:1 (Pearce 2011: 247). If
we bear these problems in mind, it may be that we need a ‘ground-​up’ approach to
the role of gender in the burial record. Rather than making broad, province-​wide
characterizations, we should look at how gender was constructed on a more regional,
cemetery-​by-​cemetery basis.
When gender analysis is attempted at the local level within cemeteries, there are indi-
cations that gender did play a part in the burial ritual in certain cases. For example, at
the cemetery at Brougham, Cumbria, the analysis of the provision of meat on the funer-
ary pyre revealed a slight gender and age distinction (Cool 2006: 218, table 18.5). While
sheep bones were predominantly found associated with adult males, bird bones (usu-
ally chicken) were more commonly found with adult females, and there was little evi-
dence for meat found with burials of individuals aged less than 18 years. Similarly, at
the late Iron Age–​mid-​Roman cemetery at Wallington Road, Baldock, Hertfordshire,
colour-​coated ceramic beakers were more commonly found with female burials (Pearce
2011: 238–​239).
From an examination on a cemetery-​by-​cemetery basis, there are also occasional
glimpses of gendered differentiation in the spatiality of some burials, most particularly
in rural contexts. Pearce’s analysis (1999) of the spatial organization of burials in rural
cemeteries in Hampshire showed that, on sites with four or more burials, the sample
was biased to either male or female burials. For example, Snell’s Corner and Burntwood
Farm showed a bias towards female burials, while Odiham, Owslesbury, and Winnall
were predominantly male burials (Pearce 1999: 156, table 1). Burials may also have been
 Gender in Roman Britain    369

utilized to gender space within rural contexts. An examination of the spatiality of the
infant burials at the rural domestic sites of Shiptonthorpe and Burnby Lane, Hayton,
East Yorkshire (Millett and Gowland 2015) suggests that the placement of infant buri-
als may have been associated with a gendering of domestic space. At Shiptonthorpe, for
example, there was a main cluster of seven individual infants (38–​40 gestational weeks)
at the eastern end of the main domestic building, one beneath its main hearth, one just
outside the wall to the south, and two others beside the northern extension, adjacent to
the watering hole. At Hayton, where occupation could be traced back to the late Iron
Age, clusterings of infant burials were found within, or associated with, domestic struc-
tures throughout the Roman period, but in the late Iron Age and late Roman periods
of occupation there was a strong preference for infant burial within or beside domestic
spaces with an easterly orientation. This pattern of easterly orientation has been identi-
fied by Scott (1991) as being particularly associated with women and childbirth, sug-
gesting that the placement of the infant burials may be specifically selected to maintain
a physical and symbolic connection between the infant and the mother/​family (Millett
and Gowland 2015). It may be that the focus on the larger Romano-​British urban
cemeteries in the archaeological record has hidden the more nuanced and localized
expressions of gendered identity. With a reappraisal of the existing burial evidence, par-
ticularly in rural contexts, and new theoretical approaches to identifying the expression
of gender, we may gain a more rounded picture of how gendered identity was expressed
in death.

Dress, Adornment, and Personal Appearance

In past societies, as now, the body is often the centre of identity construction
(Montserrat 2000: 15), and cultural constructions of gendered identities are estab-
lished, interpreted, and reinforced through the manipulation of appearance to cre-
ate a social identity (Lesick 1997: 38; Sørensen 2000:124–​143; Rosten 2008). Clothing,
dress accessories, and items used to care for the appearance of the body (so-​called toilet
items) are used to communicate an image, one that often relates to cultural ideas of
gender roles, but also relates to other aspects of identity, such as age, social status, and
ethnicity (Swift 2011: 206). One example from Roman Britain is that of the tombstone
of Regina (RIB 1065), a native woman commemorated on a mid–​late-​second-​century
ad funerary monument at Arbeia, South Shields, who is shown wearing a floor-​length,
fringed ‘Gallic’ coat (for images, see Croom 2000: figures 22 and 23; for further discus-
sion, see Wild 1968; Carroll 2012), a form of dress that appears to have predominated
in the northern provinces (Wild 1968; Rothe 2009). Regina is also depicted adorned
in jewellery and with a jewellery casket (representing her gender and social status)
and spindle and distaff (representing her idealized gendered role and feminine virtue)
(Lovén 1998; Allason-​Jones 2005). This image depicts Regina as both ‘being’ her gender
identity, through her dress and ornamentation, and ‘doing’ her gendered role, through
the imagery of the spindle and distaff.
370

370    Melanie Sherratt and Alison Moore

Further evidence for gender in Roman Britain comes through analysis of the small
finds, particularly that found in association with osteologically sexed skeletal material.
Not all small items of personal adornment were gender specific—​for example, rings
and brooches were worn by both men and women, but other items may have had more
gendered connotations. For example, cross-​bow brooches and belt-​sets, common in the
later Roman period, are frequently associated with elite males, both military and civilian
(Swift 2011: 207). Similarly, earrings and beaded necklaces and bracelets are often found
in association with female burials, as, for example, Burial 291, a female excavated from
the eastern cemetery of Roman London who was interred with multiple bracelets, rings,
and necklaces, and with eleven coins placed in a casket at her feet (Barber and Bowsher
2000: 153–​155, 165–​169) (see also Cool, this volume, for further discussion of jewellery
and Moore, this volume, for the association between jewellery and age). The fabric of
some small finds may also have gendered associations. Jet, for example, has been found
predominantly in female burials (Allason-​Jones 1996), although jet items have occa-
sionally be found, usually unworn, associated with male burials such as the Catterick
‘transgressive’ burials (Graves 951, 752) (for a full discussion, see Cool 2002a, b; 2011).
Similarly, the function of an item may also have gendered associations—​for example,
high-​bowed brooches used to fasten coarse material, such as overgarments, may have
been used by both sexes, while flatter brooches may have been used for thinner fabrics,
such as female under-​shifts (Allison 2015: 108).
Toilet items and care of the body were closely associated with Roman ideas of femi-
nine virtue and elite status. However, not all toilet items were closely associated with
either gender; for example, nail cleaners, tweezers, and cosmetic grinders were com-
monly used by both men and women (Crummy and Eckardt 2003; Eckardt 2005; Swift
2011: 208–​209). However, some items, particularly caskets, mirrors, unguent bottles, and
hairpins, may have gendered associations, being found associated with osteologically
sexed female burials (Cool 1991; 2002a: 145; 2007; Allason-​Jones 1995; Swift 2011: 208).
For example, a late-​fourth-​century ad burial of a young female aged 18–​23 years at York
was interred with a blue glass jug (perhaps for perfume and cosmetics), a glass mirror,
and items of jewellery including jet and ivory bracelets, earrings, pendants, and beads
(Leach et al. 2009). Hairpins in particular have been strongly associated with female
gender identity, as the dressing of hair was often used as pictorial and literary ‘short-
hand’ for female identity in the Roman period (Revell 2015). Elaborate hairpins, some
depicting Venus bathing or female heads, have been found in association with female
burials in Britain (Cool 1991; Allason-​Jones 1995), and sometimes these hairpins were
also of precious metals—​for example, the silver cubic-​headed hairpin decorated with
facets found with a female, aged c. 40 years, at the late Roman cemetery of Poundbury,
Dorset (Farwell and Molleson 1993: 96).
However, while funerary contexts link the physical body directly, ‘reading’ gender
specifically from artefacts can be problematic. First, we run the risk of assuming that
items of personal adornment with gendered connotations in the twenty-​first century
held the same connotations in the past. It is important to consider that the artefacts pre-
sent in a burial context represent an identity that has been consciously assembled, not
 Gender in Roman Britain    371

by the individual within the grave, but by those involved in the funerary process (van
Driel-​Murray 1995: 4; Keegan 2002: 105; Rosten 2008: 203). Furthermore, the percentage
of burials with grave goods is small, particularly into the later Roman period, and what
artefacts survive—​owing to the nature of their material—​may be representing only a
small, perhaps elite, section of society (Rosten 2008: 200). It is rare that clothing sur-
vives in Romano-​British burial contexts, so it is difficult to rely on clothing to help iden-
tify the gender of the individual and osteoarchaeological techniques are not foolproof,
particularly in assigning the correct biological sex in the very young and the elderly
(Aykroyd et al. 1999; Appleby 2010: 148). So far, the majority of archaeological examina-
tions of gender in Roman Britain have been through large-​scale cemetery excavations,
which can lead to a simplistic ‘male–​female’ viewpoint, disguising the complexity and
the potential variability of gender expression. While the Catterick ‘transgressive’ burials
have opened debate as to variable or alternative expressions of gendered identity (for
further discussion, see Matthews 1994; Dowson 2000, 2006; Voss 2000; Pinto and Pinto
2013), we need to consider how this variability was expressed in wider Roman society.
An item that an Italio-​Roman may have considered ‘effeminate’ may have been part of
male gender expression elsewhere in the Roman Empire, such as the earrings some-
times worn by Syrian and North African males (Matthews 1994). It is interesting to note
that the Catterick burials are from an area with a strong military presence in the Roman
period, and it may be that what is being represented is a different ‘cultural’ expression,
along the lines of the north African-​style pottery found in certain military contexts in
northern Britain (see Swan 1992, 1999).

Economic Activity
Another way to approach gender is through ideas of labour and gendered divisions of
labour. Revell (2015) argues that ‘economic activity can be a powerful way to structure the
ideals of gender, both in terms of the activities appropriate for various genders, but also
in the ways these activities might be deemed of unequal value to the society, which then
feeds into the creation of gender hierarchies’. Unfortunately, little work has been done into
the place of work within the creation of gender identity and gender roles in the Roman
world, partly because of our unspoken assumptions as to the ‘value’ of work and to Roman
cultural assumptions of gender behaviours, exacerbated by the nature of the evidence that
has come down to us. As Revell (2015) suggests, economic engagement (that is, male work)
has been seen as having a wider social value than care-​giving (female work); as women
have been stereotyped as care-​givers, their contribution to economic production in the
past has been underestimated. Treggiari (1979) and Kampen (1981) analysed iconographic
evidence from the Mediterranean world and found that, while some non-​elite women
were noted as having occupations outside of the home, it was only rarely recorded pictori-
ally or in the written evidence. For example, Treggiari (1979: 66) noted a 6:1 bias towards
male occupations being recorded, a statistic that she puts down to work not being seen
as an integral part of a Roman female’s identity (for further discussion and examples, see
372

372    Melanie Sherratt and Alison Moore

also Knapp 2011: 80–​90). Similarly, Revell’s analysis (2015) of funerary monuments in the
agricultural and cloth-​producing regions of Gallia Belgica shows that, while men were
depicted as being involved in all levels of economic activity, women were represented as
the passive consumers of the wealth being produced, shown at banquets, at their toilet, or
within family scenes. This pictorial representation, Revell (2015) argues, underlines the
power of the dominant male ideology and shows how gender differentiation was negoti-
ated and gender hierarchies were legitimized.
For Roman Britain the evidence for gendered divisions of labour is scarce. Excavations
at the military cemetery at Brougham have recorded a possible gender association with
artefacts associated with drinking. Large copper vessels for mixing drinks were asso-
ciated with female graves, while male burials were associated with glass drinking ves-
sels, suggesting a gendered division between preparation and consumption (Cool 2004,
2011). A further gender division of activity has been recorded through the bioarchaeo-
logical evidence. One of the most effective ways to grind corn is to ‘bed’ the quern stone
at ground level with the operator sitting or squatting alongside (Cool 2006: 74). Skeletal
evidence from the Roman cemetery at Poundbury, Dorset, showed that nearly three
times more women than men showed skeletal modification to the tibia (lower leg) asso-
ciated with prolongued periods of squatting (known as ‘squatting facets’), suggesting
that squatting down to grind grain may primarily have been a woman’s task (Farwell and
Molleson 1993: 200; Cool 2006: 74). Similarly, the patterning of osteoarthritic changes to
the vertebrae differed between males and females at Poundbury, with males showing a
tendency to a greater degree of severity of osteoarthritic changes associated with drag-
ging or lifting activities, while females recorded more localized changes to the lower
back, indicating that their activities involved carrying loads in the arms (Farwell and
Molleson 1993: 201–​202), which may be indicative of a gendered division of labour.
Changes within the patterning of grave-​good provision may, dependent upon context,
also be indicative of transformations within patterns of economic activity. For exam-
ple, Moore (2010) analysed a sample of burials of males and females from urban and
rural contexts from southern Roman Britain and identified a division in the burial treat-
ment (provision of grave goods and coffins) of elderly males (50+ years) and younger
males (30–​49 years) in rural contexts, not recorded between men from the urban group.
She argues that this difference in burial treatment for elderly males may be linked to
how ageing men were perceived in relation to their economic activity. While older men
would retain important knowledge and experience, the ill-​health and physical degen-
eration associated with old age may have meant that they could no longer engage fully
with the labour-​intensive work required in a predominantly agricultural society such as
Roman Britain. As such, the perceived ‘economic value’ of their work labour changed,
represented by a transition in their treatment in burial (Moore 2010: 172–​173).

Social, Familial, and Domestic Structures

Within the context of the Roman world family structure was often idealized as hus-
band, wife, children, and extended familia, and is represented as such on funerary
 Gender in Roman Britain    373

reliefs and in the epigraphic record. However, how the impact of the conquest of Rome
on the provinces and the corresponding cultural changes may have disrupted or reor-
ganized gender relations is still relatively under-​explored (Revell 2010; although see
Millett 1990: 60, for the influence of women on disseminating change; Woolf 1998: p.
ix, on masculinity; and Mattingly 2011: 94–​121 on disempowerment of women). Revell
(2015) suggests that a high level of army recruitment, in, for example, Batavia or north-​
western Iberia, may have led to a pronounced gender imbalance, with a correspond-
ing overall female prominence, or the restructuring of the social hierarchies between
married and unmarried women. It would be instructive to identify the possibility of
changing gender relations within the militarized zone of Roman Britain; as women and
children have been identified as being present in forts in Britain (e.g. van Driel Murray
1995; Allason-​Jones 1999; see also Allison 2013), it may be that these women, both local
and from across the empire, may have had to create a new model of feminine iden-
tity and behaviour that incorporated differing ethnicities’ ideas of femininity (Revell
2015). Another impact of cultural change on gender relations may have been in the gen-
der differentiation of the pace of acceptance of new behaviours and material cultures
(Revell 2015). For example, was the early adoption of ‘toilet instruments’ by Romano-​
Britons identified by Hill (1997) first driven by women, as a new method of expressing
female identity or by men, altering their appearance to adapt within the changed social
dynamics of the first century ad? It may be, as Revell (2015) argues, that these types of
questions are not suited to a general, province-​wide, approach and need detailed, local
case studies fully to understand the relationship between cultural change and gender
relations, but we do need to begin to address these questions to the material culture
record of Roman Britain to develop a way forward to a deeper understanding of gender
roles and identities.
Another area in which we may find clues as to gendered behaviours is within the spa-
tiality of urban and domestic structures. It has been argued that the forum was an engen-
dered space, where the buildings were designed for use by men (Richlin 1997; Revell
2009:  161), while bathhouses with separate suites, such as at London and Bath, may
point to gender segregation in bathing (Revell 2009: 117–​178). While there was no obvi-
ous gender separation in the structure or layout of the Roman house (Wallace-​Hadrill
1988: 50–​52), unlike Greek houses (e.g. Goldberg 1999; Nevett 2001), it has been argued
that the spatial use of the house was organized temporally, with men and women using
the space at different times (Laurence 1994: 127–​128). Similarly, Allison (1999) examined
the distribution of artefacts in Pompeiian houses, and identified a possible distribu-
tion of ‘feminine’ material culture (for example, items associated with spinning, weav-
ing, and needlework) associated with smaller rooms off the atrium, although, as Allison
(1999: 352–​353) admits, these items may also have been used by slaves (see also Allison
2007 and 2015 on the use of gendered approaches to material culture of sites). Although
little work of this nature has been done in Roman Britain, Hingley (1990) attempted to
introduce the concept of gendered space in Iron Age and Roman period houses, argu-
ing that space was segregated through a concept of public (male) and private (female)
demarcation. Although such ‘structuralist’ approaches have been criticized as treating
gender categories as fixed and static (Sørensen 2000: 147), they do allow ‘hidden’ aspects
374

374    Melanie Sherratt and Alison Moore

of gendered behaviour to be considered within the Roman archaeological record and,

eventually, to be accepted as a legitimate and vital area of study.

Future Directions for Gender Studies

in Roman Britain

As discussed above, much work still needs to be done to understand gender identity and
gender roles in Roman Britain and consideration of questions of the economic roles of
men and women, gender relations in the face of cultural change, and the gendered uti-
lization of space all hold potential in advancing our understanding of gender. Another
fruitful area of study is bioarchaeology (for Romano-​British examples, see, e.g. papers
by Redfern 2003; Gowland 2004; Fuller et al. 2006; Gowland and Redfern 2010; Redfern
et al. 2010, 2012, 2015). Through a detailed study of skeletal remains, it is possible to
understand whether nutritional markers showed a gender distinction, thereby indi-
cating whether cultural practices such as preferential food allocation were gendered.
For example, Bonsall’s analysis (2014) of the teeth of a skeletal sample from Ancaster,
Lincolnshire, and Winchester, Hampshire, showed no statistically significant gendered
difference between males and females in the rate of tooth loss, caries, or periodontal
disease, suggesting that men and women shared a broadly similar diet. Bioarchaeology
can also be used to infer gender-​related activities and divisions of labour (see Holliman
2011: 152–​156). Muscular-​skeletal markers and patterns of degenerative joint disease can
provide information on who was performing which tasks regularly, while the study of
trauma patterning can inform on gender-​related violence (for examples, see Robb 1997;
Redfern 2013). It is through amalgamating all the strands of evidence, from material cul-
ture through to the bioarchaeology, that we can gain a deeper and nuanced understand-
ing of gendered identity in Roman Britain.

Conclusion

The study of gender in Roman Britain has developed quickly since its inception in the late
1980s and early 1990s, and we are now reasonably confident in identifying certain items
of material culture, such as toilet items, crossbow brooches, and jet jewellery, as having
a ‘gendered’ identity. However, gender as an area of legitimate study is still running the
risk of being ‘ghettoized’ as the search for women in the past, and we need to think about
and develop more detailed approaches to identifying and understanding masculinity and
‘third’ genders. By broadening our approaches away from reliance on the cemetery evi-
dence to encompass questions of economic activity, gendered reaction to cultural change,
spatiality, and bioarchaeology, it will be possible to gain a more coherent and rounded
 Gender in Roman Britain    375

picture of how gender was created and expressed in the province. Roman Britain holds
great potential for informing us on gender identity; it is up to us to ask the right questions.

References
Allason-​Jones, L. (1989). Women in Roman Britain. London: British Museum Press.
Allason-​ Jones, L. (1995). ‘”Sexing” Small Finds’, in P. Rush (ed.), Theoretical Roman
Archaeology: Second Conference Proceedings. Worldwide Archaeology Series 14. Aldershot:
Avebury, 22–​32.
Allason-​Jones, L. (1996). Roman Jet in the Yorkshire Museum. York: Yorkshire Museum.
Allason-​Jones, L. (1999). ‘Women and the Roman Army in Britain’, in A. Goldsworthy and I.
Haynes (eds), The Roman Army as a Community. JRA Supplementary Series 34. Portsmouth,
RI: Journal of Roman Archaeology, 41–​52.
Allason-​Jones, L. (2005). Women in Roman Britain. York: Council for British Archaeology.
Allason-​Jones, L. (2012). ‘Women in Roman Britain’, in S. L. James and S. Dillon (eds),
Companion to Women in the Ancient World. Oxford: Wiley-​Blackwell, 467–​477.
Allison, P. E. (1999). ‘Artefact distribution and spatial function in Pompeian houses’, in B.
Rawson and P. Weaver (eds), The Roman Family in Italy: Status, Sentiment, Space. Oxford:
Oxford University Press, 321–​354.
Allison, P. E. (2007). ‘Engendering Roman Domestic Space’, in R. Westgate, N. Fisher, and J.
Whitley (eds), Building Communities: House, Settlement and Society in the Aegean World
and Beyond. Proceedings of a Conference held at Cardiff University, 17–​21 April 2001. British
School at Athens Studies 15. London: British School at Athens, 343–​350.
Allison, P. E. (2013). People and Spaces in Roman Military Bases. Cambridge:  Cambridge
University Press.
Allison, P. E. (2015). ‘Characterizing Roman Artefacts to Investigate Gendered Practices in
Contexts without Sexed Bodies’, American Journal of Archaeology, 119/​1: 103–​123.
Appleby, J. E. P. (2010). ‘Why we Need an Archaeology of Old Age and a Suggested Approach’,
Norwegian Archaeological Review, 43/​2: 145–​168.
Aykroyd, R., Lucy, D., Pollard, M. A., and Roberts, C. A. (1999). ‘Nasty, Brutish but Necessarily
Short: A Reconsideration of the Statistical Methods Used to Calculate Age at Death from
Adult Human Skeletal and Dental Age Indicators’, American Antiquity, 61/​1: 55–​70.
Baker, P. (2003). ‘A Brief Comment on the TRAC Session Dedicated to Interdisciplinary
Approaches to the Study of Roman Women’, in G. Carr, E. Swift, and J. Weekes (eds),
TRAC 2002: Proceedings of the Twelfth Annual Theoretical Roman Archaeology Conference,
Canterbury 2002. Oxford: Oxbow Books, 140–​146.
Barber, B., and Bowsher, D. (2000). The Eastern Cemetery of Roman London, Excavations 1983–​
1990. London: Museum of London Archaeology Service.
Bonsall, L. (2014). ‘A Comparison of Female and Male Oral Health in Skeletal Populations from
Late Roman Britain: Implications for Diet’, Archives of Oral Biology, 59/​12: 1279–​1300.
Butler, J. (1999). Gender Trouble: Feminism and the Subversion of Identity. 10th anniversary edi-
tion. New York and London: Routledge.
Carroll, M. (2012). ‘“The Insignia of Women”: Dress, Gender and Identity on the Roman
Funerary Monument of Regina from Arbeia’, Archaeological Journal, 169/​1: 281–​311.
Cocks, A. H. (1921). ‘A Romano-​British Homestead in the Hambleden Valley’, Archaeologia,
71: 141–​198.
376

376    Melanie Sherratt and Alison Moore

Conkey, M. W., and Spector, J. (1984). ‘Archaeology and the Study of Gender’, in M. Schiffer
(ed.), Advances in Archaeological Method and Theory 7, 1–​38. London: Academic Press.
Cool, H. E. M. (1991). ‘Roman Metal Hair Pins from Southern Britain’, Archaeological Journal,
147: 148–​182.
Cool, H. E.  M. (2000). ‘The Parts Left Over:  Material Culture into the 5th Century’, in T.
Wilmott and P. Wilson (eds), The Late Roman Transition in the North. BAR British Series
299. Oxford: Archaeopress, 47–​65.
Cool, H. E. M. (2002a). ‘Bottles for Bacchus?’, in M. Aldhouse-​Green and P. Webster (eds),
Artefacts and Archaeology: Aspects of the Celtic and Roman World. Cardiff: University of
Wales Press, 132–​151.
Cool, H. E. M. (2002b). ‘An Overview of the Small Finds from Catterick’, in P. R. Wilson (ed.),
Cataractonium: Roman Catterick and its Hinterland: Excavations and Research 1958–​1997,
Part II. Council for British Archaeology Research Report 129. York:  Council for British
Archaeology, 24–​43.
Cool, H. E. M. (2004). The Roman Cemetery at Brougham, Cumbria: Excavations 1966–​67.
London: Society for the Promotion of Roman Studies.
Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge
University Press.
Cool, H. E. M. (2007). ‘The Small Finds in their Regional Context’, in D. Miles, S. Palmer,
A. Smith, and G. P. Jones, Iron Age and Roman Settlement in the Upper Thames
Valley: Excavations at Claydon Pike and Other Sites within the Cotswold Water Park. Thames
Valley Landscapes Monograph 27. Oxford: Oxford Archaeology, 342–​350.
Cool, H. E.  M. (2011). ‘Funerary Contexts’, in L. Allason-​Jones (ed.), Artefacts in Roman
Britain: Their Purpose and Use. Cambridge: Cambridge University Press, 293–​312.
Croom, A. (2000). Roman Clothing and Fashion. Stroud: Tempus.
Crowe, F. (2001). ‘Women, Burial Data and Issues of Inclusion: The Problems and Potentials
of Romano-​British Cemeteries’, in S. Dixon (ed.), Childhood, Class and Kin in the Roman
World. London: Routledge, 144–​162.
Crummy, N., and Eckardt, H. (2003). ‘Regional Identities and Technologies of the Self: Nail
Cleaners in Roman Britain’, Archaeological Journal, 160: 44–​69.
Davidson, C. (2000). ‘Gender Imbalances in Romano-​British Cemetery Populations: A Revaluation
of the Evidence’, in J. Pearce , M. Millet, and M. Struck (eds), Burial, Society and Context in the
Roman World. Oxford: Oxbow Books, 231–​237.
Díaz-​Andreu, M. (2005). ‘Gender Identity’, in M. Díaz-​Andreu, S. Lucy, S. Babić, and D. N.
Edwards (eds), The Archaeology of Identity: Approaches to Gender, Age, Status, Ethnicity and
Religion. London: Routledge, 13–​42.
Dowson, T. (2000) (ed.). ‘Why Queer Archaeology? An Introduction’, World Archaeology,
32: 161–​165.
Dowson, T. (2006). ‘Archaeologists, Feminists and Queers: Sexual Politics in the Construction
of the Past’, in P. L. Geller and M. K. Stockett (eds), Feminist Anthropology: Past, Present and
Future. Philadelphia: University of Philadelphia Press, 89–​102.
Eckardt, H. (2005). ‘The Social Distribution of Roman Artefacts: The Case of Nail-​Cleaners
and Brooches in Britain’, Journal of Roman Archaeology, 18: 139–​160.
Farwell, D. E., and Molleson, T. L. (1993). Excavations at Poundbury 1966–​1980. Volume II: The
Cemeteries. Dorchester: Dorset Natural History and Archaeological Society.
Fuller, B. T., Molleson, T. I., Harris, D. A, Gilmour, L. T., and Hedges, R. E. M. (2006). ‘Isotopic
Evidence for Breastfeeding and Possible Adult Dietary Differences from Late/​Sub-​Roman
Britain’, American Journal of Physical Anthropology, 129/​1: 45–​54.
 Gender in Roman Britain    377

Gilchrist, R. (1999). Gender and Archaeology: Contesting the Past. London: Routledge.

Goldberg, M. Y. (1999). ‘Spatial and Behavioural Negotiation in Classical Athenian City
Houses’, in P. M. Allison (ed.), The Archaeology of Household Activities. London: Routledge,
142–​161.
Gowland, R. (2004). ‘The Social Identity of Health in Late Roman Britain’, in B. Croxford, H.
Eckardt, J. Meade, and J. Weekes (eds), TRAC 2003: Proceedings of the Thirteenth Theoretical
Roman Archaeology Conference, Leicester 2003. Oxford: Oxbow Books, 135–​146.
Gowland, R., and Redfern, R. (2010). ‘Childhood Health in the Roman World: Perspectives
from the Centre and Margin of the Empire’, Childhood in the Past, 3/​1: 15–​42.
Hill, J. (1997). ‘The End of One Kind of Body and the Beginning of Another Kind of Body?
Toilet Instruments and Romanization in Southern England during the 1st Century ad’, in
A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow Books,
96–​107.
Hill, J. D. (2001). ‘Romanisation, Gender and Class:  Recent Approaches to Identity in
Britain and their Possible Consequences’, in S. James and M. Millett (eds), Britons and
Romans: Advancing an Archaeological Agenda. Council for British Archaeology Research
Report 125. York: Council for British Archaeology, 12–​18.
Hingley, R. (1990). ‘Domestic Organisation and Gender Relations in Iron Age and
Romano-​ British Households’, in R. Samson (ed.), The Social Archaeology of Houses.
Edinburgh: Edinburgh University Press, 125–​147.
Holliman, S. E. (2011). ‘Sex and Genderi Bioarchaeological Research:  Theory, Method and
Interpretation’, in S. C. Agarwal and B. A. Glencross (eds), Social Bioarchaeology. Blackwell
Studies in Global Archaeology 14. Chichester, West Sussex: Wiley-​Blackwell, 149–​182.
Kampen, N. B. (1981). Image and Status: Working Women in Ostia. Berlin: Gebr. Mann Verlag.
Keegan, S. L. (2002). Inhumation Rites in Late Roman Britain: The Treatment of the Engendered
Body. BAR British Series 333. Oxford: Archaeopress.
Knapp, A. B. (1998). ‘Boys Will Be Boys: Masculinist Approaches to a Gendered Archaeology’, in
K. Hays-​Gilpin and D. S. Whitley (eds), Reader in Gender Archaeology. London: Routledge,
365–​373.
Knapp, R. (2011). Invisible Romans. London: Profile Books.
Laurence, R. (1994). Roman Pompeii: Space and Society. London: Routledge.
Leach, S., Lewis, M., Chenery, C., Muldner, G., and Eckardt, H. (2009). ‘Migration and Diversity
in Roman Britain:  A  Multidisciplinary Approach to the Identification of Immigrants in
Roman York, England’, American Journal of Physical Anthropology, 140/​3: 546–​561.
Lesick, K. S. (1997). ‘Re-​Engendering Gender: Some Theoretical and Methodological Concerns
on a Burgeoning Archaeological Pursuit’, in J. Moore and E. Scott (eds), Invisible People and
Processes: Writing Gender and Childhood into European Archaeology. Leicester: Leicester
University Press, 31–​41.
Little, B. J., and Shackel, P. A. (2007). Archaeology as a Tool of Civic Engagement. Walnut Creek,
CA: Altamira.
Lovén, L. L. (1998). ‘Wool-​Working and Female Virtue’, in L. L. Lovén and A. Strömberg (eds),
Aspects of Women in Antiquity. Jonsered: Aströms, 85–​95.
McKinley, J. (2000). ‘The Analysis of Cremated Bone’, in M. Cox and S. Mays (eds), Human
Osteology in Archaeology and Forensic Science. London: Greenwich Medical Media, 403–​421.
Matthews, K. (1994). ‘An Archaeology of Homosexuality? Perspectives from the Classical
World’, in S. Cottam, D. Dungworth, S. Scott, and J. Taylor (eds), TRAC 94: Proceedings of the
Fourth Annual Theoretical Roman Archaeology Conference, Durham 1994. Oxford: Oxbow
Books, 118–​132.
378

378    Melanie Sherratt and Alison Moore

Mattingly, D. (2011). Imperialism, Power and Identity:  Experiencing the Roman Empire.
Princeton: Princeton University Press.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Millett, M., and Gowland, R. (2015). ‘Infant and Child Burial Rites in Roman Britain: A Study
from East Yorkshire’, Britannia 46, 171–198.
Montserrat, D. (2000). ‘Reading Gender in the Roman World’, in J. Huskinson (ed.),
Experiencing Rome: Culture, Identity and Power in the Roman Empire. London: Routledge,
153–​181.
Moore, A. (2010). ‘Age and Identity in Funerary Contexts: The Elderly in Southern Roman
Britain’, in A. Moore, G. Taylor, E. Harris, P. Girdwood, and L. Shipley (eds), TRAC
2009:  Proceedings of the Nineteenth Annual Theoretical Roman Archaeology Conference,
Michigan and Southampton 2009. Oxford: Oxbow Books, 105–​119.
Morelli, A. (2009). Roman Britain and Classical Deities: Gender and Sexuality in Roman Art.
Oxford: Archaeopress.
Nevett, L. C. (2001). House and Society in the Ancient Greek World. Cambridge: Cambridge
University Press.
Pearce, J. (1999). ‘The Dispersed Dead: Preliminary Observations on Burial and Settlement
Space in Rural Roman Britain’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC
98: Proceedings of the Eighth Annual Theoretical Roman Archaeology Conference, Leicester
1998. Oxford: Oxbow Books, 151–​162.
Pearce, J. (2011). ‘Representations and Realities: Cemeteries as Evidence for Women in Roman
Britain’, Medicina nei Secoli, 23(1): 227–​254.
Pinto, R., and Pinto, L. C. G. (2013). ‘Transgendered Archaeology: The Galli and the Catterick
Transvestite’, in A. Bokern, M. Bolder-​Boos, Stefan Krmnicek, D. Maschek, and S. Page (eds),
TRAC 2012: Proceedings of the Twenty-​Second Theoretical Roman Archaeology Conference,
Frankfurt 2012. Oxford: Oxbow Books, 169–​182.
Redfern, R. (2003). ‘Sex and the City: A Biocultural Investigation into Female Health in Roman
Britain’, in G. Carr, E. Swift, and J. Weekes (eds), TRAC 2002:  Proceedings of the Twelfth
Annual Theoretical Roman Archaeology Conference, Canterbury 2002. Oxford: Oxbow Books,
147–​170.
Redfern, R. C. (2013). ‘Violence as an Aspect of the Durotriges Female Life Course’, in S. Ralph
(ed.), The Archaeology of Violence: Interdisciplinary Approaches. Albany, NY: State University
of New York Press, 63–​97.
Redfern, R. C., Hamlin, C., and Beavan Athfield, N. (2010). ‘Temporal Changes in Diet: A Stable
Isotope Analysis of Late Iron Age and Roman Dorset, Britain’, Journal of Archaeological
Science, 37: 1149–​1160.
Redfern, R. C., Millard A. R., and Hamlin, C. (2012). ‘A Regional Investigation of Subadult
Dietary Patterns and Health in Late Iron Age and Roman Dorset, England’, Journal of
Archaeological Science, 39/​5: 1249–​1259.
Redfern, R. C., De Witte, S. N., Pearce, J., Hamlin, C., and Dinwiddy, K. E. (2015). ‘Urban–​
Rural Differences in Roman Dorset, England: A Bioarchaeological Perspective on Roman
Settlements’, American Journal of Physical Anthropology, 157/1: 107–120.
Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge
University Press.
Revell, L. (2010). Romanization:  A  Feminist Critique, in A. Moore, G. Taylor, E. Harris, P.
Girdwood, and L. Shipley (eds), TRAC 2009: Proceedings of the Nineteenth Annual Theoretical
 Gender in Roman Britain    379

Roman Archaeology Conference, Michigan and Southampton 2009. Oxford:  Oxbow

Books, 1–​10.
Revell, L. (2015). Ways of Being Roman:  Discourses of Identity in the Roman West.
Oxford: Oxbow Books.
Richlin, A. (1997). ‘Gender and Rhetoric: Producing Manhood in the Schools’, in W. J. Dominik
(ed.), Roman Eloquence: Rhetoric in Society and Literature. London: Routledge, 99–​110.
Robb, J. (1997). ‘Violence and Gender in Early Italy’, in D. L. Martin and D. W. Frayer (eds),
Troubled Times: Violence and Warfare in the Past. Amsterdam: Gordon and Breach, 111–​144.
Rosten, J. (2008). ‘Appearance, Diversity and Identity in Roman Britain’, in C. S. Colburn and
M. K. Heyn (eds), Reading a Dynamic Canvass: Adornment in the Ancient Mediterranean
World. Newcastle-​upon-​Tyne: Cambridge Scholars Publishing, 194–​222.
Rothe, U. (2009). Dress and Cultural Identity in the Rhine-​Moselle Region of the Roman Empire.
BAR International Series 2038. Oxford: Archaeopress.
Scott, E. (1991). ‘Animal and Infant Burials in Romano-​ British Villas: A Revitalization
Movement’, in P. Garwood, D. Jennings, R. Skeates, and J. Toms (eds), Sacred and Profane:
Proceedings of a Conference on Archaeology, Ritual and Religion, Oxford 1989. Oxford
Committee for Archaeology Monograph 32. Oxford: Oxbow Books, 115–​121.
Scott, E. (1993) (ed.). Theoretical Roman Archaeology: First Conference Proceedings. Aldershot:
Avebury.
Scott, E. (1995). ‘Women and Gender Relations in the Roman Empire’, in P. Rush (ed.),
Theoretical Roman Archaeology:  Second Conference Proceedings. Aldershot:  Avebury,
174–​189.
Scott, E. (1997). ‘Introduction:  On the Incompleteness of Archaeological Narratives’, in J.
Moore and E. Scott (eds), Invisible People and Processes: Writing Gender and Childhood in
European Archaeology. Leicester: Leicester University Press, 1–​12.
Scott, E. (2001). ‘Unpicking a Myth:  The Infanticide of Female and Disabled Infants in
Antiquity’, in D. Davies, A. Gardner, and K. Lockyer (eds), TRAC 2000: Proceedings of the
Tenth Annual Theoretical Roman Archaeology Conference, London 2000. Oxford:  Oxbow
Books, 142–​151.
Sørensen, M-​L. S. (2000). Gender Archaeology. Malden, MA: Polity Press.
Swan, V. G. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman
Pottery Studies, 5: 1–​33.
Swan. V. G. (1999). Ethnicity, Conquest and Recruitment: Two Case Studies from the Northern
Military Provinces. Journal of Roman Archaeology Supplementary Series 72. Portsmouth,
RI: Journal of Roman Archaeology.
Swift, E. (2011). ‘Personal Ornament’, in L. Allason-​Jones (ed.), Artefacts in Roman Britain: Their
Purpose and Use. Cambridge: Cambridge University Press, 194–​218.
Treggiari, S. (1979). ‘Lower Class Women in the Roman Economy’. Florilegium, 1: 65–​86.
van Driel-​Murray, C. (1994). ‘A Question of Gender in a Military Context’, Helinium, 34/​2:
342–​362.
van Driel-​Murray, C. (1995). ‘Gender in Question’, in P. Rush (ed.), Theoretical Roman
Archaeology:  Second Conference Proceedings. Worldwide Archaeology Series 14.
Aldershot: Avebury, 3–​21.
van Driel-​ Murray, C. (1997). ‘Women in Forts?’ Jahresbericht der Gesellschaft Pro
Vindonissa, 55–​61.
Voss, B. L. (2000). ‘Feminisms, Queer Theories and the Archaeological Study of Past
Sexualities’, World Archaeology, 32: 180–​192.
380

380    Melanie Sherratt and Alison Moore

Wallace-​Hadrill, A. (1988). ‘The Social Structure of the Roman House’, Papers of the British
School at Rome, 56: 43–​97.
Wild, J. P. (1968). ‘Clothing in the North-​West Provinces of the Roman Empire’, Bonner
Jahrbücher, 168: 166–​240.
Woolf, G. (1998). Becoming Roman:  The Origins of Provincial Civilisation in Gaul.
Cambridge: Cambridge University Press.
 Chapter 19

Deviancy i n L at e
Romano-​B rit i sh  Bu ria l

Belinda Crerar

Introduction

No single explanation for the treatment of the deceased applies to all human societies.
Ritualized disposal of the dead has been recognized archaeologically in Europe from
at least the Upper Palaeolithic (Bello and Andrews 2009: 14–​15) and, from this point
onwards, behaviour in response to death has varied dramatically between different cul-
tures at different times.
The identity of the deceased and their social position are a key theme in research that
addresses death and burial in past cultures. One aspect that has attracted increased
attention in recent debates is that of ‘deviancy’—​a term used to describe those indi-
viduals who were shunned by their peers or excluded outright from society (Murphy
2008a; Aspöck 2009; Reynolds 2009; Crerar 2012; Tucker 2012). In one crucial respect,
deviancy differs from other aspects of identity (such as sex/​gender, age, or health), since
it depends on other people’s responses to an individual rather than anything intrinsic
to the individual themselves. Consequently, it cannot be detected through osteologi-
cal or artefactual markers from a single burial: for example, a skeleton showing signs of
kyphosis tells us only that that individual suffered from that condition, not that society
was prejudiced against him or her because of it (Shay 1985: 22). Reasons for stigmatizing
certain persons are highly socially specific and may rely on factors that are difficult to
reconstruct historically or archaeologically, such as local superstitions. Since the study
of deviancy is gaining renewed attention, it is important to reconsider here the method-
ologies that have been used in identifying deviancy archaeologically and the theoretical
models employed to support these processes, drawing on broader archaeological and
anthropological debate.
The history of scholarship on ‘deviancy’ has followed two largely distinct paths, each
approaching the issue from opposite angles. On one side, theoretical discussions based
382

382   Belinda Crerar

on anthropological and ethnological case studies have looked at the treatment of known
socially stigmatized persons in living communities and how burial practices were
adapted to reflect these individuals’ outsider status, if at all (Frazer 1932–​3; Hertz 1960
[1907]; Saxe 1970). Theories have then been extrapolated to explain how social deviants
might have been treated in any past society. Alternatively, archaeologists making dis-
coveries of specific burials that diverge from the expected norm for a region and period
have often proposed social exclusion as an explanation for their apparent aberrance
(MacDonald 1979; Matthews 1981; Boylston et al. 2000; Anderson 2001).
Both approaches have certain value, one for its methodology and the other for the
close analysis of the material remains. However, despite the great potential for overlap,
they have rarely been combined in a conscious and self-​aware manner. A brief overview
of the primary themes of both approaches is offered here to highlight areas of mutual
benefit, but also to draw attention to the dangers of applying theoretical models whole-
sale without due consideration of specific archaeological evidence.

Past Approaches to ‘Deviancy’

The New Archaeology movement of the 1960s and 1970s introduced the idea that social
‘deviancy’ should be considered as a unified phenomenon to archaeological discourse.
Theorists of this period argued that the treatment of individuals following death was
directly dependent on their social identity during life (which could include manner of
death) and was enacted by the burial party according to prescribed systems (Saxe 1970;
Binford 1971; Trigger 2006: 386–​483). Consequently, differences between burials within
the same society could be explained by differences in identity and social position of the
deceased. The term ‘deviant’ was employed to refer to a particular social state whereby
individuals were perceived to have transgressed social codes in a way that would have
affected their social identity, being ultimately reflected in their manner of burial. As
summarized by Saxe (1970: 10-​11):

… “deviance” defines the ego as having breached the rights/​duties relationships with
alter egos and hence brings an end to normal reciprocity, a ‘deviant’ life and/​or death
would elicit only the social persona culturally congruent with that deviance, that is,
one that lacks the right (among others) to “normal” treatment.

It has sometimes been argued that the use of the term ‘deviant’ by scholars of this
period stemmed from their enthusiasm for quantitative and scientific approaches
towards archaeological material and neutrally described less-​numerous burial rites
without interpretative commentary (Aspöck 2009: 11). However, examples of ‘deviant’
identities given—​most notably disabled persons and criminals—​betray an association
between ‘deviancy’ and negative social position even at this early stage of the discipline
(Saxe 1970: 118–​119).
 Deviancy in Late Romano-British Burial    383

According to the New Archaeology model, a ‘deviant’ social identity in life was
believed to manifest archaeologically in a burial that eschewed all other indicators
of identity normally employed by the society in question, since the deviancy of the
individual overshadowed or rendered invalid all other aspects of identity. Saxe (1970:
10–​11) proposed that ‘deviant’ burials expressed only the identity of deviance itself and
therefore they could be recognized archaeologically by few attributes and low invest-
ment of labour. Though writing more than a decade later, Shay closely aligned herself
with this theoretical position when discussing the issue of ‘deviancy’. She, however,
expanded the definition of the term, arguing that ‘deviancy’ should be understood
as a sliding scale from ‘the very saintly to the most sinful acts’ (Shay 1985: 223–​226).
Following the model originally proposed by Saxe, Shay (1985: 266) maintained that
individuals at the negative end of this scale possessed a ‘shallow social persona’ (that
is, one comprising few social identities), which was expressed in their burial by a low
number of ‘funerary components’. Those deemed to possess a ‘positive deviant iden-
tity’ were afforded a ‘complex social persona’, which manifested in the provision of a
high number of ‘funerary components’. While most would now agree that the concept
that social identity can be deduced through quantifiable grave offerings is overly sim-
plistic, Shay’s analysis offered a new direction for the debate by speculating that differ-
ential treatment in burial need not stem solely from negative perceptions surrounding
the deceased. The importance of her work, and of others who have also argued for
‘positive deviancy’ (e.g. Meyer-​Orlac 1997; Pader 1980, 1982; Philpott 1991), is particu-
larly pertinent now, since the more recent scholarship on the subject displays an over-
whelming and almost automatic tendency to equate ‘deviancy’ with negative social
identity.
New Archaeology (or processualism, as it became known) offered ample scope for
discussions of ‘deviant burials’, as it was keen to demonstrate how variations in social
systems maintained or undermined social cohesion. However, in the late 1970s and
1980s, archaeologists began to question the validity of the systematized model of human
society that processualism had proposed. As a result, more central aspects of social
interactions and identity required renewed attention, and the study of ‘deviancy’ was
pushed to one side in favour of increasing our understanding of normative patterns of
behaviour. Consequently, while other research on burial and identity advanced signifi-
cantly from the processual model, the limited scholarship on ‘deviancy’ at this time con-
tinued to turn towards processualist thought for a theoretical scaffold. Even among the
most prominent post-​processualist theorists, there remained a tangible adherence to
Saxe’s initial propositions:

Those dead whose lives or deaths symbolise central social values may be placed in
opposition to the deviant dead whose actions have threatened the social order. In
medieval and post-​medieval Britain, for example, deviants such as witches, executed
criminals, suicides and women dying in childbirth were among the social groups
whose treatment in death set them apart from the worthy dead. The symbolic asso-
ciations of the disposal contexts of these two groups can be understood in positive
384

384   Belinda Crerar

or negative terms: the sanctified ground in the churchyard was contrasted with the
limbo of evil spirits at the crossroads. (Parker Pearson 1993: 206-​207)

The differences between this and earlier theoretical work is that, having rejected the
correlation between social identity and the material complexity of burial, anthropolo-
gists of the post-​processualist era were reticent about how to recognize the deviant dead
archaeologically. Consequently, many theoretical and anthropological papers decline to
define precisely what practices they considered to be ‘deviant’ through an acknowledge-
ment that ‘deviancy’ is dependent on normality and normality is particular to an indi-
vidual society.
At the same time as these theoretical debates were occurring, European archae-
ologists were uncovering human burials that defied their expectations for appropriate
disposal of the dead, specifically those in non-​supine postures, buried outside of des-
ignated cemeteries or showing evidence for human manipulation of the corpse or skel-
eton. Their responses to this material followed a fairly consistent theme that stretched
back to at least the late nineteenth century. The earliest reports tended to explain prone
burials as mistakes in the handling of closed coffins or even insobriety on the part of the
undertaker (Rolleston 1869: 477). Decapitated burials were almost invariably linked to
fear of the dead (Royce 1883: 76–​77; Hollingworth and O’Reilly 1925; Moir and Maynard
1931; Collingwood 1937: 239; Calkin 1947, 1953; Wright 1953: 123; Dewar 1958). These ideas
permeated discussions on these burials to the extent that, in 1977, Liversidge claimed
that decapitation was ‘a well-​known way of laying a ghost’ (Liversidge 1977: 35). No theo-
retical models were advanced to support these interpretations, and one is left with the
sense that they were made on an entirely intuitive basis.
During the 1960s and 1970s alternative explanations were proposed, but they main-
tained the tendency to connect such burials with negative social positions, indifference,
or even hostility towards the deceased. An important example is MacDonald’s theory
(1979) of ‘vicarious human sacrifice’ relating to burials in the large Roman cemetery
at Lankhills, Hampshire. MacDonald argued that seven decapitated burials from this
cemetery were deliberately positioned near to burials that were either rich, military, or
‘ritually unusual’. A high proportion of the decapitated individuals were elderly females,
and he concluded from this that less productive members of society—​‘anyone who was
not valued highly’—​were ritually decapitated for the benefit of more socially prominent
individuals (MacDonald 1979: 415–4​17).
Such conclusions were invariably published in excavation reports and focused on the
evidence from one site alone. This also applies to Matthews’s conclusion (1981) that the
decapitated burials at a cemetery in Dunstable were the victims of a massacre by a local
tyrant. As such, they seldom, if ever, fit the evidence for these rites when discovered
elsewhere. Furthermore, these theories rarely considered wider burial patterns, even
within the cemeteries that they discuss. Boylston et al. (2000) speculated that decapita-
tion at a cemetery in Bedford was performed when individuals died from sudden ill-
ness and was intended to mimic the death of a warrior. However, in their report there
is no comparison with pathology, trauma, and mortality patterns among other burials
 Deviancy in Late Romano-British Burial    385

within the cemetery to evaluate whether decapitated burials constituted a united and
distinct group.
Perhaps because of the general neglect of the subject in post-​processualist theory, ele-
ments of processualism and latent assumptions about negative social identity are still
apparent in the more recent archaeological work on the subject, without sufficient scru-
tiny. Furthermore, this work has developed a greater sense of certainty about who would
have been chosen for ‘deviant’ burial and how they may be recognized archaeologically:

[deviant persons] can include criminals, women who died during childbirth, unbap-
tised infants, people with disabilities, and supposed revenants, to name but a few.
Such burials can be identified in the archaeological record from an examination of
the location and external characteristics of the grave site. Furthermore, the position
of the body in addition to its association with unusual grave goods can be a further
feature of non-​normative burials. (Murphy 2008b: p. xii)

Despite the work of Pader (1980, 1982) and Shay (1985) on the possible variability of
‘deviant’ identities, the idea that ‘deviant’ equals ‘negative’ has seeped into the uncon-
sciousness of the discipline at large. The most recent publications on the subject are keen
to see atypical burial practices as representing evidence for a ‘darker attitude towards
corpses’ (Taylor 2008: 92). This has resulted in what can be viewed as a ‘checklist’ of
archaeological signifiers that are seen as evidence for negative social identity, including:

• prone posture;
• decapitation;
• amputated limbs;
• bound limbs;
• burial in a peripheral or isolated space;
• plaster burials;
• boulders on the corpse or stone coverings to the grave;
• exotic or amuletic grave goods (especially nails in the grave);
• very deep graves;
• very shallow graves;
• ‘strange’ body positions.1

Interpretations about the people who received these burial treatments frequently
conclude that they had, in some way, affronted social rules or ideals, either deliberately
or involuntarily, for example:

• criminals;
• murder victims;
• the mentally disabled;
• the physical disabled/​diseased;
• unbaptized infants;
386

386   Belinda Crerar

• infants of any kind;
• the young;
• the old;
• persons killed in battle;
• women who died in childbirth;
• practitioners of magic;
• mors immatura: the untimely/​unmarried dead;
• suicides;
• sacrificial victims.2

There is a danger when studying a subject such as social marginality that discussions
will veer into subjectivity. Archaeological remains that ‘look strange’ to modern eyes
and evidence for practices that contravene modern standards of acceptability have all,
in the past, been explained as evidence for deviant behaviour in Roman Britain with-
out sufficient consideration of Romano-​British cultural values. Similarly, the distinction
between a ‘deviant’ burial rite and a ‘minority’ burial rite often appears to lie in its famili-
arity to later observers: burial in a stone coffin is termed a ‘minority’ rite whereas decapi-
tated burial, despite being as, if not more, common in Romano-​British cemeteries, is
termed ‘deviant’. Of greatest concern is the assumption that there will be something
socially abhorrent about individuals buried in idiosyncratic or unfamiliar ways, and it is
the job of the archaeologist to discover what it was, rather than to question whether this
is an appropriate conclusion in the first place (Tsaliki 2008: 2).

Deviant Burials and Deviant

Persons: A Justified Connection?

Why do criminals, the disabled, and persons associated with magic so often top the list of
those considered necessary to bury in different ways? The Roman period witnessed the
first society in Britain to be characterized by a large-​scale interaction of multiple cultural
groups resulting in variegated horizontal and vertical social organization as well as exten-
sive cultural exchange and acculturation, all overseen by a complex, bureaucratic, and,
to some degree, impersonal administration. Social positions were highly negotiable, and
production and consumption of material culture increased to unprecedented levels, with a
greater propensity to disposability than before. This can make the Roman period feel very
familiar to modern western Europeans in some ways (Reece 1988). While many aspects
of Romano-​British life such as technological developments, trade, and manufacture can
be approached directly through the material remains, social interactions and interper-
sonal behaviour are more abstruse. Ethnographic studies rarely provide appropriate com-
parative data since they typically involve small-​scale, pre-​industrial communities where
interactions occur on an intimate communal level and the people themselves are largely
 Deviancy in Late Romano-British Burial    387

sheltered from external global networks and influences (Binford 1971). However, the impe-
rialism, materialism, urban expansion, and industrial enterprise of eighteenth-​and nine-
teenth-​century Europe offer an attractive alternative paradigm. This line of enquiry may
have been encouraged by the history of British archaeology itself: whether consciously or
not, scholars of the early twentieth century tended to see Rome’s interest in its provinces
as a counterpart to the imperial ambitions of Victorian Britain (Haverfield 1905; Hingley
2000) (see also Hingley, Millett, and Wilson, this volume, for discussion of development
of Romano-​British archaeology). It is then a deceptively short step to apply other aspects
of this period to Romano-​British culture, such as notions of etiquette, taste, hierarchy,
intra-​societal relations, and appropriate religious behaviour. In a discussion of the eastern
cemetery of Roman London, Barber and Hall (2000: 119) argue that the individuals buried
there were excluded from ‘polite society’ on account of wealth (see also Henig 1984: 205).
While the phrase was no doubt intended as a colloquialism for the urban elite, the con-
notations of using a phrase typically associated with more recent eras in regard to Roman
Britain reveals worrying subconscious infiltrations from anachronistic social models into
our interpretations of Roman social behaviour.
A further issue is undoubtedly the superficial resemblance of the dominant late
Romano-​British burial practices to more recent concepts of ‘decent burial’—​that is,
interment soon after death in an appropriately sized rectilinear grave in an area of land
designated for this use, usually within a wooden coffin with few grave goods. Once
again, it is a short step from recognizing similarities in material remains to conclud-
ing conformity of motivation. Affirmation that parallels are legitimate is often sourced
from the supposed influence of a fledgling Christianity in the late Roman period, which
is exploited to argue for articulation between Roman and modern western religious
views on death (Green 1977; Thomas 1981; Watts 1998; Taylor 2010). These sentiments
are visible in the repeated claim that, despite their fourth-​century proliferation coincid-
ing with the rise of Christianity in Roman Britain, and without a clear understanding of
their meaning, decapitated and prone burials were ‘pagan’ rites (Charmasson 1968: 142;
Davies et al. 1986: 107; Woodward 1993: 227, 236; Cooke 1998: 250; Watts 1998: 45–​46;
2005: 147; Williams 1999: 102). It seems that this stems, not from an attempt to assess the
religious complexities of fourth-​century Britain, or to associate these rites with specific
pagan beliefs about the afterlife, but rather from an automatic expectation that they can-
not possibly be Christian. Attempting to understand Roman society in familiar terms is
a precarious position to start from, since it leads to overlooking differences or rational-
izing them in anachronistic ways. This fosters an inclination to push evidence for prob-
lematic or unfamiliar practices to the peripheries of what are assumed to be societal
norms. It is thus understandable why interpretations of burials that appear unfamiliar
to modern eyes should revolve around concepts of the illegal, immoral, or macabre.
Though Salway (1981: 707) is perhaps overly pessimistic in his assessment of studying
decapitated burials, his final sentiments provide a valuable warning:

The truth is, that we do not know either the origin of [decapitated burial] nor what
it meant. We can only record it, with the observation that it ought to remind us of
388

388   Belinda Crerar

just how much that is very alien to our ways of thinking lay beneath the superficially
modern and familiar appearance of the Roman world.

To state an obvious point, but one that is frequently ignored, a rite that appears ‘bizarre’
to a modern audience was not necessarily considered so by those who practised it. Saxe
(1970: 175–​176) records that the Ashanti, inhabitants of an independent kingdom on the
Gold Coast of West Africa, defleshed their kings before rearticulating their bones (often
incompletely or incorrectly) with gold wire—​a practice that may well be described by
modern commentators as ‘bizarre’ but was presumably not considered so by the Ashanti
themselves. Furthermore, for the members of the Babenberg and Hapsburg–​Lorraine
royal dynasties, death in foreign lands warranted rites including boiling of the corpse in
vinegar or wine, defleshing, evisceration, and ‘skeletal mutilations’ as a reflection of high
status and preferential treatment—​treatment that in other contexts may be considered a
mark of disgrace to the deceased. As Weiss-​Krejci (2008: 171–​172, table 10.1) points out,
this does not prevent the rite from being ‘deviant’ in the sense that the unusual circum-
stances surrounding death (in this case, death abroad) have the effect of requiring differ-
ential burial treatment, but it does shake our confidence in making connections between
minority rites and a negative stigma. Decapitation, prone posture, or piling boulders
over a corpse certainly appear to be ‘bizarre’ to modern eyes, largely because they are no
longer part of contemporary mortuary practice, but we must not approach the study of
these rites with the assumption that those who practised them shared our prejudices. The
use of the same term—​‘deviant’—​to describe both material remains and social persona,
when these may not be related, is one of the main confusing factors in the current debate
and may actually hinder lines of enquiry that question their association. It is preferable,
therefore, to explain any archaeological remains in purely descriptive terms and reserve
the use of ‘deviant’ to the interpretation of social persona, if deemed relevant.

Approaching ‘Deviancy’
in Romano-​British Archaeology

Archaeology is rapidly revealing the complexity of Romano-​British society, and it is,

therefore, essential to tread carefully when determining what behaviour was contrary to
societal norms and what was merely less well established. A considerable amount of recent
archaeological work has focused on aspects of, and attitudes towards, different social
groups in Roman Britain (e.g. Gowland 2001, 2004, 2009; Keegan 2002; Roberts and Cox
2003; Allason-​Jones 2005; Redfern 2005; Eckardt et al. 2009; Leach et al. 2009, 2010; Lewis
2010; Moore 2010; Pitts and Griffin 2012). These studies are invaluable to the investiga-
tion of deviancy, because they highlight not only culture-​specific attitudes of the period,
but also the diversity inherent in Romano-​British society. They also provide a framework,
rooted in evidence from Roman Britain itself, for recognizing social difference.
 Deviancy in Late Romano-British Burial    389

In the simplest terms, if we seek to understand ‘deviancy’ within a society, there needs
to be a ‘norm’ for it to oppose. However, identifying such a ‘norm’ is rarely straightfor-
ward. In her doctoral thesis, Aspöck (2009) demonstrated the variety of burial rites that
may be practised by a single society and still conceived of as within the realm of normal-
ity. This has serious implications for any attempt to define ‘deviant’ burials in opposition
to a defined ‘norm’, because the ‘norm’ is rarely, if ever, uniform. How, then, do we distin-
guish ‘deviant’ from ‘minority’? Although the late Romano-​British period saw greater
homogeneity in funerary behaviour than previous eras, minority practices were still
widespread. Some people were buried in lead or stone coffins; some had coins placed in
their hands or mouths; others were buried next to, above, or below other individuals in
the same grave; some were buried in ditches or along field boundaries. Yet these burials
are not generally considered ‘deviant’, although they are numerically less abundant than
those that lack these characteristics.
It is often remarked that, were it not for the decapitation itself, there is nothing to
distinguish decapitated burials from others of the late Roman period (Black 1986: 225;
Cooke 1998: 250; Booth et al. 2010: 480–​481). This raises the question of whether it is the
techniques of archaeological enquiry that have characterized these burials as those of
social outcasts by consistently focusing on their differences from, rather than similari-
ties to, contemporaneous inhumations. We have already discussed how presuppositions
held by researchers could have influenced interpretation of ‘deviant’ burial practices, but
unsatisfactory analytical approaches have also compounded the effect. Burials show-
ing certain ‘deviant’ characteristics are invariably removed from their original context
within a wider cemetery to be analysed in isolation (e.g. Harman et al. 1981; Philpott
1991). This has the result of making them appear conceptually detached from other
funerary practices by virtue of the methodology used in their analysis.
The following discussion sets out an alternative model for investigating burial rites
that have traditionally been classed as ‘deviant’—​one that relies on contextual analysis of
these burials alongside contemporaneous interments that did not receive the same treat-
ment. Decapitated burials from the late Roman period are taken as the example. This
methodology is designed to establish identity relationships between decapitated and
non-​decapitated persons so as to determine whether additional evidence exists to sup-
port a conclusion of social ostracism for the former. A case study approach is adopted,
focused on the region of Britain that roughly constitutes modern Cambridgeshire. This
reduction in sample size from the whole of Britain to a single region allows for all burials
within the sample to be analysed according to the same criteria.

Decapitated Burials in Cambridgeshire

Cambridgeshire has produced at least 30 late Roman inhumation sites. Those consulted
here are listed in Table 19.1. A total of 59 decapitated burials out of 628 inhumations
have been recorded. Aspects of social identity that can be recognized, albeit imperfectly,
390

Table 19.1 Sites consulted
Cemetery Reference

Babraham Timberlake et al. (2007)

Castle Hill
Chesterton Lane Corner
Colne Fen, Earith
Earith Camp Ground
Foxton
Girton College
Guilden Morden
Jeavon’s Lane, Cambourne
Jesus Lane
King’s Dyke
Knapwell Plantation
Knobbs Farm (2004)
Knobbs Farm (2008)
Knobbs Farm (2009)
Linton Village College
Lower Cambourne
Lynch Farm
Market Deeping
Milton Barrow
Orton Longueville
Prickwillow Road
Stanground North
The Parks, Godmanchester
Vicar’s Farm
Water Newton
Watersmeet
Watson’s Lane
West Fen Road, Ely
Wittering
Alexander and Pullinger (1999)
Cessford (2007)
Regan (2003)
Regan et al. (2004)
Goode and Bardill (1995); Maynard et al. (1997)
Hollingworth and O’Reilly (1925)
Fox and Lethbridge (1926); Lethbridge (1936)
Wright et al. (2009)
Alexander et al. (2004)
Challands (1977)
Wright et al. (2009)
Wills (2004)
Armour (2008)
Armour and Morley (2009)
Gilmore (2009)
Wright et al. (2009)
R. Jones (1975)
Trimble (2000)
T. Reynolds (1997)
Wilson et al. (1975: 252)
Atkins and Mudd (2003)
Meadows (n.d.)
A. Jones (1998, 2003)
Lucas and Whittaker (2001)
Casa Hatton and Wall (1999, 2006)
Nicholson (2006)
Lucas and Hinman (1996)
Regan (2001)
McKenna (1993)
 Deviancy in Late Romano-British Burial    391

through skeletal remains such as age, sex, and health, as well as aspects of funerary rites,
including coffin and grave-​good allocation and posture, have been collated for all of
the inhumation burials in this area and the profile of decapitated and non-​decapitated
burials compared. These aspects of an individual’s identity have been analysed quanti-
tatively, and the results presented in the following graphs (Figures 19.1 to 19.5). For this
approach, the data have been divided into three groups:

• group 1: decapitated burials (n = 59);

• group 2: non-​decapitated burials from cemeteries with decapitated burials (n = 377);
• group 3: non-​decapitated burials from cemeteries without decapitated burials (n = 192).

The division of the non-​decapitated sample into two groups is designed to explore
whether broader differences in mortuary behaviour governed burial practices in ceme-
teries that both did and did not contain decapitated burials, and the question of whether
their presence in the former was due to different systems of funerary behaviour prac-
tised by certain communities within the case-​study region. This allows for a greater level

60.00
Decapitated burials

Non-decapitated burials
50.00 from cemeteries with
decapitated burials

Non-decapitated burials
from cemeteries without
40.00 decapitated burials
% of inhumations per group

30.00

20.00

10.00

0.00
Neonate Infant Juvenile Subadult Adult Mature Unknown Unknown
(<1yr) (1–5 yrs) (6–13 yrs) (14–17 yrs) (18+ yrs) (45+ yrs) immature
Age

Figure 19.1  Comparative age of decapitated and non-​decapitated inhumations along the Fen
Edge.
Source: © B. Crerar.
392

392   Belinda Crerar

45.00
Decapitated burials

40.00 Non-decapitated burials

from cemeteries with
decapitated burials
35.00
Non-decapitated burials
from cemeteries without
decapitated burials
30.00
% of inhumations per group

25.00

20.00

15.00

10.00

5.00

0.00
Male/?male Female/?female Unknown sex (adult) Immature
Sex

Figure 19.2  Comparative sex of decapitated and non-​decapitated inhumations along the Fen
Edge.
Source: © B. Crerar.

of analysis than a simple ‘decapitated versus non-​decapitated burials’ comparison. The

following results are derived from a more detailed study that formed part of the author’s
doctoral research (Crerar 2012). This can be referred to for more in-​depth discussion of
both the methodology and the analysis.

Age
Among the adult and mature categories the decapitated burials are more closely
aligned to the profile of group 2 than group 3, which suggests conformity between the
proportion of adult and mature individuals decapitated and the overall proportion
of these age groups within the same cemeteries (Figure 19.1). The results for imma-
ture burials seem to show that decapitation within these age groups was common.
However, the quantitative data as presented here are slightly misleading, as this is not
true of the whole region: all but four of the decapitated immature individuals were
discovered at one site, The Parks cemetery, Godmanchester. This implies that the high
number of immature individuals was not a regional phenomenon but a result of the
 Deviancy in Late Romano-British Burial    393

80.00
Decapitated burials

Non-decapitated burials
70.00
from cemeteries with
decapitated burials

60.00 Non-decapitated burials

from cemeteries without
decapitated burials
% of inhumations per group

50.00

40.00

30.00

20.00

10.00

0.00
Grave goods No grave goods Unknown
Grave-good allocation

Figure 19.3  Comparative grave-​good allocation for decapitated and non-​decapitated inhuma-
tions along the Fen Edge.
Source: © B. Crerar.

practices of one particular community. It is important to note that this cemetery also
contained more immature individuals overall than any of the others in the region, at
27 per cent of the buried population. Consequently, the propensity to inter the young
at The Parks cemetery site correlates with the increased number of decapitated imma-
ture individuals also found here. Similarly, overall, the age profile of the decapitated
individuals can be seen to reflect the general age profile of the cemeteries in which
they are found.

Sex
Figure 19.2 shows the relative proportion of males and females in each of the three
groups. Unfortunately, a high number of ‘unknown’ cases make conclusions drawn
from these results tenuous. From what can be deduced it appears that sex was not a
determining factor for the decapitation rite, as the proportions of males and females
among the decapitated burials closely follows those of groups 2 and 3. Slightly more of
394

394   Belinda Crerar

90.00
Decapitated burials

80.00 Non-decapitated burials

from cemeteries with
decapitated burials
70.00
Non-decapitated burials
from cemeteries without
decapitated burials
60.00
% of inhumations per group

50.00

40.00

30.00

20.00

10.00

0.00
Coffin No coffin Unknown
Coffin allocation

Figure 19.4  Comparative coffin allocation for decapitated and non-​decapitated inhumations
along the Fen Edge.
Source: © B. Crerar.

the decapitated burials were female than those in the non-​decapitated categories, but, as
males dominated overall, this does not appear to point towards a distinct preference for
female subjects.

Grave Goods
Here ‘grave goods’ are defined as any object (worn or unworn) deliberately placed within
the grave at the time of burial that cannot reasonably be related to structural elements
of the grave furniture. This includes both inorganic finds and organic deposits such as
evidence for food offerings. The quantitative data for burial with grave goods seem to
indicate a higher likelihood of decapitated burials being interred with grave offerings
compared to non-​decapitated individuals (Figure 19.3). The data also suggest that bur-
ial with grave goods was correspondingly higher among group 2 burials compared to
group 3, suggesting that burial with grave goods was more commonly practised by com-
munities who also practised decapitated burial. This perhaps implies that offering grave
 Deviancy in Late Romano-British Burial    395

90.00
Decapitated burials

80.00 Non-decapitated burials

from cemeteries with
decapitated burials
70.00
Non-decapitated burials
from cemeteries without
decapitated burials
60.00
% of inhumations per group

50.00

40.00

30.00

20.00

10.00

0.00
Prone Supine On side Unknown
Posture

Figure 19.5  Comparative posture of decapitated and non-​decapitated inhumations along the
Fen Edge.
Source: © B. Crerar.

goods and decapitation were conceptually linked, both perhaps signifying a greater
tendency of these communities to engage in ancillary minority rites during the burial
process. However, the slightly greater number of unknown cases for group 3 may have
affected these conclusions.
Consideration of the type of grave goods offered and how this relates to those found
with non-​decapitated individuals reveals further correlations. At the cemetery at
Babraham, it seems that the offering of a vessel at the head end of the grave consti-
tuted a distinct burial rite among the population in the later phases of the cemetery’s
use. The allocation of this rite to a decapitated male indicates that the act of decapita-
tion did not exclude this individual from receiving additional rites afforded to non-​
decapitated individuals. There is also another way to view this pattern. A similar ritual
process is observed from the 2009 excavations at Knobbs farm, where five decapitated
burials were found to have miniature vessels positioned above the shoulders, paral-
leled by five non-​decapitated burials who also received a miniature vessel next to the
skull (Armour and Morley 2009). In the case of one decapitated individual [965], the
vessel chosen was a face-​pot, suggesting a meaningful association between grave good
396

396   Belinda Crerar

and location. In isolation, a face-​pot placed above the shoulders of a decapitated per-
son may appear to be a highly distinctive and symbolic act relating to the practice of
decapitation and the specialness of this burial. However, when it is considered that the
placement of vessels was also applied to non-​decapitated burials, the interpretation
must be modified. It is then necessary to question whether the allocation of vessels
near the head for both decapitated and non-​decapitated burials does indeed reflect the
former being treated like the latter, or instead implies a desire to treat non-​decapitated
persons in a way that connected them to decapitated individuals.

Coffins
Where data have survived, there is a slight bias towards the burial of decapitated individ-
uals without coffins compared to the non-​decapitated burials from both groups (Figure
19.4). However, if the high number of unknown cases from group 2 can be interpreted as
a lack of evidence for coffins, then the proportions become more or less even across all
categories.

Posture
The data for posture contain a high number of unknown cases. Included in this cat-
egory are all burials for which no specific posture is mentioned in the relevant report.
However, it can be reasonably assumed that the majority of these were supine, since the
rarity and interest of prone or side posture typically warrant comment from authors,
while the more numerous supine pose does not. Therefore, despite the high number of
‘unknown’ cases, we may be fairly confident in the relative proportions of prone burials
between the three groups.
The majority of the decapitated burials (85 per cent) were buried supine extended
or, in two cases (burial [F.118] from Watson’s Lane and burial [24] from Guilden
Morden), with legs slightly flexed. However, Figure 19.5 suggests that a greater
proportion of decapitated burials than non-​decapitated were buried prone. While
there does not seem to be a difference in the proportion of prone burials among
groups 2 and 3, which stand at 3.9 per cent and 3.1 per cent respectively, prone buri-
als in group 1 account for 15.5 per cent of the total. This implies a demonstrable link
between prone and decapitated burial. This conclusion was also reached through
the analysis of decapitated and prone burials in Oxfordshire (Harman et al. 1981).
However, this cannot be taken as evidence for interpreting these burials as ‘deviant’,
because the implications of burying someone face-​down are as equally obscure as
removing their head and would benefit from a detailed study along similar lines to
that conducted here. It is, nonetheless, an interesting connection, which suggests,
 Deviancy in Late Romano-British Burial    397

like the allocation of grave goods, that additional minority rites were often applied
to the burials of decapitated persons.

Health, Disease, and Disability

Where pathological information has been recorded, only one of the decapitated
burials from Cambridgeshire shows evidence for disability and none appears to
have suffered a physically violent death. A  decapitated male [161] from Jesus Lane
showed gross morphological changes to both hips consistent with septic arthropa-
thy, which would have severely impeded movement and almost certainly resulted in
a pronounced limp. The two other decapitated burials from this cemetery showed no
evidence for poor health, yet three non-​decapitated individuals (burials [19], [70],
and [73]) also suffered from conditions that may have resulted in limps. As such, it
cannot be concluded that the poor health of individual [161] resulted in his decapita-
tion as this is not applied to other infirm persons, even within the same cemetery.
Two other decapitated burials (adult [A.2] from Guilden Morden and mature adult
[F.118] from Watson’s Lane) suffered from severe osteoarthritis, which may have
affected their physical capabilities, though this may not have been considered unu-
sual for adults and mature members of society. Furthermore, juvenile [12] from The
Parks, Godmanchester, may have suffered from rickets, though, as discussed above,
it is likely that young age, rather than physical health, was the determining factor in
decapitation at this site.
There is little evidence overall that decapitation was related to poor health in this
region. To compound this, there are several cases from Cambridgeshire of conspicu-
ously disabled individuals who were not decapitated. One striking example is the bur-
ial of the hydrocephalic neonate buried in a lead coffin with pipe-​clay figurines from
Arrington (Taylor 1993). The wealth expended on this burial also implies that physi-
cal disability, even congenital abnormalities, was not automatically a cause for different
or stigmatizing burial rites. Further examples are recorded from the communities that
practised decapitation, such as male [3423] from Foxton, whose cranial impact wounds
imply that he was bludgeoned to death and yet who was buried in a mausoleum along
with a female and a neonate (Maynard et al. 1997: 32). This too suggests that those who
suffered from trauma or pathologies during their lives were not necessarily treated dif-
ferently in burial.

Spatial Positioning
Where decapitated burials were discovered within formal cemeteries, there appears
to be no evidence that they were segregated from other burials. At Babraham, in par-
ticular, burials are arranged in clusters around focal graves, with later inhumations
398

398   Belinda Crerar

Decapitated burials
Non-decapitated burials

Figure 19.6  Guilden Morden cemetery, showing distribution of decapitated burials, adapted
from Fox and Lethbridge (1926: figure 1) and Lethbridge (1934: site plan).
Source: © B. Crerar.

positioned in loose rows towards the north-​east of the site. The five decapitated buri-
als at this site are found in both the clusters and the later rows (Timberlake et al. 2007).
Guilden Morden represents the only site among those analysed where arguments
for the peripheral location for decapitated burials may be put forward, but even here
the evidence is tenuous. Six decapitated burials have been discovered from two sea-
sons of excavation at Guilden Morden, the first in 1924 and the second in 1935 (Fox and
Lethbridge 1926; Lethbridge 1936). The plans from both are shown in Figure 19.6. Prone
and decapitated male [A.6] is positioned in the centre of the excavated area, surrounded
by other burials. Further inhumation burials extend towards the east, west, and north,
while towards the south are cremations. It is possible that this line marking the division
between cremations and inhumations was viewed as a boundary. If this is the case, one
burial [A.6] might legitimately be said to lie on the periphery of the part of the cemetery
reserved for inhumations, although two other decapitated burials [A.2] and [A.8] are
integrated in a dense area of inhumations to the west. On the other side of the excavated
area, burial [B.3] is positioned alongside a ditch that possibly marked the limit of the
cemetery, thus also locating [B.3] close to the periphery. However, neither burial [B.3]
 Deviancy in Late Romano-British Burial    399

nor [A.6] is isolated in its location, and, given the central locations of other decapitated
burials at this site, it seems unlikely that the aim was to deny these individuals burial
within the main cemetery area.

Discussion

When contextualized analysis is applied to the study of decapitated burials, an

important factor becomes apparent. In the case-​study region, no clear differences
between decapitated and non-​decapitated burials can be discerned for factors that
relate to recognizable aspects of lived identity, such as age, sex, and health. One pos-
sible pattern is that decapitation was rarely afforded to the very young (with the
exception of at The Parks cemetery), although, since immature burials are rare in
both groups 2 and 3, this may be related not to the rite of decapitation per se but to
more general social rules for the burial of these age groups (see also A. Moore and
Gowland, this volume).
Patterns become visible when factors that correspond to additional elements of funer-
ary ritual are considered—​that is, a greater prevalence of grave goods with decapitated
burials, a greater likelihood of prone posture, and a decline in the likelihood of coff-
ined burial. This suggests that the likelihood of a person being decapitated after death
depended more on his or her association with a particular social group that was more
inclined to engage in funerary customs involving multiple minority rites rather then
on his or her own personal identity. That these additional minority rites were applied to
non-​decapitated as well as decapitated persons suggests that the burial of the latter was
conducted alongside, rather than apart from, contemporary perceptions on practices
relating to death and the treatment of the dead. In addition, absence of evidence to sug-
gest that decapitated burials of this period were spatially separated from contempora-
neous burials again implies incorporation of decapitated burials within the governing
funerary structures. The consequence of this must be that decapitated burials should
be viewed as an accepted and ‘normative’ aspect of late Romano-​British funerary cus-
toms rather than indicative of social exclusion resulting from fear or hostility towards
the deceased.

Conclusion

The case study of the Fen Edge indicates that, in order to gain a better understanding
of how decapitated burial was perceived by the communities who practised it, looking
at this rite alone is not as valuable as understanding the dominant conditions govern-
ing burial at the sites where it is found, and how decapitated burials compare to these
broader trends.
400

400   Belinda Crerar

There is little doubt that social stigmas existed in Roman Britain. Communities, as
much as individuals, depend on ‘others’ to solidify their own self-​identity: one cannot
identify oneself with one thing without the existence of alternatives. Whether or not
this extended to different funerary rites for outsiders is a difficult question to answer,
particularly in the absence of written accounts. Those that exist for Roman Britain are
entirely silent on this issue. We are, therefore, dependent on the archaeological record
alone, but this must be approached with caution and without pre-​formed conclusions.
The re-​emerging interest has done much to highlight the value of minority burial rites
for understanding social concerns and beliefs. However, Aspöck’s phrase ‘the relativ-
ity of normality’ applies equally to Romano-​British society as it does to those further
removed from western history. Uncritical conclusions that certain rites reflected fear
and negativity towards the deceased automatically assume that such acts occupied the
peripheries of social life and can tell us little or nothing about common social patterns.
However, repositioning these practices within the broader context of contemporaneous
behaviour brings us closer to understanding whether they did indeed stand apart from
common modes of thought or whether they merely reflect the complexity and diversity
that existed during that time.

Acknowledgements

I would like to thank the editors of the Oxford Handbook of Roman Britain for the invita-
tion to contribute to this volume. Thanks also to Lacey Wallace and Johanna Hanink for
their comments on earlier drafts of this chapter.

Notes
1. See Black (1986); Isserlin (1997); Aspöck (2008); Taylor (2008, 2010); Tsaliki (2008);
Reynolds (2009).
2. Hertz (1960 [1907]); Hawkes and Dunning (1961); Saxe (1970); Parker Pearson (1993);
Meyer-​Orlac (1997); Murphy (2008b); Taylor (2008, 2010); Tsaliki (2008); Reynolds
(2009).

References
Alexander, J., and Pullinger, J. (1999). ‘Roman Cambridge: Excavations on Castle Hill, 1956–​
1988’, Proceedings of the Cambridge Antiquarian Society, 88.
Alexander, M., Dodwell, N., and Evans, C. (2004). ‘A Roman Cemetery in Jesus Lane,
Cambridge’, Proceedings of the Cambridge Antiquarian Society, 93: 67–​94.
Allason-​Jones, L. (2005). Women in Roman Britain. 2nd edn. York:  Council for British
Archaeology.
 Deviancy in Late Romano-British Burial    401

Anderson, T. (2001). ‘Two Decapitations from Roman Towcester’, International Journal of

Osteoarchaeology, 11/​6: 400–​405.
Armour, N. (2008). ‘Knobbs Farm, Somersham: Phase 5B Investigations’. Unpublished Report
815, Cambridge Archaeological Unit.
Armour, N., and Morley, I. (2009) ‘Knobbs Farm, Somersham: Phase 5B Investigations’.
Unpublished draft report, Cambridge Archaeological Unit.
Aspöck, E. (2008). ‘What Actually is a “Deviant Burial”? Comparing German-​Language and
Anglophone Research on “Deviant Burials”’, in E. M. Murphy (ed.), Deviant Burial in the
Archaeological Record. Oxford: Oxbow Books, 17–​34.
Aspöck, E. (2009). ‘The Relativity of Normality: An Archaeological and Anthropological Study
of Deviant Burials and Different Treatment at Death’. Unpublished Ph.D. thesis, University
of Reading.
Atkins, R., and Mudd, A. (2003). ‘An Iron Age and Romano-​British Settlement at Prickwillow
Road, Ely. Excavations 1999–​2000’, Proceedings of the Cambridge Antiquarian Society,
92: 5–​55.
Barber, B., and Hall, J. (2000). ‘Digging up the People of Roman London: Interpreting Evidence
from Roman London’s Cemeteries’, in I. Haynes, H. Sheldon, and L. Hannigan (eds), London
Under Ground: The Archaeology of a City. Oxford: Oxbow Books, 102–​120.
Bello, S., and Andrews, P. (2009). ‘Patterns in Human Burial Practice’, in R. Gowland and C. J.
Knüsel (eds), Social Archaeology of Funerary Remains. Oxford: Oxbow Books, 14–​29.
Binford, L. R. (1971). ‘Mortuary Practices: Their Study and Potential’, in J. Brown (ed.),
Approaches to the Social Dimensions of Mortuary Practices, Memoirs of the American
Archaeology Society, 25, 6–​29.
Black, E. W. (1986). ‘Romano-​British Burial Customs and Religious Beliefs in South-​East
England’, Archaeological Journal, 143: 201–​239.
Booth, P., Simmonds, A., Boyle, A., Clough, S., Cool, H., and Poore, D. (2010). The Late Roman
Cemetery at Lankhills, Winchester:  Excavations 2000–​2005. Oxford:  Oxford Archaeology
Monograph 10.
Boylston, A., Knüsel, C. J., Roberts, C. A., and Dawson, M. (2000). ‘Investigation of a Romano-​
British Rural Ritual in Bedford, England’, Journal of Archaeological Science, 27: 241–​254.
Calkin, J. B. (1947). ‘Two Romano-​British Burials at Kimmeridge’, Proceedings of the Dorset
Natural History and Archaeological Society, 69: 33–​41.
Calkin, J. B. (1953). ‘Some Archaeological Discoveries in the Isle of Purbeck’, Proceedings of
Dorset Natural History and Archaeological Society, 74: 48–​54.
Casa Hatton, R., and Wall, W. (1999). A Late Roman Cemetery beside the A1 near Durobrivae,
(Water Newton): Archaeological Recording. Cambridgeshire County Council: Archaeological
Field Unit Report 165.
Casa Hatton, R., and Wall, W. (2006). ‘A Late Roman Cemetery at Durobrivae, Chesterton’,
Proceedings of the Cambridge Antiquarian Society, 95: 5–​24.
Cessford, C. (2007). ‘Middle Anglo-​Saxon Justice:  The Chesterton Lane Corner Execution
Cemetery and Related Sequence, Cambridge’, Archaeological Journal, 164: 197–​226.
Challands, A. (1977). ‘The King’s Dyke Burials’, Durobrivae:  A Review of Nene Valley
Archaeology, 5: 27–​30
Charmasson, J. (1968). ‘Une nécropole du Bas-​Empire:  La Brèche, Commune de Laudun
(Gard)’, Cahiers ligures de Préhistoire et d’archéologie, 17: 131–​150.
Collingwood, R. G. (1937). ‘Roman Britain in 1936: Sites Explored’, Journal of Roman Studies,
27: 223–​245.
402

402   Belinda Crerar

Cooke, N. (1998). ‘The Definition and Interpretation of Late Roman Burial Rites in the Western
Empire’. Unpublished Ph.D. thesis, University College, London.
Crerar, B. (2012). ‘Contextualising Deviancy: A Regional Approach to Decapitated Inhumation
in Late Roman Britain’. Unpublished Ph.D. thesis, University of Cambridge.
Davies, S. M., Stacey, L. C., and Woodward, P. J. (1986). ‘Excavations at Alington Avenue,
Fordington, Dorchester, 1984–​85: Interim Report’, Proceedings of the Dorset Natural History
and Archaeological Society, 107: 101–​110.
Dewar, H. S. L. (1958). ‘Replies, 159. Romano-​British Cemetery at Charlton Mackrell’, Notes and
Queries: Somerset, Dorset, 27: 206–​207; 180–​181.
Eckardt, H., Chenery, C., Booth, P., Evans, J. A., Lamb, A., and Müldner, G. (2009). ‘Oxygen and
Strontium Isotope Evidence for Mobility in Roman Winchester’, Journal of Archaeological
Science, 36: 2816–​2825.
Fox, C., and Lethbridge, T. C. (1926). ‘The La Tène and Romano-​British Cemetery at Guilden
Morden, Cambridgeshire’, Proceedings of the Cambridge Antiquarian Society, 27: 49–​7 1.
Frazer, J. G. (1932–​3). The Fear of the Dead in Primitive Religion: Lectures Delivered on the William
Wyse Foundation at Trinity College, Cambridge, 1932–​1933. 3 vols. London: Macmillan.
Gilmore, N. (2009). ‘Excavation, Evaluation and Watching Brief at Linton Village College,
Linton, Cambridgeshire’. Unpublished Report 1058, Oxford Archaeology East.
Goode, J. L., and Bardill, M. (1995). ‘St Neots to Duxford Gas Pipeline 1994: Human Remains’,
in John Price Engineering Archaeological Services, St Neots to Duxford 900MM Gas Pipeline
1994:  Archaeological Report, Volume 3. Unpublished report, Engineering Archaeological
Services, 857–​890.
Gowland, R. (2001). ‘Playing Dead’, in G. Davies, A. Gardner, and K. Lockyear (eds), TRAC
2000: Proceedings of the Tenth Annual Theoretical Roman Archaeology Conference, London
2000. Oxford: Oxbow Books, 152–​168.
Gowland, R. (2004). ‘The Social Identity of Health in Late Roman Britain’, in B. Croxford,
H. Eckardt, J. Meade, and J. Weekes (eds), TRAC 2003:  Proceedings of the Thirteenth
Annual Theoretical Roman Archaeology Conference, Leicester 2003. Oxford: Oxbow Books,
135–​146.
Gowland, R. (2009). ‘Aging the Past: Examining Age Identity from Funerary Evidence’, in R.
Gowland and C. J. Knusel (eds), Social Archaeology of Funerary Remains. Oxford: Oxbow
Books, 143–​154.
Green, C. S. (1977). ‘The Significance of Plaster Burials for the Recognition of Christian
Cemeteries’, in R. Reece (ed.), Burial in the Roman World. CBA Research Report 22.
London: Council for British Archaeology, 46–​53.
Harman, M., Molleson, T. I., and Price, J. L. (1981). ‘Burials, Bodies and Beheadings in Romano-​
British and Anglo-​Saxon Cemeteries’, Bulletin of the British Museum of Natural History
(Geology), 35/​3: 145–​188.
Haverfield, F. (1905). The Romanization of Roman Britain. London: Published for the British
Academy by H. Frowde.
Hawkes, S. C., and Dunning, G. C. (1961). ‘Soldiers and Settlers in Britain, Fourth to Fifth
Century’, Medieval Archaeology, 5: 1–​70.
Henig, M. (1984). Religion in Roman Britain. London: B. T. Batsford.
Hertz, R. (1960 [1907]). Death and the Right Hand, trans. R. Needham and C. Needham.
London: Cohen and West.
Hingley, R. (2000). Roman Officers and English Gentlemen: Of the Imperial Origins of Roman
Archaeology. London: Routledge.
 Deviancy in Late Romano-British Burial    403

Hollingworth, E. J., and O’Reilly, M. M. (1925). The Anglo-​Saxon Cemetery at Girton College,
Cambridge. Cambridge: Cambridge University Press.
Isserlin, R. M.  J. (1997). ‘Thinking the Unthinkable:  Human Sacrifice in Roman Britain’,
in K. Meadows, C. Lemke, and J. Heron (eds), TRAC 96: Proceedings of the Sixth Annual
Theoretical Roman Archaeology Conference, Sheffield 1996. Oxford: Oxbow Books, 91–​100.
Jones, A. (ed.). (2003). Settlement, Burial and Industry in Roman Godmanchester: Excavations
in the Extra-​Mural Area, The Parks 1998, London Road 1997–​8 and Other Investigations. BAR
British Series 346. Oxford: Archaeopress.
Jones, R. (1975). ‘The Romano-​British Farmstead and its Cemetery at Lynch Farm, near
Peterborough’, Northamptonshire Archaeology, 10: 94–​137.
Keegan, S. L. (2002). Inhumation Rites in Late Roman Britain: The Treatment of the Engendered
Body. BAR British Series 333. Oxford: Archaeopress.
Leach, S., Lewis, M., Chenery, C., Eckardt, H., and Müldner, G. (2009). ‘Migration and
Diversity in Roman Britain: A Multidisciplinary Approach to Immigrants in Roman York,
England’, American Journal of Physical Anthropology, 140: 546–​561.
Leach, S., Eckardt, H., Chenery, C., Müldner, G., and Lewis, M. (2010). ‘A “Lady” of
York: Migration, Ethnicity and Identity in Roman York’, Antiquity, 84: 131–​145.
Lethbridge, T. C. (1936). ‘Further Excavations in the Early Iron Age and Romano-​British
Cemetery at Guilden Morden’, Proceedings of the Cambridge Antiquarian Society,
36: 109–​120.
Lewis, M. (2010). ‘Life and Death in a civitas Capital: Metabolic Disease and Trauma in the
Children from Late Roman Dorchester, Dorset’, American Journal of Physical Anthropology,
142/​3: 405–​416.
Liversidge, J. (1977). ‘Roman Burials in the Cambridge Area’, Proceedings of the Cambridge
Antiquarian Society, 67: 11–​38.
Lucas, G., and Hinman, M. (1996). ‘Archaeological Excavations of an Iron Age Settlement
and Romano-​British Enclosures at Watson’s Lane, Little Thetford, Ely, Cambridgeshire’.
Unpublished Report 194, Cambridge Archaeological Unit.
Lucas, G., and Whittaker, P. (2001). ‘Vicar’s Farm, Cambridge: Post-​Excavation Assessment
Report’. Unpublished Report 425, Cambridge Archaeological Unit.
MacDonald, J. L. (1979). ‘Religion’, in G. Clarke, Pre-​Roman and Roman Winchester Part II: The
Roman Cemetery at Lankhills. Winchester Studies 3. Oxford: Clarendon Press, 404–​433.
McKenna, R. (1993). ‘A Romano-​British Decapitated Burial at Church Farm, Wittering’.
Unpublished report, Peterborough City Museum and Maxwell Art Gallery.
Matthews, C. L. (1981). ‘A Romano-​British Inhumation Cemetery at Dunstable (Durocobrivae)’,
Bedfordshire Archaeological Journal, 15: 1–​73.
Maynard, D. J., Cleary, R., Moore, R., Brooks, I. P., and Price, J. (1997). ‘Excavations
at Foxton, Cambridgeshire 1994’, in J. Price, I. P. Brooks, and D. J. Maynard (eds), The
Archaeology of the St Neots to Duxford Gas Pipeline 1994. BAR British Series 255.
Oxford: Archaeopress, 21–​39.
Meadows, I. (n.d.). ‘Development sites in Peterborough:  Nene Valley Research Committee
Unpublished Annual Report 1989–​1990’, 8–​9.
Meyer-​Orlac, R. (1997). ‘Zur Problematik der Sonderbestattungen in der Archäologie’, in
K. F. Rittershofer (ed.), Sonderbestattungen in der Bronzezeit in Östlichen. Mitteleuropa
Internationale Archäologie 37. Espelkamp: Verlag Marie Leidorf GmbH, 1–​10.
Moir, J. R., and Maynard, G. (1931). ‘The Roman Villa at Castle Hill, Whitton, Ipswich’,
Proceedings of the Suffolk Institute of Archaeology, 21: 240–​260.
404

404   Belinda Crerar

Moore, A. (2010). ‘Age and Identity in Funerary Contexts: The Elderly in Southern Roman
Britain’, in A. Moore, G. Taylor, E. Harris, P. Girdwood, and L. Shipley (eds), TRAC
2009:  Proceedings of the Nineteenth Annual Theoretical Roman Archaeology Conference,
Michigan and Southampton 2009. Oxford: Oxbow Books, 105–​119.
Murphy, E. M. (2008a) (ed.). Deviant Burial in the Archaeological Record. Oxford:
Oxbow Books.
Murphy, E. M. (2008b). ‘Introduction’, in E. M. Murphy (ed.), Deviant Burial in the
Archaeological Record. Oxford: Oxbow Books, pp. xii–​xviii.
Nicholson, K. (2006). ‘A Late Roman Cemetery at Watersmeet, Mill Common, Huntingdon’,
Proceedings of the Cambridge Antiquarian Society, 95: 57–​90.
Pader, E.-​J. (1980). ‘Material Symbolism and Social Relations in Mortuary Studies’, in P. Rahtz,
T. Dickinson, and L. Watts (eds), Anglo-​Saxon Cemeteries 1979:  The Fourth Anglo-​Saxon
Symposium at Oxford. BAR British Series 82. Oxford: Archaeopress, 143–​160.
Pader, E.-​J. (1982). Symbolism, Social Relations and the Interpretation of Mortuary Remains.
BAR International Series 130. Oxford: Archaeopress.
Parker Pearson, M. (1993). ‘The Powerful Dead:  Archaeological Relationships between the
Living and the Dead’, Cambridge Archaeological Journal, 3/​2: 203–​229.
Philpott, R. (1991). Burial Practices in Roman Britain:  A  Survey of Grave Treatment and
Furnishing A.D. 43–​410. BAR British Series 219. Oxford: Archaeopress.
Pitts, M., and Griffin, R. (2012). ‘Exploring Health and Social Well-​Being in Late Roman
Britain: An Intercemetery Approach’, American Journal of Archaeology, 116/​2: 253–​276.
Redfern, R. C. (2005). ‘A Gendered Analysis of Health from the Iron Age to the End of the
Romano-​British Period in Dorset, England (Mid–​Late 8th Century ad to the End of the 4th
Century AD)’. Unpublished Ph.D. thesis, University of Birmingham.
Reece, R. (1988). My Roman Britain. Oxford: Cotswolds Studies 3.
Regan, R. (2001). ‘West Fen Road, Ely, Cambridgeshire’. Unpublished assessment report,
Cornwell Field Cambridge Archaeological Unit.
Regan, R. (2003). ‘An Archaeological Excavation at Colne Fen, Earith, Langdale Hall, Sites V
and VI’. Unpublished Report 537, Cambridge Archaeological Unit.
Regan, R., Evans, C., and Webley, L. (2004). ‘The Camp Ground Excavations, Colne Fen,
Earith: Assessment Report’. Unpublished Report 654, Cambridge Archaeological Unit.
Reynolds, A. (2009). Anglo-​Saxon Deviant Burial Customs. Oxford: Oxford University Press.
Reynolds, T. (1997). ‘MILEW III and IV Report Draft No 1’. Unpublished report, Cambridge
County Council Archaeological Field Unit.
Roberts, C., and Cox, M. (2003). Health and Disease in Britain from Prehistory to the Present
Day. Stroud: Sutton.
Rolleston, G. (1869). ‘Researches and Excavations Carried on in an Ancient Cemetery at
Frilford near Abingdon, Berks, in the Years 1867–​68’, Archaeologia, 42: 417–​485.
Royce, D. (1883). ‘Finds on or near to the Excursion of the Society at Stow-​on-​the-​Wold’,
Transactions of the Bristol and Gloucestershire Archaeological Society, 7: 69–​80.
Salway, P. (1981). Roman Britain. Oxford: Oxford University Press.
Saxe, A. (1970). ‘Social Dimensions of Mortuary Practices’. Unpublished Ph.D.  thesis,
University of Michigan.
Shay, T. (1985). ‘Differential Treatment of Deviancy at Death as Revealed in Anthropological
and Archaeological Material’, Journal of Anthropological Archaeology, 4: 221–​241.
Taylor, A. (1993). ‘A Roman Lead Coffin with Pipeclay Figurines from Arrington,
Cambridgeshire’, Britannia, 24: 191–​225.
 Deviancy in Late Romano-British Burial    405

Taylor, A. (2008). ‘Aspects of Deviant Burial in Roman Britain’, in E. M. Murphy (ed.), Deviant
Burial in the Archaeological Record. Oxford: Oxbow Books, 91–​114.
Taylor, A. (2010). ‘The Deviant Dead: Roman Britain’s Unusual Burials’, Current Archaeology,
244 (July), 20–​27.
Thomas, C. (1981). Christianity in Roman Britain to ad 500. London: Batsford.
Timberlake, S., Dodwell, N., and Armour, N. (2007). ‘The Roman Cemetery, The Babraham
Institute, Cambridgeshire:  An Archaeological Excavation’. Unpublished Report 754,
Cambridge Archaeological Unit.
Trigger, B. G. (2006). A History of Archaeological Thought. 2nd edn. Cambridge: Cambridge
University Press.
Trimble, D. (2000.) Archaeological Excavations Undertaken along the Route of the Market
Deeping Bypass. Archaeological Project Services Report 2000/​93: volume 3, appendices 7–​13.
Tsaliki, A. (2008). ‘Unusual Burials and Necrophobia: An Insight into the Burial Archaeology
of Fear’, in E. M. Murphy (ed.), Deviant Burial in the Archaeological Record. Oxford: Oxbow
Books, 1–​16.
Tucker, K. (2012). ‘ “Whence this Severance of the Head?” The Osteology and Archaeology of
Human Decapitation in Britain’. Unpublished Ph.D. thesis, University of Winchester.
Watts, D. (1998). Religion in Late Roman Britain: Forces of Change. London: Routledge.
Watts, D. (2005). Boudicca’s Heirs: Women in Early Britain. London: Routledge.
Weiss-​Krejci, E. (2008). ‘Unusual Life, Unusual Death and the Fate of the Corpse: A Case Study
from Dynastic Europe’, in E. M. Murphy (ed.), Deviant Burial in the Archaeological Record.
Oxford: Oxbow Books, 169–​190.
Williams, H. (1999). ‘Identities and Cemeteries in Roman and Early Medieval Britain’, in P.
Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC 98: Proceedings of the Eighth Annual
Theoretical Roman Archaeology Conference, Leicester 1998. Oxford: Oxbow Books, 96–​107.
Wills, J. (2004). ‘Knobbs Farm, Somersham, Cambridgeshire:  Phase 4—​An Archaeological
Excavation’. Unpublished Report 632, Cambridge Archaeological Unit.
Wilson, D. R., Wright, R. P., Hassall, M. W. C., and Tomlin, R. S. O. (1975). ‘Roman Britain in
1974: Sites Explored’, Britannia, 6: 220–​283.
Woodward, A. (1993). ‘Discussion’, in D. E. Farwell and T. I. Molleson, Excavations at
Poundbury 1966–​80. Volume 2: The Cemeteries. Dorset Natural History and Archaeological
Society Monograph 11. Dorchester:  Dorset Natural History and Archaeological Society,
216–​239.
Wright, J., Leivers, M., Seager Smith, R., and Stevens, S. J. (2009). Cambourne New
Settlement:  Iron Age and Romano-​ British Settlement on the Clay Uplands of West
Cambridgeshire. Wessex Archaeology Report 23. Salisbury: Wessex Archaeology.
Wright, R. P. (1953). ‘Roman Britain in 1952:  Sites Explored’, Journal of Roman Studies,
43: 104–​128.
406

Chapter 20

Cl othing and I de nt i t y

H. E. M. Cool

The concept of identity is one that students of Roman Britain have increasingly found use-
ful. Three of the ten papers in a 2001 volume setting out research frameworks in the field
had it in their titles, and another had it as a subheading (James and Millett 2001). After
sporadic appearances in the Theoretical Roman Archaeology Conference proceedings, it
featured regularly from 2003, reaching a high point in 2006, when it appeared in a third of
the titles. As the concept can mean different things to different people, it is appropriate to
observe that I follow Gardner, who was, in turn, following Giddens and Jenkins as seeing
identity as a framing device (Gardner 2007). Put simply, the way you dress tends to be gov-
erned by the group you belong to, be that family, religious affiliation, class, age, gender, and
so on. For most individuals, the group pressures are likely to be a combination of different
aspects of these. Defining identity in this way also makes it easy to see how the identity of
an individual can and does shift with time and in certain circumstances.
Appearance is frequently central to a person’s identity, providing immediate signals
about who the person is and what role(s) he or she carries out. Appearance is made up
of many things, and so in this chapter jewellery, hairstyles, and grooming generally will
be considered, in addition to clothes. To cover this for half a millennium in the allotted
space has meant that fine detail has been omitted, so a short overview of useful sources
will be found at the end of the chapter. Where points are made without further refer-
ence, it provides useful works where they can be explored.
In the first part of the chapter the sources and nature of the evidence for appear-
ance are considered. These condition how we should or can approach and interpret it.
The second part explores how the evidence can be used to investigate communities in
Britain and how this changed with time.

The Sources of the Evidence

Roman textiles survive because their burial conditions were either very dry, like Egypt,
or damp, like Britain. The anaerobic conditions at Vindolanda produced not only the
 Clothing and Identity    407

famous writing tablets, but also a very important corpus of textile remains. Such burial
conditions mean that the original colours may be hard to make out by eye, but dye anal-
ysis can help. Occasionally, cloth is preserved in graves, generally in coffins with lead
linings. Cloth impressions are found on ceramic articles such as tiles, and small areas of
textiles can be preserved by mineral replacement when they are in contact with metal
fittings. Mineral replacement on fittings from clothes in which a corpse has been buried
can be especially helpful. Unfortunately the main burial rite during the period when
such fittings were most in use was cremation; and by the fourth century, when inhuma-
tion was widely practised, most clothes did not need them (see Weekes, this volume, for
discussion of burial practices). So, unlike the sixth and seventh centuries, when mineral
replacement has been put to such good use in reconstructing costume (Walton Rogers
2007), this is not an option for the late pre-​Roman Iron Age or Roman periods.
From these sources, we know about the direction of the spin of the thread (Z: right
hand, S: left hand), different weaving patterns, the number of warp (vertical) and weft
(horizontal) threads per centimetre, and the type of fibre used. Thread numbers provide
information about fabric quality. Different spin directions and weaving patterns can be
associated with different regions and influences (see Wild 1970). For example, diamond
twills were being produced by the weavers in areas that were to become the north-​west-
ern provinces prior to their absorption into the empire, and continued to be regularly
produced. Basket and half-​basket weaves were a Mediterranean style that seems to have
been introduced afterwards. Wool and flax were the main fibres used, but the British
evidence is biased towards wool because linen, being a vegetable fibre, does not survive
well in anaerobic conditions.
There is an assumption that most cloth found in Britain was made there, but there
is documentary evidence that clothes could be acquired from considerable distances.
One Vindolanda letter may reflect the journey of a garrison member to Gaul to acquire
clothing supplies (Bowman and Thomas 1994: 255; Liu 2012: 21). Both provinces shared
common traditions of spin and weaving patterns, and, from the fabric alone, it would
not be possible to distinguish between origins.
In some cases an exotic origin is obvious from either the fibre or the dye. The bi-​col-
oured silk damask from a grave at Holborough (Jessup 1954; Wild 1970: 101), and the silk
damask shot through with gold thread from a burial at Spitalfields (Swain and Roberts
1999: 15), are obvious cases of import. Silk was also found in a grave at Butt Road,
Colchester (Wild 1983: 148; Crummy et al. 1993: 129). The threads of the two damasks
had been spun, indicating that the silk was woven within the empire, and a Syrian origin
has been suggested. Those of the Colchester fragment had not been spun, so the fabric
itself could have originated in China.
It is clear that less exotic fabrics were also brought to Britain. At Fordington, a child
was buried in a garment with woollen stripes dyed with true ‘Tyrian purple’ shellfish dye
(Crowfoot 2002: 158–​159; Walton Rogers 2002: 159). The thread count of a tiny fragment
from the burial was over 100 to the centimetre; both this and the dye point to a garment
coming from the Mediterranean.
It is worth reflecting on the manufacturing process. In antiquity it is generally con-
sidered that there was a spinning to weaving time ratio of at least five to one (Wild 2002:
408

408   H. E. M. Cool

8). Using a drop spindle, a spinner can produce between 35 and 50 metres of thread per
hour depending on the weight of the whorl (Mårtensson et al. 2009: 393). Using the
same authors’ calculations for the amount of thread needed to weave one square metre
of cloth, combined with the thread count for the commonest fabric type at Vindolanda
(14 × 12 (Wild 2002: 14)), it is possible to calculate that simply spinning the thread would
have taken between fifty-​three and seventy-​six hours; weaving time would then have
been a further ten to fifteen hours: all for a mere square metre of cloth. Prior to mecha-
nization during the Industrial Revolution, textile production occupied a considerable
amount of time. The products, as a consequence, were made to last. This was not the
stuff of rapid seasonal fashion changes, and was the antithesis of modern western prac-
tices where clothes can be mass produced, short-​lived fashion. The opportunity to dis-
tinguish oneself by wearing the latest or most extreme fashion was limited.
Another point to notice is that the textile industry, prior to mechanization, had gener-
ally been the preserve of women. Textile production probably developed as a gendered
activity because, unlike high temperature industries that are inherently dangerous
and also frequently have processes that cannot be interrupted, textile work is compat-
ible with childcare. It can be stopped and started without damaging the product, and
the environment is a safe one (see Barber 1994 generally, and 29–​30 especially here).
Identifying how widespread textile production was in Roman Britain is difficult,
because, unlike the periods immediately before or after, loom weights and other tools
for warp-​weighted looms are scarce. It seems probable that it was the two-​beam vertical
loom that was favoured at the time and that leaves little archaeological trace. From finds
of spindle whorls and weaving tablets it seems that domestic production was probably
widespread. In the later fourth century, spindle whorls start to be made in shale and
bone. These are turned artefacts, more effort has gone into their production, and they
regularly occur in the burials of rich women (Booth et al. 2010: 275). This suggests that
the way in which textile production was viewed may have been changing, with its accou-
trements becoming status symbols.
So, even before we move onto how the clothes might be a vehicle through which iden-
tity was expressed, it is important to realize that a large part of a woman’s identity might
be connected to cloth production. The idealized picture of the Roman matron in charge
of the textiles in her household is one frequently encountered in the literary sources.
Livia, wife of Augustus, may well have been a scheming, ambitious woman not averse
to murder, but her public face was that of the austere mistress of the house producing
her husband’s clothing, while his daughter and granddaughter were taught to spin and
weave (Suetonius, Lives of the Caesars II.63; II.73). This is reminiscent of how embroi-
dery played a defining role in the creation of female identities in post-​medieval Britain
(Parker 1984). The textile industries of antiquity have often been something of a minor-
ity interest for archaeologists, surfacing only when they are mentioned in official sources
like the Notitia Dignitatum. To the people we study, however, they were very important
and carried many layers of meaning beyond the purely functional.
Evidence for clothing styles comes primarily from pictorial representations, aided
by actual garments preserved in Egypt, and so we know most about the Mediterranean
 Clothing and Identity    409

world. Evidence available for the north-​western provinces is biased towards those fami-
lies with sufficient money and interest to set up figured grave monuments for their mem-
bers. Clothing in both regions was based on simple rectangular pieces of cloth draped
and fastened to shape. There was a minimum of tailoring and sewing, as clothes were
woven to shape on the loom. Classic Mediterranean region garments included tunics,
which were rectangles with an opening for the head centrally and the sides sewn up,
leaving gaps for the arms at the tops. In the earlier Roman period, tunics were wide, so
that, when gathered by a belt or girdle, the folds in the fabric would form rudimentary
sleeves. As time progressed, the width of the tunic decreased, and sleeves started to be
deliberately fashioned. The basic tunic was a unisex garment, belted at the waist for men
and below the bust for females, with the female version being worn longer. Women also
wore tube dresses, which were either rectangles seamed up one side or tubes woven as a
single piece on a two-​beam loom. Both could be fastened at the shoulders by brooches.
Within the north-​western provinces there is evidence for a style of tunic, which Wild
(1968: 168) has termed the Gallic coat. This was similar to the basic tunic but appears
to have had sleeves from an earlier period. Again, it was a unisex garment but does not
normally appear to have been worn belted by either sex. Outer wear consisted of cloaks
and capes (male) and mantles (female), the difference being that the former could be
hooded.
Colour and pattern were woven into the cloth, as embroidery was virtually unknown.
Coloured cloth would normally be of wool, as that can be dyed effectively, whereas
vegetable fibres, such as linen, do not take colour easily. Tunics were woven with verti-
cal stripes running down from the shoulder. In later antiquity figured tapestry-​woven
motifs were popular, at least on the clothing of the elite. To what extent these were seen
in Roman Britain is unclear. At Lullingstone, the painted figures in the house church do
appear to show this style of fashion (Wild 1987). At Poundbury, wall paintings from a
fourth-​century mausoleum seem to show the older styles with shoulder stripes (Sparey-​
Green 1993: 135–​140; Hartley et al. 2006: 207, no. 193, fig. 37). Though it has been argued
that the men depicted might be intended to show the Apostles, their appearance sug-
gests they might be portraits of a family group. Certainly, the very square-​cut faces
would have been similar to that of one of the burials (Farwell and Molleson 1993: 166).
So here we may have a snapshot of some of the men of a later Romano-​British family and
their clothing.
Even without complex tailoring, such garments can still encode a lot of messages to
viewers from the same society. The toga was the outward symbol of the male citizen and
would immediately impress the onlooker with the degree of the resources the individual
commanded. Wearing this uncomfortable garment required servants or slaves to ensure
the elaborate folds were appropriately arranged and the white wool was laundered.
Additional information would be available from the clavii (woven stripes) on tunics.
Someone of senatorial rank wore the broad stripe. Boys wore the toga praetexta with a
purple edge, exchanging it for a pure white one when they came of age. Whether such
rules were always observed is another question. At the quarries at Mons Claudianus,
in Egypt, many fabrics have wide purple stripes, and this is not the sort of site where
410

410   H. E. M. Cool

large numbers of senators are to be expected. As with sumptuary laws, which stated only
those of equestrian and senatorial rank or higher were allowed to wear gold rings, it is
possible that such codes of dress were frequently ignored. After all, there is no point in
codifying proper dress by law unless there is regular transgressive behaviour.
The other important material was leather. Properly tanned leather using vegetable
extracts was probably the best thing about being part of the Roman Empire for many
people. It was waterproof in a way that prehistoric leather had not been. This property
means that we know a great deal about Roman-​period footwear because it survives in
anaerobic conditions. Prehistoric shoes, by contrast, are extremely rare, because then
the leather was preserved by smoking and greasing. They were not waterproof in the
same way and do not normally survive. They appear to have been simple one-​piece con-
struction, whereas Roman-​period shoes had a wide range of styles and construction
techniques. Shoes wear out far more quickly than clothes, and Roman ones were subject
to regular changes of fashion. This can be tracked in the sort of detail we cannot trace in
other types of clothing. Even on sites where the leather itself does not survive, it is pos-
sible to see the adoption of Roman-​style footwear from the iron hobnails. Not all shoe
styles were nailed, but sufficient numbers were to provide us with a useful index. On
some sites it is possible to use them to show the arrival of new footwear types, alongside
other evidence of changing ways of presenting the self relating to hairstyles and the like
(Cool 2007: 346).
Hairstyles are another aspect of appearances that can be tracked. Coinage showed the
latest fashions and can be seen as a pattern book. Independent evidence that these fash-
ions were followed by women in Roman Britain is provided by the insular fashions in
hairpins, where lengths at different times mirror what would have been necessary for
the fashionable styles.

The Blue Britons

What was the national dress of the Britons in the late pre-​Roman Iron Age? We face the
problems of the non-​survival of prehistoric leather and the absence of images of dressed
individuals. Unlike some northern countries, we have no late Iron Age dressed bog bod-
ies to provide a corrective. We do have Caesar’s famous description of the Britons, which
says they painted themselves blue with woad and wore their hair long with a moustache
(Gallic Wars V.14). As this is stated alongside the information that they lived on milk and
meat and wore skins, it has to be read with some caution. The latter traits, after all, are
typical of a Roman author wishing to depict savages.
There is, however, a regular association in the literary sources between the Britons
and blue. Propertius (The Elegies II.18 B, 1–​4) writing in the mid 20s bc, harangues an
unfortunate girl for wearing woad like a Briton (see Jackson 2010: 12). Martial (Epigrams
XI.53), at the end of the first century ad, is amazed that such a paragon as Claudia Rufina
could have sprung from blue British stock. Though it is a literary device to denote
 Clothing and Identity    411

strangeness, it may have been founded on the observation that the Britons used face or
body colour in a way that other races the Romans came into contact with did not.
There is some archaeological evidence to support this idea in the form of small cop-
per alloy pestles and mortars. These cosmetic sets were for grinding mineral pigments
and not for preparing woad (Jackson 2010: 13), so more colours than blue may have been
available. They are insular and started to be made in the first century bc, so using make-​
up of whatever colour was obviously a widespread enough habit for the type to be devel-
oped. In so far as the evidence from a small number in graves can be trusted, this was a
make-​up style primarily associated with women.
As with quite a lot of types of small metal items with an origin in the later Iron Age,
a proliferation of types and an expansion in the area of use can be seen from the later
first century ad. This style of make-​up clearly spread to the rest of the province, with
occasional finds in Scotland. It continued to be favoured in the second century, probably
falling from favour in the third (Jackson 2010: 66–​68). Naturally, British women were
not the only ones to use make-​up. Elsewhere, small glass bottles have been found to con-
tain powdered cosmetics, which can be interpreted as rouge (e.g. Carazzetti and Biaggio
Simona 1988: 79; Goethert 1989: 276, no. e;), but the ubiquity of these sets in Britain sug-
gests they were serving a widespread and specialized need.
Traditions in personal grooming, starting in the late Iron Age and going into the
Roman period, are demonstrated by the small toilet sets that consist of tweezers, a small
cupped implement, and a tool with a bifid end, called a nail cleaner. These appear to
have been a continental fashion, adopted by the Catuvellaunian elite a little prior to the
conquest, which then rapidly found favour among a wider social and spatial base. This
was a fashion that continued to the end of the Roman period in Britain, with many dif-
ferent types evolving. Interestingly, although the idea of tool kit for grooming was an
import from the continent, in Britain it took on a life of its own, as, unlike elsewhere,
the nail cleaner continued to be an important element (Eckardt and Crummy 2008: 24).
The presence of razors in the graves of adult males suggests they were, at least in part,
clean-​shaven. One cremation burial, at the King Harry Lane site at Verulamium, had
both a razor and shears for cutting hair and perhaps the trimming of the moustaches
mentioned by Caesar (Stead and Rigby 1989: 370, no. 384).
Brooches become increasingly visible during the later Iron Age but, again, mainly
in southern England. It is interesting to ponder whether this indicates changing cos-
tume or depositional habits. The latter has been proposed (Jundi and Hill 1998), with
the suggestion that it reflects that items pertaining to the individual were now thought
appropriate for all forms of structured deposition, in a way they had not been before
(see Smith, this volume, for further discussion of ritual deposition). The authors suggest
that the proliferation of personal ornaments and grooming tools argues for important
changes in the way people expressed their social identities.
While this may be so, if the increasing visibility is to do with depositional practices,
then, logically, the earlier rarity cannot be assumed to mean they did not exist. After all,
the Iron Age saw an extremely low deposition rate of metal items generally. What one
can say is that various features, which may be thought to be quintessentially associated
412

412   H. E. M. Cool

with Roman Britain, were already in place in parts of southern England at the time of the
conquest. If brooches are taken as an index, then, initially, it was of limited geographi-
cal impact. It is possible to assign bow brooches to four broad bands and this is done in
Table 20.1, based on the data in Mackreth (2011). Those of the late Iron Age were in use
up to a decade or so after the invasion and include both British and continental forms.
The Colchester Derivatives are the British two-​piece series that was developing in the
50s and continued in use into the second century, and the early Roman forms are the
ones, such as Trumpets and Headstuds, that were developing in the 60s to 70s and that
continued in use into the third quarter of the second century. The middle Roman ones

Table 20.1 Brooches from England and Wales by region and broad date band

Region Late Iron Colchester Early Mid 4th Plate Pen-​ Total
Age [LFA] Derivative Roman Roman Century annular
[CD] [ER] [ER] [X4] [PEN]

North-​ 27 34 112 31 20 70 52 346

west
[NW]
North-​ 63 51 207 105 45 195 110 776
east [NE]
West 194 268 166 25 16 73 74 816
Midlands
[WM]
East 728 458 276 50 16 214 163 1,905
Midlands
[EM]
East 1,258 1,019 277 124 72 822 147 3,719
Anglia
[EA]
South-​ 1,271 945 441 79 33 469 276 3,514
west [SW]
South 1,419 322 108 38 34 298 111 2,330
central
[SC]
South-​ 625 137 44 42 44 154 59 1,105
east [SE]
Total 5,585 3,234 1,631 494 280 2,295 992 14,511

Note: The first four columns are bow brooches. Mackreth (2011: 168) notes that the plate brooch
figures are biased towards East Anglia because of historic issues over data collection. Plate brooches
have therefore been omitted from Figure 20.1(a). Labels used in Figures 20.1(a–​c) are given in [square
brackets].
Source: Mackreth (2011). © Hilary Cool.
 Clothing and Identity    413

belong to later second and third centuries, and, in the fourth century, there are the cross-
bows. As can be seen (Table 20.1 and Figure 20.1(a)), late Iron Age brooch-​wearing was,
overwhelmingly, a phenomenon of the south-​east, south-​central, and East Anglian area.
The popularity elsewhere was to come later, after the conquest.

Visible Females

The increasing visibility of women is a noticeable feature of the later first century ad,
though not one often commented on. Personal ornaments always form a sizeable part of
any small finds assemblage, and some of these are known, by grave assemblages, depic-
tions in painting and on stone, and the literary sources, to be female ornaments. These
include earrings, hairpins, most bracelets, and strings of small beads. Given the number
of hairpins found, it seems women were adopting new hairstyles that needed new tools
to produce (Stephens 2008). Insular copper-​alloy forms were in use by early in the sec-
ond century and had probably started in production earlier. During the second century,
there were regional styles covering the south and the Midlands, as well as more general
types, in metal and bone, common to all areas. Equivalent regional copper alloy groups
have not been identified in the north, but the early forms made in bone are found in
those areas there that had any interest in acquiring more than the most basic aspects of
Romano-​British material culture. There is no comparable artefact in the late Iron Age,
so this is a widespread shift in women’s appearances.
It is also worth considering the numerous brooches of the first and second centuries.
These have not generally been considered gendered artefacts, but there are changes in
the later first century that suggest it might be worth reconsidering this. The early Roman
forms, as summarized in Table 20.1, which had developed by the final quarter of the first
century, regularly have headloops. These were there to facilitate the joining-​together of
pairs of brooches by chains (Allason-​Jones 2005: figure 16). This feature is a new one,
as the late Iron Age to early Roman forms have no provision for this. Very occasion-
ally, mid to late first-​century bc brooches are found with an additional ring through
the springs, which enables them to be chained, but these tend to be made in precious
metals, like the golden ones in the Winchester hoard (Hill et al. 2004: 12) and a silver
pair from Great Chesterford (Fox 1958: plate 40b), and are thus very rare. At the large
pre-​conquest cemetery at King Harry Lane, Verulamium, well over 200 bow brooches
were found with the burials. Only one of these (Stead and Rigby 1989: 342, no. 269.3) had
any evidence that it might have been joined to any of the other brooches in the grave, as
it retained a fine iron chain wrapped around the bow. The normal pattern for most late
Iron Age people would thus seem to have been not to wear chained brooches.
It is women who need pairs of brooches, as can be seen on the classic tube dresses of
the Mediterranean world, and in the style known as Menimane’s costume, worn in the
Rhineland. There the dress, with a brooch on each shoulder, was worn over a tight-​fitting
sleeved bodice (Wild 1968: 199–​207). Men needed only single brooches to pin cloaks on
414

–0.8 –0.6 –0.4 –0.2 0.0 0.2 –1.0 –0.5 0.0 0.5
CD
WM 0.2 0.5
SWEA 0.5
ER EM NE RH
–0.0 Pen 0.0 H EA
SW EM
LFA 0.0
NW NW HPH 0.0
SC –0.2
PH
SE WM
NE MR
–0.5 –0.4 Ha –0.5
X4
–0.5
–0.6 SC
–1.0
–1.0 –0.8
–1.0 SE
–1.0 –0.5 0.0 –1.0 –0.5 0.0 0.5

(a) The types of brooches compared to
the regions they were found in.
Type labelling as per Table 20.1.
–1.0 –0.5
(b) The pin fixing arrangement on
Colchester Derivative brooches.
Ha—spring Harlow type;
RH—spring, rear hook;
PH—spring, Polden Hill types;
H—hinged;
HPH—hinged variant of Polden Hill.
0.0

0.2 PHS ET EA
NE EHS SE
EM
WM PCD
0.0 0.0
PT SC
North-west
–0.2 NW PW North-east
EW SW
–0.4
–0.5
–0.6
ECD North-west East Midlands
–0.8
–1.0 West Midlands East Anglia
–1.0
South-west
South
–1.0 –0.8 –0.6 –0.4 –0.2 0.0 0.2 central

South-east
(c) The use of colour on the early Roman South-west
brooch forms.
ECD—enamelled Colchester Derivative;
PCD—plain Colchester Derivative;
ET—enamelled Trumpet,
PT—plain Trumpet; (d) Key to the regional coding
EHS—enamelled Headstud; used in Table 20.1 and Figures 20.1(a-c).
PHS—plain Headstud;
EW—enamelled Wroxeter;
PW—plain Wroxeter.

Figure 20.1  Chronological and regional distributions of brooches.

Note: The area and type labels are those given in Table 20.1 when not otherwise specified. Figures 20.1 (a–​c) are
Correspondence Analysis plots that show which regions have higher proportions of which types.
Source: data derived from Mackreth (2011). © Hilary Cool.
 Clothing and Identity    415

one shoulder, and sometimes not even that. Depictions of early military cloaks show
them being fastened with a row of toggles down the front (Sumner 2009: figures 63–​65).
We have no useful depictions of first-​and second-​century Romano-​British women to
show us whether styles of dress had changed, but something must have altered to make it
worth bronzesmiths’ time to start to produce brooches so ideally suited to female needs.
A further point needs to be made about these headloop brooches. This was a British
fashion. Look at any continental collection, and any brooch with a headloop will be a
British type. As such, it is immediately recognized as foreign by modern scholars. The
headstud brooch from Augst, described as ‘Englische Emailbügelfibeln’ (Riha 1979: 158,
variante 5.17.4), can be noted as an example. Presumably contemporary people in Augst,
seeing someone wearing such a brooch, would also have marked them as foreign (for
Britons in the empire, see also Ivleva, this volume). Whatever was driving these changes
in Britain, it was not external influences and the desire to emulate. A new visual iden-
tity was evolving for Romano-​British women, and, possibly, this is what is driving the
upsurge in numbers of cosmetic and toilet sets in the later first century.
Interestingly, this particular identity may have been relatively short-​lived. The demise
of the headloop brooches in the mid to later second century strongly suggests a major
change in women’s fashions in Britain. Possibly it is at this point that they started wear-
ing a version of the Gallic coat, which would not have needed brooches for fastening.

A Sense of Place

This proliferation of small metal objects intimately associated with the presentation
of the individual gave rise to many different regional types. The regional distributions
of the cosmetic and toilet sets and the hairpins are relatively well understood. With
brooches we are at present merely scratching the surface, and Mackreth’s corpus (2011)
will be a rich source to mine. We know that there are regional types, but the detailed
work to see how these interact has yet to start. Plouviez (2008) and McIntosh (2011)
have looked at the broad generic level, but, given the strongly chronological patterns
seen in Figure 20.1(a), it is best to look within these at subsets of contemporary data.
This is done for the Colchester Derivatives in Figure 20.1(b), where it can be seen that
different spring-​fixing mechanisms are favoured in different regions. This is very useful
for archaeologists looking at regionality, but possibly of less use for exploring identity
issues, as these are features that would have been hidden from view.
Of more interest is the use of colour. Enamel starts to be used on brooches lavishly
from the later first century, again a major discontinuity with what had gone before.
Figure 20.1(c) shows the plot of the early Roman bow brooch types divided according
to whether they were coloured or plain. In southern England, colour was favoured in
the west and plain brooches in the east. Further north, the Midlands favoured colour,
but the north beyond that did not. It is useful to speculate whether this was related to
favoured fabric types in those areas. Coloured brooches would look better and be more
416

416   H. E. M. Cool

noticeable on plain fabrics. Conversely, shiny plain metal brooches would be ideal on
patterned material. Much work would be needed to explore the inter-​relations of this
with the contemporary enamelled plate brooches, but considering brooch colour could
well be a fruitful avenue for regional identity studies. Brooches themselves are only
small artefacts and, in themselves, of limited impact for signalling the desire to present a
particular style of appearance. Thinking of them as part of an ensemble is probably more
helpful.
Comparing regional patterns across artefact types can also be fruitful. The south-​
west was a region that seems to have been far more object-​oriented than many parts
of Britain, and it frequently produced very diagnostic regional types of commonly
used artefacts. As well as in brooches, this can be seen in hairpins (Cool 1991: figures
15–​17) and toilet sets (Eckardt and Crummy 2008: figures 29–​31). So the absence of
a distinctive type of cosmetic set in that region is curious, especially when regional
types can be picked up elsewhere (Jackson 2010: maps 4–​15). By comparison with
other objects, it seems likely that, had cosmetic sets been important and widely used
by the women of the West Country, there would have been regional types. Does this
indicate that women there did not readily adopt the make-​up the sets imply, and if
not why not? Was it only appropriate for certain types of people or activities that had
no resonance in the west? Whatever made the classical authors associate Britons and
blue was based on observations made decades prior to the conquest, when the main
point of contact was with the communities of the south-​east. Perhaps south-​western
people would have viewed the blue Britons of the south-​east as alien beings, just like
the classical authors did.

Soldiers and Authority

To the outsider, the military community would have been one that was immediately sig-
nalled by their appearance. During the first to third centuries, the presence of soldiers
on a site can easily be identified, because military fittings, such as pieces of armour and
belt fittings, are common finds. On the ground, the soldier would have been noticea-
ble by many other signals. One of these would have been that his tunic was belted. For
the early to middle Roman period there is no evidence that men in Britain wore belts if
they were not in the army. On the street, the soldier might also have been recognizable
through the style of his cloak. ‘Taking the cloak’ and ‘wearing the cloak’ were synonyms
for ‘going to’, and ‘being at’, war (Speidel 2012: 9). Sculpture regularly shows soldiers
wearing a hooded semi-​circular cloak known as paenula (Sumner 2009: 73–​80). The
identity of the soldier, even off-​duty, would have been plain to all.
As time progressed other aspects would have served to differentiate military and
civilian. The number of brooches in use decreases markedly from the mid-​second cen-
tury (Table 20.1). Proportionately, the mid-​Roman ones (Knee and P brooches) are a
commoner feature of northern assemblages than they are of southern ones (see Figure
 Clothing and Identity    417

20.1(a)), and this reflects a continued use among the military. They thus provide useful
markers of the possible presence of soldiers in southern towns on policing duty.
By the fourth century, large crossbow brooches had become symbols of authority in
the empire in a way that brooches had not been before, and were worn by both officers
and officials. Indeed, Collins (2010: 73) has gone as far as to suggest that the wearing of
any type of brooch may have signified wealth and social prestige. By the fourth century,
officials too were wearing military-​style belts, as can be seen from them forming part
of the ceremonial attributes of the Comes Sacrarum Largitonium (Count of the Sacred
Largesses), shown on the Notitia Dignitatum (Guest 2005: plate 10).
Some finds from Britain suggest that crossbows and belts carried the same signifi-
cance here. In a few places, like Lankhills (Booth et al. 2010: 278–​290), men have been
found buried with the crossbow and belt set, which suggests they were being used in
the same way as in the rest of the empire. That the brooches themselves seemed to have
retained an aura of authority is indicated by evidence that individual examples contin-
ued to be in use long after their conventional dating suggests (Booth et al. 2010: 279,
284). The evidence relating to the belts is more equivocal. In places like the south-​west,
late belt equipment of the Hawkes and Dunning (1961) styles occurs with such regularity
on sites as to cast doubt on the fact that all men who wore belts were either soldiers or in
authority (Cool 2010: 8). Insular forms developed, and the suspicion must be that many
more men were wearing belts than had ever been the situation before. In the changed
world of the fourth century, frontiers and garrisons may have been in different places
than they were in the second, and so there may have been an actual militarization of new
areas. It seems very probable, though, that this proliferation of belt equipment could
reflect more a militarization of male costume. In Britain, for whatever reason, there
seems to have been a desire to create a form of military identity that was not the same
as that in neighbouring countries. It was not only belt equipment that developed local
forms. Swift (2000: 211) has shown how the types of crossbow brooches in use in Britain
diverged markedly from the continental pattern, and there is evidence of manufacture
at Wroxeter (White 2007: 71, fig. 25). It seems very possible that it is not a military iden-
tity that is being projected, but rather a British interpretation of a continuing ‘Roman’
identity. This has interesting implications for the fifth century as these forms of belt fit-
tings and brooches continue in use then.

Problems in Perceiving Identities

In a data-​rich environment like Roman Britain, archaeology is very good at recon-

structing broad patterns of the sort that have been outlined so far. We can, for example,
track the spread of bead necklaces through the population. The glass beads of the late
Iron Age tend to be large and polychrome and may not have been used for necklaces.
From the mid-​second century much smaller, monochrome glass beads start to appear.
It is very noticeable that the earliest of these tend to occur in large numbers on military
418

418   H. E. M. Cool

sites, such as in the fortress baths drain deposit within the legionary fortress at Caerleon
(Brewer 1986: 147–1​52). So it seems a reasonable supposition to assume that this fashion
was introduced by the women who were part of the military community and who came
from other parts of the empire. By the fourth century, such beads are widespread finds,
which indicates that it was a fashion that had spread to other social groups.
What archaeology alone is less good at doing is identifying in what circumstances it
would have been appropriate to wear such necklaces. Some indications can be gained
by looking at the associations in graves. From these, it is clear that these are a female
fashion and that many of the associations are with children and adolescents. It would
be tempting to suggest that this was primarily a fashion governed by age, but it has to be
viewed against a regularly recurring pattern that indicates that young people generally,
and especially young girls, were thought to be in special need of things in their graves
(Martin-​Kilcher 2000; Gowland 2001) (see also A. Moore and Gowland, this volume).
In such circumstances caution has to be exercised before we could assume that older
women generally stopped wearing jewellery in their early twenties.
Identity does not tend to come from a single source. The main influence may change
at different stages of life or even just for particular events. All of this is likely to have been
reflected in appearance. Religious observance is a good example of this. Male citizens
tended to wear the toga only for certain formal ceremonies, one of which was sacrifice
to the gods, when a fold would have reverently been pulled up to cover the head. There
are many altars in Britain that commemorate dedications by senior officers on behalf of
their troops. At the initial dedication and at subsequent ceremonies, the commander
would most likely have been in a toga rather than his normal uniform, but this leaves no
archaeological trace.
There are also the problems that the same item can have different meanings in differ-
ent contexts. A type of massive jet bracelet made of segmented beads was just another
female ornament until one was found with an undoubtedly male skeleton in a burial at
Catterick. This individual could be interpreted as a gallus, or castrated priest, of the god-
dess Cybele who would have adopted female attire and ornamentation (see Crerar, this
volume, for discussion of ‘deviant’ burials). Interrogation of other grave associations
also revealed that this, apparently female, ornament was associated with other male
graves that showed unusual characteristics (Cool 2011: 301). That, in at least some cases,
the type represents priestly regalia seems probable, though it is still a matter of debate
whether that would have been recognized at the time as primarily a religious item.
The fact that it is the textile elements of appearance that can contribute so much to
a given identity causes particular problems. Staying with religion, we can think of the
hijab that Muslim women wear now. There is a very wide range of styles, from the simple
headscarf to the burqa, that provide information about ethnicity and/​or the sect that is
followed. Equally, the different habits of Christian monks and nuns always used to pro-
vide information about the type of Rule that was followed. All of this, though, is cloth-​
based, and equivalent signals from Roman Britain are not now available. The nearest
we can get back to a similar situation is by extrapolation. We know that Christianity
was established in Britain by the late fourth century. We can notice that there is a major
 Clothing and Identity    419

change in women’s hairstyles then, because the number of hairpins often falls in small
find assemblages (Cool 2000: 48). We can observe that, as in many patriarchal mono-
theistic religions, the early Church fathers were frequently obsessed with what women
did with their hair. We can thus suggest that there may well have been a move towards
women habitually wearing a head covering in the way they had not been doing before;
and this could well have been as a result of Christian teaching. What we cannot do is
observe this directly.
Acknowledging areas where there are problems should not distract us from the cen-
tral fact that we have great deal of data that can be used to explore many different aspects
of clothing and identity in Roman Britain, with this chapter merely scratching the sur-
face of the subject. Integrating the different strands is important, and, sometimes, there
are so many data that we are spoilt for choice. So let us finish with one of those happy
instances.
The young woman discussed above, who was buried in Spitalfields in exotic silk, is
just such a case. We know she spent her childhood in Rome, because of the lead iso-
topes in her teeth (Montgomery et al. 2010: 217–​219). We know she was rich, because she
had been provided with a lead-​lined stone sarcophagus. We can suspect she followed a
Bacchic form of religion looking to an afterlife, because of the grave goods (Cool 2002).
Which aspect came to the fore when she chose to dress in silk damask? Very possibly it
was a combination of all and more, but trying to force her into current archaeological
fashions is a sterile occupation. What would people have seen when they looked at her?
A dress made of such glorious shimmering cloth sparkling with gold would have been
beautiful. Would any other reason have been needed in life or death for its choice?

Overview of Sources

The works cited below provide key starting points for anyone wishing to explore the data
on which this chapter is based. They all have useful bibliographies.
For cloth, Wild (1970) is the foundation upon which all subsequent work on tex-
tiles in the north-​western provinces is based. His 2002 paper provides an update for
the Romano-​British textile industry. The whole of the magisterial collection of papers
considering textiles in prehistoric and Roman Europe edited by Gleba and Mannering
(2012) can be read with advantage. Especially relevant here are the papers dealing with
the British Isles by Heckett (2012), DeRoche (2012), and Wild (2012). For the relationship
between women and textiles generally Barber (1994) is full of useful insights, though the
focus of the study is on the prehistoric period.
Wild (1968) remains the best exploration of clothing in the north-​western provinces.
Croom (2000) provides an accessible overview of clothing in the whole empire, with a
good colour plate section that shows both the ancient sources and modern reconstruc-
tions of how the garments would look. It also includes an illustrated summary of hair-
style fashion. For military clothing, the best starting points are Sumner’s study (2009)
420

420   H. E. M. Cool

of military dress and the collection of essays edited by Nosch (2012). The former is
beautifully illustrated and has an important section on the colour of the garments. The
best overviews for prehistoric and Roman footwear are the essays by Groenman-​van
Waateringe (2001) and van Driel-​Murray (2001a) in Goubitz et al. (2001); van Driel-​
Murray’s paper (2001b) is also useful and possibly more accessible for many people.
The two key texts for brooches are by Bayley and Butcher (2004) and Mackreth (2011).
The most comprehensive overview of metal personal ornaments, other than brooches,
remains my doctoral thesis (Cool 1983, now available online), with the hairpins updated
in Cool (1991). The thesis was completed prior to the boom in metal detecting that has
brought many new precious metal jewellery items to light and so may usefully be aug-
mented by Johns (1996). The main typology for bone hairpins is that of Crummy (1979),
for glass beads that of Guido (1978), and earrings that of Allason-​Jones (1989). The start-
ing point for the metal elements of military equipment is Bishop and Coulston (1993
[second edition  2006]). Swift (2000) is an important review of crossbow brooches,
belt-​fitting, bracelets, and beads in fourth-​century western Europe. The report on the
material found at the Lankhills cemetery provides useful updates on these and other
categories from fourth-​to fifth-​century Roman Britain (Booth et al. 2010). For toilet and
other grooming equipment, the starting point is Eckardt and Crummy (2008). Swift’s
paper (2011) provides additional guidance on the bibliography for personal ornamen-
tation. There are also introductory and specialist bibliographies on <www.barbicanra.
co.uk>, which are updated from time to time.

References
Primary Sources
Julius Caesar, The Gallic Wars, trans. H. J. Edwards. Cambridge, MA: Harvard University Press,
1917.
Martial, The Epigrams, trans. D. R. Shackleton-​Bailey. Cambridge, MA: Harvard University
Press, 1993.
Propertius, The Elegies, trans. A. S. Kline <http://​www.poetryintranslation.com/​PITBR/​Latin/​
PropertiusBkTwo.htm#Toc201112270> (accessed 17 October 2013).
Suetonius, Lives of the Caesars, trans. J. C. Rolfe. Cambridge, MA: Harvard University Press,
1914.

Secondary Sources
Allason-​Jones, L. (1989). Earrings in Roman Britain. BAR British Series 201. Oxford: Archaeopress.
Allason-​Jones, L. (2005). Women in Roman Britain. New edn. York: Blackwell.
Barber, E. W. (1994). Women’s Work:  The First 20,000 Years. New  York and London:  W.
W. Norton & Co.
Bayley, J., and Butcher, S. (2004). Roman Brooches in Britain: A Technological and Typological
Study based on the Richborough Collection. Research Report of the Committee of the Society
of Antiquaries of London 68. London: The Society of Antiquaries of London.
 Clothing and Identity    421

Bishop, M. C., and Coulston, J. C. N. (1993)[2006]. Roman Military Equipment from the Punic
Wars to the Fall of Rome. 2nd edn. London: B. T. Batsford Ltd. Oxford: Oxbow Books.
Booth, P., Simmonds, A., Boyle, A., Clough, S., Cool, H. E. M., and Poore, D. (2010). The Late
Roman Cemetery at Lankhills, Winchester:  Excavations 2000–​2005. Oxford Archaeology
Monograph 10. Oxford: Oxford Archaeology.
Bowman, A. K., and Thomas, J. D. (1994). The Vindolanda Writing Tablets. Tabulae
Vindolandenses II. London: British Museum Press.
Brewer, R. J. (1986). ‘The Beads and Glass Counters’, in J. D. Zienkiewicz, The Legionary Fortress
Baths at Caerleon: The Finds. Cardiff: National Museum of Wales, 145–​156.
Carazzetti, R., and Biaggio Simona, S. (1988). Vetri Romani del Cantone Ticino. Locarno: Cittá
di Locarno Museo Civico e Archaeologico.
Collins, R. (2010). ‘Brooch Use in the Fourth and Fifth Century Frontier’, in R. Collins and L.
Allason-​Jones (eds), Finds from the Frontier. CBA Research Report 162. York: Council for
British Archaeology, 64–​77.
Cool, H. E. M. (1983). ‘A Study of the Roman Personal Ornaments Made of Metal, Excluding
Brooches, from Southern Britain’, Ph.D.  thesis, University of Wales <http://​ethos.bl.uk/​
Home.do> (accessed 2 June 2014).
Cool, H. E. M. (1991). ‘Roman Metal Hair Pins from Southern Britain’, Archaeological Journal,
147 (1990), 148–​182.
Cool, H. E.  M. (2000). ‘The Parts Left Over:  Material Culture into the 5th Century’, in T.
Wilmott and P. Wilson (eds), The Late Roman Transition in the North. BAR British Series
299. Oxford: Archaeopress, 47–​65.
Cool, H. E. M. (2002). ‘Bottles for Bacchus?’, in M. Aldhouse-​Green and P. Webster (eds), Artefacts
and Archaeology: Aspects of the Celtic and Roman World. Cardiff: University of Wales Press,
132–​151.
Cool, H. E. M. (2007). ‘The Small Finds in their Regional Context’, in D. Miles, S. Palmer,
A. Smith, and G. P. Jones, Iron Age and Roman Settlement in the Upper Thames
Valley: Excavations at Claydon Pike and Other Sites within the Cotswold Water Park. Thames
Valley Landscapes Monograph 27. Oxford: Oxford Archaeology, 342–​350.
Cool, H. E. M. (2010). ‘A Different Life’, in R. Collins and L. Allason-​Jones (eds), Finds from the
Frontier. CBA Research Report 162. York: Council for British Archaeology, 1–​9.
Cool, H. E.  M. (2011). ‘Funerary Contexts’, in L. Allason-​Jones (ed.), Artefacts in Roman
Britain: Their Purpose and Use. Cambridge: Cambridge University Press, 293–​312.
Croom, A. (2000). Roman Clothing and Fashion. Stroud: Tempus.
Crowfoot, E. (2002). ‘Textiles from the Lead Coffin’, in S. M. Davies, P. S. Bellamy, M. J. Heaton,
and P. J. Woodward, Excavations at Alington Avenue, Fordington, Dorchester, Dorset, 1984–​
87. Dorset Natural History and Archaeological Society Monograph 15. Dorchester: Dorset
Natural History and Archaeological Society, 158–​159.
Crummy, N. (1979). ‘A Chronology of Romano-​British Bone Pins’, Britannia, 10: 157–​164.
Crummy, N., Crummy, P., and Crossan, C. (1993). Excavations of Roman and Later Cemeteries,
Churches and Monastic Sites in Colchester, 1971–​88. Colchester Archaeological Report
9. Colchester: Colchester Archaeology.
DeRoche, D. (2012). ‘England: Bronze and Iron Ages’, in M. Gleba and U. Mannering (eds),
Textiles and Production in Europe: From Prehistory to AD 400. Ancient Textiles Series, vol-
ume 11. Oxford: Oxford University Press, 444–​450.
Eckardt, H. and Crummy, N. 2008. Styling the Body in Late Iron Age and Roman Britain.
Montagnac: Monographies Instrumentum 36.
422

422   H. E. M. Cool

Farwell, D. E., and Molleson, T. I. (1993). Excavations at Poundbury 1966–​ 80. Volume
2:  The Cemeteries. Dorset Natural History and Archaeological Society Monograph 11.
Dorchester: Dorset Natural History and Archaeological Society.
Fox, C. (1958). Pattern and Purpose. Cardiff: National Museum of Wales.
Gardner, A. (2007). An Archaeology of Identity: Soldiers and Society in Late Roman Britain.
Walnut Creek, CA: Left Coast Press.
Gleba, M., and Mannering, U. (2012) (eds). Textiles and Production in Europe: From Prehistory
to ad 400. Ancient Textiles Series, volume 11. Oxford: Oxbow Books.
Goethert, K. (1989). ‘Zur Körper–​und Schönheitspflege in frührömischer Zeit’, in A. Haffner
(ed.), Gräber–​Spiegel des Lebens. Mainz am Rhein: Verlag Philipp von Zabern, 275–​288.
Goubitz, O., van Driel-​Murray, C., and Groenman-​van Waateringe, W., Stepping through Time.
Zwolle: Stichting Promotie Archeologie.
Gowland, R. (2001). ‘Playing Dead:  Implications of Mortuary Evidence for the Social
Construction of Childhood in Roman Britain’, in G. Davies, A. Gardner, and K. Lockyear
(eds), TRAC 2000:  Proceedings of the Tenth Theoretical Roman Archaeology Conference,
London 2000. Oxford: Oxbow Books, 152–​168.
Groenman-​van Waateringe, W. (2001). ‘Prehistoric Footwear’, in O. Goubitz, C. van Driel-​
Murray, and W. Groenman-​van Waateringe (eds), Stepping through Time. Zwolle: Stichting
Promotie Archeologie, 379–​396
Guest, P. (2005). The Late Roman Gold and Silver Coins from the Hoxne Treasure.
London: British Museum Press.
Guido, M. (1978). The Glass Beads of the Prehistoric and Roman Periods in Britain and Ireland.
Research Report of the Committee of the Society of Antiquaries of London 35. London: The
Society of Antiquaries of London.
Hartley, E., Hawkes, J., Henig, M., and Mee, F. (2006) (eds). Constantine the Great:  York’s
Roman Emperor. York and Aldershot: York Museums and Gallery Trust/​Lund Humphries.
Hawkes, S. C., and Dunning, G. C. (1961). ‘Soldiers and Settlers in Britain, Fourth to Fifth
Century with a Catalogue of Animal-​Ornamented Buckles and Related Fittings’, Medieval
Archaeology, 5: 1–​70.
Heckett, E. W. (2012). ‘Scotland and Ireland’, in M. Gleba and U. Mannering (eds), Textiles and
Production in Europe: From Prehistory to ad 400. Ancient Textiles Series, volume 11. Oxford:
Oxbow Books, 428–​442.
Hill, J. D., Spence, A. J., La Niece, S., and Worrell, S. (2004). ‘The Winchester Hoard: A Find of
Unique Iron Age Gold Jewellery from Southern England’, Antiquaries Journal, 84: 1–​22.
Jackson, R. (2010). Cosmetic Sets of Late Iron Age and Roman Britain. British Museum Research
Publication 181. London: British Museum.
James, S., and Millett, M. (2001) (eds). Britons and Romans:  Advancing an Archaeological
Agenda. CBA Research Report 125. York: Council for British Archaeology.
Jessup, R. F. (1954). ‘Excavation of a Roman Barrow at Holborough, Snodland’, Archaeologia
Cantiana, 68: 1–​61.
Johns, C. (1996). The Jewellery of Roman Britain. London: UCL Press.
Jundi, S., and Hill, J. D. (1998). ‘Brooches and Identities in First Century Britain: More than Meets
the Eye?’, in C. Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 97: Proceedings of the Seventh
Theoretical Roman Archaeology Conference, Nottingham 1997. Oxford: Oxbow Books, 125–​137.
Liu, J. (2012). ‘Clothing Supply and the Military: A Look at the Inscriptional Evidence’, in M.-​L.
Nosch (ed.), Wearing the Cloak: Dressing the Soldier in Roman Times. Ancient Textiles Series
10. Oxford: Oxbow Books, 19–​28.
 Clothing and Identity    423

McIntosh, F. (2011). ‘Regional Brooch-​Types in Roman Britain:  Evidence from Northern

England’, Archaeologia Aeliana [Fifth Series], 60: 155–​182.
Mackreth, D. F. (2011). Brooches in Late Iron Age and Roman Britain. Oxford: Oxbow Books.
Mårtensson, L., Nosch, M.-​ L., and Andersson Strand, E. (2009). ‘Shape of
Things: Understanding a Loom Weight’, Oxford Journal of Archaeology, 28: 373–​398.
Martin-​Kilcher, S. (2000). ‘Mors immatura in the Roman World:  A  Mirror of Society and
Tradition’, in J. Pearce, M. Millett, and M. Strück (eds), Burial, Society and Context in the
Roman World. Oxford: Oxbow Books, 63–​77.
Montgomery, J., Evans, J., Chenery, S., Pashley, V., and Killgrove, K. (2010). ‘“Gleaming, white and
deadly”: Using Lead to Track Human Exposure and Geographic Origins in the Roman Period in
Britain’, in H. Eckardt (ed.), Roman Diasporas. Portsmouth, RI: Journal of Roman Archaeology
Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 199–​226.
Nosch, M.-​L. (2012) (ed.). Wearing the Cloak: Dressing the Soldier in Roman Times. Ancient
Textiles Series 10. Oxford: Oxbow Books.
Parker, R. (1984). The Subversive Stitch. London and New York: Women’s Press [repr. 2010].
Plouviez, J. (2008). Counting Brooches: Ways of Examining Roman Brooch Assemblages in
London and Beyond’, in J. Clark, J. Cotton, J. Hall, R. Sherris, and H. Swain (eds) Londinium
and Beyond. CBA Research Report 156. York: Council for British Archaeology, 171–​176.
Riha, E. (1979). Die römischen Fibeln aus Augst und Kaiseraugst. Forschungen in Augst
3. Augst: Amt für Museen und Archäologie des Kanrons Basel-​Landschaft.
Sparey-​Green, C. (1993). ‘The Mausolea Painted Plaster’, in D. E. Farwell and T. I. Molleson,
Poundbury. Volume II:  The Cemeteries. Dorchester:  Dorset Natural History and
Archaeological Society, 135–​140.
Speidel, M. A. (2012). ‘Dressed for the Occasion: Clothes and Context in the Roman Army’, in
M.-​L. Nosch (ed.), Wearing the Cloak: Dressing the Soldier in Roman Times. Ancient Textiles
Series 10. Oxford: Oxbow Books, 1–​12.
Stead, I. M., and Rigby, V. (1989). Verulamium: The King Harry Lane Site. English Heritage
Archaeological Report 12. London: English Heritage.
Stephens J. (2008). ‘Ancient Roman Hairdressing:  On (Hair)pins and Needles’, Journal of
Roman Archaeology, 21: 111–​132.
Sumner, G. (2009). Roman Military Dress. Stroud: History.
Swain, H., and Roberts, M. (1999). The Spitalfields Roman. London: Museum of London.
Swift, E. (2000). Regionality in Dress Accessories in the late Roman West. Millau: Monographies
Instrumentum 11.
Swift, E. (2011). ‘Personal Ornaments’, in L. Allason-​ Jones (ed.), Artefacts in Roman
Britain: Their Purpose and Use. Cambridge: Cambridge University Press, 194–​218.
van Driel-​Murray, C. (2001a). ‘Footwear in the North-​Western Provinces of the Roman
Empire’, in O. Goubitz, C. van Driel-​Murray, and W. Groenman-​van Waateringe, Stepping
through Time. Zwolle: Stichting Promotie Archeologie, 337–​376.
van Driel-​Murray, C. (2001b). ‘Vindolanda and the Dating of Roman Footwear’, Britannia,
32: 185–​197.
Walton Rogers, P. (2002). ‘Dye Tests on Textile Fragments from Lead Coffin’, in S. M. Davies,
P. S. Bellamy, M. J. Heaton, and P. J. Woodward, Excavations at Alington Avenue, Fordington,
Dorchester, Dorset, 1984–​87. Dorset Natural History and Archaeological Society Monograph
15. Dorchester: Dorset Natural History and Archaeological Society, 159.
Walton Rogers, P. (2007). Cloth and Clothing in Early Anglo-​Saxon England. CBA Research
Report 145. York: Council for British Archaeology.
424

424   H. E. M. Cool

White, R. (2007). Britannia Prima. Stroud: Tempus.

Wild, J. P. (1968). ‘Clothing in the North-​West Provinces of the Roman Empire’, Bonner
Jahrbucher, 68: 166–​240.
Wild, J. P. (1970). Textile Manufacture in the Northern Roman Provinces. Cambridge: Cambridge
University Press.
Wild, J. P. (1983). ‘Textile Fragments from the Later Butt Road Cemetery’, in N. Crummy, The
Roman Small Finds from Excavations in Colchester 1971–​9. Colchester Archaeological Report
2. Colchester: Colchester Archaeology, 147–​148.
Wild, J. P. (1987). ‘The Dress of the Lullingstone Figures’, in G. W. Meates, The Lullingstone
Roman Villa. Volume II—​ the Wall Paintings and Finds. Kent Archaeological Society
Monograph 3. Maidstone: Kent Archaeological Society, 40–​41.
Wild, J. P. (2002). ‘The Textiles Industries in Roman Britain’, Britannia, 33: 1–​42.
Wild, J. P. (2012). ‘England: Roman Period’, in M. Gleba and U. Mannering (eds), Textiles and
Production in Europe: From Prehistory to AD 400. Ancient Textiles Series, volume 11. Oxford:
Oxford University Press, 451–​456.
 Chapter 21

Cem eteries an d Fu ne ra ry
Practi c e

Jake Weekes

Introduction

The archaeology of Romano-​British cemeteries and funerary practice in the early

twenty-​first century is a study of the entire funerary process (see Pearce 1997, 1998, 2001,
2002, 2013, 2015; also Fitzpatrick 2001; Cool 2004; Weekes 2008). Much emphasis has
been placed on reconstructing cremation rituals prior to cremation burial (McKinley
1989, 1994a–​c; 2001, forthcoming), which has revitalized study and led the way in devel-
oping the funerary sequence method.
Our understanding of Romano-​British funerals is based primarily on archaeologi-
cal evidence. The documentary evidence for Roman funerals is not directly appli-
cable to the Romano-​British context; this material can generally serve here only as
indirect evidence of traditions that Roman armies and administration and their
associates possibly brought with them to the new province. The unequivocal mate-
rial evidence for Roman-​style funerals in Britain tends to be epigraphic and art his-
torical, like the London tombstone of first-​century procurator Cn Julius Classicianus
(RIB 12).
My selection of examples is—​of necessity—​limited, with Robert Philpott’s examples
(1991) augmented by some well-​known and published sites and surveys (see Table 21.1;
for further detail the reader is referred to the relevant publications).
I suggest six overlapping ‘stages’ of the funerary process as useful indices for com-
parative analysis:  selection, preparation, location, modification, deposition, and
commemoration.
426

426   Jake Weekes

Table 21.1 Some significant and published case studies

Site Reference(s)

Baldock, Hertfordshire Stead and Rigby (1986)

The Lexden Tumulus, near Colchester Foster (1986); see also Crummy (1997)
Folly Lane, St Albans Niblett (1999, 2001)
Welwyn Garden City Stead (1967)
King Harry Lane, St Albans Stead and Rigby (1989)
Westhampnett, West Sussex Fitzpatrick (1997, 2001)
Stanway, Essex Crummy et al. (2007)
Pepper Hill, Springhead, Kent Booth et al. (2011)
St Dunstan’s, Canterbury Bennett (1987); Weekes (2011); Diack and Weekes
(forthcoming)
Colchester (various) Crummy et al. (1993)
Butt Road, Colchester Crummy et al. (1993)
Eastern Cemetery, London Barber and Bowsher (2000)
Western Cemetery, London S. Watson 2003
St Pancras, Chichester Down and Rule (1971)
Southwark Mackinder (2000)
Kelvedon, Essex Rodwell (1988)
Lankhills, Winchester Clarke 1979; Booth et al. (2010)
Poundbury, Dorset Farwell and Molleson (1993)
Cirencester (especially Bath Gate) McWhirr et al. (1982)
London Road, Gloucester Simmonds et al. (2008)
York (various) RCHM (1962); Wenham (1968)
Brougham, Cumbria Cool (2004)
Various continental examples Catalano et al. (2008)
(including Rome and Italy)

Selection

John Pearce, by estimating the likely population of late Iron Age and Roman Britain
and its death rate based on comparative figures for pre-​industrial societies, has dem-
onstrated that the very idea of individual burial was probably a late Iron Age and
 Cemeteries and Funerary Practice    427

Romano-​British novelty in many areas, still afforded only to a proportion of the dead
(Pearce 1999, 2008, 2013), at the same time noting that there are certain to have been
important regional differences. Simon Esmonde Cleary has also commented that buri-
als are more common in the later than the earlier period, in the south-​east of the prov-
ince, and in the towns rather than rural settings: ‘a reversal of the actual population
distribution’ (Esmonde Cleary 2001: 127). Cemetery populations then, already ‘selected’
by the very fact of death, were also selected from a wider pool of the dead in being given
an archaeologically visible funeral.
Whatever the selection process, burials and cemeteries are where we actually come
‘face to face’ with many of the people we are interested in. Developments in osteological
techniques (see Cirencester, Lankhills, Poundbury, and London case studies; but also
Redfern 2003, Gowland and Knüsel (eds) 2006, and Duday 2009, for advances) and
other archaeological sciences can reveal metric and non-​metric skeletal traits, mortal-
ity rates, and even supposed ethnic markers. Recent advances in stable isotope analysis
have provided a more solid scientific basis for investigating the origins of certain groups,
however, such as those deduced at the Winchester cemetery of Lankhills (Clarke 1979;
see also Pearce 2010; Booth, forthcoming; note also Cool’s timely reminder (2010) of
the precarious use of finds typology in delineating foreign ‘groups’ within cemeter-
ies). A particularly striking example is the Roman woman buried at Spitalfields in the
fourth century, her origin denoted by a signature lead isotope (Swain and Roberts 2001;
Montgomery et al. 2010). We can also infer variations in diet, through relative age and
stature and the trace signatures of stable isotopes. Congenital factors or infectious dis-
eases are clearly apparent in many cases, and even the occupations of the dead in life are
suggested by such factors as high lead content, bone and muscle developments from
repetitive strain, and ante-​, peri-​, and post-​mortem cut marks on the frames of soldiers,
gladiators, deviants, and others. Some healed trauma denote considerable care and
attention to the injured party (for example, at Cirencester), while certain peri-​mortem
injuries have become notorious (Wheeler 1943: 61 ff., plate 7A; Redfern 2011). One of the
most important reasons for understanding more about the lives and deaths of the dead,
and the thing that most concerns us here, is that we might begin to detect the particular
funerary choices (selection again) made in relation to those particular circumstances.
Every society has its concepts of ‘good’ and ‘bad’ deaths (see Middleton 1982), inform-
ing the types of funeral afforded to the deceased. Preparing for death can be a drawn-​
out process commencing long before physical death occurs. At Roman York, the son of
Titus Flavius Flavinus ordered his tombstone ‘to be made before Flavinus died’ (RIB 675;
RCHM 1962: 124 [no. 78]; compare the complicated preparations and revisions of one P.
Vesonius Phileros of Pompeii (Lepetz et al. 2011)). For others, death is a sudden shock to
those around them, and the remains of some are never recovered, in which case a ceno-
taph must suffice. Archaeologists have begun to infer as much from burials that seem to
have everything except human remains, but the Chester tombstone of an Optio lost at
sea (RIB 492; see Mason 2001: 113–​114) provides a clear example of the practice.
As well as, and in conjunction with, the type of death, culturally specific values con-
cerning status and gender of the deceased, their age at death, and so on will all play their
428

428   Jake Weekes

part in selecting the appropriate funerary rites (see Parker Pearson 1999; cf. Metcalf and
Huntington 1992). In the late Iron Age in south-​east England, we begin to see increas-
ingly aristocratic individuals singled out for Gallic-​influenced cremation funerals, as at
Baldock, Folly Lane, and Lexden. Elite funerals continue in Roman Britain (see Strück
2001; Pearce, this volume), but ‘special’ cases of a different sort include the many infants
and children interred within the post settings of buildings, and sometimes explicitly
propitious locations, like the temple complex at Springhead in Kent (see Philpott 1991:
97–​102; Esmonde Cleary 2001: 135–13​6; Moore 2009).
And there were others in the province for whom certain types of death were perhaps
designated in society’s response to ‘otherness’, or prescribed for deviance, or simply pre-
dictable, given their occupation. Decapitation was a widespread phenomenon (Philpott
1991: 77 ff.), although it is often unclear whether the operation occurred before or after
death, raising questions as to whether the headless were victims of a punitive violence
or ritual overkill. Others lie prone in the grave, or are found both prone and decapi-
tated; and yet, interestingly, some of these were also buried in coffins. Were they the
recipients of a deliberately ‘bad death’ at the hands of one faction (a judicial system),
given a ‘proper’ funeral by another group (family)? A prone burial placed in the backfill
of a cenotaph grave at Lankhills conjures ideas of sacrifice or revenge murder or both,
for which Iron Age precedents can be found (see also Crerar, this volume, for a discus-
sion on deviant burials). The latter might include the many burials in pits, ditches, and
even caves (see Whimster 1981: ch. 1), or the dramatic overkill of the invasion-​period
man found lying prone in Lindow Moss in Cheshire (Stead et al. 1986). The early Roman
period produces skulls from Colchester and skulls from the Walbrook in London, while
there are early and late ‘burials’ in wet places, shafts, wells (as at Baldock), and pits (see
Esmonde Cleary 2001: 134 ff.). The later examples are sometimes considered a sign of a
breakdown in society, but such interpretation might reflect the unconscious influence
of colonial mentality in response to perceived ‘otherness’ in the archaeological record.
Some burials are just frankly suspicious (see Bennett et al. 1982: 43–​46). For many, then,
defined in death by their perceived status and/​or how and why they died, the ‘normal’
funerary process could have been severely compromised and foreshortened. For many
others, it is clear that the process could be far more drawn out and intricate.

Preparation

In the case of a death good enough to warrant what was understood to be a ‘good’
funeral, the very first rituals of the sequence must have included some form of laying-​
out of the deceased. A fruitful but understudied topic here is the broadly typical limi-
tations imposed by the natural breakdown of the corpse itself (see, e.g., Iserson 2002;
Jackson and Jackson 2004). Beginning immediately at death, this alternative body clock
must affect the structure of laying-​out periods and other funerary manipulations and
modifications of the human remains, as well as the amount of time available for other
 Cemeteries and Funerary Practice    429

preparations. Culturally specific responses to the onset of rigor mortis are of interest,
but also processes such as livor mortis (discoloration of the body due to resettling of
the blood), autolysis, and putrefaction, all variously affected by environmental condi-
tions (Byers 2002: ch. 5; Dupras et al. 2012: 69 ff.). Following a primary flaccidity, rigor
mortis might begin between two and six hours after death, initially affecting the eyes,
mouth, and jaw. This could impact on the timing of ritual actions described in Roman
documentary sources, for example (see Adam 1835: 415, 416; Toynbee 1971: 44; Philpott
1991: 211 ff.), such as closing of the eyes and removal of rings and placement of coins in
the mouth (also attested archaeologically). Even the findings of forensic entomologists
could be admissible, since we have reports of Roman boys employed to drive away flies
during protracted periods of laying-​out associated with public funerals (see Adam 1835:
417–​418).
In the late Iron Age a large mortuary enclosure at Folly Lane, St Albans, was prob-
ably used for elite display at the laying-​out stage, with a central subterranean ambulatory
chamber, perhaps for viewing and probably making offerings to the deceased. The bur-
ial pit in the Lexden Tumulus, near Colchester, could also have been used in this fashion
and, like the Folly Lane example, was apparently partially destroyed as part of the cer-
emony. Also near Colchester, the floruit of a major complex of mortuary enclosures at
Stanway, including chambers and burials, is dated at c. ad 60, suggesting continuity of
this elite practice (cf. enclosures recently discovered in Kent (Allen et al. 2012; Stevenson
2013)). Ditched enclosures with central burials marked several burial foci at King Harry
Lane, St Albans. Here Niblett (2001: 101–​102) suggests that some inhumations within the
ditches could represent initial burial of the corpse prior to cremation. Mortuary struc-
tures at the Iron Age cremation cemetery at Westhampnett may also have been for lay-
ing-​out before cremation nearby. Were cemetery structures of the Roman period used
for such purposes? It is a function (not excluding other functions) that could at least be
afforded by many cemetery structures, and certainly the more definite mausolea, tem-
ple mausolea, tombs, and walled cemeteries (see Jessup 1959; Meates 1979; Black 1987:
202–​210; Strück 2001; and various examples at York, Poundbury, Southwark, and East
London).
Much more direct evidence comes from cremated material that preserves personal
ornaments/​accessories such as burnt hairpins, melted items of glass, and metal fused
to bones and hobnails. These objects therefore certainly derive from the pyre (see
McKinley 1989, 1994a, b, 2004, forthcoming; Cool 2004: 438; Weekes 2008: 147–​148) but
may well ‘belong’ earlier in the funeral, as they are equally likely to have been placed on
the corpse to adorn or clothe it during the laying-​out stages. Burnt or un-​burnt unguen-
taria (or balsamaria: small glass phials usually thought to be for anointing the body or
bones with oil (see, e.g., Jones 2002: plate 7)), found in many Romano-​British cremation
burials from as early as the first century (Philpott 1991: 117–​118), are often thought to
have contained aromatics or oils as a complement or aid to cremation (e.g. Cool 2004:
441). However, whatever they contained they could have served an earlier purpose,
as an aromatic and/​or cosmetic either prior to cremation (Weekes 2008: 148; Lepetz
et al. 2011) or equally in funerals leading to inhumation, where unguentaria are again
430

430   Jake Weekes

commonly found (Philpott 1991: 118–​119; e.g. the East London cemetery). These little
bottles are worthy of more scrutiny in comparative studies, as they may mark a particu-
lar provincial attempt at ‘Romanness’ (cf. Catalano et al. 2008: 81; Lepetz et al. 2011).
While cremation is now seen by many archaeologists as a means to unlock prepa-
ration rites, commentators have only tentatively considered the evidence for the early
stages of inhumation funerals (e.g. Booth et al. 2010: 470 ff.). Even given culturally rela-
tivist caveats, however, it seems most likely that most corpses were dressed or shrouded
prior to being transported to the cemetery (or pyre), and therefore that clothing and
dress accessories could have important laying-​out connotations per se. It is the presence
of a coffin or bier in an inhumation burial, however, that offers the most hope of recon-
structing the laying-​out ceremony. Some late Iron Age (see Stevenson 2013), and many
Romano-​British inhumations (Whimster 1981: 43; Philpott 1991: 53) preserve traces of
wooden coffins in the form of linear stains or patterns of nails within the burial pit (at
Kelvedon, Essex, and Pepperhill, Kent, aspects of coffin design could be reconstructed
from such evidence), and sometimes contain extant coffins of lead or stone (for exam-
ple, at London, Lankhills, Poundbury, York, and elsewhere; wooden coffins with lead
linings were an interesting feature at Poundbury). Some inhumations (at East London,
for example) have been interpreted as containing the bier on which the deceased was
transported to the grave, while burnt bone veneers and iron tacks recovered from cre-
mation deposits are now also thought to derive from decorated wooden biers or coffins
on which bodies were carried to and placed on the pyre (Cool 2004).
The fact that many of the lead coffins, at least, were extravagantly decorated again
suggests that they were for viewing by the living during the laying-​out and subsequent
phases, but the key point here is that the presence of a coffin has significant implications
for certain ‘grave goods’ found in close association with the body within extant or recon-
structed coffin spaces. All such objects could have been placed in the coffin at the grave
side, but it seems likely that deposition within the coffin would most often have taken
place during laying-​out, along with arrangements for clothing, dress accessories, and
probably abandonment of unguentaria and the placing of coins in the mouth, for example
The coffin has its own stratigraphy. The East London cemetery, for example, pro-
duced a number of inhumations where the body had been placed on top of a bed of
crushed chalk within the coffin, just one of the substances (others include burnt clay
and sand (see Philpott 1991: 93)) used in this way in Roman Britain, sometimes inter-
preted as a method of controlling the effects or agents of putrefaction. In predominantly
fourth-​century ‘plaster burials’, which seem to have been reserved for a particular group
(Sparey Green 1977: 48; Philpott 1991: 90–​97), the body was, at some stage prior to
burial, encased (within the coffin) either wholly or partially by gypsum plaster or lime.
In some cases this process preserved detailed evidence of the weave of a shroud, as at
Butt Road, Colchester, Poundbury, and York (see Philpott 1991: 93; Crummy et al. 1993;
shrouds can also be inferred from certain burial postures (Booth et al. 2010: 474 ff.)). The
sheer weight and size of many massive stone coffins (as within mausolea at Poundbury
and in burials at Cirencester) suggest that they served as both sarcophagus and grave
elaboration.
 Cemeteries and Funerary Practice    431

Poundbury and York burials within lead-​lined and lead coffins astonishingly pre-
served another aspect of laying-​out ceremonial: hair styling. A York woman’s auburn
hair is reported to have been held in a bun at the back of the head by two jet hairpins,
remarkably similar to the styling of more than one woman’s hair in the Poundbury cem-
etery. A young man of the late Roman period at Poundbury was laid to rest with care-
fully combed and finger-​waved hair, in ringlets.
Again, the posture of the body in an inhumation (at least where a coffin, bier, or
shroud is present) might well have been arranged during laying-​out: for example, lying
supine with the hands crossed on the chest, or perhaps lying prone, or, for that mat-
ter, with a replaced decapitated head, either in correct anatomical position or perhaps
between the knees or at the feet (Philpott 1991: 77 ff., 81 ff). The changing nature of the
corpse is also significant in interpreting burial posture. If a laying-​out period lasted long
enough (more than, say, thirty-​six hours (Iserson 2002)), the body could naturally have
entered a new phase of flaccidity prior to burial. Is it perhaps notable then that Roman
sources report a ritual of reopening of eyes and replacement of rings (Adam 1835: 415)
before the lighting of the pyre? In coffined inhumation burials, an understanding of sec-
ondary flaccidity could suggest alternative interpretations for at least some minor varia-
tions in posture. Even without taking into account the extreme physical demonstrations
of grief centred on a closed coffin in some cultures, trundling along a track on a cart and
being lowered in a long, heavy box into a grave could easily prove enough to dislodge a
carefully arranged arm or positioned head; the post-​depositional effects of decomposi-
tion combined with gravity and post-​depositional disturbance must also be considered
(see Duday 2006: 34–3​6).
Many other things would need to be prepared and set in motion at these early stages
of the funerary sequence, such as arrangements for cremation and/​or burial (pyre spe-
cialists, grave diggers, pyre and cemetery location, fuel, and construction), communica-
tion with or employment of mourners, procurement of coffins, biers, pyre, and grave
offerings. And what of enclosure and mound construction, construction of tombs, bar-
rows, mausolea, walled cemeteries, funerary monuments, monumental stone coffins,
sarcophagi and cists, and tombstones? Clearly some of these elements would have to
have been already prepared prior to death or would require a secondary burial tech-
nique of some sort, or could be added after burial of the remains, like funerary mon-
uments, tombstones, and barrows. Much of this activity could have been mediated or
made possible through contributions to a collegium (funeral ‘club’).

Location

In all human societies there are places for the living, and places for the dead (see Parker
Pearson 1999: especially 124 ff.). The next stage of the funerary process would be to
remove the corpse to the place where it will soon belong, the burial place or cemetery,
perhaps modifying the body en route. This, in an urban setting at least, either chimed
432

432   Jake Weekes

with, or was a response to, the imposition of Roman ritualized space, the legal necessity
for burials to be carried out beyond the ceremonial boundary of the settlement as set out
in the Twelve Tables (Cicero, De Legibus ii, 23, 58; see Toynbee 1971: 48). This brought
about the Roman conduct of pompa, processions with public expressions of grief, which
appear on funerary reliefs (Toynbee 1971: plate 11), perhaps the most obvious early pub-
lic example of funerary ‘Romanness’ that could have been seen in both neighbouring
Gaul and newly conquered areas of Britannia. Some form of journey or procession to
the cemetery and/​or cremation site beyond the accepted boundary or pomerium of early
Romano-​British urban centres was in any case a necessity (see Philpott 1991; Esmonde
Cleary 2001), and Romano-​British cemeteries flanking the routes out of towns is a gen-
eral pattern. These often became extra-​mural cemeteries, although the picture we get
of suburbs can actually be more complex, with particular cemetery plots going in and
out of use side by side with other activities like industry. The findings from numerous
excavations of individual cemetery plots within the East London cemetery are still the
most complete example we have of urban cemetery morphology, and even this is a par-
tial view. The small towns of Roman Britain are only partially understood in this regard
but present a similar pattern in microcosm (Esmonde Cleary 2001), as, for example, at
Baldock, Hertfordshire, Kelvedon in Essex, Ospringe, Kent (Whiting et al. 1931), and at
various vici in the upland, ‘military’ zone (like Brougham, Cumbria).
Most rural burial places again appear to conform to the notion of separateness, with
small cemetery plots being situated at peripheral locations or focused in places asso-
ciated with liminality, like burial monuments already ancient in the Iron Age (for
example, Mill Hill, in Kent (Parfitt 1995)). The Lexden Tumulus is perhaps particularly
noteworthy in its apparent emulation of such barrow mounds (see also Stevenson 2013;
Helm 2014); the building of new large earthen mounds continued to be a statement in
Roman Essex and elsewhere (Strück 2001), mirroring monuments in Belgic Gaul (see
Toynbee 1971: 179–1​88). On the other hand, rural Romano-​British cemetery plots could
often contain just a few less lavish burials of various types, perhaps associated with a
particular family or settlement. Large-​scale archaeological projects frequently encoun-
ter more than one such cemetery plot lying within ancient landscapes of hollow ways
and field boundaries, their related settlements either known or presumed to lie beyond
the excavation limits, as at Mucking, Essex (British Museum archive), or, more recently,
on the Isle of Thanet (Jon Rady, personal communication). In areas already dominated
by pre-​conquest inhumation, then, essentially prehistoric rites could continue to be
monumentalized with enclosures, mounds, and so on (Philpott 1991; see Strück 2001;
Booth, forthcoming), and in newly garrisoned areas, new cemeteries grew outside forts
and associated vici, as at Gloucester and Brougham. These new burial places would have
changed the local relativity of the living and the dead, impacting on cultural geography.
We have seen that in the late Iron Age there are hints that some initial journeys may
not have been to pyre or burial but to a place for more public laying-​out or even modi-
fication through excarnation. Our understanding of the location of cremation places
in relation to Romano-​British cemeteries is only slightly less limited, reliant on pyre
method and accidents of survival and discovery. In the case of a pyre built directly on
 Cemeteries and Funerary Practice    433

the ground surface (Flächenbustum (Strück 1993)), there is little hope that ephemeral
traces will remain, but evidence of pyres with pits beneath (Grubenbusta (Strück 1993))
is building. Single examples in the East London cemetery and Southwark, and multiple
instances at Pepper Hill, testify that at least some cremations were carried out within
or next to the cemetery, and large spreads of apparently redeposited pyre material at
East London also seem to point to this. Alternatively, cremation may have taken place
at a reusable or permanent facility (ustrinum (see Polfer 2001)), several of which are
suggested in continental settings by more or less structural features in association with
concentrations of pyre and other materials. The best-​known British example is that at
Trentholme Drive, York (Wenham 1968), but we have not yet had the opportunity to
excavate such a feature/​structure under modern conditions. Many ustrina may have left
only ephemeral traces, requiring special preservation conditions (see Lepetz et al. 2011),
or no trace at all. One interpretation of cremation burial deposits where the remains
of more than one individual are present is that they result from reuse of ustrina, but
alternative interpretations relating to depositional practice can be offered. There are also
unanswered questions as to whether certain under-​pyre pits in fact constitute a bur-
ial type in themselves, the bustum, where the pit beneath the pyre was considered the
grave (Strück 1993). This archaeological definition relies on a particular translation of a
Roman source (Festus, De Significatu Verborum).
Intra-​cemetery organization of burials is evident from the late Iron Age, and seems
to continue in Romano-​British cemeteries with various types of grouping, clustering,
and other foci, which can be diagnosed from burial type, coffin, or container type as
well as cemetery structures, shared alignments of burials, patterns of burials respect-
ing others, and so on. An alignment of three adjacent third-​century amphora burials at
Crundale, near Canterbury (site archive), is as good an example as any, or the multiple
inhumations in perpendicular relation to a boundary ditch (Feature 12) at Lankhills.
Other specialized burial foci were architecturally emphasized as mausolea, tombs, or
other structures, as at Southwark, East London, Poundbury, York, and Brougham. And
burial location often combines with further funerary differentiation; for example, the
Poundbury mausolea were clearly a focus for burials in lead-​lined coffins. Further intra-​
cemetery divisions include the exclusion from cemeteries or the peripheral location of
typically late Roman prone or decapitation burials, as seen at Poundbury, Lankhills, and
Butt Road. Although a rural context is common for such burials, they are still typically
found at the edge of even small cemeteries (Philpott 1991: 71 ff., 81 ff.), perhaps underlin-
ing a liminal status.
There are, of course, exceptions to the rule of spatial separation of the living and
the dead—​namely, the common instances of neonates and infants found in what we
might consider odd locations within both towns and rural settlements. Many seem to
be ‘foundation’ deposits, crossing over between the ritual and religious sphere, per-
haps because of their immaturity, lacking a name and only partially ‘belonging’. Burials
of the remains of both children and adults, sometimes modified, within wells, shafts,
pits, ditches, quarries, or caves also appear to have sacrificial connotations (Esmonde
Cleary 2001).
434

434   Jake Weekes

Further alternative spatial arrangements for the dead of note include burial practices
of certain rural elites, who often seem to have kept the dead closer, either within mau-
solea, barrows, or walled cemeteries in close proximity to, or even adjoining or within,
lived space (see Esmonde Cleary 2001: 130–​132; Millett 2006; Moore 2009). Then there
are the ‘backland’ inhumations, apparently permitted in the rear of occupied plots at late
Roman Ilchester and elsewhere (Esmonde Cleary 2001: 129, fig.14.2).

Modification

There is potential indirect evidence of excarnation in late Iron Age and Roman Britain
in the many ‘stray’ body parts of the archaeological record (Esmonde Cleary 2001), and,
in the Roman period, plaster, decapitation, embalming (if present (Philpott 1991: 92)),
and so on, all represent forms of modification, but without doubt the most significant
method is the developing tradition of cremation.
At a glance a general historical shift from early Romano-​British cremation to later
inhumation, at least in the south-​east of the province, is undeniable (Philpott 1991: 53 ff.).
This sea change can be seen as part of a wider continental shifts of emphasis at certain
times, apparently stemming from Rome itself (see Morris 1992). Yet regional and local
trajectories are more complex (see Philpott 1991: 54–​57; Booth, forthcoming). Much of
Britain retained already age-​old traditions of inhumation well into and throughout the
Roman period, and, even within the south-​eastern lowland zone where cremation dom-
inated in the early Roman period, we typically find both treatments coexisting within
the same cemeteries. There is also plenty of evidence for later cremation (late third and
fourth centuries) and first-​and second-​century inhumation (Philpott 1991: 50–5​2; see
Clarke 1979: 350–​351).
We must, therefore, realize the potential for many late Iron Age rites, including
the ‘invisible’ ones like possible excarnation, to continue into and even throughout
the Roman period and beyond, perhaps subtly changing through the introduction of
Romano-​British components. The inhumation burials of the Durotriges of south-
ern Dorset (see Poundbury) have long been recognized as a case study in this regard
(Whimster 1981: 37–​59; Philpott 1991: 6–​7, 54–​55), but others, such as crouched burial
in the Midlands and the north are discussed by Philpott (1991: 56–​57) and augmented
with fresh evidence by Paul Booth (forthcoming). He also picks up on growing evi-
dence of a late Iron Age Kent inhumation tradition, such as early burials at Pepper Hill,
Springhead, and elsewhere.
Any distinctly Roman ritual was, therefore, played out in an already complex
milieu. Herodian gives a rare account of a Roman funeral in the province: the crema-
tion of Septimius Severus at York in ad 211 (see Wenham 1968: 25; McKinley 2001: 43),
although cremated bones of the emperor were buried at Rome. Cn Julius Classicianus,
who, from his tombstone (RIB 12), seems to have been buried in Londinium in the first
century, was probably cremated too. However, the study of late Iron Age, Roman, and
 Cemeteries and Funerary Practice    435

Romano-​British cremation is primarily archaeological (see Philpott 1991: 48–​49) and

thanks largely to the work of Jacqueline McKinley, who has inverted previously estab-
lished views that cremation destroys evidence of ritual (Philpott 1991: 53–​54). Cremated
material is either recovered in situ from pyre sites or found redeposited in burials, layers,
or pre-​existing features such as ditches, pits, and so on (see McKinley 2001, forthcoming;
Weekes 2008). We can reconstruct much of the technology and materials used, includ-
ing something of maintenance methods and the types of wood used for fuel, sometimes
apparently selected for aromatic qualities (see Davis and de Moulins 2000; J. Watson
2004) or grass for kindling (see Kreuz 2001). We can also deduce pyre-​side rituals, not-
ing, in addition to materials that could easily have been present in the laying-​out context,
food and drink offerings such as charred fruits or animal bone or the charred containers
that held them. In certain late Iron Age funerals we know that pyre ritual allowed for
highly conspicuous consumption; witness the unidentified mass of objects in copper
alloy and silver destroyed by the expensive pyre at Folly Lane (cf. contemporary conti-
nental examples (Fitzpatrick 2001)). A similar picture derives from the list of materials
from an early Roman pyre at Holborough, Kent (Jessup 1955, 1959). Perhaps more typi-
cal late Iron Age and Romano-​British pyre sites have been excavated at Westhampnett,
East London, Southwark, and Pepper Hill. In the latter three cases, Celtic beans and
pulses, pine nuts, and grapes, respectively, were scattered as part of the ceremony, maybe
local variations on the same ritual theme. Conversely, in a third-​century funeral in the
northern military zone (Brougham: possibly therefore an intrusive rite), someone was
cremated alongside, or with (or on?) a horse, whose equally cremated bones were found
mixed with those of the human.
Methods of recovery of the cremated bone from the pyre may have varied, but analysis
of deposits, along with ethnographic and historical accounts, suggests that hand picking
of the material from cooled or cooling pyre material is most typical. Bone mineralized
by pyre temperatures in excess of 800 degrees Celsius is whitened, aiding identification.
The sorts of burnt materials often found mixed with the bone are typically animal bone
or remnants of bone artefacts (like Brougham veneers), but also clothing or personal
items of metal or glass that may have been in direct association with the corpse during
laying-​out and/​or placement on the pyres; as in many cases, this material adheres to
cremated bone. While plainly not bone, other items could still easily have been retrieved
via the same method, either accidentally or deliberately. Alternatively, a burial can con-
sist of unsorted pyre material redeposited en masse (as at Folly Lane), sometimes leav-
ing telltale scorching on the interior of the burial pit or other feature into which it was
apparently tipped while still hot (McKinley 1989: 73).
As well as modification of the body and pyre offerings, there is evidence that other
objects sometimes required modifications as part of the funerary process. The decap-
itated pipe-​clay Venus figurine from a cremation burial at St Dunstan’s Terrace,
Canterbury, is thus interpreted. It has also been noted that many of the ceramic vessels
supplied as grave goods had been, at some stage, rendered unusable for the living at least
(see Biddulph 2002, 2005). Deliberate breakage of objects is a typical, cross-​cultural way
of consigning objects to an afterlife of strictly symbolic functionality.
436

436   Jake Weekes

Deposition

With some sites we can even try to reconstruct depositional aspects of the laying stage
(as at Folly Lane, with offerings to the deceased in the subterranean chamber, the subse-
quent destruction and backfilling of the chamber, and the inclusion within its backfill of
fragments of broken pottery, probably derived from funerary feasting; cf. the ‘amphorae
pavement’ at Clemency (Metzler et al. 1991; see Fitzpatrick 2001: 18–​20)). Deposition in
cremation funerals is evident from those materials left at the pyre site. Pyre pits some-
times contain intact, un-​burnt, or only scorched objects that appear to represent some
sort of ‘closure’ of the feature, as at Southwark, where a number of ceramic tazze (incense
burners) and lamps had been placed in a spent pyre (cf. the numerous lamps, deliber-
ately smashed, and token cremated bone deposit in ‘burial’ 50 at Baldock, or un-​burnt
contents of ‘Grave 911’ in Cemetery 2 at Mucking (Weekes 2008: 152)). Various types of
redeposited pyre material that are often found in other features (McKinley 1997, 2001,
forthcoming) could also be significant mortuary deposits, so words such as ‘dump’ or
‘debris’ are inadequate to describe them (Weekes 2008).
Late Iron Age funerals also used burials themselves as contexts for display, however,
and could contain quite stark combinations of objects that seem to allude to both ‘native’
roots and current connections with Gaul and Rome: the cross-​cultural ‘DNA’ of creole
Romano-​British depositional practice.
The most obvious forerunner to early Romano-​British cemeteries in the south-​
east is that at King Harry Lane, St Albans. Here cremation burials were of a type well
known in north-​west Gaul and Italy and actually seem to set the tone for more wide-
spread traditions in the lowland zone (cf. Baldock, Colchester, London, Canterbury,
Chichester, and so on). This type of cremation burial is typically centred on a discrete
deposit of cremated bone fragments, either loose or in a bag or more durable container,
either in isolation or accompanied by accessory vessels and other items, in a pit (these
could be of various shapes, sizes, and designs). Further burial types, more typical of the
Rhine frontier, are also beginning to be recognized in Britannia (Pearce 1999). These
Brandschuttgräber contain unsorted pyre material either with or without a discrete
deposit of sorted cremated bone. Those with sorted cremated bone (perhaps in a sepa-
rate container) are called Brandschüttungsgräber; those with only mixed bone and pyre
material, Brandgrubengräber. In truth, all such classifications are just a starting point for
comparison.
Loose or bagged cremation deposits would seem to be a common Iron Age trait. A
marked concentration of the material suggests original deposition in bags, long since
decayed. Sometimes, as with a burial from Mill Hill, Deal, Kent, a brooch lying on top
may originally have fastened the bag (Parfitt 1995). Un-​urned (see Philpott 1991: 45–4​7)
and urned cremation burials are often apparently contemporaneous in Roman period
cemeteries, but interesting local patterns also emerge, as, for example, at Pepper Hill,
Kent; here an early local tradition of un-​urned burial seems to have been superseded
 Cemeteries and Funerary Practice    437

by the much broader practice of placing the remains in a coarse ware jar. The jar tradi-
tion, while widespread, was not ubiquitous either, as various beakers, bowls, and flag-
ons could be used (cf. Baldock, Colchester, London, Canterbury, Chichester, and so on).
Further afield, in burials of elites and those of suspected soldiers and colonists, there are
instances of specially made lead ossuaria, a recently discovered example at Colchester
being linked with the grave of M. Favonius Facilis, a first-​century centurion of the XX
Legion whose tombstone had long been known (RIB 200). Urns of porphyry, presum-
ably housing remains of well-​off newcomers or local elites, are known from early cem-
eteries at London, while, perhaps primarily among rural elites, there is a definite pattern
of placing the cremated bone in glass jars and bottles (Philpott 1991: 26 ff).
In upland Britain, or where raw materials were more plentiful in the lowland zone, we
might see Roman period examples of loose or bagged cremation deposits placed in stone
cists or hollowed-​out blocks, often an adaptation of long-​standing traditions: inhuma-
tion in various forms of cists and stone-​lined graves (see Philpott 1991: 9–​10; 56–​57; prac-
tices in Scotland have been summarized by Collard and Hunter 2000). The stone cist is
a notable feature of the second-​and third-​century cremation cemetery at Brougham,
Cumbria, for example (cf. Wilmott 2010). Cists (especially of tile) are also found in the
south-​east, housing urned or un-​urned cremated bone along with accessory vessels and
other accessories (Philpott 1991: 9), but here burial assemblages were often placed in
boxes or caskets (for example, Baldock and Chichester) or large amphorae (for example,
Canterbury and East London). Other forms of grave modification—​for example, gabled
cists above inhumations formed of large tiles at York and Chester—​have Roman and
Gallic correlates and could be more specifically linked with or copied from incomers
(e.g. Toynbee 1971: plate 24; cf. examples in Les Dossiers D’Archéologie 330, Rome et Ses
Morts; Catalano et al, November/​December 2008).
Various forms of box-​shaped cists and stone-​lined inhumation graves were clearly
designed around predispositions for burial posture. Iron Age and early Roman inhu-
mations generally are found in a spectrum of crouched or supine postures and grave
orientations, but Romano-​British inhumations became increasingly supine and aligned
west–​east (Philpott 1991: 1), at Baldock, Butt Road, Poundbury, and Lankhills, for exam-
ple. The southern Durotriges, with their distinct late Iron Age inhumation tradition,
seem to have adopted a supine posture in their creolized Romano-​British practice
(see Philpott 1991: 54–​55). However, broadly contemporaneous crouched and supine
Romano-​British inhumations have recently been noted at London Road, Gloucester.
Beyond more extreme variations in posture (that is, prone burials), there are also a range
of carefully positioned or apparently casually discarded bodies and cremation depos-
its/​containers in alternative locations, such as wells or pits (see Esmonde Cleary 2001),
to consider. These burials occasionally form part of a sequence of structured deposits,
and in the case of culturally Iron Age but chronologically conquest period Lindow Man
(Stead et al. 1986)—​as with many infant burials—​the body may have been a specialized
installation in itself (see also A. Moore and Gowland, this volume). More commonly,
we consider the various items placed in the burial pit (the grave goods) as accessories to
the installation of the human remains. Some generalizations can be made. For instance,
438

438   Jake Weekes

ceramic vessels associated with food and drink consumption and storage (platters,
dishes, beakers, flagons, jars, and amphorae) are clearly an important category from the
late Iron Age to the late Roman period. Attempts have been made to derive a typical
furnishing associated with a ‘symbolic meal’, the most influential being that of Philpott
(1991: 35; see also Biddulph 2005, driven by correspondence analysis), but, while there
would certainly be a general propensity to include these sorts of items, there is consider-
able variability in the numbers and types of vessels from burial to burial and no typical
suite. Many other types of object are found in late Iron Age and Romano-​British burials,
and some seem to represent degrees of personalization (see Weekes, forthcoming (a)),
from occasional board games (for example, Stanway) to unique items in otherwise quite
‘typical’ burials that resist statistical analysis, like a wooden sheathed miniature sword at
Cramner House, St Dunstans, Canterbury.
It is interesting to note that contemporary inhumation and cremation burials often
contain the same sorts of grave goods. Although the conceptual framework for position-
ing of objects in relation to the body is lost during cremation, there is plenty of evidence
of more abstract positioning, such as shoes, often still on the feet of the inhumed, being
placed in a cremation burial either side of a bowl containing the cremation deposit at St
Dunstan’s, Canterbury (cf. Black 1987: 216).
Later Roman inhumation burial would appear to be characterized by a depletion or
often complete absence of grave goods (at least archaeologically visible ones), but such
aspects cannot be relied upon for dating of burial contexts. Some ‘unfurnished’ burials
might turn out on radiocarbon dating to be either late Iron Age or sub-​Roman.
Food items placed in burials provide another developing area of study (see White
2007). For example, in the case of un-​burnt (but cooked?) chickens found in London
burials, where the carcass is typically found in the top of a cremation container, a ques-
tion is raised as to whether these offerings were made at burial or afterwards. We can
get glimpses of the final gestures at the graveside, such as a ‘valedictory toast’, suggested
by broken pieces of the same vessel within the coffin and in the grave backfill (cited in
Booth, forthcoming).

Commemoration

Mausolea, tombs, tiled vaults, and so on could facilitate revisiting of the dead, and many
more humble graves had various forms of steps, plinths, boarding, shuttering, and vault-
ing (Philpott 1991: 69–​70), possibly aiding continued access. Certain cremation burials
strongly suggest that contents could have been revisited or even manipulated, removed,
and placed/​replaced following initial burial, such as amphora burial at Warwick Square,
London.
Newcomers to Britain in the conquest period could have brought an inclination for
continuing graveside feasts like the Cena novendialis, held on the ninth day after the
funeral (see Toynbee 1971: 51), and cremation and inhumation burials with contrivances
 Cemeteries and Funerary Practice    439

allowing for continued access from the surface via lead pipes, flue tiles, and so on would
have allowed the dead to ‘participate’. Examples, perhaps the burial of military person-
nel, are well known from Colchester and Caerleon (see Toynbee 1971: plates 14, 22, and
23). Such devices are typical of many burials in the Italian homeland (Toynbee 1971; for
numerous examples, see Les Dossiers D’Archéologie 330, Rome et Ses Morts (Catalano
et al, November/​December 2008), but especially Buccellato et al. 2008: 18; cf. Lepetz
et al. 2011).
The suggestion of widespread revisiting also gives a new twist to the interpretation
of the many lids, dishes, and other large flat objects such as tiles and quern stones
used as lids on containers of cremated bones and on amphorae and cists. As well
as acting as grave markers, such items could have covered the contents at ground
level, many of which would have been long since removed as a result of cessation
of maintenance, or post-​deposition processes like ploughing (but see Weekes 2007).
Remnants of funerary feasting are seen in negative features, like the pit filled with
numerous flagons and specialized animal burials in the East London cemetery or the
Aschengruben, cremation burial-​like features found to contain animal rather than
human bone (Wigg 1993). However, in Britain there has recently been an increasing
archaeological recognition that ephemeral traces of cemetery surfaces might remain
to be discovered given the right conditions (see Ortalli, forthcoming). These are the
horizons at which access to burials was maintained and mounds and other structures
built and on to which offerings to the dead could be placed (see Bennett 1987: 68;
Simmonds et al. 2008: 9).
While elite burials in pre-​Roman and Roman Britain would have been clearly signi-
fied in the landscape through the provision of enclosures, barrows, monuments, mau-
solea, tombs, and vaults, grave markers could be an appropriate and affordable level of
commemoration for the less distinguished in death (see Pearce 2011). Some dead rela-
tives or friends may have been located in the cemetery via the protruding neck of an
amphora, or a prominent—​if quickly overgrown—​tile or wooden cover. Those provided
with stone markers have bequeathed the source material for study of a Roman intrusive
rite par excellence: the mortuary epigraphic habit. We have not only notable Roman
administrative and military examples such as those already cited—​but also the eques-
trian tombstones of imported auxiliaries, along with stones set up to the native inhabit-
ants. The twenty-​nine inscribed tombstones at Brougham, for example, demonstrate an
interesting mix of a few foreigners among what seem to be ethnically British vicani. The
inscriptions use Latin and adopted formulae, such as the D-​(is) M-​(anibius): ‘To the
Spirits of the Departed’. However, as with all creole culture, and as Fitzpatrick rightly
notes, ‘the situation is rather more complex and subtle’ (Fitzpatrick 2004:  408). The
same could be said of relief sculptures on certain stones, from the equestrian and mar-
tial themes of the auxiliaries to variously provincial-​looking symbolic people, objects,
and even non-​representational details of more ‘native’ work. Like any products of the
mortuary sphere, these formed a language that people could use in multivalent ways,
and they require a considered, hermeneutical approach if we are to disentangle their
potentially diverse meanings and aesthetics.
440

440   Jake Weekes

Interpretation and Translation

The late twentieth and early twenty-​first centuries have gradually seen the demise of
the overly classical approach to the Roman provinces and, it is hoped, the associated
‘Romanization’ paradigm. Recent theorists have, to various extents, applied more rela-
tivist and postcolonial attitudes, recognizing, as anthropologists, a two-​way process of
cultural change on the part of both the colonizers and the colonized (e.g. Webster 2001).
Key to developing a post-​processual debate in mortuary studies has been the recogni-
tion that the funerary context is not a straightforward index of either status or afterlife
beliefs (see Ucko 1969; cf. Pearce 2001, 2013, this volume). Certainly we would now rec-
ognize that a contextual archaeology is required when it comes to noting instances of
possible ‘Roman’ iconography and relating them to other evidence of afterlife beliefs (cf.
Alcock 1980; Black 1987; and components of Philpott’s ‘Romanized’ funerary practices
(1991)). Funerals and funerary traditions are texts affording diverse approaches to their
many connotations. We begin to see the Romano-​British funeral as a social context for
display and the use of metaphoric and metanymic symbols and objects in structurating
(cf. Giddens 1984) a creole culture. Far more promising, therefore, is Pearce’s exploration
(1999, 2008) of burial furnishing as a medium for provincial constructs of Romanitas
and, in particular, Humanitas.
Some recent forays into the overall meanings of Roman and provincial funerals have
been inspired by earlier concepts of ‘reversal’ (Scheid 1984; cf. Lepetz et al. 2011) and
liminality (Turner 1969; cf. Weekes, forthcoming (b)). Weekes (forthcoming (b)) in fact
experiments with a combination of both Scheidian and Turneresque approaches, the
sociological outcome of which can actually be unproblematically summed up by an ear-
lier assertion by Merrifield:

Burial ceremonies were essentially rites of transition, in which the mourners were
often as much concerned to sever the connection between the deceased and …
earthly life, as to ensure … acceptance in the life to come. (Merrifield 1987: 71)

It must also be recognized, however, that symbolism of the funerary context is multi-​
vocal (see Turner 1977 [1967]), a polysemic milieu incorporating various complex signs
that could include creole conceptions of Humanitas; for example, both alongside and
articulated through personal and individual commemorations or alongside the diverse
nuances created for various participants in the funeral (cf. Williams 2004).
Certainly no two funerals will ever be wholly the same, as a brief reflection on our
own experience will confirm. This is because funerals paradoxically combine a sense
of tradition with opportunities for specialization and particular expressions. From
a broadly humanist perspective (deterministic models are available elsewhere:  see
Biddulph 2012), I have suggested that the medium through which funerary traditions
develop and change is best described as translation (Weekes, forthcoming (a)). In a
structure/​agent dialectic model, the perceived existing funerary tradition represents a
 Cemeteries and Funerary Practice    441

structure that must be translated by agents into each particular funerary event. Each
funeral in turn is perceived within the context of tradition and translated by prospective
agents into future funerals. In this way a single occurrence of a funerary feature, such as
the placement of an unguent bottle atop the cremated remains in the grave, can go on to
inform a more widespread tradition, while currently traditional facets may equally be
lost in translation, and disappear from tradition.
Funerary traditions can be seen as folk art and drama, vital social contexts that
allow the working and playing-​out of cultural, regional, local, and personal identities
and expressions, through oral traditions, theatre, symbolic modification of the body,
and other objects (sculpture) and installations. These events are clearly of considera-
ble potential for understanding the development of a distinct Romano-​British culture,
if only we can reconstruct and interpret them. With the recognition of the diachronic
funerary process and its synchronic possibilities of meaning, this study area is no longer
in its infancy.

Abbreviation
RIB  R
 . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Oxford:
Clarendon Press, 1965–​to date.

References
Primary Sources
Cicero, De Res Publica and De Legibus, trans. C. W. Keys. Loeb Edition. London and
Cambridge, MA: Harvard University Press, 1928.
Festus, De Verborum Significatu quae supersunt cum Pauli Epitome, trans. W. M. Lindsay.
Hildesheim: Georg Olms Verlagsbuchhandlung, 1965.

Secondary Sources
Adam, A. (1835). Roman Antiquities: Or, An Account of the Manners and Customs of the Romans,
ed. Revd J. R. Mayor. 12th edn. London: T. Cadell.
Alcock, J. (1980). ‘Classical Religious Belief and Burial Practice in Roman Britain’,
Archaeological Journal, 137: 50–​85.
Allen, T., Donnelly, M., Hardy, A., Hayden, C., and Powell, K. (2012). A Road through the
Past:  Archaeological Discoveries on the A2 Pepperhill to Cobham Road-​Scheme in Kent.
Oxford Archaeology Monograph No 16. Oxford: Oxford Archaeology.
Barber, B., and Bowsher, D. (2000). The Eastern Cemetery of Roman London: Excavations 1983–​
1990. MoLAS Monograph 4. London: Museum of London Archaeology Service.
Bennett, P. (1987). ‘Cranmer House, London Road’, in S. S. Frere, P. Bennett, J. Rady, and S.
Stow (eds), Canterbury Excavations:  Intra-​and Extra-​Mural Sites, 1949–​55 and 1980-​84.
Maidstone: Kent Archaeological Society, 56–​73.
442

442   Jake Weekes

Bennett, P., Frere, S. S., and Stow, S. (1982). Excavations at Canterbury Castle.
Canterbury: Canterbury Archaeological Trust.
Biddulph, E. (2002). ‘One for the Road? Providing Food and Drink for the Final Journey’,
Archaeologia Cantiana, 122: 101–​111.
Biddulph, E. (2005). ‘Last Orders:  Choosing Pottery for Funerals in Roman Essex’, Oxford
Journal of Archaeology, 24/​1: 23–​45.
Biddulph, E. (2012). ‘On Cultural Selection:  Examining the Process of Cultural Evolution
through Funerary Evidence’, in M. Duggan, F. McIntosh, and D. Rohl (eds), TRAC
2011: Proceedings of the Twenty-​First Theoretical Roman Archaeology Conference, Newcastle
2011. Oxford: Oxbow Books, 76–​90.
Black, E. W. (1987). ‘Romano-​British Burial Customs and Religious Beliefs in South-​East
England’, Archaeological Journal, 143: 201–​239.
Booth, P. (forthcoming). ‘Some Recent Work on Romano-​British Cemeteries’, in J. Pearce and
J. Weekes (eds), Death as a Process: Funerals in the Roman World.
Booth, P., Simmonds, A., Boyle, A., Clough, S., Cool, H. E. M., and Poore, D. (2010). The Late
Roman Cemetery at Lankhills, Winchester: Excavations 2000–​2005. Oxford Archaeology
Monograph 10. Oxford: Oxford Archaeology.
Booth, P., Champion, T., Foreman, S., Garwood, P., Glass, H., Munby, J., and Reynolds, A.
(2011). On Track:  The Archaeology of the High Speed 1 Section 1 in Kent. Oxford:  Oxford
Archaeology.
Buccellato, A., Catalano, P., and Pantano, W. (2008). ‘La Site et la necropole de Castellaccio’,
in P. Catalano, J. Scheid, and S. Verger (Coordination Scientifique), Rome et ses Morts:
L’Archéologie Funéraire dans L’Occident Romain. Les Dossiers d’Archéologie 330. Dijon:
Editions Faton, 14–​19.
Byers, S. (2002). Introduction to Forensic Anthropology: A Textbook. London: Allyn and Bacon.
Catalano, P., Scheid, J., and Verger, S. (Coordination Scientifique) [and various authors].
(2008). Rome et ses Morts: L’Archéologie Funéraire dans L’Occident Romain. Les Dossiers
d’Archéologie 330. Dijon: Editions Faton.
Clarke, G. (1979). The Roman Cemetery at Lankhills. Winchester Studies 3. Oxford: Clarendon
Press.
Collard, M., and Hunter, F. (2000). ‘A Roman Burial from Cramond (Edinburgh) Rediscovered’,
Proceedings of the Society of Antiquaries of Scotland, 130: 525–​535.
Cool, H. E. M. (2004). The Roman Cemetery at Brougham, Cumbria. Excavations 1966–​67.
Britannia Monograph Series 21. London: Society for the Promotion of Roman Studies.
Cool, H. (2010). ‘Finding the Foreigners’, in H. Eckhardt (ed.), Roman Diasporas: Archaeological
Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology
Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 27–​44.
Crummy, P. (1997). City of Victory. Colchester: Colchester Archaeological Trust.
Crummy, N., Crummy, P., and Crossan, C. (1993). Excavations of Roman and Later
Cemeteries and Monastic Sites in Colchester, 1971–​88. Colchester Archaeological Report
9. Colchester: Colchester Archaeological Trust.
Crummy, P., Benfeld, S., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An Élite
Burial Site at Camulodunum. Britannia Monograph Series 24. London: Society for the
Promotion of Roman Studies.
Davis, D., and de Moulins, D. (2000). ‘The Plant Remains’, in B. Barber and D. Bowsher,
The Eastern Cemetery of Roman London:  Excavations 1983–​1990. MoLAS Monograph
4. London: Museum of London Archaeology Service, 368–​379.
 Cemeteries and Funerary Practice    443

Diack, M., and Weekes, J. (forthcoming). ‘A Romano-​British Cemetery at St Dunstan’s Terrace,

Canterbury’. Canterbury Archaeological Trust Occasional Paper.
Down, A., and Rule, M. (1971). Chichester Excavations 1. Chichester: Chichester Civic Society
Excavations Committee.
Duday, H. (2006). ‘L’Archéothanatologie ou l’archéologie de la mort’ [Archaeothanatology
or the Archaeology of Death], in R. Gowland and C. Knüsel (eds), Social Archaeology of
Funerary Remains. Oxford: Oxbow Books, 30–​56.
Duday, H. (2009). The Archaeology of the Dead: Lectures in Archaeothanetology, trans. Anna
Maria Cipriani and John Pearce. Oxford: Oxbow Books.
Dupras, T. L., Schultz, J. J., Wheeler, S. M., and Williams, L. J. (2012). Forensic Recovery of
Human Remains: Archaeological Approaches. London: Taylor and Francis Group.
Esmonde Cleary, S. (2001). ‘Putting the Dead in their Place: Burial Location in Roman Britain’,
in J. Pearce, M. Millet, and M. Strück (eds), Burial Society and Context in the Roman World.
Oxford: Oxbow Books, 127–​142.
Farwell, D. E., and Molleson, T. L. (1993). Excavations at Poundbury 1966–​80. Volume II: The
Cemeteries. Dorset Natural History and Archaeological Society Monograph Series 11.
Dorset: Dorset Natural History and Archaeological Society.
Fitzpatrick, A. P. (1997). Archaeological Excavations on the Route of the A27 Westhampnett
Bypass, West Sussex. Volume 2: The Late Iron Age, Romano-​ British and Anglo-​
Saxon
Cemeteries. Wessex Archaeology Report 12. Salisbury: Wessex Archaeology.
Fitzpatrick, A. P. (2001). ‘Ritual, Sequence and Structure in Late Iron Age Mortuary Practices in
North-​West Europe’, in J. Pearce, M. Millet, and M. Strück (eds), Burial Society and Context
in the Roman World. Oxford: Oxbow Books, 15–​29.
Fitzpatrick, A. P. (2004). ‘The Tombstones’, in H. E. M. Cool, The Roman Cemetery at
Brougham, Cumbria. Excavations 1966–​67. Britannia Monograph Series 21. London: Society
for the Promotion of Roman Studies.
Foster, J. (1986). The Lexden Tumulus: A Reappraisal of an Iron Age Burial from Colchester,
Essex. BAR British Series 156. Oxford: Archaeopress.
Giddens, A. (1984). The Construction of Society:  Outline of the Theory of Structuration.
Cambridge: Polity Press.
Gowland, R., and Knüsel, C. (2006) (eds). Social Archaeology of Funerary Remains.
Oxford: Oxbow Books.
Helm, R. (2014). ‘Outside the Town: Roman Industry, Burial and Religion at St, Augustine
House, Canterbury’. Canterbury Archaeological Trust Occasional Paper 10. Canterbury:
Canterbury Archaeological Trust.
Iserson, K. V. (2002). ‘Rigor mortis and Other Post-​ Mortem Changes’, in Macmillan
Encyclopedia of Death and Dying http://​www.encyclopedia.com/​doc/​1G2-​3407200244.html
(accessed 10 September 2012).
Jackson, R. W., and Jackson, J. M. (2004). Forensic Science. London: Pearson Prentice Hall.
Jessup, R. F. (1955). ‘Excavations of a Roman Barrow at Holborough, Snodland’, Archaeologia
Cantiana, 68: 1–​68.
Jessup, R. F. (1959). ‘Barrows and Walled Cemeteries in Roman Britain’, Journal of the British
Archaeological Association, 22: 1–​32.
Jones. M. J. (2002). Roman Lincoln: Conquest, Colony and Capital. Stroud: Tempus.
Kreuz, A. (2001). ‘Functional and Conceptual Archaeobotanical Data from Roman
Cremations’, in J. Pearce, M. Millett, and M. Strück (eds), Burial Society and Context in the
Roman World. Oxford: Oxbow Books, 45–​51.
444

444   Jake Weekes

Lepetz, S., Van Andringa, W, Duday, H., Joly, D., Malagoli, C., Matterne, V., and Tuffreau-​Libre,
M. (2011). ‘Publius Vesonius Phileros vivos monumentum fecit: Investigations in a Sector of
the Porta Nocera Cemetery in Roman Pompeii’, in M. Carroll and J. Rempel (eds), Living
Through the Dead: Burial and Commemoration in the Classical World. Oxford: Oxbow
Books, 110–​133.
Mackinder, A. (2000). A Romano-​British Cemetery on Watling Street: Excavations at 165 Great
Dover Street, Southwark, London. MoLAS Archaeology Studies 4.  London:  Museum of
London Archaeology Service.
McKinley, J. I. (1989). ‘Cremations: Expectations, Methodologies and Realities’, in C. A.
Roberts, F. Lee, and J. Bintliff (eds), Burial Archaeology: Current Research, Methods and
Developments. BAR British Series 211. Oxford: Archaeopress, 65–​76.
McKinley, J. I. (1994a). The Anglo-​Saxon Cemetery at Spong Hill, North Elmham. Part VIII: The
Cremations. East Anglian Archaeology 69. Gressenhall: Norfolk Archaeological Unit.
McKinley, J. I. (1994b). ‘A Pyre and Grave Goods in British Cremation Burials: Have we Missed
Something?’, Antiquity, 68: 132–​138.
McKinley, J. I. (1994c). ‘Bone Fragment Size in British Cremation Burials and its Implications
for Pure Technology and Ritual’, Journal of Archaeological Science, 21: 339–​342.
McKinley, J. I. (1997). ‘The Cremated Human Bone from Burials and Cremation Related
Contexts’, in A. P. Fitzpatrick, Archaeological Excavations on the Route of the A27
Westhampnett Bypass, West Sussex. Volume 2: The Late Iron Age, Romano-​British and Anglo-​
Saxon Cemeteries. Wessex Archaeology Report 12. Salisbury; Wessex Archaeology, 55–​72.
McKinley, J. I. (2001). ‘Phoenix Rising: Aspects of Cremation in Roman Britain’, in J. Pearce, M.
Millett, and M. Strück (eds), Burial Society and Context in the Roman World. Oxford: Oxbow
Books, 38–​44.
McKinley, J. I. (2004). ‘The Human Remains and Aspects of Pyre Technology and Cremation
Rituals’, in H. E. M. Cool, The Roman Cemetery at Brougham, Cumbria. Excavations 1966–​67.
Britannia Monograph Series 21. London: Society for the Promotion of Roman Studies, 283–​309.
McKinley, J. I. (forthcoming). ‘“How did it go?” … Putting the Process back into Cremation’, in
J. Pearce and J. Weekes (eds), Death as a Process: Funerals in the Roman World.
McWhirr, A., Viner, L., and Wells, C. (1982). Cirencester Excavations II. Romano-​British
Cemeteries at Cirencester. Cirencester: Cirencester Excavation Committee.
Mason, D. J. P. (2001). Roman Chester: City of the Eagles. Stroud: Tempus.
Meates, G. W. (1979). The Lullingstone Roman Villa. Volume I: The Site. Monograph Series of the
Kent Archaeological Society 1. Maidstone: Kent Archaeological Society.
Metcalf, P., and Huntington, R. (1992). Celebrations of Death: The Anthropology of Mortuary
Ritual. Cambridge: Cambridge University Press.
Merrifield, R. (1987). The Archaeology of Ritual and Magic. London: Batsford.
Metzler, J., Waringo, R., Bis, R., and Metzler-​Zens, N. (1991). Clemency et les tombes de
l’aristocratie dans Gaule Belgique. Dossiers d’ archéologie du Musee national d’ histoire de l’
art 1. Luxembourg.
Middleton, J. (1982). ‘Lugbara Death’, in M. Bloch and J. Parry (eds), Death and the Regeneration
of Life. Cambridge: Cambridge University Press, 134–​154.
Millett, M. (2006) (ed.). Shiptonthorpe, East Yorkshire: Archaeological Studies of a Romano-​
British Roadside Settlement. Yorkshire Archaeology Report 5. Leeds: Yorkshire
Archaeological Society/​East Riding Archaeological Society.
Montgomery, J., Evans, J. A., Chenery, S. R., Pashley, V., and Killgrove, K. (2010). ‘Gleaming,
White and Deadly: Using Lead to Track Human Exposure and Geographic Origins
 Cemeteries and Funerary Practice    445

in the Roman Period in Britain’, in H. Eckardt (ed.), Roman Diasporas: Archaeological

Approaches to Mobility and Diversity in the Roman Empire. Journal of Roman Archaeology
Supplementary Series 78. Portsmouth, RI: Journal of Roman Archaeology, 199–​226.
Moore, A. (2009). ‘Hearth And Home: The Burial of Infants within Romano-​British Contexts’,
Childhood in the Past, 2: 33–​54.
Morris, I. (1992). Death Ritual and Social Structure in Classical Antiquity.
Cambridge: Cambridge University Press.
Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium. Britannia
Monograph Series 14. London: Society for the Promotion of Roman Studies.
Niblett, R. (2001). ‘Funerary Rites in Verulamium during the Early Roman Period’, in J.
Pearce, M. Millett, and M. Strück (eds), Burial Society and Context in the Roman World.
Oxford: Oxbow Books, 97–​104.
Ortalli, J. (forthcoming). ‘Archaeology and Funerary Cult:  The Stratigraphy of Soils in the
Cemeteries of Emilia Romagna (Northern Italy)’, in J. Pearce and J. Weekes (eds), Death as a
Process: Funerals in the Roman World.
Parfitt, K. (1995). Iron Age Burials from Mill Hill, Deal. London: British Museum Press.
Parker Pearson, M. (1999). The Archaeology of Death and Burial. Stroud: Sutton Publishing.
Pearce, J. (1997). ‘Death and Time: The Structure of Late Iron Age Mortuary Ritual’, in A. Gwilt
and C. Haselgrove (eds), Reconstructing Iron Age Societies: New Approaches to the British
Iron Age. Oxbow Monograph 71. Oxford: Oxbow Books, 174–​180.
Pearce, J. (1998). ‘From Death to Deposition: The Sequence of Ritual in Cremation Burials of
the Roman Period’, in C. Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 98: Proceedings
of the Seventh Annual Theoretical Roman Archaeology Conference Nottingham 1997.
Oxford: Oxbow Books, 99–​111.
Pearce, J. (1999). ‘Case Studies in a Contextual Archaeology of Burial Practice in Roman
Britain’. Unpublished Ph.D. thesis, University of Durham.
Pearce, J. (2001). ‘Burial, Society and Context in the Provincial Roman World’, in J. Pearce, M.
Millett, and M. Strück (eds), Burial, Society and Context in the Roman World. Oxford: Oxbow
Books, 1–​12.
Pearce, J. (2002). ‘Ritual and Interpretation in Provincial Roman Cemeteries’, Britannia,
33: 373–​377.
Pearce, J. (2008). ‘Burial Evidence from Roman Britain: The Un-​Numbered Dead’, in Pour
une Archéologie du Rite. Nouvelles Perspectives de L’Archéologie Funéraire, Études Réunies
par John Scheid. Collection de l’École Française de Rome 40. Rome: École Française de
Rome, 29–​42.
Pearce, J. (2010). ‘Burial, Identity and Migration in the Roman World’, in H. Eckardt (ed.),
Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire.
Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI: Journal of Roman
Archaeology, 79–​98.
Pearce, J. (2011). ‘Marking the Dead: Tombs and Topography in the Roman Provinces’, in M.
Carroll and J. Rempel (eds), Living through the Dead: Burial and Commemoration in the
Classical World. Oxford: Oxbow Books, 134–​158.
Pearce, J. (2013). Contextual Archaeology of Burial Practice: Case Studies from Roman Britain.
BAR British Series S588. Oxford: Archaeopress.
Pearce, J. (2015). ‘A Civilised Death? The Interpretation of Provincial Roman Grave Good
Assemblages’, in J. R. Brandt, H. Roland, and M. Prusac (eds), Death and Changing Rituals:
Function and Meaning in Ancient Funerary Practices. Oxford: Oxbow Books, 223–48.
446

446   Jake Weekes

Philpott, R. (1991). Burial Practices in Roman Britain: A Study of Grave Treatment and
Furnishing, ad 43–​410. BAR British Series 219. Oxford: Tempus Reparatum.
Polfer, M. (2001). ‘Reconstructing Funerary Rituals: The Evidence of Ustrina and Related
Archaeological Structures’, in J. Pearce, M. Millett, and M. Strück (eds), Burial, Society and
Context in the Roman World. Oxford: Oxbow Books, 30–​37.
RCHM (1962). Royal Commission on Historical Monuments in England’, An Inventory of the
Historical Monuments in the City of York. Volume 1: Eboracum. Roman York. London: RCHM.
Redfern, R. C. (2003). ‘Sex and the City: A Biocultural Investigation into Female Health in
Roman Britain’, in G. Carr, E. Swift, and J. Weekes (eds), TRAC 2002. Proceedings of the
Twelfth Annual Theoretical Roman Archaeology Conference, University of Kent at Canterbury,
5–​6 April 2002. Oxford: Oxbow Books, 147–​169.
Redfern, R. C. (2011). ‘A Re-​appraisal of the Evidence for Violence in the Late Iron Age Human
Remains from Maiden Castle Hillfort, Dorset, England’, Proceedings of the Prehistoric
Society, 77: 111–​138.
Rodwell, K. A. (1988). The Prehistoric and Roman Settlement at Kelvedon, Essex. CBA Research
Report 63/​ Chelmsford Archaeological Trust Report 6. London: Council for British
Archaeology.
Scheid, J. (1984). ‘Contraria facere: Renversements et déplacements dans les rites funéraires’,
Annali dell’ Istituto Orientale di Napoli, 6: 117–​139.
Simmonds, A., Márquez-​Grant, N., and Loe, L. (2008). Life and Death in a Roman City:
Excavation of a Roman Cemetery with a Mass Grave at 120–​122 London Road, Gloucester.
Oxford Archaeology Monograph 6. Oxford: Oxford Archaeology.
Sparey Green, C. (1977). ‘The Significance of Plaster Burials for the Recognition of Christian
Cemeteries’, in R. Reece (ed.), Burial in the Roman World. CBA Research Report 22.
London: Council for British Archaeology, 46–​53.
Stead, I. M. (1967). ‘A La Tène III Burial at Welwyn Garden City’, Archaeologia, 101: 1–​62.
Stead, I. M., and Rigby, V. (1986). Baldock: The Excavation of a Roman and Pre-​Roman
Settlement, 1968–​72. Britannia Monograph Series 7. London: Society for the Promotion of
Roman Studies.
Stead, I. M., and Rigby, V. (1989). Verulamium: The King Harry Lane Site. English Heritage
Archaeological Report 12. London: English Heritage.
Stead, I. M., Bourke, J., and Brothwell, D. (1986). Lindow Man: The Body in the Bog. Ithaca,
NY: Cornell University Press.
Stevenson, J. (2013). Living by the Sword: The Archaeology of Brisley Farm, Ashford, Kent. Spoilheap
Monograph Series No 6. Portslade, Guildford, and London: Spoilheap Publications.
Strück, M. (1993). ‘Busta in Britannien und ihre Verbindungen zum Kontinent: Allgemeine
Uberlegungen zur Herleitung der Bestattungssitte’, in M. Stück (ed.), Römerzeitliche Gräber
als Quellen zu Religion Bevölkerungstruktur und Sozialgeschichte. Archäologische Schriften
des Instituts für Vor-​und Frühgeschichte der Johannes Gutenberg-​Universistät Mainz Band
3. Mainz: Johannes Gutenberg-​Universität, 81–​94.
Strück, M. (2001). ‘High Status Burials in Roman Britain (First–​Third Century ad): ​Potential
Interpretation’, in J. Pearce, M. Millett, and M. Strück (eds), Burial, Society and Context in the
Roman World. Oxford: Oxbow Books, 85–​96.
Swain, H., and Roberts, M. (2001). The Spitalfields Roman. 2nd edn. London:  Museum of
London.
Toynbee, J. M. C. (1971). Death and Burial in the Roman World. London and Southampton:
Thames and Hudson.
 Cemeteries and Funerary Practice    447

Turner, V. (1969). The Ritual Process: Structure and Anti-​Structure. Chicago: Aldine Publishing

Company.
Turner, V. (1977 [1967]). The Forest of Symbols: Aspects of Ndembu Ritual. Ithaca, NY: Cornell
University Press.
Ucko, P. J. (1969). ‘Ethnography and the Archaeological Interpretation Of Funerary Remains’,
World Archaeology, 1: 262–​280.
Watson, S. (2003). An Excavation in the Western Cemetery of Roman London: Atlantic House,
City of London. London: Museum of London.
Watson. J. (2004). ‘Wood Preserved by Iron Nails’, in H. E. M. Cool, The Roman Cemetery at
Brougham, Cumbria. Excavations 1966–​67. Britannia Monograph Series 21. London: Society
for the Promotion of Roman Studies, 272–​273.
Webster, J. (2001). ‘Creolizing the Roman Provinces’, American Journal of Archaeology, 105/​
2: 209–​225.
Weekes, J. (2007). ‘A Specific Problem? Detection, Protection and Excavation of Romano-​
British Cemeteries through Competitive Tendering’, in B. Croxford, N. White, and R. Roth
(eds), TRAC 2006:  Proceedings of the Sixteenth Annual Theoretical Roman Archaeology
Conference. Cambridge 2006. Oxford: Oxbow Books, 183–​191.
Weekes, J. (2008). ‘Classification and Analysis of Archaeological Contexts for the
Reconstruction of Early Romano-​British Cremation Funerals’, Britannia, 34: 145–​160.
Weekes, J. (2011). ‘A Review of Canterbury’s Roman Cemeteries’, Archaeologia Cantiana
CXXXI, 23–​42.
Weekes, J. (forthcoming (a)). ‘Evidence for Personalisation of Cult in early Romano-​British
Cremation Burials in South-​East England’, in R. Häeussler, A. C. King, G. Schörner, and
F. Marco Simón (eds), Religion in the Roman Empire: The Dynamics of Individualisation.
Oxford: Oxbow Books.
Weekes, J. (forthcoming (b)). ‘Postscript:  on Polysemy, Liminality and communitas’,
in J. Pearce and J. Weekes (eds), Death as a Process:  Funerals in the Roman World.
Oxford: Oxbow Books.
Wenham, L. P. (1968). The Romano-​British Cemetery at Trentholme Drive, York. Ministry of
Public Buildings and Works Archaeological Report 5. London: HMSO.
Wheeler, R. E. M. (1943). Maiden Castle, Dorset. Reports of the Research Committee of the
Society of Antiquaries of London 12. Oxford: Society of Antiquaries.
Whimster, R. (1981). Burial Practices in Iron Age Britain. BAR British Series 90(i–​ii). Oxford:
Archaeopress.
White, N. C.  C. (2007). ‘Catering for the Cultural Identities of the Deceased in Roman
Britain: Potential and Problems’, in B. Croxford, N. Ray, R. Roth, and N. White (eds), TRAC
2006:  Proceedings of the Sixteenth Annual Theoretical Roman Archaeology Conference.
Cambridge 2006. Oxford: Oxbow Books, 115–​132.
Whiting, W., May, T., and Hawley, W. (1931). Report on the Excavation of the Roman Cemetery at
Ospringe, Kent. London: Society of Antiquaries.
Wigg, A. (1993). ‘Zu Funktion und Deutung der Aschengruben’, in M. Strück (ed.),
Römerzeitliche Gräber als Quellen zu Religion Bevölkerungstruktur und Sozialgeschichte.
Archäologische Schriften des Instituts für Vor-​ und Frühgeschichte der Johannes
Gutenberg-​Universistät Mainz Band 3. Mainz: Johannes Gutenberg Universität, 111–​116.
Williams, H. M. R. (2004). ‘Potted Histories: ​Cremation, Ceramics and Social Memory in Early
Roman Britain’, Oxford Journal of Archaeology, 23/​4: 417–​427.
Wilmott, T. (2010). ‘Birdoswald Roman Cemetery’, English Heritage Research News, 14: 16–​19.
448

Chapter 22

Identit y a nd t h e
M ilitary C ommu ni t y
in Rom an Bri ta i n

Ian Haynes

Introduction

In his powerful study Identity and Violence, the Nobel Laureate Amartya Sen explores
the harm done when a group is labelled according to some ‘allegedly predominant iden-
tity that drowns other affiliations’ (Sen 2006: p. xv). It may be argued that those associ-
ated with the military community in Roman Britain have themselves suffered in this
manner; the diverse factors that contributed to each individual’s identity have often
been overlooked as attention has focused on the army as an instrument of violence.
The need for greater nuance when discussing identity has been recognized in Romano-​
British studies for some time (Hill 2001), and the importance of stressing factors such
as age and gender widely noted, but it has only really been in the last decade that these
approaches to the military community have attracted more general attention in the field
(see also Allason-​Jones, this volume, for discussion of military culture).
Identity is an elastic term, open to such a range of uses that it runs a real risk of mean-
ing everything and nothing: individuals can adopt or emphasize different identities in
different settings. A Roman soldier, for example, could be a subordinate, a superior, a
colleague, father, son, husband, brother, a guild member, and/​or a patron. The way he
identified himself might have been influenced by a range of factors, from the obviously
professional, such as the status of his unit, to the non-​institutional—​his ethnic origin or
the language he spoke off-​duty, or the gods he chose to worship. Yet for all its looseness,
the concept of identity can be helpful when challenging the perception of the Roman
army in Britain as a homogenous offshoot of a monolithic institution. While Roman
soldiers have probably attracted more archaeological attention en masse than any other
group in Roman Britain, this has sometimes only reinforced the view that there is little
 Identity and the Military Community in Roman Britain    449

more that we can learn from them, that their identities are somehow less complicated
and less interesting than those elsewhere in provincial society. Yet, not only is this belief
mistaken; it also fails to take account of the fact that the study of the soldier cannot
be fruitfully detached from the study of those who lived around him, his family and
dependants. These too were an integral part of the larger military community for which
we have so much information, information that allows us to look at the issue of identity
in fresh and multifaceted ways.
Before we explore this theme, however, it is worth emphasizing one point that is too
readily overlooked: the Roman army in Britain, around which the province’s military
communities coalesced, shared many characteristics with the armies of other provinces,
but it had its own particularities too. While soldiers still represented a small propor-
tion of the total population, the army of Britain was exceptionally large by provincial
standards, a fact that probably had substantial implications for the way in which people
employed and responded to military identities. A significant part of this army was sta-
tioned on the most heavily defended frontier barrier in the empire, a further indication
of the distinctive circumstances in which it found itself.
This chapter attempts to unpick some of the evidence for identity groups associated
with the Roman army in Britain. In so doing it begins by recognizing that the military
community comprised far more than just the soldiers themselves. It goes on to note
some of the differentials that may have mattered most to the soldiers and civilians who
formed part of that community, considering evidence for status, gender, cult practice,
and ethnic affiliation, before concluding with a brief consideration of the veterans as a
distinct identity group.

The Military Community

Recent research has explored the link between military identities and notions of community
(Goldsworthy and Haynes 1999; James 1999). So powerful has this idea become that a major
recent study of Roman Britain, David Mattingly’s An Imperial Possession (2006), has actu-
ally been structured around the idea of addressing the province’s story through the prism
of three distinct community groups, with the ‘military community’ as the first of these.
In many respects, this development is a healthy one, so long as it does not lead to
another form of disengagement between the study of the military and the study of the
rest of the province’s population. The value of this line of enquiry lies in its recognition
that the lives of men, women, and children of diverse statuses and cultural backgrounds
were intimately bound up with the army in Britain and warrant study in their own right.
Thus the story is one not simply of soldiers, but of a far wider range of people. The sec-
ond advantage of this way of thinking is that it forces us to look beyond the important
foundational body of work that has built up on the official structures of the army as an
imperial institution. It considers those in the military community as agents, not simply
as cogs in a machine.
450

450   Ian Haynes

Representatives of many different status groups lived in and around the forts
and fortresses of Roman Britain. A  dedication discovered at High Rochester in
Northumberland (RIB 1271) neatly conveys something of the complex social relations
that could exist in the military milieu. It records that the freedman Eutychus made
a dedication on his own part and on the part of his dependants for the welfare of his
patron Rufinus, the Tribune, and Rufinus’ wife, Lucilla. We learn elsewhere that Lucilla
was herself a clarissima femina, a woman of senatorial rank (RIB 1288). High Rochester
was an outpost fort beyond Hadrian’s Wall, yet even here we see a spectrum of statuses
among the non-​combatants, from the dependants (including slaves?) of a freedman,
himself a former slave, to a woman in the social elite.
That Eutychus had the resources, education, and inclination to commission such a
dedication reminds us that, even within the army, a largely overlooked miscellany of
servants serviced the needs of the military community. These had their own hierarchies,
with the calones who served private soldiers at the bottom. For Horace, these were the
lowest form of life, so low indeed that he depicts them as bit players in the degrading
torment of adulterers in the afterlife, urinating on the damned (Sermones 1.2.44). Yet
some troopers clearly cherished their servants; Numerianus a trooper of ala I Asturum,
for example, set up a tombstone to his freedman Victor the Moor at South Shields
(RIB 1064).
Alongside a fresh appreciation of what written sources tell us about the range of indi-
viduals who lived in the orbit of the army, there has come a renewed commitment to
understanding what other archaeological evidence might reveal about these diverse
individuals. Here a major breakthrough came with Carol van Driel-​Murray’s work
(1995) on the shoes from Vindolanda. Despite the fact that the famous writing tablets
from this site make very few references to those women and children who were not part
of the commanding officer’s family—​the group to which Rufinus and Lucilla belonged
at High Rochester—​analysis of the shoes found within the walls of the first-​century forts
indicate that many were present there.
The heightened awareness of the presence of non-​combatants has, in turn, had an
impact on the way that military installations are examined. There have been several
attempts to engender Roman forts, notably by Pim Allison (Allison et al. 2004; Allison
2005, 2006), and important discussions of the challenges inherent in reading the data
associated with them (Allason-​Jones 1995; James 2006). At the same time, a grow-
ing interest in extramural settlements and annexes—​the areas previously assumed
to be occupied by civilian followers—​has led to a surge in fieldwork at these too. Such
research is most effectively realized where, as recently at Vindolanda, work both within
and beyond the walls continues in tandem, demonstrating that a range of identities were
found to act across both zones (A. Birley 2013). Work on the canabae, extramural settle-
ments associated with legionary fortresses, demonstrates similar processes at work, albeit
on a larger scale. For example, recent fieldwork at Caerleon, home of the Legio II Augusta,
shows just how many official buildings lay outside the fortress walls (Guest et al. 2012).
Even within the territoria of their forts and fortresses—​the lands the army managed
directly—​it is clear that there was scope for members of the military community to play
 Identity and the Military Community in Roman Britain    451

out different identities across a range of settings. And, in articulating these roles, they
would have used material culture differently and both consciously and unconsciously.
This makes analysis of the data challenging enough, but it is not the only problem, for it
is vital to remember that the military community did not simply consist of those living
in the shadow of ramparts and barracks. The reach and impact of Rome’s military infra-
structure meant that soldiers, former soldiers, and their dependants, extended families,
and associates were found far beyond these relatively easy-​to-​identify installations.
These deliberations, in turn, raise a further question, one of fundamental importance
to any analysis of identity and the Roman army. Archaeologists most frequently iden-
tify the presence of the military community through the presence of certain diagnostic
buildings of types commonly used in the first three centuries ad, but to what degree can
other sources of evidence be used to indicate its presence?
What is increasingly clear is that what is often grouped together under the title of
‘military equipment’ is a problematic category and certainly not a reliable indicator of
the presence or absence of soldiers. A recent review of Portable Antiquities Scheme
data revealed that many examples of such material were found in rural settings far from
known military dispositions, and noted that, while there are many possible explanations
for this, such as the mobility of soldiers, the degree to which a distinction between mili-
tary and civilian objects can be made must be questioned (Worrell and Pearce 2012: 38–3​
9).
Nor, casting the net wider beyond the presence or absence of individual finds, is it
possible to speak with confidence of ‘military assemblages’, notional groupings of arte-
facts that indicate the presence of soldiers (Allason-​Jones 2001). These cannot be dem-
onstrated convincingly in the first three centuries ad, and, if anything, it is even harder
to identify them later. As Hilary Cool (2010) notes in her discussion of finds from the
fourth-​century frontier, the world in which later soldiers operated was ‘a different life’
in material terms, one we are only now beginning to study systematically (Collins and
Allason-​Jones 2010; Coulston 2010). Difficulties in characterizing the late period are
compounded by the still greater variety to be found within and between military bases
from the late third century onwards (Gardner 2007: 16, 262–​263), and the growing use of
towns to house troops.
Traditional treatments of Roman armies describe, generally usefully, their complex
rank structures and organization. There is no need to repeat these treatments here, but
there is a need to reflect more deeply on what archaeology reveals about their impor-
tance for military identity. At one level, as a brilliant study by Simon James (1999) has
demonstrated, a powerful ‘bottom-​up’ pressure was at work that helped to foster an
‘imagined community’, a shared sense of identity among soldiers across the provinces,
even among those widely separated in time and space. This was reinforced by the sense
of shared experience common in what sociologists term ‘occupational communities’,
groups whose professional circumstances are characterized by a distinctive status and,
commonly, by potentially more dangerous or socially marginal work (Haynes 1999: 9).
In the absence of regulations and mass production of military equipment under the
early and high empire, this sense of soldierly identity helped to popularize certain forms
452

452   Ian Haynes

of material expression, shared forms of equipment, and dress. This went far beyond the
sword and its accoutrements, the quintessential tool and symbol of the Roman soldier,
to touch on a range of items of clothing. Military fashions absorbed provincial traditions
and, not infrequently, the styles of Rome’s enemies too, disseminating them remarkably
widely (James 2011: 189). The speed with which such fashions could travel should not
be underestimated, but they contributed to the evolution of the shared articulation of
military identity. The force of this dynamic is powerfully conveyed by metalworkers’
moulds from forts on Hadrian’s Wall (Allason-​Jones and Dungworth 1997). These show
that multiple small local producers operating in the military milieu could, potentially,
produce objects and adornments virtually indistinguishable from those worn across an
area stretching from the Tyne to the Euphrates.
Yet the compelling materiality of a shared soldierly identity must be considered
against other forms of distinction. Surviving evidence for the terms and conditions of
service in the legions, the alae and cohorts of the auxilia, the fleet and the still poorly
understood numeri, suggest that, at least until the early third century, there were sig-
nificant status differences between these troop types and the men who served within
them. Familiar images of Roman soldiers, themselves often derived from Roman
public art, tend to suggest notably different forms arms and armour distinguished
the appearance of different types of soldier. The actual archaeological evidence from
Britain suggests that, while some of these differences—​for example, the particular
association of lorica segmentata (plate armour) with legionaries—​may be question-
able (Maxfield 1986), the difference between, say, legionary and auxiliary soldiers
would have been clearly signalled to contemporary viewers, notably by the shield
types and missile weapons employed (Bishop and Coulston 2006: 254–​259). Over
time, however, the two troop types may have become less obviously distinguishable in
appearance—​a product not just of converging military identities, but also of changes
in wider society. Unlike entry to the auxilia, entry to the legions was restricted to citi-
zens; as more citizens joined the auxilia, the difference between the statuses of the
troop types was eroded. Yet, even after the early third century ad, when the edict of
Caracalla extended citizenship to all freeborn subjects of the empire, some marks
of distinction were preserved. Coins minted by the usurper Carausius (286–​93 ad)
sought to claim the allegiance of named local legions, but not auxiliary formations
(Casey 1994: 92–​96; Coulston 2004: 137).
Other forms of distinction were ultimately to prove at least as important. The vast
majority of Rome’s cavalry were in the auxilia and, in the earliest years of Britain’s pro-
vincial history, would have been overwhelmingly non-​citizens. Yet from the beginning
they asserted their special status as horse soldiers with spectacular confidence. A dispro-
portionately high number of the military figural tombstones from Britain were erected
by cavalrymen. In the first century ad, their memorials are characterized by impressive
set pieces, with the deceased galloping over the body of a cowering, battered, or even
beheaded barbarian. The style, which can be paralleled on the Rhine frontier, has the
particular virtue of foregrounding not only the soldier but also the horse, the particular
mark of his military stature. While these monuments fall from fashion in Britain, the
 Identity and the Military Community in Roman Britain    453

horsemen’s status did not decline; rather, they superseded legionary infantry soldiers as
the most potent force within later Roman armies.
For the individual soldier it is clear that his military affiliation mattered greatly; along-
side his rank, a soldier’s heirs would always make sure that his regiment was recorded
on a tombstone. While movement between regiments is well documented for men who
reached the level of centurion or troop commander, most would have spent their lives
within one regiment, and, after the mid-​second century, many regiments would have
retained their headquarters at the same location for generations. Though soldiers would
indeed have identified with the wider ‘community of the soldiers’, it is equally important
to note that membership of a particular ‘regimental community’ would have been a vital
part of his military identity too (James 2001: 87).

Cult Practice and Identity

Evidence for cult practice in and around Roman military installations gives us some
sense of how these disparate affiliations operated. Here it is possible to explore not only
the official dimensions of military identity, but also other community groupings with
which individuals chose to identify. The results allow for a certain level of what might be
termed ‘social mapping’, as they expose patterns of association that do not simply repli-
cate the official structures defined by army administration.
To begin at the broadest level, it is most commonly argued that a sense of shared
identity across Rome’s armies was actively fostered by participation in shared rituals.
Some of the commonly cited evidence in support of this belief is more slender than is
usually acknowledged, and the arguments for a shared religious calendar across the
empire’s armies are actually based on a single document, the feriale Duranum, found
not in Britain but in Syria (Fink et al. 1940). While it is quite possibly correct to assume
that this papyrus is the sole surviving exemplar of what was once a widely distributed
and consulted document, it is important to note that by no means all the entries found
within it can be paralleled in the relatively rich body of evidence we have for cult in
the army of Britain. Yet there are indications of shared practice. Thus the Natalis Urbis
Romae Aeternae, celebrated annually on 21 April, appears both in the Syrian calendar
and on an altar dedicated to Dea Roma on the same day by duplicarii (double-​pay sol-
diers) at High Rochester (RIB 1270).
Dedications to Jupiter Optimus Maximus, the greatest god of the Roman pantheon,
also appear to point to a practice commonly followed by military communities in
Britain and beyond. These dedications are very frequently linked in academic literature
to the vows military units are believed to have repeated annually in early January, and,
though the text on the inscriptions does not state this explicitly, this may well be the
correct explanation. It is an explanation that fits well with the fact that these dedica-
tions are commonly fronted by commanding officers, who as the senior figures in their
communities would have played the leading role in such attestations. Furthermore, the
454

454   Ian Haynes

fact that at Maryport in Cumbria—​site of the largest number of such altar finds—​we
can find groups of three or four altars given by the same named commander does sug-
gest a regular cycle of dedication. The widespread belief, however, that this cycle was
accompanied by the ritualized burial of an old altar every time a new one was erected
(Wenham 1939) have been disproved following the re-​excavation of the Maryport site
(Haynes and Wilmott 2012).
That Jupiter dedications did nonetheless contribute to a shared sense of identity
within the military community is certainly suggested by other evidence. First, we can
see that those non-​soldiers who lived alongside forts sometimes made their own com-
munal dedications to the god. Thus the vikani consistentes castello Veluniate (the people
of vicus (village) of the fort at Carriden, in Falkirk) made vows to Jupiter too, but their
dedication was set up by one Aelius Mansuetus, who gives no military rank, but was
presumably a leading member of the vicus community (RIB 3503). Among the military
professionals, a shared sense of identity bound up with Jupiter veneration is suggested
from rather different types of finds:  belt plates and baldric fittings. Many surviving
examples, including a fine third-​century baldric phalera from Carlisle with the prayer
optime maxime con(serva numerum omnium) (‘ask that the god, the greatest and
the best, keep safe the men of the wearer’s unit’) laid out in an openwork design framing
Jupiter’s eagle. Directly comparable pieces are known from the finds of belt fragments
from across western Europe, as well as the Danubian and African provinces (Bishop and
Coulston 2006: 162).
The impact of Christianity on this and other aspects of military culture in Britain
remains hard to gauge, but evidence from outside the province suggests that it may well
have had an impact on military dress and symbolism in the years following Constantine’s
triumphal entry to Rome in ad 312. We may plausibly suggest a growing Christian pres-
ence within the army of Britain before this, and Gildas claims that two men from the
Urbs Legionum (Caerleon) were martyred under Diocletian (De Excidio Britanniae
10). There is some evidence for church-​building at fort sites after Constantine’s conver-
sion. The strange apsidal building discovered in the principia (headquarters building) at
Vindolanda dating to c. ad 400 or later (Birley et al. 1999: 20–​21) is best understood as
one such structure.
It is quite likely that emerging forms of Christianity may have been able to incorpo-
rate some forms of pre-​Christian military cult practice, but others may have proved
more problematic. Of these, one may have been the cult of the standards—​a practice
bound up with regimental identity and condemned c. ad 200 by the Christian polemi-
cist Tertullian (Ad Nationes 1.12). Tertullian was not writing in a British setting, he was a
North African of Montanist persuasion who was notably less inclined towards accom-
modation with pagan tradition than many of his contemporaries may have been. He
was, nevertheless, clearly not wholly exaggerating when he described the cult of the
standards as a form of veneration that was embraced by the soldiery more passionately
even than the worship of Jupiter himself.
The totemic power of the standards carried by the regiments of Britain’s army is elo-
quently illustrated by the frequency with which these symbols are repeatedly depicted
 Identity and the Military Community in Roman Britain    455

on building inscriptions. Thus, the work of Legio XX Valeria Victrix is instantly recog-
nizable because of the legion’s tendency to mark it with a running boar. Alongside their
famous eagle standards, the British legions carried images of such potent beasts, both
real and mythical, which were deemed to be bound up with a genius (loosely translated
as protective spirit) and worthy of cult in their own right.
Though they did not carry eagle standards, the auxiliary forces had their own animal
totems; thus the altars of cohors IIII Gallorum are often carved with the figure of a stork,
while cohors VI Nerviorum seems to have toted a bull. These standards were housed, as
were the legionary standards, in shrines located above the strong room in the headquar-
ters building, the most secure and central position within a fort. They were placed both
physically and spiritually at the heart of the military community.
There has been some debate as to whether there were further cult practices that served
to reinforce regimental identity within the auxilia. Writing in 1978, Eric Birley argued
that it was possible to identify what he termed ‘approved cults’, cults not originally of the
notional Roman pantheon which nonetheless received special attention from a given
regimental community. In its original form, this argument was further bound up with
a belief that some units continued to draw recruits from their place of origin, an argu-
ment that must now be challenged (Haynes 2013: 298). But a range of evidence recov-
ered at Vindolanda after Birley had composed his essay does pose interesting questions
about the degree to which some cults were favoured in official settings. While it was long
believed that only the Capitoline deities and the standards could be honoured by shrines
within the fort walls (Helgeland 1978)—​unnecessarily, in fact, given evidence to the con-
trary from other provincial territories—​evidence from Vindolanda makes it clear that
temples to other gods could be accommodated within the fortifications. The discovery
of a shrine to Dolichenus demonstrates that this god, many of whose known worship-
pers were associated with the army, was permitted space in what might be thought of as
a highly regulated area (Birley and Birley 2010).
A different kind of special relationship is suggested by a dedication from Vindolanda
to Dea Gallia, the divine personification of Gaul, by cives Galli … concordesque Britanni
(A. R. Birley 2008). A. R. Birley (2008: 176–​177) has argued that the Gaulish and British
citizens recorded here as operating in concord to make the dedication are possibly sol-
diers; my own suspicion is that they are in fact civilians—​for no formal titles or regimen-
tal affiliations are given. Whoever they were, the act of dedication was clearly intended
to foster harmony and/​or celebrate shared endeavours by distinct groups within the
community, possibly after tensions had broken out between them.
Cult dedications can also serve to demonstrate the unofficial subdivisions within
institutions that may have proved much more important to the identities of individu-
als than their official status. Three altars from Birrens in Dumfriesshire show how this
may have worked: all were erected by men within the auxiliary unit cohors II Tungrorum
equitata and all date from the mid-second century ad, yet each is dedicated not only to
a different god or gods, but by men from different regions. One group came from Raetia
(RIB 2100), and the others from Germania—​the Vellauvian district (RIB 2107) and the
territory of the Condrusi (RIB 2108) respectively. Yet, while each of these dedications
456

456   Ian Haynes

consciously asserts a particular regional identity, it also subconsciously attests to the

power of those forces that bound the military community together. Each dedication
follows a familiar epigraphic style, each is composed in Latin, and each is designed to
receive offerings and/​or commemorate the discharge of a vow. Each, therefore, adheres
essentially to the quintessential attributes of Roman cult in the western provinces. And,
in this respect, it is notable that, while some of these dedications were to gods associated
with the men’s homelands, not all of them were.
The range of religious dedications left by members of the military community in
Britain is broad and suggests an array of allegiances and affiliations. Perhaps the best
known is the cult of Mithras, which is particularly strongly associated with the army
in Britain, even though the god draws worshippers from a notably broader constitu-
ency in other provinces. The drama that accompanied the discovery of some of the god’s
temples, the fascination with the mysterious initiation rites of the cult, and the intrigu-
ing synthesis between Persian belief and western practice have meant that Mithras has
received more than his due in Romano-​British studies. In reality, the small number of
temples—​each capable of accommodating only a relatively small community of wor-
shippers and in a number of cases experiencing extended periods of disuse—​suggests
that only a very small proportion of men in the military community were active adher-
ents, and the cult appears to have been closed to women. This is not, however, to say
that the cult was without influence; successive commanding officers at Carrawburgh,
for example, thought it appropriate to erect altars to the god—​even though it is not clear
that they were themselves initiates, certainly none of them describes himself with an
initiation rank.
A very different pattern of dedication is associated with Veteris/​Di Veteres, a deity or
deities, who may have been perceived as an individual, a group, a male, or a female.
Sixty-​one altars are known to have been set up to this enigmatic figure(s), who may have
had Germanic antecedents, and yet who appears to have been worshipped predomi-
nantly in and around Hadrian’s Wall. In sharp contrast to the surviving dedications to
Mithras, altars associated with this deity are often extremely basic and rough cut, with
almost half not even recording the name of the devotee. But a couple of dedications
were clearly set up by military personnel, including one from Carvoran by Iulius Pastor,
an imagnifer (carrier of the emperor’s image) in cohors II Delmatarum (RIB 1795). The
names that do survive and the simplicity of most of the altars do appear to suggest a cult
most popular among lower-​status worshippers, but sadly it is impossible to demonstrate
whether or not these were all soldiers or a mix of soldiers and civilians.
Amid all this diversity, it is important to recall that, by definition, inscriptions reflect
the activities and sympathies of those touched by the ‘epigraphic habit’ (the practice of
writing in stone), and that these individuals are not typical of the mass of the popu-
lation. Inscriptions also cluster in certain regions and times, overwhelmingly to the
period between the mid-​first century ad and the late third. It is, therefore, vital not to
confine any search for the relationship between cult, community, and identity to them.
Fortunately, there is plenty of other archaeological information. When looked at with
due care and attention, the area in and around military installations is often found to be
 Identity and the Military Community in Roman Britain    457

replete with evidence for ritual practice. While the old adage that archaeologists are fre-
quently disposed to describe as ritual that which they cannot otherwise explain should
not be forgotten, there are fascinating indicators of practices that both complement
and conflict with the better-​known ones that are documented in the epigraphic record.
Excavations of a third-​century barrack range at South Shields, for example, revealed that
a stone with a phallus graffito had been placed facing into the wall of three separate con-
tubernium suites (Croom 2001: 68). These suites, which comprise two rooms, are gen-
erally understood to have housed the eight-​man team/​tent group that constituted the
smallest element of an infantry unit. With its potent apotropaic symbolism, it appears
that each graffito may have been situated to protect this key group.
No less compellingly, other sites have yielded traces of a complex array of structured
deposits—​striking mixed assemblages of military equipment, human remains, and
other items—​that find their most obvious parallels in indigenous votive practices. And
such parallels certainly offer a very tempting way to interpret some of the pits discovered
at the site of Newstead in southern Scotland, where the nature of the deposits unearthed
points to something more than just a clearing-​up of the site by the fort’s garrison (Clarke
and Jones 1996; Clarke 2001). Such discoveries challenge those who have sought to study
military identity in Roman Britain in isolation from the late Iron Age background.
There has been relatively little systematic work on the cemeteries adjacent to Roman
military installations, but the work that has taken place signals again that it is dangerous
to assume a monolithic ‘military’ identity. While it is notable that a very high proportion
of surviving Romano-​British tombstones do record members of the military commu-
nity, giving the names of either the soldiers or veterans themselves (almost invariably
with their regimental titles inscribed), and/​or the families, freedmen, and slaves with
whom they shared their lives, the evidence from the graves themselves reflect significant
variation. It is the diversity of funerary practice rather than a sense of similarity that
confronts the excavator. True, there are some grounds for believing that the military
practised cremation rather longer than most of the civil population in Roman Britain,
but there is no neat set of grave goods or burial rite that was universal to soldiers and
otherwise distinguished them from the population.
The relationship of burial to identity is again, of course, a complex one. The dead,
it has been rightly noted, do not bury themselves, so the evidence that survives from
the graves tells us more about those who oversaw the funeral than it does about the
deceased. Importantly, those cemeteries that have been explored do appear to contain
the burials not only of men of military age, but of a broader demographic. While the
conjectured age and sex ratios may not always follow what is assumed to have been the
pattern in the broader population, it is clear that older men, women, and children are
present in many of the military cemeteries studied to date. It is not unreasonable to read
family plots into some of the data, as, for example, at Birdoswald in Cumbria (Wilmott
2010). Here, a mid-second-​century cinerary urn containing a Type 3 earring and a mail
chatelaine was cut into the ground as close as possible to an earlier urn—​the pair lying
together in a circular enclosure. Changes in the latest phases of the Birdoswald cem-
etery, sometime after ad 370, may suggest another aspect of family groupings too: in
458

458   Ian Haynes

sharp contrast to the cremation practice that had characterized the cemetery through
much of its use, two inhumation burials were inserted across the entrance to a major
funerary enclosure. An act that appears to imply at once continuity and change in the
use of the area—​the deliberate location of the burials and striking difference in rite—​
makes it tempting to suggest that the later (Christian) occupants of the adjacent fort site
both connected themselves with the cemetery’s use by their predecessors, but disassoci-
ated themselves from the beliefs that sustained them.
Of all military cemeteries in Roman Britain, it is the one that has been most stud-
ied from the perspective of identity, the cemetery at Brougham, also in Cumbria, that
may also prove to be the least representative (Cool 2004). An inscription from the site
refers to a unit of Stratonician cavalry (RIB 780), and it appears that an ‘irregular’ unit
was stationed there in the third century. In a complex statistical analysis of the material
recovered in the 1966 excavations, Cool has argued that there appear to be certain cor-
relations between artefact types buried with certain age/​sex groups. Thus, for example,
deep samian bowls appear only in adult burials (Cool 2004: 451). No less fascinatingly,
she argues that some of the rites—​such as horse cremation—​and some of the artefacts—​
such as bucket pendants—​reflect the trans-​Danubian ethnic identity in the unit (Cool
2004: 464–​466). Importantly though, she adds that some of the artefacts this group saw
fit to inter with their dead were neither from the Danube region nor from the north of
Britain, but had been picked up by the regiment, perhaps on its new posting.

Ethnic Identity

Cool’s work at Brougham is one of a number of cases where the presence of distinctive
artefact types has been associated with ethnic identity groups within the wider military
community, but it is particularly compelling because it brings this evidence together
with the testimony of very unusual funerary practice. More generally, it is notably diffi-
cult to determine the degree to which—​and the contexts within which—​ethnic identity
was exhibited within the military community.
This may seem a surprising statement given that the vast majority of regiments sta-
tioned in Britain—​most of the alae, cohorts, and numeri—​actually carry ethnic titles.
But it is too readily forgotten that these are the names the Roman authorities gave these
units when they were first raised. Thus they recall a Roman notion of ethnic identity,
which was not necessarily the label the men themselves would have chosen. The differ-
ence between imposed names chosen by outsiders (exonyms) and those used by a group
of itself (endoymns) is one of considerable import in all societies, but it has a particular
potency in the Roman Empire. Furthermore, whatever the label, any ethnic or cultural
links within a unit would have undergone revision as the ranks were replenished by men
who were drawn from what the authorities deemed to be the nearest source of suita-
ble manpower, not necessarily from the place where the regiment was originally raised
(Haynes 2013: 157–1​60).
 Identity and the Military Community in Roman Britain    459

It is commonly asserted, largely on the basis of epigraphic evidence from some of

the Danubian provinces and the North African site of Lambaesis, that Rome’s armies
moved—​over time—​towards a practice of local recruitment. Yet it must be noted that
there is no direct evidence that this became the rule in Britain and the number of British
recruits known to have actually served there is very small indeed. Conversely, parts of
Gaul and Germany are likely to have played a significant role not simply in supplying
the greater part of the units and men sent to Britain in the original Claudian invasion
force, but also in the centuries that followed. Roman commentators make clear that they
saw these regions as particularly rich in human capital (e.g., Cicero, Phil. 5.2.5–​6). Their
contribution to the pool of recruits may be set alongside other regions such as Thrace,
North Africa, and Syria, which provided units but, it appears, a much less consistent
supply of manpower (Haynes 2013: 103–​120).
As noted above, ethnicity does not lend itself readily to archaeological study. Some of
the alleged examples of the survival of ‘ethnic’ fighting practices within the Roman army
should be treated with considerable caution. Thus the image of a curved falx sword on
official inscriptions set up by cohors I Aelia Dacorum on Hadrian’s Wall (RIB 1909 and
RIB 1914), a century after the regiment had first arrived in Britain, is much more likely to
indicate that the falx was deemed a convenient symbol for a Roman idea of ‘Dacianess’,
rather than a weapon the soldiers carried into battle. Evidence for different ethnic iden-
tity groups is actually less likely to come from official documents and military equip-
ment, and more through the subtler evidence for daily routine. Some pottery forms, in
particular, may reflect the arrival of soldiers or soldiers’ families from different regions.
For example, Swan (1992) has proposed that the appearance of North African-​style
cooking wares in Northern Britain may be associated with African legionaries serving
in Legio VI. Elsewhere, the distinctive so-​called Frisian or Housesteads Ware found out-
side several forts on the northern frontier clearly follows forms of pottery found beyond
the lower Rhine (Jobey 1979). Its distribution quite possibly attests the presence of irreg-
ular soldiers drawn from what is now the Netherlands, and indeed units of this type,
such as the cuneus Frisiorum, are attested epigraphically on Hadrian’s Wall. This evi-
dence is powerfully suggestive then, but we should hesitate before attempting to conflate
any one type of data with ethnic identity; cooking methods and dietary practices can
spread swiftly between disparate communities, and most pots can be relatively easily
copied (see also Iveleva, this volume, for discussion of ethnicity).

Veterans

One further manifestation of military identity may be usefully discussed. This is that of
the veteran. That it exists as a category deemed worthy of stressing on funerary inscrip-
tions and law codes suggests that it had a particular importance. Veterans certainly
appear as a notable identity group in Tacitus’ account of Colchester in the build-​up to
the Boudiccan Revolt of ad 60/​61. By the time of the revolt, the fortress at Colchester
460

460   Ian Haynes

had been substantially remodelled, turning it into a city for veterans—​it had become
Britain’s first colonia. Tacitus (Annals 12.32; 14.31) explains that, as such, the city was
designed to function both as security against revolt and as a device for habituating the
native population to their new legal situation. It failed on both counts, partly, the his-
torian claims, because of mismanagement and cruelty by Roman administrators more
generally, but also because of the antipathy specifically felt towards the veterans them-
selves locally. Given this, it is perhaps ironic that veterans are so often seen not only as
a distinctive identity group, but also as powerful and effective evangelists for ‘Roman’
culture in the transformation of Britain. There are certainly some areas where a distinct
veteran presence may be suggested archaeologically, such as at Colchester prior to the
Boudiccan revolt, and plausibly in the emergence of small ‘villa’ buildings that resemble
military quarters (Black 1994). Yet, overall, it is likely that the degree to which veterans
constituted a distinct identity group varied, depending on time and place. The find-​spots
of veterans’ funerary inscriptions and of diplomas granted to discharged auxiliaries sug-
gest that some stayed close to forts and fortresses, preserving perhaps their soldierly
connections more conspicuously in a military milieu, while others settled further afield
in areas where their smaller numbers may have led them to blend in more readily with
other country dwellers. How many of these individuals would, after twenty-​five years of
service in the army, have had the energy or the inclination to serve as agents of cultural
change must be debated.
In conclusion, we must acknowledge that the particular character of the Roman army
in Britain was distinguished by a distinctive network of occupational communities. The
role of the soldier as a man licensed to kill on behalf of the emperor marked him and his
dependrnts out in provincial society. Bound up in the cosmopolitan culture of the army
and privileged through the assured supply of cash and corn, soldiers and their stations
can be strikingly conspicuous in the archaeological record, but it does not follow from
this that the people in and around these forts and fortresses constituted a homogenous
group. Rather, the rich evidence that survives paints a vivid and varied picture, and
shows just how varied and multi-​layered identities in Roman Britain could have been.

Abbreviation
RIB  R
 . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I:
Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Primary Sources
Cicero, Philippics I–​II, ed. J. T. Ramsey. Cambridge Greek and Latin Classics.
Cambridge: Cambridge University Press, 2003.
Gildas, Gildas Volume 7, ed. M. Winterbottom. Chichester: Phillimore and Co., 1978.
 Identity and the Military Community in Roman Britain    461

Horace, Satires Book 1, ed. E. Gower. Cambridge Greek and Latin Classics.
Cambridge: Cambridge University Press.
Tacitus, Annals I–​III, ed. J. Jackson. Loeb Classical Library. Cambridge, MA:  Harvard
University Press, 1979.
Tacitus, Annals IV–​VI, XI–​XII, ed. J. Jackson. Loeb Classical Library. Cambridge, MA: Harvard
University Press, 1986.
Tacitus, Annals XII–​XVI, ed. J. Jackson. Loeb Classical Library. Cambridge, MA: Harvard
University Press, 1981.
Tertullian, Ad Nationes I, ed. Q. Howe, 2007 www.tertullian.org/​works/​ad_​nationes.htm.

Secondary Sources
Allason-​ Jones, L. (1995). ‘“Sexing”’ Small Finds’, in P. Rush (ed.), Theoretical Roman
Archaeology: Second Conference Proceedings. Aldershot: Avebury, 22–​32.
Allason-​Jones, L. (2001). ‘Material Culture and Identity’, in S. James and M. Millett (eds),
Britons and Romans:  Advancing an Archaeological Agenda. CBA Research Report 125.
York: Council for British Archaeology, 19–​25.
Allason-​Jones, L., and Dungworth, D. B. (1997). ‘Metalworking on Hadrian’s Wall’, in W.
Groenman-​van Waateringe, B. L. van Beek, W. J. H. Willems, and S. L. Wynia (eds), Roman
Frontier Studies 1995:  Proceedings of the XVIth International Congress of Roman Frontier
Studies. Oxford: Oxbow Books, 317–​321.
Allison, P. M. (2005). ‘Mapping Artefacts and Activities within Roman Military Forts’, in Z.
Visy (ed.), Limes XIX: Proceedings of the XIXth Congress of Roman Frontier Studies held in
Pécs, Hungary, September 2003. Pécs: University of Pécs, 833–​846.
Allison, P. M. (2006). ‘Mapping for Gender:  Interpreting Artefact Distribution in Roman
Military Forts in Germany’, Archaeological Dialogues, 13/​1: 1–​48.
Allison, P. M., Fairbairn, A. S., Ellis, S. J. R., and Blackall, C. W. (2004). ‘Extracting the Social
Relevance of Artefact Distribution in Roman Military Forts’, Internet Archaeology 17 <http://​
intarch.ac.uk/​journal/​issue17/​allison_​index.html>.
Birley, A. (2013). ‘The Fort Wall:  A  Great Divide?’, in R. Collins and M. Symonds (eds),
Breaking down Barriers: Hadrian’s Wall in the 21st Century. Journal of Roman Archaeology
Supplementary Series 93. Portsmouth, RI: Journal of Roman Archaeology, 85–​104.
Birley, A., and Birley, A. R. (2010). ‘A dolichenum at Vindolanda’, Archaeologia Aeliana, 5th ser.,
39: 25–​52. http://​www.tertullian.org/​articles/​howe_​adnationes1.htm
Birley, A. R. (2008). ‘Cives Galli de(ae). Galliae concordesque Britanni:  A  Dedication at
Vindolanda’, L’Antiquité classique, 77: 171–​187.
Birley, E. (1978). ‘The Religion of the Roman Army: 1895–​1977’, Aufstieg und Niedergang der
römischen Welt, II 16/​2: 1506–​1541.
Birley, R., Birley, A., and Blake, J. (1999). The 1998 Excavations at Vindolanda: The Praetorium
Site. Interim Report. Bardon Mill: Vindolanda Trust.
Bishop, M. C., and Coulston, J. C. N. (2006). Roman Military Equipment: From the Punic Wars
to the Fall of Rome. 2nd edn. Oxford: Oxbow Books.
Black, E. W. (1994). ‘Villa-​Owners:  Romano-​British Gentlemen and Officers’, Britannia,
25: 99–​110.
Casey, P. J. (1994). Carausius and Allectus: The British Usurpers. London: Batsford.
Clarke, S. (2001). ‘Probably Ritual. Assemblage Interpretation at the Newstead Military
Complex: Towards a More Holistic Approach’, in A. T. Smith and A. Brookes (eds), Holy
462

462   Ian Haynes

Ground: Theoretical Issues Relating to the Landscape and Material Culture of Ritual Space
Objects. BAR International Series 956. Oxford: Archaeopress, 73–​83.
Clarke, S., and Jones, R. (1996). ‘The Newstead Pits’, Journal of Roman Military Equipment
Studies 5: 109–​124.
Collins, R., and Allason-​Jones, L. (2010) (eds). Finds from the Frontier: Material Culture in the
4th–​5th centuries. CBA Research Report 162. York: Council for British Archaeology.
Cool, H. E. M. (2004). The Roman Cemetery at Brougham, Cumbria: Excavations 1966–​1967.
Britannia Monograph Series 21. London: Society for the Promotion of Roman Studies.
Cool, H. E. M. (2010). ‘A Different Life’, in R. Collins and L. Allason-​Jones (eds), Finds from the
Frontier: Material Culture in the 4th–​5th Centuries. CBA Research Report 162. York: Council
for British Archaeology, 1–​9.
Coulston, J. C. N. (2004). ‘Military Identity and Personal Self-​Identity in the Roman Army’,
in L. De Ligt, E. A. Hemelrijk, and H. W. Singor (eds), Roman Rule and Civic Life: Local
and Regional Perspectives. Proceedings of the Fourth Workshop of the International Network
Impact of Empire. Amsterdam: J. C. Gieben, 132–​152.
Coulston, J. C. N. (2010). ‘Military Equipment of the “Long” 4th Century on Hadrian’s Wall’, in
R. Collins and L. Allason-​Jones (eds), Finds from the Frontier: Material Culture in the 4th–​
5th Centuries, CBA Research Report 162. York: Council for British Archaeology, 50–​63.
Croom, A. T. (2001). ‘Some Finds from the 1997–​98 Excavations at South Shields Roman Fort’,
Arbeia Journal, 6–​7: 68–6​9.
Fink, R. O., Hoey, A. S., and Snyder, W. F. (1940). ‘The Feriale Duranum’, Yale Classical Studies,
7: 1–​222.
Gardner, A. (2007). An Archaeology of Social Identity:  Soldiers and Society in Late Roman
Britain. Walnut Creek, California: Walnut Creek.
Goldsworthy, A., and Haynes, I. P. (1999) (eds). The Roman Army as a Community. Journal
of Roman Archaeology Supplementary Series 34. Portsmouth, RI:  Journal of Roman
Archaeology.
Guest, P., Luke, M., and Pudney, C. (2012). Archaeological Evaluation of the Extramural
Monumental Complex (‘the Southern Canabae’) at Caerleon, 2011. Cardiff Studies in
Archaeology Specialist Report, vol. 33. Cardiff:  Cardiff University Archaeology and
Conservation.
Haynes, I. P. (1999). ‘Introduction: The Roman Army as a Community’, in A. Goldsworthy
and I. P. Haynes (eds), The Roman Army as a Community. Journal of Roman Archaeology
Supplementary Series 34. Portsmouth, RI: Journal of Roman Archaeology, 7–​13.
Haynes, I. P. (2013). Blood of the Provinces: The Roman Auxilia and the Making of Provincial
Society from Augustus to the Severans. Oxford: Oxford University Press.
Haynes, I. P., and Wilmott T. (2012). ‘The Maryport Altars: An Archaeological Myth Dispelled’,
Studia Universitatis Babes-​Bolyai, Historia, 57/​1: 25–​37.
Helgeland, J. (1978). ‘Roman Army Religion’, Aufstieg und Niedergang der römischen Welt, II 16/​
2: 1470–​1505.
Hill, J. D. (2001). ‘Romanisation, Gender and Class:  Recent Approaches to Identity in
Britain and their Possible Consequences’, in S. James and M. Millett (eds), Britons and
Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for
British Archaeology, 12–​18.
James, S. T. (1999). ‘The Community of Soldiers: A Major Identity and Centre of Power in the
Roman Empire’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC 98: Proceedings
 Identity and the Military Community in Roman Britain    463

of the Eighth Annual Theoretical Roman Archaeology Conference, Leicester 1998.

Oxford: Oxbow Books, 14–​25.
James, S. T. (2001). ‘Soldiers and Civilians: Identity and Interaction in Roman Britain’, in S.
James and M. Millett (eds), Britons and Romans: Advancing an Archaeological Agenda. CBA
Research Report 125. York: Council for British Archaeology, 77–​89.
James, S. T. (2006). ‘Engendering Change in our Understanding of the Structure of Roman
Military Communities’, Archaeological Dialogues, 13/​1: 31–​36.
James, S. T. (2011). Rome and the Sword: How Warriors and Weapons Shaped Roman History.
London: Thames and Hudson.
Jobey, G. (1979). ‘Housesteads Ware: A Frisian Tradition on Hadrian’s Wall’, Archaeologia
Aeliana, 5th ser. 7: 127–​143.
Mattingly, D. J. (2006). An Imperial Possession:  Britain in the Roman Empire.
London: Allen Lane.
Maxfield, V. A. (1986). ‘Pre-​Flavian Forts and their Garrisons’, Britannia, 17: 59–​72.
Sen, A. (2006). Identity and Violence:  The Illusion of Destiny. London and New  York:  W.
W. Norton & Co.
Swan, V. G. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman
Pottery Studies, 5: 1–​33.
van Driel-​Murray, C. (1995). ‘Gender in Question’, in P. Rush (ed.), Theoretical Roman
Archaeology: Second Conference Proceedings. Avebury, 3–​21.
Wenham, L. P. (1939). ‘Notes on the Garrisoning of Maryport’, Transactions of the Cumberland
and Westmorland Antiquarian and Archaeological Society. 2nd ser. 39: 19–​30.
Wilmott, T. (2010). ‘Birdoswald Roman Cemetery’, Hadrian’s Wall Archaeology, 1: 43–​46.
Worrell, S., and Pearce, J. (2012). ‘Finds Reported under the Portable Antiquities Scheme’,
Britannia, 43: 1–​39.
464

Chapter 23

Roman M ilita ry C u lt u re

Lindsay Allason-​J ones

Introduction

While there is considerable debate as to the number of people who lived in Britain
throughout the Roman occupation, it is clear that a significant proportion of the popu-
lace lived in a military milieu, either as soldiers, dependants, or suppliers. The Claudian
invasion in ad 43 probably involved 40,000 men in four legions plus auxiliary units
(Salway 1981: 75), while the frontier along Hadrian’s Wall in the third century is likely to
have had a military strength of over 15,000 (Breeze and Dobson 2000: 146). Even a con-
servative estimate of army personnel plus associated families and traders has suggested
a military community of 125,000 (Millett 1990: 181–18​6), and at times of military activity
this could have been doubled. Indeed, compared to the other provinces, Britain seems
to have had a remarkably high proportion of military to civilian in all periods. This may
have been due to the problems involved in controlling an island, part of which remained
unconquered, and which was some distance from Rome. It may also have been partly
due to Rome’s conflicting opinions about its army: there was pride in the military
advances that had made the Roman Empire a great global power but there was also an
awareness that few of the serving soldiers, whether in the legions or in the auxiliary
units, were natives of mainland Italy (Forni 1953), and, as a consequence, their loyalties
were uncertain (Cheesman 1914: 7; Haynes 2013: 3). When the units were fighting wars
of conquest, they were to be applauded, but in peacetime they were expensive, difficult
to control, and a constant threat to the stability of the empire. Britain was thus an ideal
place to garrison troops: they would be well placed to suppress rebellion, but, should an
ambitious military leader in the province plan to advance on Rome, his rebellion would
be hindered by the delays in transporting men and supplies across the Channel. This
resulted in some units being garrisoned in Britain for generations, leading to a military
population with very mixed origins, some of whom may never have seen active war ser-
vice. In these circumstances, the difference between military and civilian is likely to have
become blurred, as soldiers married local girls or, having insufficient military activity
 Roman Military Culture    465

to occupy them, engaged in trade or other non-​military activities on their own account.
This blurring is evident in the material culture to be found in the military zone.
The Roman army consisted of legions and auxiliary units. Each legion was made up
of 5,000 infantrymen plus a small cavalry attachment and was divided into 10 cohorts of
480 men. Each cohort was subdivided into 6 centuries of 80 men, with the first cohort a
double unit of 800 men divided into 5 larger centuries. In the early years of the Roman
occupation of Britain, the auxiliary units were either cohorts of infantry or alae of cav-
alry; both could be made up of 1,000 or 500 men, referred to respectively as milliary or
quingenary units (for general discussions of the organization of the Roman army and its
deployment in Britain, see Mattingly 2006: 87–​252; Breeze 2007).
The legions of Roman Britain had their bases at the fortresses of Caerleon, York, and
Chester, while the auxiliary units guarded the frontier. The exact number of auxiliary
units that served in Britain during the Roman period is as yet unknown, but it is thought
that 63 or 64 units were stationed in the province in the mid-​second century ad, the
period for which most records survive. The names of the auxiliary units indicate where
each unit was originally raised but may not reflect the ethnic make-​up of the soldiers
in that unit. A  case in point is Nectovelius, a Brigantian tribesman from Yorkshire,
who is recorded as serving at Mumrills in Scotland in the Second Cohort of Thracians
(RIB 2142), a unit whose origins can be traced to Thrace (southern Bulgaria/​European
Turkey). This was a professional army that boys joined as volunteers between the ages of
17 and 25. An ambitious recruit could, with hard work and a modicum of luck, follow a
well-​trodden career path through the ranks.
This concept of a military career would have been completely alien to the Britons,
who had no standing armies, each tribe preferring to levy troops as and when required.
Salway (1981: 76) has suggested that, ‘since the vast majority of the British troops are
likely to have been farmers, the needs of agriculture must have made lengthy campaigns
almost impossible’. In times of war the key soldiers of the Iron Age were the charioteers,
aristocratic landowners for whom military action would have conferred kudos. This
basic difference in the structure of the Roman and British military forces led to different
tactics in the field. The Roman troops excelled in close-​order fighting, relying on well-​
disciplined, well-​drilled, heavily armoured and armed soldiers; the British preferred the
skirmish, when chariots not only brought the troops to the scene of the battle but also
caused ‘confusion in the enemy ranks by the panic induced by their rushing horses and
hurled missiles’ (Salway 1981: 77).

Armour and Weaponry

The difference in approach to battle tactics is reflected in the armour and weaponry used
by the British tribal warrior and the Roman soldier. The Roman army developed a range
of military equipment, both field equipment and personal weaponry, with each item
being designed for a specific purpose and improved and adapted through the centuries
466

466   Lindsay Allason-Jones

(Bishop and Coulston 2006). Theoretically, it should be possible to tell whether an item
was used by a legionary or an auxiliary, but in practice this is less straightforward than
might be expected (Maxfield 1986; Bishop and Coulston 2006: 255–25​8). Nor are the
weapons as easily dated as we might wish (see, e.g., Marchant 1990), and many items
may have been used for generations. The artefacts used by the cavalry are more straight-
forward as regards the identification of horse equipment, but can still occur in ambigu-
ous spaces, such as road surfaces or in otherwise civilian contexts (Bishop 1988, 2011).
The identification of Roman armour and weapons, and the heated debates as to their
owners, manufacture, and use, can be followed in the volumes of the Journal of Roman
Military Equipment Studies: these studies remind us that this is an area of research where
evidence must be gathered from the whole Roman Empire and not just from a single
province.
A first-​century ad assessment states that ‘the Britons are unprotected by armour.
There are very many cavalry. The cavalry do not use swords nor do the wretched Britons
mount in order to throw javelins’ (Bowman and Thomas 1994; Tab. Vindol. no. 164).
Archaeology confirms that spears, shields of different shapes, swords, and occasion-
ally slings were used, although the different tribes may have had their preferences for
a particular weapon. The weapons found are mostly made to the highest quality and
beautifully decorated, usually in the La Tène style; see, for example, the Witham and
Battersea shields (Jope 1971; Stead 1985). Archaeology further supports the idea that the
Britons did not use complex body armour, as few artefacts that can be confidently iden-
tified as Iron Age armour have been found—​a rare example being the Waterloo helmet
(Aldhouse-​Green 1995: 44). Other helmets found in graves, such as the Hallaton cav-
alry helmet (Sharp and James 2012: 38–​41) is of Roman manufacture and raise questions
as to whether they belonged to a Briton who fought with the Roman invaders or were
acquired by conquest, theft, purchase, or accidental discovery. Their inclusion in ritual
contexts is not unusual; indeed, weapons and armour of Iron Age date are invariably
found in funerary or votive contexts, indicating that such possessions were linked to
their owners’ religious beliefs and feelings of identity and self-​worth and so would not
be discarded lightly (Milne et al. 1997).
Roman weaponry and armour, while important for their owners at the time, do not
seem to have been awarded the same level of respect. The evidence is building that much
armour was destined to be recycling material when it was damaged or was of no fur-
ther practical use to its owner. Parcels of lorica squamata found at Piercebridge, folded
up into neat little parcels for melting down in small crucibles, are an obvious example
(Cool and Mason 2008: D11.40), while the fused mass of military bronzes from Great
Chesters suggests that quite large melting pots were used for this purpose, as well as the
small, egg-​cup-​sized clay crucibles that are found in both military and civilian contexts
(Allason-​Jones 1996: no. 46). Metal analysis also reveals that Roman military fittings of
copper alloy were melted down and made into new pieces in fort and vicus workshops
or even turrets (Allason-​Jones and Dungworth 1997; Dungworth 1997, 1998). The x-​ray
of a block of lorica sections found at Caerleon in 2010 shows how the iron strips were
stripped of their copper alloy fittings in advance of recycling (Chapman 2012).
 Roman Military Culture    467

The discovery of the Corbridge Hoard in 1964 created great excitement among schol-
ars of the Roman military, because it confirmed that the images on Trajan’s Column
depicted Roman legionaries accurately (Daniels 1968; Allason-​Jones and Bishop 1988).
The sections of twelve lorica segmentata cuirasses with their fittings and leather straps
survived well enough to allow H. Russell Robinson to work out for the first time how
lorica segmentata worked, and most reconstructions of legionaries by museums and re-​
enactment groups have used this evidence ever since (Robinson 1975; Bishop 2002). The
iron tools (pickaxe, chisel, saws, shears, crowbar, pulley block, key and iron lamp com-
plete with hanging bracket), plus the wooden tankard, glass gaming set, broken win-
dow glass, melon beads, feathers, fragments of writing tablets, and papyrus seemed to
fit badly with the armour, pilum, forty-​seven spearheads and three catapult bolts; yet the
collection taken as a whole gives a more holistic view of military culture than was appre-
ciated at the time.
All the objects in the Corbridge Hoard were military by association—​association
with the other objects in the box and association from the fact that they were found
on the site of a fort. Ideas as to why the box came to be where it was found and why it
was filled with such an eclectic mix of artefacts ranged from enemy looting, Roman
looting, military tidying before the abandonment of the fort, deliberate hiding
of material in the face of enemy attack, and even the production of raw materials
needed for medical supplies (Daniels 1968; Davies 1970; Allason-​Jones and Bishop
1988). The problem with Corbridge as a provenance is that it was not always a fort
(Bishop and Dore 1988), and the contents of the box could just as easily be a hoard
buried by a civilian scrap-​metal merchant as deliberate tidying-​up by military com-
mand. Would the immediate presence of soldiers be presumed if the hoard had been
found on a villa site, in a town, or on a Romano-​British settlement site, or in those
circumstances would the material be accepted as scrap for recycling or disposal or
regarded as loot?
Cemetery excavations at Petty Knowes, in Northumberland, have shown that the
soldiers from the fort at High Rochester were dressed only in tunics and boots when
cremated, with their grave goods confined to coins for Charon’s fee (Charlton and Day
1984). Throughout the empire it was rare for a Roman soldier to be buried with his
armour and weapons, one of the few exceptions in Britain being at Canterbury, Kent
(Bennett et al. 1982: 185–​190; Bishop 2011: 117); the practice, however, did become more
common in the later period, as Germanic and trans-​Danubian influences began to affect
military burial customs (Bishop 2011: 228). The equipment of retiring soldiers in Britain
appears to have gone back into the general stock or was handed on to specific, favoured
individuals, as can be seen from examples adapted for multiple users; for example, the
helmet cheekpiece from South Shields that was cut down to fit a smaller head (Allason-​
Jones and Miket 1984: no. 3.723) or the helmet from London that has the names of four
individuals inscribed on its neckguard (Robinson 1975: pl. 54). There is, however, docu-
mentary evidence that, at some periods, a soldier’s armour and weaponry were consid-
ered to be personal financial assets, and his heirs received the cash value of his military
equipment on his death (Gilliam 1967).
468

468   Lindsay Allason-Jones

Some items of Roman military equipment have been found in a votive context, but
this is only a tiny proportion of the quantity of armour and weaponry that must have
been utilized over the centuries of Roman occupation. A clear example is the catapult
washer found among the other votive offerings at Bath (Cunliffe 1984:  81), although
the Fulham sword and the Bosham helmet may also be examples (Bishop 2011: 119).
However, the reference by Florus (Epitome II.24) to a general under Augustus depos-
iting captured weaponry in water rather than burning it implies that making a votive
offering of captured equipment was the exception rather than the rule.
In 1989, Mike Bishop summed up the various ways a piece of military equipment
could end up in the archaeological record: concealment, burial rites, dedications, bat-
tles, booty, and chance (further developed in Bishop 2011). To this might be added a
catastrophic event, such as the fire in a barrack block at South Shields (Hodgson 2005),
or abandonment because the item was broken and could not be recycled or reused.
Because of this selectivity in the deposition of military equipment, little has survived.
If, for example, all the objects traditionally identified as military equipment that were
found in the excavations of turrets on Hadrian’s Wall are added together, the tally is
eight spearheads, two shield bosses, one arrowhead, two ballista balls, and three chapes
(Allason-​Jones 1988). This is not a great deal of evidence to suggest that Hadrian’s Wall
in the second century ad was a military frontier. The list of weapons and armour from
the forts along the Wall is equally short, reminding us that a soldier’s life was not made
up solely of battles. Even in peacetime, when there was no enemy action, a serving sol-
dier would have considered himself to be a proper soldier; he would still have received
a soldier’s salary, identified himself with the military, and followed a military routine,
but he would not have spent all his waking hours wearing full armour or carrying all
his weapons. Pieces of armour are usually easily identifiable as being military, so, on the
rare occasions when a body is found in full panoply of lorica segmentata, it is quickly
labelled as being that of a soldier. If a soldier was from one of the auxiliary units of the
later period when the armour is less obvious (Bishop and Coulston 2006) or if he died
when off duty, wearing just a tunic, belt, and sandals, then his military identity could be
more difficult to deduce.

Personal Belongings

Needles, nail-​cleaners, and tweezers are items that are rarely recognized as evidence for
a military presence, yet several examples have been found in contexts that were solely
military (Allason-​Jones 1988). Modern soldiers expect to have to repair their clothes,
keep their nails clean, and remove splinters from their fingers; indeed, in recognition
of this necessity, British recruits during the First and Second World Wars were given a
small sewing kit (known as a ‘housewife’) with the rest of their military issue.
Some artefacts seem specifically designed to confuse archaeologists. For example, it is
impossible to assign a simple stud to military or civilian use unless it is found attached to
 Roman Military Culture    469

an irrefutably military object such as a shield. The evidence of bell-​shaped studs, which
have been found on daggers (as at Carnuntum), on box lock plates, and on doors, is an
interesting case study (Allason-​Jones 1985). If the Romans could use such a distinctive
object for a variety of uses, the attempts of modern archaeologists to categorize such
objects can and will lead to some inaccurate assumptions.
Brooches are equally ambiguous. For many years the archaeological profession
proved remarkably resistant to the notion that soldiers might have worn brooches,
but, once this idea was accepted, great efforts were put into identifying which brooches
were military and which civilian, with several further attempts to assign brooches to
gender, and all have failed. Knee brooches, for example, are invariably referred to as a
military type, yet there are plenty of examples that come from towns and villas, even
a few from female graves (Neal 1974: figure 55, no. 25; Snape 1993: 5.4; Bishop 1996: no.
310). An enamelled plate brooch from Turret 18b has exact parallels from civilian con-
texts at Canterbury and Knaith Park in Lincolnshire, as well as the military contexts of
South Shields and Wallsend forts (Allason-​Jones 1988: no. 18b.1). This may be because
recruits brought brooches into the army from their previous civilian lives or soldiers
took their military brooches into retirement with them, but is more likely to be because
people, whether soldier or civilian, male or female, bought and used the brooches that
most attracted them. Even large crossbow brooches, used on occasions as indicators
of German mercenaries or civil-​service officials, have been found in association with
female jewellery from The Wincle, in Cheshire (Johns et al. 1980; Collins 2010).
The fort of Elginhaugh in Scotland, which was occupied for only a few years between
the mid-​ad 70s and early 90s, produced an extraordinary selection of luxury goods,
including furniture fittings (Hanson 2007). The army was not living in deliberate dis-
comfort even in a frontier zone. The individual soldiers did not know that they were
going to be at Elginhaugh for only a decade and, even if they had done, ten years is a
long time in a person’s life: they would have wanted to make themselves as comfortable
as possible. Even more so the officers, whose postings in Britain may have been merely
an unavoidable stage in a political career. They would bring with them the luxuries they
knew they, personally, could not live without, with the result that the assemblage at
Elginhaugh includes a lead candelabra support in the form of a small boy and a very
large Minerva head couch decoration. One would look for these in vain under military
equipment, yet they form part of the assemblage of a front-​line Roman fort.
A high percentage of the military sites in Britain had even shorter periods of occupa-
tion than Elginhaugh. As the army moved north from the point of invasion, march-
ing camps were built. These could be occupied for a single night or reused by following
troops or on the return journey. Such sites can be found throughout the country, includ-
ing the north of Scotland, although it is not always easy to differentiate between march-
ing camps and construction camps used as bases while permanent forts were being
erected (Welfare and Swan 1995; Jones 2012). Very few artefacts have been found in asso-
ciation with these sites; this may be because of their short duration of occupation or
because few have been comprehensively excavated. Certainly our knowledge of what
a soldier carried on the march does not come from the camps but from literature (e.g.
470

470   Lindsay Allason-Jones

Vegetius, Epitome of Military Science), sculptural evidence (e.g. Trajan’s Column), or

excavations elsewhere in the empire (e.g. Kalkriese: Schlüter and Wiegels 1999).

Daily Life

Once established in a fort, each group of men living in a contubernium were respon-
sible for feeding themselves, using whatever they were issued with, supplemented by
whatever they could acquire for themselves. Stone querns have been found at all the
forts excavated in Britain so far, indicating that soldiers were supplied with grain and
expected to grind it themselves. Vivien Swan has commented that pottery, despite
appearing in large amounts on forts and extramural settlements, would not have trav-
elled well, ‘and only the sturdiest vessels, such as mortaria, could have been carried
into the region [of Hadrian’s Wall] without significant breakages’ (Swan 2008: 49). She
has identified the suite of pots that soldiers would have had in the different centuries
of Roman occupation—​for example, carinated bowls, shallow cooking dishes, and tall,
multi-​purpose jars in the first and second centuries—​and she has identified a number
of military pottery production sites (Swan 2008: 49). Legionaries and auxiliaries who
came from the other provinces of the empire would have brought with them, not only
recipes and tastes in food, but also the pottery styles needed to cook their food in what
they would have considered the proper manner. This is particularly visible in regard to
African troops (Swan 1992) and the Frisians stationed at Housesteads (Jobey 1979); as a
result, no military pottery assemblage is precisely the same as another.
The samian assemblage at Housesteads includes facet-​cut samian from the vicus but
not from the fort, and Brenda Dickinson’s report suggests a different range of potters’
wares was available to the two communities (Rushworth 2009: 488–4​99). The products
of some samian producers are to be found both within the fort and outside it, such as
Cinnamus II, Criciro, Doveccus, and Sacer, but these are in the minority. As Dickinson
has noted, the Central Gaulish wares in both fort and vicus tend to be ‘the work of the
commoner Lezoux potters whose bowls are widespread in Britain’ (Rushworth 2009:
488–​499), while the East Gaulish wares from the vicus reflect more unusual sources,
such as Trier and Rheinzabern.
Comparison of the window glass at Housesteads has shown that more was found
in the vicus excavations than in the fort, suggesting that it was more common for the
civilian buildings to have had glazed windows than the military buildings. The lim-
ited window glass found within the fort may, however, reflect the nature of the build-
ings excavated or a preference for shutters (Allason-​Jones 2013: 76). An assessment of
the vessel glass by Denise Allen also revealed that the pieces from within the fort were
much smaller than those found outside, possibly reflecting their discovery in the ram-
part backing material, which had been derived from domestic rubbish. It is also notice-
able that the vicus produced more unusual vessel glass, such as fragments of drinking
beakers with painted arena scenes as well as some etched products, than has so far been
 Roman Military Culture    471

found within the fort walls (Rushworth 2009). The research on Housesteads has indi-
cated that not only can each fort assemblage be unique; there can also be subtle differ-
ences between a fort and its extramural settlement, not to mention between the various
buildings within that fort.

Epigraphic Evidence

There is considerable evidence for the military community in Britain through inscrip-
tions—​epigraphy being largely a military habit in the provinces—​and artefacts, but the
most informative source is undoubtedly the Vindolanda writing tablets (Bowman and
Thomas 1983, 1994, 2003). Mostly written at the end of the first century ad, a genera-
tion before the northern frontier was established along the line of Hadrian’s Wall, the
tablets provide a picture of a bureaucratic institution in which every aspect of life was
regimented and recorded. Daily reports, for example, followed a set formula, as can be
seen on Tablet 574: ‘15 April. Report of the Ninth Cohort of Batavians. All who should
be are at duty stations, as is the baggage. The optiones and curators made the report.
Arcuittius, optio of the century of Crescens, delivered it’ (Bowman and Thomas 2003:
20–2​1; Tab.Vindol. no. 574). Such reports would have been essential throughout the
empire to ensure efficiency. Occasionally there were glitches, as can be seen in a writing
tablet from Carlisle on which Docilis wrote to his prefect, Augurinus: ‘as you ordered,
we have attached below all the names of lancers who were missing lances, either who did
not have fighting lances, or who (did not have) the smaller subarmales, or who (did not
have) regulation swords’ (Tomlin 1998).
A particularly detailed report confounds the usual presumption that a fort always
contained its full garrison. This report on the First Cohort of Tungrians, dated 18 May,
states that, of the 752 men on the complement, 46 are away guarding the governor and
337 are at Coria (Bowman and Thomas 1994: 90–​98; Tab. Vindol. no. 153). Also listed
are those unavailable for duty because of ill health, with six being ‘sick’, six wounded,
and ten ‘suffering from inflammation of the eyes’. As well as soldiers on duty elsewhere
and those otherwise out of action, there were also soldiers on leave. The handwriting on
the applications for leave (commeatus) indicate that, although they are very formulaic,
the applicants wrote the requests out in full instead of filling in forms with spaces for
names and dates, indicating literacy was not confined to the upper ranks (Bowman and
Thomas 1994: 109–​117).

Elusive Artefacts

The Vindolanda writing tablets, both the ink tablets and the wax tablets, show that in the
Roman army making lists and writing letters were common. There are lists of supplies
472

472   Lindsay Allason-Jones

for day-​to-​day use and for special occasions, and detailed accounts were kept (see, in
particular, Bowman and Thomas 2003; Tab.Vindol. no. 581). Soldiers wrote letters to
officers asking to be recommended for promotion or transfer (Bowman and Thomas
1994; Tab.Vindol. no. 250), to their friends and family giving news or asking for clothing
to supplement their issue (Bowman and Thomas 1994; Tab. Vindol. no. 346), and to each
other to keep in touch. A small writing tablet, which is the size of a modern postcard, is
not adequate, however, for a long report; for more complex correspondence or report
writing papyrus or linen was required. That such reports were written or received in
Britain is confirmed by the discovery at Housesteads of a circular bronze document case
lid, complete with handle (unpublished, but see also Pugsley 2003: 95), as well as the tiny
fragments of papyrus found in the Corbridge Hoard (Allason-​Jones and Bishop 1988).
The Vindolanda tablets not only form part of the material culture of the frontier zone;
they also provide evidence of the artefacts that do not survive in the archaeological
record and can often be forgotten. For example, the tablets themselves required wax; if
we presume a military population on Hadrian’s Wall of around 15,000 and postulate each
soldier had one wax tablet—​accepting that some soldiers may have preferred ink tablets
while others may have owned several wax tablets—​and calculating about 4 ounces of
wax per tablet, then approximately 350 pounds of wax would have been needed for the
writing tablets alone at any one time. Add to this the use of candles, for which there are
no figures relating to size or numbers used and only the evidence of a few candlesticks
to confirm that there were candles, as well as the use of wax for sealing vessels and for
medicine, then we must presume that a considerable supply of wax was required to keep
any one Roman fort functioning, and that somewhere in the vicinity there were consid-
erable numbers of beehives (Allason-​Jones 2008).
More candles, however, are likely to have been made of tallow—​the rendered fat of
sheep or oxen. Excavations of military sites produce many animal bones; these ani-
mals would have been primarily a source of food and leather as well as providing the
raw material for bone objects, such as handles and combs. However, in the past people
used every bit of an animal—​nothing was wasted—​and tallow, fat, sinew, blood, and so
on were all important commodities. Lists from Vindolanda refer to the buying of tal-
low (Bowman and Thomas 1994; Tab. Vindol. no. 184) as well as sinew (Bowman and
Thomas 1994; Tab. Vindol. no. 343), and barrels and baskets would have been needed to
transport these tradable goods.

Civilians

Many of the items listed in the Vindolanda tablets, such as curtains, cushions, furniture,
dog collars, cart axles, and hunting nets, would also have been used by the communities
that lived outside the forts. The population of these extramural settlements would have
been made up of retired soldiers, traders and merchants, and the families of the soldiers.
Since the 1980s the concept that there were women present within and around forts has
 Roman Military Culture    473

become widely accepted, although this agreement has not been arrived at without a
struggle (Campbell 1978, 2010; Saller and Shaw 1984; Roxan 1991; Allason-​Jones 1995,
1998, 1999, 2003, 2005, 2012, forthcoming; van Driel-Murray 1995; Phang 2001; Allison
2006a, b, 2008, 2009). These women were not just wives, daughters, or general ‘campfol-
lowers’; tombstone evidence indicates that some soldiers had their widowed mothers
with them. At Risingham, for example, one of the outpost forts north of Hadrian’s Wall
that otherwise has very limited evidence for a civilian population, Dionysius Fortunatus
buried his mother Aurelia Lupula (RIB 1250). At York, Emilia Theodora outlived her
son Valerius Theodorianus, of Nomentum, who lived 35 years, 6 months (RIB 677).
There were also the sisters of soldiers in the military zone: Vacia, for example, buried her
brother Aelius Mercurialis, a cornicularius, at the frontier fort of Great Chesters (RIB
1742). Mothers-​in-​law (RIB 594) and even nieces (RIB 1830) were included in these com-
munities as well as slaves of both sexes (Speidel 1989; Justinian, Digest 49.15.6)
As the take-​home pay of a soldier, whether a legionary or an auxiliary, was not lavish
(Brickstock 2010, 2011), the female relations of a soldier would have had to work in order
to supplement the family’s income, particularly if they had to pay rent for their accom-
modation. In a vicus there will have been a number of jobs in the service industries,
providing meals, doing laundry, and so on, although there is little concrete evidence
for who was performing these tasks. There would also have been opportunities in trad-
ing, industry, and agriculture. We might assume, therefore, that women in a military
milieu would have had some spare cash to pay for artefacts to adorn themselves and
their homes. As a result, there is little to differentiate the material culture of these fami-
lies from those who lived in towns, on farms, or in villas.

Conclusions

The overall picture of the military areas of Britain is one of constant movement, as gar-
risons were changed, units were seconded to other duties, supplies were transported, and
individual soldiers went on leave. Instead of the traditional view of regimented units of
soldiers, all with regulation kit, constantly battling the enemy across the frontier, the arte-
facts offer a different image of a community that, while obsessed with bureaucracy, had
its fair share of problems, with soldiers arguing (Bowman and Thomas 1994; Tab. Vindol.
nos. 256; 321), conducting business on their own account as well as officially (Bowman and
Thomas 1994; Tab. Vindol. no. 310), owing each other money (Bowman and Thomas 1994;
Tab. Vindol. no. 312), committing felonies and misdemeanours (Bowman and Thomas
1994; Tab. Vindol. nos 655; 656), and complaining of unjust punishment (Bowman and
Thomas 1994; Tab. Vindol. no. 344). Soldiers lost and damaged their equipment, and
repairs can often be seen, some more efficient than others. There were people from all
over the known world living in the communities within and around the forts, and they all
had individual ideas of what material possessions they needed and a range of methods for
acquiring them. The mixed nature of these communities does not support the traditional
474

474   Lindsay Allason-Jones

view that it is possible to differentiate between military and civilian or male and female.
What is clear, however, is that each fort was different, and each fort assemblage should be
approached with the expectation that nothing will be as expected.

Ancient Sources
Florus, Epitome of Roman History, trans. E. S. Foster. Loeb Classical Library 231. Cambridge,
MA: Harvard University Press, 1929.
Justinian, Digesta seu Pandectae, ed. T. Mommsen [ed. A. Wilson]. Philadelphia,
Pennsylvania: University of Pennsylvania Press, 1870.
Vegetius, Epitome of Military Science, trans. N. P. Milner. 2nd edn. Liverpool:  Liverpool
University Press, 1996.

References
Aldhouse-​Green, M. J. (1995) (ed.). The Celtic World. London: Routledge.
Allason-​Jones, L. (1985). ‘Bell-​Shaped Studs?’, in M. C. Bishop (ed.), The Production and
Distribution of Roman Military Equipment:  Proceedings of the Second Roman Military
Equipment Research Seminar. BAR International Series 275. Oxford: Archaeopress, 95–​105.
Allason-​Jones, L. (1988). ‘Small Finds from Turrets on Hadrian’s Wall’, in J. C. Coulston (ed.),
Military Equipment and the Identity of Roman Soldiers: Proceedings of the Fourth Roman
Military Equipment Conference. BAR British Series 394. Oxford: Archaeopress, 197–​233.
Allason-​Jones, L. (1995). ‘“Sexing” Small Finds’, in P. Rush (ed.), Theoretical Roman Archaeology:
Second Conference Proceedings. Aldershot: Avebury, 22–​32.
Allason-​Jones, L. (1996). ‘Roman Military and Domestic Artefacts from Great Chesters’,
Archaeologia Aeliana, 5th ser., 24: 187–​214.
Allason-​Jones, L. (1998). ‘The Women of Roman Maryport’, in R. J. A. Wilson (ed.), Roman
Maryport and its Setting: Essays in Memory of Michael G. Jarrett. Cumberland and
Westermorland Antiquarian and Archaeological Society Extra Series 28. Maryport:
Cumberland and Westmorland Antiquarian and Archaeological Society, 105–​111.
Allason-​Jones, L. (1999). ‘Women and the Roman Army in Britain’, in A. Galsworthy and
I. Haynes (eds), The Roman Army as a Community. Journal of Roman Archaeology,
Supplementary Series 34. Portsmouth, RI: Journal of Roman Archaeology, 41–​51.
Allason-​Jones, L. (2003). ‘The Family in Roman Britain’, in M. Todd (ed.), A Companion to
Roman Britain. Oxford: Blackwell, 273–​287.
Allason-​Jones, L. (2005). Women in Roman Britain. 2nd edn. London: British Museum
Publications.
Allason-​Jones, L. (2008). ‘Finding Significance in the Finds’, in P. Bidwell (ed.), Understanding
Hadrian’s Wall. South Shields: Arbeia Society, 41–​47.
Allason-​Jones, L. (2012). ‘Women in Roman Britain’, in S. Dillon and S. L. James (eds),
Companion for Women in the Ancient World. Oxford: Wiley-​Blackwell, 467–​77.
Allason-​Jones, L. (2013). ‘Housesteads vicus: A Case Study in Material Culture and Roman
Life’, in R. Collins and M. F. A. Symonds (eds), Breaking down Boundaries: Hadrian’s Wall
in the 21st Century: Journal of Roman Archaeology, Supplementary Series 93. Portsmouth,
RI: Journal of Roman Archaeology, 71–​84.
 Roman Military Culture    475

Allason-​Jones, L. (forthcoming). ‘Women, the Military and patria potestas in Roman Britain’,
in P. Bidwell (ed.), Proceedings of the Newcastle Limes Congress of Frontier Studies 2011.
Allason-​Jones, L., and Bishop, M. C. (1988). Excavations at Roman Corbridge:  The Hoard.
English Heritage Archaeological Report 7. London: English Heritage.
Allason-​Jones, L., and Dungworth, D. B. (1997). ‘Metalworking on Hadrian’s Wall’, in W.
Groenman-​van Waateringe, B. L. van Beek, W. J. H. Willems, and S. L. Wynia (eds), Roman
Frontier Studies 1995: Proceedings of the XVIth International Congress of Roman Frontier
Studies. Oxbow Monograph 91. Oxford: Oxbow Books, 317–​321.
Allason-​Jones, L., and Miket, R. F. (1984). The Catalogue of Small Finds from South Shields
Roman Fort. Newcastle upon Tyne: Society of Antiquaries of Newcastle upon Tyne.
Allison, P. (2006a). ‘Artefact Distribution within the Auxiliary Fort at Ellingen:  Evidence
for Building Use and the Presence of Women and Children’, Bericht den Römischen
Germanischen Kommission, 87: 387–​452.
Allison, P. (2006b). ‘Mapping for Gender:  Interpreting Artefact Distribution in Roman
Military Forts in Germany’, Archaeological Dialogues, 13/​1: 1–​4.
Allison, P. (2008). ‘The Women and Children inside 1st and 2nd-​Century Forts: Comparing
the Archaeological Evidence’, in U. Brandl (ed.), Frauen und Römisches Militär; Beiträge
eines Runden Tisches in Xanten rom 7, bis 9, Juli 2005. BAR International Series 1759. Oxford:
Archaeopress, 120–​139.
Allison, P. (2009). ‘The Women in Early Forts: GIS and Artefact Distribution Analyses in 1st
and 2nd Century Germany’, in A. Morillo, N. Havel, and E. Martin (eds), Limes XX: Estudios
sobre la Frontera Romana. Madrid: Polifemo, 1193–​1201.
Bennett, P., Frere, S. S., and Stow, S. (1982). Excavations at Canterbury 1. Maidstone:  Kent
Archaeological Society.
Bishop, M. C. (1988). ‘Cavalry Equipment of the Roman Army in the 1st Century ad’, in J. C.
N. Coulston (ed.), Military Equipment and the Identity of Roman Soldiers: Proceedings of
the Fourth Roman Military Equipment Conference. BAR International Series 394. Oxford:
Archaeopress, 67–​79.
Bishop, M. C. (1989). ‘O Fortuna: A Sideways Look at the Archaeological Record and Roman
Military Equipment’, in C. van Driel-​Murray (ed), Roman Military Equipment: The Sources of
Evidence. Proceedings of the Fifth Roman Military Equipment Conference. BAR International
Series 476. Oxford: Archaeopress, 1–​11.
Bishop, M. C. (1996). Finds from Roman Aldborough. Oxbow Monograph 65. Oxford: Oxbow Books.
Bishop, M. C. (2002). Lorica Segmentata I: A Handbook of Articulated Roman Plate Armour.
Journal of Roman Military Equipment Studies Monograph 1. Chirnside: Armatura.
Bishop, M. C. (2011). ‘Weaponry and Military Equipment’, in L. Allason-​Jones (ed.), Artefacts
in Roman Britain. Cambridge: Cambridge University Press, 114–​132.
Bishop, M. C., and Coulston, J. C. (2006). Roman Military Equipment from the Punic Wars to
the Fall of Rome. 2nd edn. Oxford: Oxbow Books.
Bishop, M. C., and Dore, J. N. (1988). Corbridge: Excavations of the Roman Fort and Town,
1947–​80. London: Historic Buildings and Monuments Commission.
Bowman, A. K., and Thomas, J. D. (1983). Vindolanda:  The Latin Writing Tablets. Tabulae
Vindolandenses I. London: British Museum Press.
Bowman, A. K., and Thomas, J. D. (1994). Vindolanda:  The Latin Writing Tablets. Tabulae
Vindolandenses II. London: British Museum Press.
Bowman, A. K., and Thomas, J. D. (2003). Vindolanda:  The Latin Writing Tablets. Tabulae
Vindolandenses III. London: British Museum Press.
476

476   Lindsay Allason-Jones

Breeze, D. J. (2007). Roman Frontiers in Britain. London: Bristol Classical Press.

Breeze, D. J., and Dobson, B. (2000). Hadrian’s Wall. 4th edn. London: Penguin.
Brickstock, R. J. (2010). ‘Coins and the Frontier Troops in the 4th Century’, in R. Collins and L.
Allason-​Jones (eds), Finds from the Frontier. York: Council for British Archaeology Research
Report 162, 86–​91.
Brickstock, R. J. (2011). ‘Commerce’, in L. Allason-​Jones (ed.), Artefacts from Roman Britain.
Cambridge: Cambridge University Press, 20–​45.
Campbell, B. (1978). ‘The Marriage of Soldiers under the Empire’, Journal of Roman Studies, 61:
153–​166.
Campbell, D. B. (2010). ‘Women in Roman Forts: Residents, Visitors or Barred from Entry?’,
Ancient Warfare, 4/​6: 48–​53.
Chapman, E. (2012). ‘Taking Stock of the Block: Investigating the Armour Found during the
2010 Excavations at Caerleon’, Lucerna, 43 (July): 5.
Charlton, D. B., and Day, J. (1984). ‘The Roman Cemetery at Petty Knowes, Rochester,
Northumberland’, Archaeologia Aeliana, 5th ser., 12: 1–​31.
Cheesman, G. (1914). The Auxilia of the Roman Imperial Army. Oxford: Clarendon Press.
Collins, R. (2010). ‘Brooch Use in the 4th–​5th Century’, in R. Collins and L. Allason-​
Jones (eds), Finds from the Frontier. CBA Research Report 162. York: Council for British
Archaeology, 86–​91.
Cool, H. E. M., and Mason, D. J. (2008). Roman Piercebridge: Excavations by D. W. Harding and
Peter Scott 1969–​1981. Durham: Architectural and Archaeological Society of Durham and
Northumberland.
Cunliffe, B. (1984). Roman Bath Discovered. 2nd edn. London: Routledge and Kegan Paul.
Daniels, C. M. (1968). ‘A Hoard of Iron and Other Materials from Corbridge’, Archaeologia
Aeliana. 4th ser., 46: 115–1​26.
Davies, R. W. (1970). ‘A Note of the Hoard of Roman Equipment Buried at Corbridge’, Durham
University Journal, 31, 2/​1970: 177–1​80.
Dungworth, D. D. B. (1997). ‘Roman Copper Alloys: Analysis of Artefacts from Northern
Britain’, Journal of Archaeological Science, 24: 901–​910.
Dungworth, D. (1998). ‘EXRDF Analysis of Copper-​ Alloy Samples’, in H. E.  M. Cool
and C. Philo (eds), Roman Castleford Excavations 1974–​85 Volume I:  The Small Finds.
Wakefield: West Yorkshire Archaeology Service, 117.
Forni, G. (1953). Il reclutamento delle legioni da Augusto a Dioleziano. Milan and
Rome: Fratelli Bocca.
Gilliam, J. F. (1967). ‘The deposita of an Auxiliary Soldier (P. Columbia inv. 32)’, Bonner
Jahrbucher, 67: 5–​9.
Hanson, W. S. (2007). Elginhaugh: A Flavian Fort and its Annexe. Britannia Monograph Series
23. London: Britannia.
Haynes, I. (2013). Blood of the Provinces. Oxford: Oxford University Press.
Hodgson, N. (2005). ‘Destruction by The Enemy? Military Equipment and the Interpretation
of a Late-​Third-​Century Fire at South Shields’, Archäologie de Schlachtfelder—​Militaria aus
Zerstörunghorizonten. Akten der 14 Internationalen Roman Military Equipment Conference.
Carnuntum y Jahrbuch, 207–​216.
Jobey, I. (1979). ‘Housesteads Ware: A Frisian Tradition on Hadrian’s Wall’, Archaeologia.
Aeliana, 5th ser., 7: 127–​143.
Johns, C., Thompson, H., and Wagstaff, P. (1980). ‘The Wincle, Cheshire, Hoard of Roman
Gold Jewellery’, Antiquaries Journal, 60: 48–​58.
 Roman Military Culture    477

Jones, R. H. (2012). Roman Camps in Britain. Stroud: Amberley.

Jope, E. M. (1971). ‘The Witham Shield’, British Museum Quarterly, 35/​1–​4: 61–6​9.
Marchant, D. (1990). ‘Roman Weapons in Great Britain: A Case Study: Spearheads, Problems
in Dating and Typology’, Journal of Roman Military Equipment Studies, 1: 1–​6.
Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409.
London: Penguin.
Maxfield, V. (1986). ‘Pre-​Flavian Forts and their Garrisons’, Britannia, 17: 59–​72.
Millett, M. (1990). The Romanization of Britain. Cambridge: Cambridge University Press.
Milne, G., Bates, M., and Webber, M. D. (1997). ‘Problems, Potential and Partial Solutions: An
Archaeological Study of the Tidal Thames, England’, World Archaeology, 29/​1: 130–1​46.
Neal, D. S. (1974). The Excavation of the Roman Villa in Gadebridge Park, Hemel Hempstead,
1963–​8. Society of Antiquaries of London Research Report 31. London:  Society of
Antiquaries of London.
Phang, S. E. (2001). The Marriage of Roman Soldiers (13 bc–​ad 235). Boston: Brill.
Pugsley, P. (2003). Roman Domestic Wood. BAR International Series 1118. Oxford: Archaeopress.
Robinson, H. R. (1975). Armour of Imperial Rome. London: Arms and Armour Press.
Roxan, M. M. (1991). ‘Women on the Frontiers’, in V. A. Maxfield and M. J. Dobson (eds),
Roman Frontier Studies 1989. Exeter: University of Exeter Press, 462.
Rushworth, A. (2009). Housesteads Roman Fort—​ the Grandest Station:  Excavation and
Survey at Housesteads, 1954–​95, by Charles Daniels, John Gillam, James Crow and Others.
London: English Heritage.
Saller, R. P., and Shaw, B. D. (1984). ‘Tombstones and Roman Family Relations in the Principate:
Civilians, Soldiers and Slaves’, Journal of Roman Studies, 74: 124–​155.
Salway, P. (1981). Roman Britain. Oxford: Clarendon Press.
Schlüter, W., and Wiegels R. (1999) (eds). Rom, Germanien und die Ausgrabungen von Kalkriese.
Osnabrücker Forschungen zu Altertum und Antike-​Rezeption I. Bramsche.
Sharp, H., and James, S. (2012). ‘Reconstructing the Hallaton Helmet’, Current Archaeology,
264: 38–​41.
Snape, M. E. (1993). Roman Brooches from North Britain. BAR British Series 235.
Oxford: Archaeopress.
Speidel, M. (1989). ‘The Soldiers’ Servants’, Ancient Society, 20: 238–2​47.
Stead, I. M. (1985). The Battersea Shield. London: British Museum Publications.
Swan, V. G. (1992). ‘Legion VI and its Men: African Legionaries in Britain’, Journal of Pottery
Studies, 5: 1–​33.
Swan, V. G. (2008). ‘Builders, Suppliers and Supplies in the Tyne–​Solway Region and Beyond’,
in P. Bidwell (ed.), Understanding Hadrian’s Wall. South Shields: Arbeia Society, 49–​82.
Tomlin, R. S. O. (1998). ‘Roman Manuscripts from Carlisle: The Ink-​Written Tablets’, Britannia,
34: 41–​51.
van Driel-Murray, C. (1995). ‘Gender in Question’, in P. Rush (ed.), Theoretical Roman
Archaeology Conference Proceedings. Aldershot: Avebury, 3–​21.
Welfare, H., and Swan, V. G. (1995). Roman Camps in England:  The Field Archaeology.
London: Royal Commission on Historical Monuments.
478
 PA RT  I I I

F OR M S OF
K N OW L E D G E
480
 Chapter 24

Changing Mat e ria l i t i e s

Andrew Gardner

Introduction: The Material Impact

of Roman Imperialism

Roman Britain stands out. Compared with much of the archaeology of preceding and
immediately subsequent periods, many of the material characteristics of the Roman
period are strikingly obvious, from the architecture of forts and towns to the abundant
pottery, coinage, and other finds common to sites of many types (Esmonde Cleary 2009:
198). That Britain’s incorporation into the Roman world saw at once the emergence of
a new material world can hardly be denied. While many of the popular forms of arte-
facts in the Roman period had their roots in Iron Age traditions, there were changes to
every sphere of material life and major innovations in technologies of all kinds, from
those to do with representation and communication to exchange and dining. Of course,
change was far from even in time or spatial distribution, and, alongside the continuity
of pre-​conquest materials and practices, this has posed the central question of Roman
archaeology in Britain: what effect did the Roman Empire have upon British societies?
In spite of well over a century of scholarly research into this question, we are, in many
ways, only beginning to sketch out an adequate answer. Indeed, the very material visibil-
ity of the Roman period is a significant factor in the slow development of interpretative
sophistication within the field, burdening scholars with both a considerable weight of
organizational work in bringing order to large assemblages of finds, and a beguiling nar-
rative of cultural change simply in terms of the apparently obvious appeal of the material
trappings of advanced civilization. The latter perspective of course seemed self-​evident
to scholars in the modern, imperial, and materialist society of the nineteenth and earlier
twentieth centuries (e.g. Frere 1988: 36; cf. Scott 1993a: 7–​8; Hingley 2000: 130–​149). It is
ironic, though, that such a simplistic interpretation of a great wealth of material culture
was also strongly reinforced by readings of a handful of objects of a very particular kind,
in the form of the writings of ancient authors (Laurence 2001: 92–​95; cf. Andrén 1998:
482

482   Andrew Gardner

9–​25; Storey 1999). Only relatively recently has the material culture of Roman Britain
begun to be understood both in sufficient detail, and with more sophisticated theoretical
approaches, to enable interpretations of it to break free from these long-​standing con-
straints. In this chapter, my aim is to show that there is now a range of ways of exam-
ining the material transformations and continuities of Britain before, during, and after
the Roman period, all of which open up new possibilities in answering the key question
of the impact of Roman imperialism upon Britain (and, indeed, of the role of Britain in
the shaping of Roman imperialism). Before looking at these, however, it is necessary to
review in a little more detail the character of the dominant answer to that question up
until the 1980s.
The historiography of the concept of ‘Romanization’ is now a fairly well-​worn topic
(e.g. Hingley 1996; Webster 2001: 209–​217; Mattingly 2011: 3–42; cf. Millett, this volume),
so what I wish to emphasize here is the way in which material culture has fitted into this
evolving paradigm. Indeed, it is rather neatly illustrative of how material things can be
interpreted in many ways to note that ‘Romanization’ emerged as a concept precisely
to deal with the abundance of provincial archaeology that was emerging in countries
like Germany and Britain in the later nineteenth century. The idea that such material—​
humble objects, lacking inscriptions—​could make a contribution to the historical study
of the Roman Empire was, in its way, revolutionary (Freeman 2007: 604; Hingley 2008:
434–43​7). Francis Haverfield and his mentor Theodor Mommsen thus opened the door
to the material study of the Roman Empire, even if for them, and many of their follow-
ers, interpretation of the objects was still constrained by the textual record, and was also
firmly linked to an essentialist view of culture. In this respect it is a simple matter to
align Roman archaeology during this period with other examples of cultural-​historical
research, this approach being common in both prehistory and various old world his-
torical archaeologies through the later nineteenth and much of the twentieth centuries
(Jones 1997: 29–​39; Trigger 2006: 216). Indeed, Haverfield himself remarks on the con-
temporary understanding of the connection between artefacts and culture at the end
of his survey of material evidence in The Romanization of Roman Britain: ‘even among
the civilized nations of the present age the recent growth of stronger national feelings
has been accompanied by a preference for home-​products and home-​manufactures and
a distaste for foreign surroundings’ (Haverfield 1912: 23). Such an approach, in which
objects were understood as fairly straightforwardly diagnostic of the norms to which
people living in Roman or native cultural realms subscribed, with anything in-​between
being aspirational of shifting affiliation from the latter to the former (Freeman 1993:
442–44​4; Jones 1997: 33–​36), was to be very persistent in Romano-​British archaeology.
Yet we must be aware that even within the ‘long sleep’ of Roman archaeology (to bor-
row a phrase from Renfrew (1982: 6–​7)), and its dreams of ‘Romanization’, there were
subtle variations in the understanding of artefacts and their role in narratives about the
Roman period. Haverfield’s principal direct successor—​thanks to the slaughter of the
First World War—​was R. G. Collingwood, who gained greater recognition as a philos-
opher than a Roman archaeologist (Freeman 2007: 536–​537). In his work in the latter
field, though, he understood that the use of objects of a certain type was not necessarily
 Changing Materialities   483

indicative of cultural norms or aspirations—​that is, that the preference for a new type of
object may be functional without being normative (Collingwood and Myres 1937: 222;
Hingley 2000: 134). He also placed an emphasis on material culture as independent of
textual history; his ‘handbook’ on The Archaeology of Roman Britain (1930) is organized
by site types and material categories, not chronology (cf. Reece 1997: 477; Freeman 2007:
604). While the later edition of this book was revised by Ian Richmond according to the
same logic (Collingwood and Richmond 1969), Richmond’s own Roman Britain (1955)
moved to a more synthetic style, which placed military history foremost in the story of
Roman Britain. Sheppard Frere’s Britannia (1987; the first edition was published in 1967)
continued in this vein, with over half of the book framed by narrative (military) his-
tory. His understanding of material culture also recapitulated the normative approach,
common also still in later Iron Age archaeology of the time (Hingley 2000: 140–​142; cf.,
e.g., Hodson 1964). Such a view can be found in most textbooks of the 1960s through
to the 1980s and even beyond (e.g. Scullard 1979; Salway 1993; there were still excep-
tions—​for example, a somewhat different approach is taken in Liversidge (1968), explic-
itly eschewing military–​historical narrative). This lengthy period of limited theoretical
development created the rather stifling environment in Roman archaeology in Britain
described by Eleanor Scott (1993b) at the beginning of the 1990s.
Indeed, it is precisely at this point—​with the publication of two or three key studies,
and the organization of the first Theoretical Roman Archaeology Conference (TRAC)—​
that we reach the pivotal period for a more decisive shift in the archaeology of Roman
Britain, bringing new perspectives on material culture to the fore. Although not par-
ticularly engaged with the contemporary discourses in mainstream archaeological
theory, Richard Reece’s teaching at the Institute of Archaeology, with its firm emphasis
on liberating material culture from text-​derived categories (an approach summarized
in the self-​published Reece (1988); cf. Reece 1993a; James 2003: 180; Woolf 2004: 419;
Millett, this volume), was instrumental in the emergence of a generation of scholars of
Roman Britain who would seek out different frameworks. One such was Martin Millett,
whose The Romanization of Britain (1990) acted as a catalyst for further change. Millett’s
approach, informed by selected aspects of processual and early post-​processual archae-
ology, as well as by some of the debates about Roman society in ancient history, was cer-
tainly novel in its explicit shift away from an event-​based narrative of Romano-​British
history into which objects were simply slotted, to a much more analytical investigation
(cf. Woolf 1991; Hingley 2001: 112). This book built upon some of the methodological
innovations that had been taking place in the 1970s and 1980s in the study of aggre-
gate data sets and spatial patterning (e.g. Hodder and Hassall 1971; Fulford and Hodder
1975), and thus began to give artefacts their appropriate weight in a synthetic account
for the first time. Yet, and in spite of efforts to consider different approaches to the sig-
nificance of artefacts, Millett’s acculturative interpretation of the significance of mate-
rial change still had at its heart a normative view of culture (Freeman 1993: 442–​445;
Jones 1997: 34–3​6). Nonetheless, the impact of Millett’s work was enormous, and the
debate provoked by it gave huge momentum to the initiation of new approaches to
Romano-​British material, particularly as played out at further TRACs. Another key
484

484   Andrew Gardner

development of this period was the publication of Richard Hingley’s Rural Settlement
in Roman Britain (1989), which turned a spotlight upon non-​villa settlement for really
the first time. This book helped cement the shift in emphasis away from a history of
Roman Britain dominated by the military narrative, a trend reinforced by the emerging
results from commercial archaeological fieldwork under the new planning framework
beginning at around the same time (Planning Policy Guidance [PPG16] began to oper-
ate in 1990, and prior to its replacement in 2010 led to many new discoveries of rural
settlement sites (Fulford and Holbrook 2011)) (see also chapters by Hingley, Millett and
Wilson, this volume).
Since the beginning of the 1990s, therefore, the study of material culture in
Romano-​British archaeology has broadened considerably—​in terms of both the range
of sites and objects considered worthy of investigation, and the theoretical and meth-
odological tools applied in doing so. Yet progress has been uneven, perhaps inevitably
given both the compressed timescale within which new ideas have been introduced
to the field (compared with mainstream archaeological theory and with other disci-
plines like anthropology) and the embedded conservatism of Roman archaeology as a
whole (evident in some reviews of the more innovative publications—​e.g. C. Thomas
1990; de la Bédoyère 2007; cf. Laurence 1999; Woolf 2004: 420). Thus, we have seen
considerable discussion of more flexible models of identity than that dictated by the
‘Romanization’ framework, in many cases influenced by postcolonial theories of the
hybrid mixing of cultures in contact situations (e.g. Webster 2001; Mattingly 2004),
and in other cases by sociological theories of identity construction (e.g. Gardner
2002; Revell 2009). At the same time, for some, ‘identity’ has become little more than
a trendy synonym for ‘Romanization’ and lacks some of the critical edge necessary in
application (Pitts 2007; Faulkner 2008; cf. Gardner 2013). Similarly, a recent move to
consider the movement of material culture around the empire, and the interaction
of ‘global’ and local cultures, through the lens of globalization (e.g. Hitchner 2008;
Pitts 2008) can fall foul of the same critique (e.g. Naerebout 2007). Other approaches,
more squarely focused around objects, have been eclectic. The most dominant strand
encompasses more and more detailed studies of the distribution of artefacts in terms of
their contexts of use, sometimes influenced by practice theory or theories of consump-
tion (e.g. Gardner 2007; Eckardt and Crummy 2008; Greene 2008; Pitts 2010; papers
in Hingley and Willis 2007; Lavan et al. 2007; Allason-​Jones 2011). However, semi-
otic approaches to meaning have not been so widely deployed (except with respect
to architecture (Grahame 2000)), while evolutionary approaches to material culture
transmission, phenomenological perspectives on human engagement with things, or
post-​humanist approaches to the agency of objects have scarcely been touched upon
(rare examples of each are González-​Ruibal 2003; Gosden 2005; Biddulph 2012). More
will be said on the potential of these kinds of approaches in the next section, but suffice
it to say here, in closing this Introduction, that, while Roman archaeology in Britain
has already moved far beyond the stultifying torpor of the ‘Romanization’ years, there
is still some way to go before we contribute actively to debates about materiality taking
place across the social sciences.
 Changing Materialities   485

Social Beings in a Material

World: From Function and Meaning
to Action and Practice

‘Materiality’ is indeed a growing area of interest in a number of disciplines, with scholars

in sociology and cultural studies (see, among others, Dant 2005; Woodward 2007) pick-
ing up particularly on developments in anthropology that have been underway for some
time (e.g. Miller 1994). Much of the discussion in this area also draws upon influential
philosophers and theorists such as Martin Heidegger, Bruno Latour, Marcel Mauss, and
Alfred Gell. The role of archaeology in this movement is ambiguous. On the one hand,
reference in the wider materiality literature to archaeological studies can be patchy (e.g.
the index to Bennett and Joyce (2010) has multiple references to anthropology, but none
to archaeology), even as archaeologists draw upon this literature themselves. On the
other hand, many of the influential members of the Material Culture group within the
Department of Anthropology at UCL (such as Victor Buchli, Daniel Miller, and Chris
Tilley) have connections with archaeology, and there is no doubt that archaeology has
a great wealth of thinking about material culture to contribute to the wider debate (cf.
Knappett 2012; Silva and Baert, forthcoming). Roman archaeology, too, potentially has
a significant part to play in this process (cf. Hodder 1993: pp. xvii–​xix; James 2003), pro-
viding it can both continue to broaden its theoretical frame of reference, and synthesize
the complexity of Roman material cultures for interdisciplinary audiences. In this sec-
tion, I will address the first of these points, highlighting some of the developments in the
theorization of materiality in archaeology and beyond, and noting where these have had
an impact on the interpretation of Romano-​British archaeology and, more importantly,
where they have not. In the following sections I will make a tentative start on the second
task, in the light of some of the questions raised by considering different perspectives.
The delayed integration of aspects of mainstream archaeological theory into Roman
studies might have had the positive effect of allowing practitioners to avoid previous
dead ends and more readily settle on relevant approaches (as explicitly argued by Scott
1993a: 7). To some extent, this has been true of the adoption of postcolonial theory,
which has had a distinctive impact in our field. However, partly because theory-​building
in archaeology is rarely straightforwardly cumulative (Trigger 2006: 5–​17, 38–​39), this
phenomenon has also exacerbated a lack of coherence in approaches to material culture.
This is most apparent with respect to the limited impact of processual—​and indeed early
post-​processual—​approaches in Roman archaeology. The New Archaeology’s empha-
sis on quantification of assemblages, pattern analysis, functional categorization of arte-
facts, and site formation process has, of course, filtered through to Roman archaeology
in various ways and over quite a prolonged period of time (e.g. Hodder and Hassall 1971;
Peacock 1982; Peña 2007). However, this influence has remained largely methodological
throughout, and has not generated the extensive theoretical debate that opposed such
486

486   Andrew Gardner

approaches to cultural-​historical archaeology and that characterized prehistoric stud-

ies in the 1960s (cf. Dyson 1989: 146). When this level of debate did open up in Roman
archaeology from the end of the 1980s, the approaches drawn into it derived largely
from later post-​processualism and postcolonial theory. This means that the rejection
of ‘Romanization’ has perhaps most strongly come from a historiographical angle, sec-
ondarily from a notion that material culture has multiple meanings, and only tangen-
tially from any substantial reanalysis of the role that material culture plays in human
life. While the ‘New’ or processual archaeologists of the 1960s and 1970s certainly did
not have all of the right ideas about this role, they at least grappled with the question,
perhaps most consistently and coherently in the ‘behavioural archaeology’ of Michael
Schiffer and colleagues. This approach has moved from the consideration of issues of
site formation and the nature of the material record to a more complete system of under-
standing human–​material relations (Skibo and Schiffer 2008; LaMotta 2012; cf. Dant
2005: 6–​8; Preucel and Mrozowski 2010: 5). Like another consistent thread emerging
from the processual era—​evolutionary approaches that see material culture as adaptive
in various ways (see, e.g., Shennan 2002)—​behavioural archaeology has had precious
little impact in Roman archaeology. The effects of the more methodological aspects
of the processual programme continue to be felt, particularly in quantitative studies
of finds (a major driver of which has surely been the processual orientation of much
commercial field practice; cf. Tyers 1996: 22–​23; Evans 2001: 27; Lucas 2001: 2), but the
omission of important aspects of the debate over the more fundamental significance of
material culture is unfortunate.
The same can be said of the next phase in the development of that debate, revolving
around approaches to style, signification, and ideology popular in the later 1970s and
early 1980s in prehistoric archaeology. As the emphasis of understandings of artefacts
shifted from their adaptive or behavioural function to their symbolic meaning dur-
ing this period—​and as some post-​processualists took this as a key plank of their plat-
form—​so a succession of neo-​Marxist, structuralist, and post-​structuralist approaches
to the material world came to the fore. These all erred towards a somewhat idealist view
of material culture, seeing it as primarily a medium for the negotiation of power (e.g.
Leone 1984; Tilley 1984), or as an expressive medium with similarities to language and
text (Tilley 1989; Hodder 1991: 121–1​55; cf. Wobst 1977)—​but in either case as actively
constituting social relations, rather than just reflecting them (Hodder 1985: 3–​11). Again,
some aspects of these ideas were taken up to a limited extent in Romano-​British archae-
ology (e.g. Millett 1990: 17, 112; Grahame 2000: 24–2​8), but, aside from a generally
increasing awareness of the potential multiplicity of material meanings (e.g. Freeman
1993: 443–​444), the real convergence between archaeological theory and Roman stud-
ies in Britain was to come only with the next stage of the debate. As frustration with
the rather non-​material approach to ‘reading’ material culture as if it were a text grew,
so the emphasis shifted to theories of practice, action, and experience, and the distinc-
tive role of material culture in the interaction between actors and social institutions,
such as identities (e.g. Buchli 1995; J. Thomas 1996; Hodder 1999: 72–​79; Dobres 2000;
cf. Olsen 2006). This perspective also ushered in a concern with agency, in reaction
 Changing Materialities   487

to the smothering determinisms (whether normative, adaptive, or structural) of pre-

vious approaches (Johnson 1989; Barrett 1993: 1–​6). Particularly as such ideas fit with
the already emerging interest in postcolonial theory in Romano-​British archaeology
(e.g. Webster 1996; Mattingly 1997), we here approach a moment of synchrony between
our field and wider currents in archaeological theory with respect to material culture.
Agency, practice, and identity became quite dominant interests in the later 1990s and
early 2000s, shaping new approaches to the contexts of use of things (e.g. Revell 2000;
Gardner 2002; Chadwick 2004; cf. Scott 1993a; Webster 1999). However, almost as soon
as this moment had arrived, those wider currents moved on more swiftly. A new debate
about materiality has emerged since the turn of the twenty-first century from which
Roman archaeology again finds itself largely detached. Whether or not these new devel-
opments are entirely constructive is an important question, but engaging with them is
certainly critical if Roman archaeology is to get closer to understanding the significance
of material transformations to the people of Roman Britain.

Materiality in Contemporary
Archaeological Thought

A striking aspect of this most recent turn in the way archaeologists approach the mate-
rial world is that an interest in ‘materiality’ has begun to draw together a rather eclectic
range of approaches and people. United primarily by a commitment to take the role of
things in human life more seriously, strands of thought deriving from post-​processual
influences such as continental philosophy and cultural anthropology (some already
noted) can be set alongside those more processual elements such as the cognitive archae-
ology closely associated with Colin Renfrew (2012) or evolutionary approaches to cul-
tural transmission (Cochrane 2011; Shennan 2012). Indeed, this potentially convergent
tendency is explicitly tackled by Ian Hodder in his recent work on ‘human–​thing entan-
glement’ (Hodder 2011, 2012; cf. Knappett 2012: 196–​199). How successfully approaches
such as the material culture studies of the UCL school, ‘symmetrical archaeology’ (influ-
enced particularly by Latour), behavioural archaeology, or evolutionary phylogenetics
can be brought into harmony is a moot point, particularly when the internal coher-
ence of some of these approaches is not particularly strong to begin with (cf. Johnson
2006: 125). However, what is of primary concern to us here is what these perspectives
might offer to the study of the archaeology of Roman Britain. In this section I will briefly
summarize the strengths and weaknesses of a couple of salient approaches, and close
with some pointers they throw up to consider against the major material trends we can
observe in Romano-​British archaeology.
The initial phase in the emergence of new approaches to materiality in archaeology
was strongly influenced by continental philosophy, primarily phenomenology (e.g.
Gosden 1994; J. Thomas 1996), but more recently the work of anthropologists has taken
488

488   Andrew Gardner

precedence. In addition to the members of the UCL ‘school’ of material culture stud-
ies already mentioned, especially Tilley in his work on meaning and experience (e.g.
1999) and Miller on the complexity of consumption (e.g. 2008), Alfred Gell, Tim Ingold,
and Marilyn Strathern have been particularly important. Gell’s notion (1998) that ‘art’
objects are not merely representational, but have a form of agency that creates effects
in viewers has been developed by a number of archaeologists and art historians (e.g.
Gosden 2005; Meskell 2005; Osborne and Tanner 2007). Ingold, meanwhile, has defined
an approach to people’s dwelling in the world that seeks to explore the varied affor-
dances for action that different materials offer (Ingold 1993, 2007; cf. Knappett 2005:
45–​57; J. Thomas 2012: 174–​175). Finally, Strathern’s work (1988) on the different ways
in which personhood is conceived in different cultures, and potentially extended to
non-​humans, including objects has inspired efforts to interpret aspects of prehistoric
material in similar terms and highlight the diversity of human self-​understandings (e.g.
Brück 2001; Fowler 2002). In all of this work, or at least its archaeological application,
there is a certain amount of slippage between the notion of some kind of object agency
as a heuristic device and the ethnography of animism as a particular world-​view (cf.
Morphy 2009; Sillar 2009). With ‘symmetrical archaeology’, influenced more by the
ideas of Bruno Latour and others in Science and Technology Studies (STS), there is
less doubt not only that things deserve to be placed alongside humans in relations of
co-​dependence, but also that they are agents in their own right (Latour 1999: 174–​215;
Webmoor and Witmore 2008; Olsen 2012; cf. Martin 2005). In part, this position is an
explicit rejection, following Latour, of modernity as an anomaly in human history that
has falsely alienated people and things, through conceptual schemas characterized by
dichotomies such as mind/​body, subject/​object, culture/​nature, and so forth (Latour
1993; Olsen 2012: 210–2​11; cf. J. Thomas 2004, 2007, for a similar view influenced more
by Heidegger’s phenomenology). The answer, in this approach, is to identify networks
of actors, human and non-​human, in the generation of particular constellations of prac-
tice; such a perspective has, however, been applied in few systematic studies thus far (but
see Martin 2005; Olsen 2010) and, like some of the other recent approaches to material-
ity, has been challenged for ignoring key somatic aspects of sensory engagement with
materials (Hurcombe 2007; Ingold 2007; Knappett 2012: 191–​192, 196). These points not-
withstanding, very little of this kind of debate has appeared at TRAC (papers by Astrid
van Oyen at TRAC 2011 (Van Oyen 2012) and 2013 are rare instances).
Other approaches to the relationships between humans and things are also signifi-
cant. While there is not space here to explore the vibrant strand of evolutionary theoriz-
ing that has produced, at times, accounts of material culture transmission that resemble
the networks of object-​actors just discussed (Gosden 2005: 198; Johnson 2011: 320–​321),
another school of thought in the processual tradition that overlaps more overtly with
recent post-​processual or interpretative work needs to be mentioned. Discussion of
material engagement within cognitive–​processual archaeology (Renfrew 2012; papers
in DeMarrais et al. 2004; Malafouris and Renfrew 2010) revolves around human cogni-
tion as a distributed phenomenon, stretching out into the material world. Things are not
merely human products, therefore, but components in the evolution of consciousness,
 Changing Materialities   489

with technological change and cognitive change intimately related. As this is a develop-
ment of processualism, there is an emphasis on the systematic study of long-​term pro-
cess, incorporating fields of expertise such as neuroscience and evolutionary biology
(Renfrew 2012: 130), which are lacking in the more particularistic interpretative studies.
The latter might be exemplified further by Hodder’s recent forays into materiality, which
also set out an approach to the interdependence of humans and things (as well as things
and things), but are perhaps distinctive in seeking to account for the messy complexity
of change in material ‘entanglements’ (Hodder 2011, 2012). Hodder advocates the idea of
object-​agency even more strongly than some of the followers of Gell and Latour (Hodder
2012: 213–​216, 219; cf. Martin 2005: 283–​284; Sillar 2009: 368), but also admits that this
takes him away from at least traditional understandings of power and institutions. This
is a crucial problem, not confined to Hodder’s work, and one that is particularly relevant
to the ways in which Roman archaeologists might engage with the theories of materiality
outlined here. The importance of recent work highlighting the interdependence of people
and things and, therefore, the significance of changes in the world of things for the nature
not just of human identities but, more fundamentally, the character of human subjectivity
in particular cultural settings is profound. However, it is neither particularly new (hav-
ing been a feature of pragmatist and symbolic interactionist approaches to social psychol-
ogy for decades—​for example, McCarthy 1984; Silva and Baert, forthcoming; cf. Gardner
2003), nor does it particularly help us deal with power in an institutional sense (Gardner
2011: 71–7​5; cf. Harding 2008: 161). The latter theme, which is surely crucial to understand-
ing Roman imperialism, requires us to consider the variety of structural forces that bear
upon human subjectivity, which exist at a variety of scales. Indeed, some have remarked
upon a lack of attention to issues of power in recent developments in Roman archaeology
(Pitts 2007: 709; Faulkner 2008; Hingley 2001: 114; Mattingly 2011: 19–​26), and this con-
cern has partly driven the move towards ‘globalization’ theory, though it may be doubted
whether this is an appropriate framework for unpicking the particular configuration of
structure and agency in the Roman world (Gardner 2013). Certainly, though, an under-
standing of the impact of changing materiality is necessary to develop this, as in other
colonial situations (cf. O’Hanlon and Washbrook 1992; Gosden 2004). How this might be
done in practice will be explored in the next sections.

Material Horizons in Roman Britain

The emergence of Roman Britain: First Century

bc–​First Century ad
The aim of this section is to outline some of the broader patterns in the material trans-
formation of Britain from the later Iron Age through to the fifth century ad. While,
of necessity, innovations will dominate the narrative, I will attempt also to highlight
490

490   Andrew Gardner

particular continuities or retrenchments of traditions. In the following section, the

implications of some of these developments on particular practices, identities, and sub-
jectivities will be examined, in the light of the points just discussed in the context of
developments in archaeological theory. One lesson, of course, from thinking more criti-
cally about the relationship between material culture and people—​including archaeolo-
gists—​is that we need to be wary of the periodizations that have structured the discipline
(Arnold 2012: 86–​88). This much is very clear from the archaeology of ‘Roman Britain’,
in which key moments of material change singularly fail to correspond to the estab-
lished chronological parameters of the period. The beginning of the incorporation of
Britain into the Roman Empire in a material sense is particularly fuzzy, as, while the
evidence of military control increases fairly conspicuously in the mid-​first century ad
(though cf. Creighton 2006: 46–​69 on pre-​43 ad evidence of military presence), other
key changes either pre-​date or post-​date that point by a significant margin. If we take
a very long view, the beginning of the material impact of the Roman world on Britain
needs to be seen in the context of the participation of Iron Age British societies in net-
works of exchange that brought elite material culture, particularly metalwork styles, to
areas in many parts of Britain (Hill 1995: 78–7​9; Champion, this volume). Such networks
are likely to have mediated the first arrivals of Mediterranean material culture and, per-
haps, played a transformative role in the meaning of objects travelling along them.
This is very evident in the case of one of the distinctive innovations that characterize
the later Iron Age, with coinage from northern Gaul appearing in the second century
bc, to be followed in the first centuries bc and ad by coinages created by leaders of sev-
eral groupings in parts of southern Britain. This later development broadly coincides
with the appearance of imports of ceramics and the commodities carried in some of
them (particularly wine and other amphora-​transported goods), evident in burial and
some settlement contexts. These new technologies do not occur in vast numbers but are
nonetheless important as material innovations, in the sense that they almost certainly
provided novel media for the negotiation of power and the formation of a greater hierar-
chy among kinship structures in some regions (Millett 1990: 12–​22, 29; Hill 1995: 78–​89;
Mattingly 2006: 58–​64; Moore 2011: 350–​351). They may also have provided leaders and
their followers with a new way of thinking about themselves both as individuals and as
members of groups. Possibly more striking material change was evident in the land-
scapes of the new political entities of this dynamic period, with the various large, semi-​
enclosed sites known as oppida being created across south-​central Britain (Millett 1990:
21–​29; Hill 1995: 70–​72, 85; Creighton 2006: 23–​24). Another conspicuous transforma-
tion, and similarly one that does not directly involve imported technology, is the ‘fibula
event horizon’ of the same period, which apparently marks a shift in the construction
of personal appearance, with much greater numbers of brooches being deposited than
previously (Hill 1995: 86; Jundi and Hill 1998). Both of these changes probably had a
more immediate effect on a wider range of people than the uses to which coinage or
imported wines were put. What that effect was is more difficult to specify. The period
from the end of the second century bc was clearly one of increasing flux, into which
Roman military presence intruded directly only under Caesar, and then Claudius nearly
 Changing Materialities   491

a century later, but it seems clear that the expansion of Roman political influence in
Gaul provides a more constant backdrop to the transformations to both individual and
institutional life in some parts of Britain. These produced a range of somewhat unstable
social formations, closely tied to charismatic authority, rather than the even pattern of
‘tribes’ traditionally supposed (on the basis, it must be admitted, of what Roman writers
wanted to see in the newly conquered province (Moore 2011: 345–​349; cf. Millett 1990:
20–2​3, 29–​35)), and this pattern of variability was to continue under more direct Roman
occupation.
The conquest and its aftermath undoubtedly brought a series of material changes
closely linked to imperial institutions, particularly those of military and then civic
administration, but these too were drawn out and uneven. The most obvious material
markers of the major military campaigns undertaken to impose control are the forts and
roads, which obviously tended to follow in the wake of, or precede a new phase of, actual
movement of troops. While often fairly transient and involved in limited forms of inter-
action with the wider indigenous population, the fortifications built to house legionary
and auxiliary soldiers were often striking impositions on the previous political landscape,
whether situated within the sprawling complex of earthworks at Camulodunum, or in the
corner of a ‘hillfort’ like Hod Hill (Millett 1990: 44–​55; Mattingly 2006: 90–​94, 128–​36).
Similarly, the roads that facilitated military logistics conspicuously marked the landscape,
no less than the requisitioning of resources and land to support the troops on campaign
and then as veteran colonists (Millett 1990: 56–​60; Witcher 1998). While the long-​term
presence of soldiers became a feature of life only in certain regions, some of their early
bases were co-​opted for urban purposes. Though the mechanisms by which administra-
tive towns were established and their relationship to military authority and personnel
remains debated, these certainly represent an architectural development that would have
confronted a greater swathe of the population of Roman Britain with a material reminder
of political change. Yet this can hardly be said to have been sudden. Like those military
sites which preceded them, the novel material environments of towns were frequently
embedded within meaningful Iron Age landscapes (Millett 1990: 69–​78; Creighton 2006:
123–​156; Rogers 2011: 47–​72), but also took some time to find articulation. Indeed, the
appearance of larger masonry houses and fully monumental public buildings in a num-
ber of civitas capitals is really a phenomenon only of the second century (see next section;
Millett 1990: 78–​91, 102–​103, 104–​111; Burnham et al. 2001: 73; Mattingly 2006: 276–​286).
Thus we can see that even in the most conspicuous material changes in settlement types
there is a mixture of rapid and generational tempos of transformation.
A similar picture can be discerned from some of the main categories of portable
objects. We have already seen how coinage was a new technology of the later Iron
Age, but the large-​scale presence of Roman troops brought with it a much larger pool
of coins, embedded in a larger-​scale set of economic and ideological structures. Yet,
while this wealth became an important draw to people setting up in the fort vici, the
spread of coinage beyond such environments, and towns, in the first century was lim-
ited (Millett 1990: 58–​59; Reece 1993b; Mattingly 2006: 497). It would be some time—​
centuries even—​before people living on many of the settlements across rural Roman
492

492   Andrew Gardner

Britain found themselves losing coins in the course of their everyday activities. A rather
greater impact, though, was made by changes in the supply and consumption of pottery.
Of course, some key developments preceded the conquest, when, alongside imported
amphorae and a range of Gallic wares, the technology of the potter’s wheel was adopted
(Hill 1995: 79; Tyers 1996: 52–​56). The immediate impact of the conquest itself was, as
in other respects, somewhat muted in the short term, as soldiers did not carry large
quantities of pottery on campaign (Tyers 1996: 56). However, imports of finewares, in
particular of samian ware, picked up rapidly later in the Claudian period, while a num-
ber of significant industries became established during the mid–​late first century ad
across southern Britain. There appears to have been considerable exchange at least of
ideas between British, Gallic, Germanic, and Mediterranean traditions in this period,
and while the military was certainly a key driver of the expansion of pottery production,
the limitation of new materials to that community that we have observed with coin-
age did not apply so obviously here. Alongside traditional pottery, new forms and fab-
rics, in significantly greater quantities, marked a change in scale of material innovation
(Millett 1990: 98–9​9, 123–​124; Tyers 1996: 56–​66; Mattingly 2006: 513–​518; cf. Evans, this
volume). While limitations of space prohibit discussion of other important categories
of artefacts here, it should be clear from the foregoing that the material transformations
linked to the initial impact of Roman imperialism upon Britain were complex, being
multi-​temporal and variable by region, class, and other factors. The same is certainly
true of later periods too.

Diverse Participations: Second–​Third Centuries ad

As we move into the mid-​Roman period, the question of ‘when did Britain become
“Roman”?’ becomes all the more pressing. This, of course, requires us to define what
becoming ‘Roman’ means, and, even if the military and administrative situation is rela-
tively clear, our material yardsticks for this remain in constant flux. Indeed, the solution
seems to be that the moment of ‘being Roman’ never fully arrives, because the refer-
ence points are also always in motion (cf. Barrett 1997). To begin again with the mate-
rial settings of life in architecture and landscape, the second and third centuries saw
urbanism take on new forms in different areas of Britain. The cityscapes of the ‘public’
towns—​the civitas capitals, coloniae, and other administrative centres—​were, in some
cases, only just becoming complete in terms of their civic identity. The forum-​basilica
complexes at Caerwent and Leicester, for example, were completed in stone towards
the middle part of the second century. Curiously, a handful of generations later, such
buildings were already undergoing changes in use—​or disuse—​as at Wroxeter (appar-
ently largely abandoned after a later third-​century fire) and Silchester (being used as a
workshop from the late third century (Millett 1990: 102–​103, 130, 137; White and Barker
1998: 73–​74, 86; Fulford and Timby 2000)). If the building type that encapsulated the
material imposition of Roman administration was thus shifting in its importance within
the Roman period, so too we must assume the power networks of Roman imperialism
 Changing Materialities   493

were dynamic over time. Meanwhile, the public towns were increasingly manifesting a
concentration of wealth in fewer hands, with many of the strip-​buildings of the earli-
est phases being succeeded by larger courtyard-​style buildings (observed in excavated
sequences in, for example, Silchester and Caerwent (see Millett 1990: 134–13​5; Mattingly
2006: 284–28​5)). In parallel to this process, the development of new forms of urban-
ism—​settlements often linked to particular economic or religious foci, and known col-
lectively as ‘small towns’—​is indicative of a more long-​term process of transformation in
the material articulation of colonial experience. The growth of sites such as Chelmsford,
Kenchester, and Water Newton from the second century indicates that settlement
nucleation, occurring for a variety of reasons and with a range of attendant material
consequences for the people living in these locations (Millett 1990: 143–​151; Burnham
1995: 12–​14; Mattingly 2006: 286–​291), was the product of a new phase of provincial cul-
ture change taking place sometime after other innovations had become embedded.
Material change in rural lifestyles—​in terms of both form and tempo—​varied by
region and class, but generally was a phenomenon of the mid-​Roman period too. The
conspicuous architectural innovation of the ‘villa’—​in reality a constellation of elements
including a range of layouts, construction materials, particular facilities such as baths
and decorative features—​is manifest in restricted areas from the later first century ad,
but became more widespread in the second and third centuries (Millett 1990: 91–​98,
117–​120; Mattingly 2006: 369–​375). These are convincingly connected to the towns and
the wealth display of those who held high status in both urban and rural contexts, but
whether they represent ‘the Roman aspirations of their owners’ (Millett 1990: 119)—​a
small but growing minority of country-​dwellers—​can be established only by a rather
more in-​depth study of the meanings of these material forms and the compromises
of colonial culture than has hitherto been undertaken (cf. Rippengal 1993). Nor is
it the case that the apparently more continuous traditions of the majority of farmers
in rural Roman Britain became fossilized through lack of opportunity (or aspiration)
and showed no material change. Again, regionality is a key feature of the archaeologi-
cal patterning, but, alongside the quite deliberate material manifestation of long-​term
rootedness (as evinced, for example, by burials in relation to prehistoric monuments
at Cotswold Community (Smith and Powell 2010: 136–​138, 165)), comparison of sites
within particular landscapes can reveal significant episodes of change. In the upper
Thames valley, for example, there are widespread instances of settlement replanning in
the early second century indicative of evolving institutional circumstances mediated by
subtle material transformation (Smith 2007; cf. Taylor 2001: 49–​52; 2007). These are as
much a part of colonial experience as the more obvious creation of ‘villa’ lifestyles.
Alongside these fairly gradual trends affecting the physical settings of Romano-​
British life, there are, of course, key thresholds in the trajectories of a wide range of arte-
facts in the mid-​Roman period. Coinage underwent almost as profound a change in its
use in the third century ad as it had in the first, thanks largely to the more direct way
it manifests the vagaries of imperial fortunes on the broadest scale. Devaluation of the
silver coinage in particular, followed by currency reform and the production of much
larger quantities of bronze coinage (both officially and as copies), along with the use of
494

494   Andrew Gardner

coins by leaders of the various breakaway polities in the mid to late third century, all con-
tributed to significant changes in the form, abundance, and ideological content of coins.
As a result of at least some of these factors, these objects became much more commonly
lost on a wide range of sites from the third quarter of the third century (Reece 2002: 46–​
57). Use of pottery also saw some interesting changes in the second and third centuries
(cf. Evans, this volume). Insular production centres saw growth in the second century,
at least partly as a result of the establishment of long-​term supply arrangements to the
garrison that was then beginning to settle down in the north; Dorset Black Burnished
Ware (BB1), a fabric with strong Iron Age technological roots, was particularly success-
ful in this context. Fine-​ware producers also emerged that, in the third century, were
well placed to take over from the demise of the samian industry, in many cases—​par-
ticularly the Oxfordshire potteries—​copying several samian forms (Tyers 1996: 66–​72).
Although a range of imports continued to appear, people were more likely to obtain all
of their pottery from sources in Britain during this period, including both local prod-
ucts and inter-​regional imports (Mattingly 2006: 516–​517). Whether this meant much in
terms of identity, and what differences existed between perceptions of very local, non-​
local, and non-​British products, is a more challenging question; a more pertinent mate-
rial shift may have been the diminution of overall quantities of new vessels available in
the mid-​third century (although this might simply be a result of problems with dating
conservative styles (Tyers 1996: 70–​7 1; cf. Going 1992: 99–​100; Cooper 1996). From these
selected examples of archaeological patterning it should be apparent that the material
world of Roman Britain had far from ‘settled down’ after the invasion period.

The Fragmenting of Roman Britain: Fourth–​Fifth

Centuries ad
The materiality of fourth-​century Britain was, as a result of the changes just discussed and
others that we will examine next, very different from that of the later first century. Which,
then, was the real ‘Roman Britain’? And what should we make of the further changes at
the beginning of the fifth century? In the last part of this outline of changing materiali-
ties, I will first look at the portable objects and then the inhabited environment, as in this
period more than others the identification of transformation in the latter is dependent on
trends in the former. With coinage and pottery, most of the fourth century is a period of
relative stability. The increased circulation of coinage apparent from the later third cen-
tury continues, with perhaps the most notable shift in distribution being the decreasing
incidence of coin finds at military sites, as the structure of army payment was changed
(Brickstock 2010). The larger pottery industries that became important in the third cen-
tury remained so, though there are some signs of stagnation in the later fourth century
(particularly a reduction in formal repertoire) as well as the emergence of a new genera-
tion of distinctive shell-​or calcite-​tempered wares produced by a series of small manu-
facturers (Millett 1990: 157–​174, 178–1​80, 224–​227; Tyers 1996: 77–​78; Mattingly 2006:
518–​519; cf. Evans, this volume). However, by sometime in the first quarter of the fifth
 Changing Materialities   495

century, much had changed. Imperial coinage supplies to Britain stopped in the first dec-
ade of that century and the pottery industries stopped production seemingly soon after
(Going 1992: 102–103; Millett 1990: 224–​227). This is not the place to go into the causes of
this apparently dramatic shift (though aspects of it that are prefigured much earlier will
be highlighted). What is striking, though, is that, as much as those causes appear to lie
in the gradual administrative fragmentation of the later Roman empire, they must also
pertain to something cultural. This can be argued simply because, while previous periods
of diminished supply had not been terminal and had even been made good with copies
of, for example, coinage (Millett 1990: 226–22​7; Reece 2002: 62), this time the techno-
logical change was definitive. However, it would be ironic if, after the quite prolonged
and gradual series of material changes outlined above, which shaped ‘Roman Britain’ in a
piecemeal fashion, the end was entirely more sudden. Is this really the case?
If we turn to the material fabric within which people lived in the fourth century, it is
apparent that the transformation of people’s engagement with those built spaces emerged
from earlier trends. The continual process of transmutation that marked the major towns
continued in the fourth century with more and more public or civic spaces changing their
function; from the basilica at Caerwent becoming a workshop in the 330s to the refuse
disposal and ‘market stalls’ in the baths complex at Canterbury in the middle part of the
century (Blockley et al. 1995: 461; Brewer and Guest, forthcoming). In a similar fashion,
changes to the uses of buildings within military sites are also commonplace in the mid-​to
late fourth century, with workshop activity in the principia at Housesteads, and butchery
in the praetorium at Binchester (Crow 2004: 96–9​8; Ferris 2010: 553–55​4). Notably, these
and other examples are derived from archaeological sequences that simultaneously indi-
cate material continuity and change: continuity in the use of various technologies and
the disposal of their products (for example, livestock, ceramics, metals), but change in
the location of activity and thus the use of certain types of built space. Within the ‘offi-
cial’ contexts of fort or town, this is surely significant for the relationship between peo-
ple performing these actions and the institutional structures of later Roman imperialism
(Gardner 2007: 243–​264). On rural sites, such processes are less pronounced. The early
fourth century sees continuing trends in settlement nucleation and the expansion of ‘vil-
las’ both numerically and, in some cases, on an individual scale, but, in the second half of
the century, numbers of sites occupied seem to decline, because the material markers of
occupation—​pottery and coins in particular—​dwindle (Millett 1990: 186–​211, 219, 223–22​
4; Mattingly 2006: 374–37​5, 534). Whatever these changes mean for the presence, absence,
or movement of people, they are indicative that the material world of Roman Britain was
in flux for some time before ‘the end’—​so, again, defining the ‘Roman-​ness’ that came to
an end is not entirely straightforward.
There are other innovations and retrenchments in the fourth century that deserve to
be noted—​including new means of personal adornment, changes in agricultural tech-
nology, the apparent end of the practice of inscribing stone monuments, and important
changes in religious practice—​but, in concluding this section, I will turn to those nota-
ble material horizons that may post-​date 410 ad but are significant in understanding the
varied responses of people across Britain to the seeming material vacuum in which they
496

496   Andrew Gardner

found themselves. For, however long the fragmentation of Roman Britain took through
the fourth century and however prolonged the use of available pots, coins, and other
objects into the fifth century might have been (see, e.g., Dark 1994: 200–20​6), there
were two further major material transformations in the fifth century that were initiated,
or accepted, by formerly Roman citizens. One was the creation, in the eastern part of
Britain, of a new, highly visible material world defined by new burial practices, styles of
metalwork and pottery, and forms of timber architecture. While these were traditionally
interpreted simply as the material culture of Anglo-​Saxon migrants, the more recent
debate as to whether the people using these objects were Germanic or Romano-​British
in origin continues, though this still risks perpetuating a misleading link between birth-
place and culture (cf. Arnold 1997: 19–​33; Hills 2003). Wherever they were born, peo-
ple using this suite of objects were creating a new materiality for Britain, but, ironically,
they were doing so with the kinds of things that had been shaped by the particular cul-
tural environment of a place just beyond the Roman frontier, which had experienced a
similar dynamic mixing of traditions as parts of Britain (Swift 2000: 88–​136; Hills 2003:
74–​75). The second horizon is the appearance of new suites of imported Mediterranean
pottery and Latin-​inscribed stones (particularly burial-​markers) in western Britain, an
area that had shown rather little material change in the Roman period as conventionally
defined. Whether this is entirely to do with the continuation of a Christian milieu here
or also has to do with the political legacy of Roman imperialism is unclear (Mattingly
2006: 534–​539; White 2007: 195–​207). What is clear, however, is that both of these mate-
rial transformations on the littoral margins of Britain give the archaeologist more to
work with than the materially invisible people living in-​between, which reminds us that
the interpretation of changing materialities is always going to be a very partial affair (cf.
Esmonde Cleary 2001: 90–​91). At the same time, the need for deeper theorization of the
relationship between people and things is made all the more apparent by the series of
remarkable changes in the fourth and fifth centuries.

Changing Materialities,
Changing People?

This survey of material trends in Roman Britain has necessarily been both selective and
simplified. Yet it gives an appropriate picture of a fluid constellation of materialities in
which something was always moving. At different moments, and in different places, tra-
ditions of using certain things were bent or broken, and new traditions established. We
must be aware, though, that this picture may be misleading in at least a couple of ways.
The first is that, in highlighting changing materialities, we perhaps miss ways of life that
did not change at all—​and we know that evidence for these can be found in the southern
parts of Britain (for example, the continuation of round-house construction in the east
Midlands (Taylor 2001: 49–​52)), as much as in the areas outside of the Roman province
 Changing Materialities   497

where they might be expected. These can certainly not be ignored, and, indeed, the per-
sistence of certain technologies within a dynamic context gives such things the poten-
tial for new meanings. A second issue has to do with the relationship between temporal
scale and concepts of novelty. Here the weakness of archaeological chronologies comes
to the fore, in terms of tracking the speed of change in relation to the pace that would
have been perceived by contemporaries. Albeit in a world of fairly slow communications
and transport of ideas and things, novel technologies may have changed their signifi-
cance within time frames that we simply cannot see. However, comparison between dif-
ferent fields of practice and different sites and regions at least gives us an insight into the
relative scale of different transformations (cf. Gardner 2012). With these points in mind,
we can begin to examine the ways in which some of the theoretical perspectives already
discussed might enable us to draw more out of the material patterns in Roman Britain.
One of the common threads among the diverse recent approaches to materiality has
been a diminution, at least, of the role of human agency in social life and an advocacy of
the power of objects. This argument, in my view, represents a troubling return to a form
of material determinism that was rejected in the earliest stages of post-​processualism
and that leaves a lot to be desired in accounting for power relationships in any context,
let alone one like the Roman Empire (cf. Gardner 2011; Hodder 2012: 208). However,
the rhetorical device of switching our perspective to that of the objects we study raises
some important issues to do with tradition and choice. These concepts are central, in
different ways, to both evolutionary and the more eclectic actor–​network/​object–​
agency/​entanglement approaches that currently dominate debate about materiality (e.g.
Gosden 2005; Shennan 2011). They are linked in as much as the weight of traditions in
artefact production and design is often seen as bearing down upon people, drawing in
different influences over generations, educating people into a particular cultural world
as they are socialized, and thus ultimately channelling their action down particular
paths. Relationships between objects have been highlighted as constituting the fabric
of particular traditions, or the assemblages of things that afford particular sets of activi-
ties, like hunting (Gosden 2005; Olsen 2010: 129–​149; Hodder 2011: 157–​159; cf. Pauketat
2001; Johnson 2004: 243–24​6). In such a fashion, choice becomes a rather ephemeral
notion, reduced to the register of selective advantage, or problem-​solving when an
established way of doing things is confronted, often when things break down or fail
(Hodder 2011: 159–1​68; Shennan 2011: 327–​328). While limited in certain crucial ways,
such a line of reasoning sheds some new light on a key problem within Romano-​British
archaeology—​albeit one only rarely explicitly recognized as such. The ‘Romanization’
paradigm was built on the premise that, when presented with the opportunity, people
would choose to be as ‘Roman’ as possible (e.g. Millett 1990: 119). One problem with
this premise is that—​partly because of the inattention to theories of material mean-
ing—​we do not really know which artefacts meant ‘Roman’ at any particular time or
place (cf. Freeman 1993: 443–​444; Mattingly 2004: 9–​10; Gosden 2005: 197–19​9, 208–20​
9). An alternative line, pursued subsequently, has been to avoid the language of being or
becoming ‘Roman’ and simply to regard the material transformations of Roman Britain
as in themselves presenting people with a limited range of choices of things—​they may
498

498   Andrew Gardner

or may not have been seen as ‘Roman’, or some other category, but they were consumed
because they were available (Cooper 1996; cf. Greene 2008). Following up the implica-
tions of this position in more depth might be one way in which some of the archaeologi-
cal patterning in Roman Britain might be accounted for by taking materiality seriously.
However, to pursue this more creatively I think we need to return to people, and to the
interaction between differently situated actors as mediated by things.
This means that, while it is vital to take seriously the role of things in shaping humans
as actors, the distinction between the motive force of human interests and intentions and
the mediating role of objects needs to be sustained if we are to give power its due. This
is true when analysing the workings of any society, but clearly is particularly important
in one defined by a particular set of power relationships and inequalities (cf. Hingley
2001: 114; Mattingly 2011: 19–​26). While very few of the material patterns outlined in the
previous section had much to do with the actors situated at the top of Roman imperial
society, they have certainly been interpreted in relation to the choices of actors situated
within particular provincial institutions—​the civic aristocracy, the military—​holding
relatively high degrees of wealth, status, and power. This much is indeed the bread-​and-​
butter of ‘Romanization’. Yet alongside the problem of what ‘Roman’ means (and closely
connected to it) can be placed the notion that changing materialities are both sources
and outcomes of tension within different groups and different individuals. Structure
(which encompasses the material world in a Giddensian sense (cf. Gardner 2007: 18,
following Giddens 1984)) is always enabling and constraining—​shaping choices, some-
what as already discussed, but also affording them in multiple ways. Just as ‘becoming
Roman’ never meant only one thing, so the changing spectrum of practices that all the
artefacts and spaces of Roman Britain were part of never simply enabled or oppressed
(cf. Witcher 2015). I would argue that this dynamic tension is the underlying cause of the
continually mobile empirical patterning that I have outlined.
On the one hand, becoming a soldier, shaped by the material environment of fort,
equipment, and state-​brokered supplies, was to become a different kind of colonial sub-
ject from a farmer, but a subject nonetheless. On the other, the farmers who acquired
new red pottery in the later first century, or new coin in the later third, need not have
thought of this as ‘Roman’ to enjoy new possibilities for action, even as they participated
in the structures of the provincial or imperial economy. Imperial structures offered free-
doms and constraints that shaped new colonial people (cf. Hill 2001: 15). But the people
also shaped the structures and perhaps, in the long run, through choosing to prioritize
the local, contributed to their fragmentation (Gardner 2007: 256–​257). I would argue,
therefore, that the archaeology of Roman Britain can show us what material change
means in terms of the detaching of people from the sources of wealth and power as the
scale and tempo of material connections increase, perhaps offset by the beguiling effects
of consumption. Yet it can also show us that structures can be remade, even inadvert-
ently levelled, at least to the point where power is deprived of persistent material expres-
sion. This is only one way of putting the pieces together, though, and the aim of this
chapter has been more broadly to show that thinking about the archaeology of Roman
Britain in terms of changing materialities—​meanings, practices, traditions, innovations,
 Changing Materialities   499

networks, tempos, agencies, structures, even genealogies and entanglements—​means

thinking about the irreducible complexity of colonial experience. As such it is entirely
appropriate to reflect upon the continued relevance of the Roman world to the materi-
alities, and the politics, of the present day.

Conclusion: A Confrontation
with Roman Materialism

The physical presence of the Roman Empire, particularly its distinctive scale and diver-
sity, has played a major part in the way subsequent generations have projected their own
political desires in material ways. From the evocation of Roman materialities directly in
architecture (for example, Edward I’s castle at Caernarfon, the monumental buildings of
imperial London or republican Washington) or portable material culture (the coins of
medieval and modern Europe), to the writing of Roman material culture as the inher-
ently superior technology of an ancestral culture (Frere 1987: 295–​312; Wilkinson 2000),
the symbolic value of Roman things in strategies of legitimation has been immense.
The process of academic, political, and popular use of images of Rome has of course
not ceased—​indeed, part of the argument for introducing theories of globalization into
Roman archaeology has been that it at least has the virtue of placing the contemporary
context that influences practitioners explicitly in the frame, rather than remaining as a
more or less well-​hidden backcloth (Hingley 2015; Witcher 2015). Among other views,
which also make their framing perspective quite explicit, are the contrasting interpreta-
tions of Rome’s material impact in Britain as indicative of the success of a free-​market
economic system (Selkirk 2006), or of the brute hand of exploitative and violent impe-
rialism (Faulkner 2000: 11–​12). Also notable is the idea that there is a political edge to
the issue of archaeological visibility—​in the sense either that archaeology is the primary
means of recovering the voices of the forgotten people of history (Hingley 1989: 1–​5;
Webster 2001: 223; Given 2004: 3–​4), or conversely that archaeologists have been right to
focus on the most conspicuous material changes, as only the people who could partici-
pate in these are important enough to make a difference in society and thus characterize
it (C. Thomas 1990; de la Bédoyère 1999: 9). The apparent fragility of the imperial system
in the fifth century is generally less exemplary now, except for those with a more nega-
tive political (Faulkner 2000) or ecological (McAnany and Yoffee 2010) view of complex
societies like Rome. Taking all of these examples into account, though, it is difficult to
see how understanding Roman materiality in Britain can avoid being politicised. Where
do new approaches to materiality fit into this matrix?
In a general sense, approaches that have emerged from more interpretative traditions,
such as those informed by Latour’s Actor–​Network Theory or the work of Alfred Gell,
tend to be fairly explicit in advocating a commitment to respect for objects as an ethical
move in a world where there are increasing debates about both biotechnology, and the
500

500   Andrew Gardner

need to confer rights on animals or environments in order to challenge human-​induced

environmental disaster (in line, perhaps, with an animist view of human–​nature balance
(Sillar 2009; Hodder 2012: 220–​221; Olsen 2012: 217–​221)). Yet in terms of the politics
of institutions, like those of an empire, such approaches have relatively little to say—​
already noted as a challenge in using them to understand the archaeology of Roman
Britain. However, other approaches to the ways in which the material world mediates
between actors or subjects and institutional networks of power certainly do invite com-
parison between the politics of materiality in the Roman and contemporary worlds.
Postcolonial and globalization approaches have been taking us in the right direction
here (Witcher 2000; Mattingly 2011), but need to take on board more of the facets of
material engagement that might be investigated, as discussed (though cf. Nesbitt and
Tolia-​Kelly (2009) and Witcher (2015), focusing on the reception of the Roman past).
While comparative work on the materiality of colonialism has become more com-
monplace (Given 2004; Gosden 2004; Stein 2005), a Romanist view of the relationship
between changing materialities and the ideologies, opportunities, and inequalities of
imperial power has yet to be written. There is no better time to re-​engage with the theo-
retical debate in the broader field of archaeology, as in this case Romanists really do have
the material to hand to drive that debate forward.

Acknowledgements

Many thanks to the editors for the invitation to contribute this chapter to the volume and
for useful comments on its content. Thanks also to Richard Hingley and Rob Witcher
for copies of forthcoming papers, and to Mark Lake, Kathryn Piquette, and Bill Sillar for
various discussions about materiality.

References
Allason-​ Jones, L. (2011) (ed.). Artefacts in Roman Britain:  Their Purpose and Use.
Cambridge: Cambridge University Press.
Andrén, A. (1998). Between Artifacts and Texts: Historical Archaeology in Global Perspective,
trans. A. Crozier. New York: Plenum Press.
Arnold, B. (2012). ‘Gender, Temporalities, and Periodization in Early Iron Age West Central
Europe’, Social Science History, 36/​1: 85–​112.
Arnold, C. J. (1997). An Archaeology of the Early Anglo-​ Saxon Kingdoms. 2nd edn.
London: Routledge.
Barrett, J. (1993). Fragments from Antiquity: An Archaeology of Social Life in Britain, 2900–​1200
bc. Oxford: Blackwell.
Barrett, J. (1997). ‘Romanization: A Critical Comment’, in D. Mattingly (ed.), Dialogues in
Roman Imperialism. JRA Supplementary Series 23. Portsmouth, RI: Journal of Roman
Archaeology, 51–​64.
 Changing Materialities   501

Bennett, T., and Joyce, P. (2010) (eds). Material Powers:  Cultural Studies, History and the
Material Turn. London: Routledge.
Biddulph, E. (2012). ‘On Cultural Selection:  Examining the Process of Cultural Evolution
through Funerary Evidence’, in M. Duggan, F. McIntosh, and D. J. Rohl (eds), TRAC
2011:  Proceedings of the Twenty-​First Annual Theoretical Roman Archaeology Conference,
Newcastle 2011. Oxford: Oxbow Books, 76–​90.
Blockley, K., Blockley, M., Blockley, P., Frere, S., and Stow, S. (1995). Excavations in the
Marlowe Car Park and Surrounding Areas, I: The Excavated Sites. Canterbury: Canterbury
Archaeological Trust.
Brewer, R., and Guest, P. (forthcoming). Caerwent Forum-​Basilica. London: Society for the
Promotion of Roman Studies.
Brickstock, R. J. (2010). ‘Coins and the Frontier Troops in the Fourth Century’, in R. Collins and
L. Allason-​Jones (eds), Finds from the Frontier: Material Culture in the 4th–​5th Centuries.
CBA Research Report 162. York: Council for British Archaeology, 86–​91.
Brück, J. (2001). ‘Monuments, Power and Personhood in the British Neolithic’, Journal of the
Royal Anthropological Institute, NS 7: 649–​667.
Buchli, V. A. (1995). ‘Interpreting Material Culture:  The Trouble with Text’, in I. Hodder,
M. Shanks, A. Alexandri, V. Buchli, J. Carman, J. Last, and G. Lucas (eds), Interpreting
Archaeology: Finding Meaning in the Past. London: Routledge, 181–​193.
Burnham, B. C. (1995). ‘Small Towns: The British Perspective’, in A. E. Brown (ed.), Roman
Small Towns in Eastern England and Beyond. Oxford: Oxbow Books, 7–​17.
Burnham, B. C., Collis, J., Dobinson, C., Haselgrove, C., and Jones, M. (2001). ‘Themes for
Urban Research, c.100 bc to ad 200’, in S. James and M. Millett (eds), Britons and Romans:
Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for British
Archaeology, 67–​76.
Chadwick, A. M. (2004). ‘“Heavier Burdens for Willing Shoulders”? Writing Different Histories,
Humanities and Social Practices for the Romano-​British Countryside’, in B. Croxford, H.
Eckardt, J. Meade, and J. Weekes (eds), TRAC 2003: Proceedings of the Thirteenth Annual
Theoretical Roman Archaeology Conference, Leicester 2003. Oxford: Oxbow Books, 90–​110.
Cochrane, E. (2011). ‘Units of Transmission in Evolutionary Archaeology and the Role
of Memetics’, in E. Cochrane and A. Gardner (eds), Evolutionary and Interpretive
Archaeologies: A Dialogue. Walnut Creek, CA: Left Coast Press, 31–​61.
Collingwood, R. G. (1930). The Archaeology of Roman Britain. London: Methuen.
Collingwood, R. G., and Myres, J. N. L. (1937). Roman Britain and the English Settlements. 2nd
edn. Oxford: Clarendon Press.
Collingwood, R. G., and Richmond, I. (1969). The Archaeology of Roman Britain. London:
Methuen.
Cooper, N. J. (1996). ‘Searching for the Blank Generation: Consumer Choice in Roman and
Post-​Roman Britain’, in J. Webster and N. J. Cooper (eds), Roman Imperialism: Post-​Colonial
Perspectives. Leicester: University of Leicester, School of Archaeological Studies, 85–​98.
Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge.
Crow, J. (2004). Housesteads: A Fort and Garrison on Hadrian’s Wall. Stroud: Tempus.
Dant, T. (2005). Materiality and Society. Maidenhead: Open University Press.
Dark, K. R. (1994). Civitas to Kingdom: British Political Continuity 300–​800. Leicester: Leicester
University Press.
de la Bédoyère, G. (1999). The Golden Age of Roman Britain. Stroud: Tempus.
de la Bédoyère, G. (2007). ‘Romanization’, British Archaeology. 92 (Jan.–​Feb.): 22.
502

502   Andrew Gardner

DeMarrais, E., Gosden, C., and Renfrew, C. (2004) (eds). Rethinking Materiality:  The
Engagement of Mind with the Material World. Cambridge:  McDonald Institute for
Archaeological Research.
Dobres, M.-​A. (2000). Technology and Social Agency:  Outlining a Practice Framework for
Archaeology. Oxford: Blackwell.
Dyson, S. (1989). ‘The Relevance for Romanists of Recent Approaches to Archaeology in
Greece’, Journal of Roman Archaeology, 2: 143–​146.
Eckardt, H., and Crummy, N. (2008). Styling the Body in Late Iron Age and Roman Britain:
A Contextual Approach to Toilet Instruments. Instrumentum Monograph 36. Montagnac:
Instrumentum.
Esmonde Cleary, S. (2001). ‘The Roman to Medieval Transition’, in S. James and M. Millett
(eds), Britons and Romans: Advancing an Archaeological Agenda. CBA Research Report 125.
York: Council for British Archaeology, 90–​97.
Esmonde Cleary, S. (2009). ‘Roman Britain: Civil and Rural Society’, in J. Hunter and I. Ralston
(eds), The Archaeology of Britain: An Introduction from Earliest Times to the Twenty-​First
Century. 2nd edn. London: Routledge, 198–​218.
Evans, J. (2001). ‘Material Approaches to the Identification of Different Romano-​British Site
Types’, in S. James and M. Millett (eds), Britons and Romans: Advancing an Archaeological
Agenda. CBA Research Report 125. York: Council for British Archaeology, 26–​35.
Faulkner, N. (2000). The Decline and Fall of Roman Britain. Stroud: Tempus.
Faulkner, N. (2008). ‘Roman Archaeology in an Epoch of Neoliberalism and Imperialist War’,
in C. Fenwick, M. Wiggins, and D. Wythe (eds), TRAC 2007: Proceedings of the Seventeenth
Annual Theoretical Roman Archaeology Conference, London 2007. Oxford:  Oxbow
Books, 63–​73.
Ferris, I. M. (2010). The Beautiful Rooms are Empty: Excavations at Binchester Roman Fort,
County Durham 1976–​1981 and 1986–​1991 (Part 2). Durham: Durham County Council.
Fowler, C. (2002). ‘Body Parts: Personhood and Materiality in the Earlier Manx Neolithic’, in Y.
Hamilakis, M. Pluccienik, and S. Tarlow (eds), Thinking through the Body: Archaeologies of
Corporeality. New York: Kluwer/​Plenum, 47–​69.
Freeman, P. W.  M. (1993). ‘Romanisation and Roman Material Culture’, Journal of Roman
Archaeology, 6: 438–​445.
Freeman, P. W. M. (2007). The Best Training-​Ground for Archaeologists: Francis Haverfield and
the Invention of Romano-​British Archaeology. Oxford: Oxbow Books.
Frere, S. S. (1987). Britannia: A History of Roman Britain. 3rd edn. London: Routledge.
Frere, S. S. (1988). ‘Roman Britain since Haverfield and Richmond’, History and Archaeology
Review, 3: 31–​36.
Fulford, M. G., and Hodder, I. R. (1975). ‘A Regression Analysis of Some Late Romano-​British
Pottery: A Case Study’, Oxoniensia, 39: 26–​33.
Fulford, M., and Holbrook, N. (2011). ‘Assessing the Contribution of Commercial Archaeology
to the Study of the Roman Period in England, 1990–​2004’, Antiquaries Journal, 91: 323–​345.
Fulford, M., and Timby, J. (2000). Late Iron Age and Roman Silchester: Excavations on the Site
of the Forum-​Basilica, 1977, 1980–​86. London: Society for the Promotion of Roman Studies.
Gardner, A. (2002). ‘Social Identity and the Duality of Structure in Late Roman-​Period Britain’,
Journal of Social Archaeology, 2/​3: 323–​351.
Gardner, A. (2003). ‘Seeking a Material Turn: The Artefactuality of the Roman Empire’, in G.
Carr, E. Swift, and J. Weekes (eds), TRAC 2002: Proceedings of the Twelfth Annual Theoretical
Roman Archaeology Conference, Canterbury 2002. Oxbow Books, 1–​13.
 Changing Materialities   503

Gardner, A. (2007). An Archaeology of Identity: Soldiers and Society in Late Roman Britain.

Walnut Creek, CA: Left Coast Press.
Gardner, A. (2011). ‘Action and Structure in Interpretive Archaeologies’, in E. Cochrane and
A. Gardner (eds), Evolutionary and Interpretive Archaeologies: A Dialogue. Walnut Creek,
CA: Left Coast Press, 63–​82.
Gardner, A. (2012). ‘Time and Empire in the Roman World’, Journal of Social Archaeology, 12/​
2: 145–​166.
Gardner, A. (2013). ‘Thinking about Roman Imperialism: Postcolonialism, Globalization and
beyond?’, Britannia, 44: 1–​25.
Gell, A. (1998). Art and Agency: An Anthropological Theory. Oxford: Clarendon Press.
Giddens, A. (1984). The Constitution of Society:  Outline of the Theory of Structuration.
Cambridge: Polity Press.
Given, M. (2004). The Archaeology of the Colonized. London: Routledge.
Going, C. J. (1992). ‘Economic “Long Waves” in the Roman Period? A Reconnaissance of the
Romano-​British Ceramic Evidence’, Oxford Journal of Archaeology, 11/​1: 93–​117.
González-​Ruibal, A. (2003). ‘Restoring Ontological Security:  Roman and Native Objects
in Early Roman Gallaecia (NW Iberia)’, in G. Carr, E. Swift, and J. Weekes (eds), TRAC
2002:  Proceedings of the Twelfth Annual Theoretical Roman Archaeology Conference,
University of Kent at Canterbury 2002. Oxford: Oxbow Books, 30–​47.
Gosden, C. (1994). Social Being and Time. Oxford: Blackwell.
Gosden, C. (2004). Archaeology and Colonialism. Cambridge: Cambridge University Press.
Gosden, C. (2005). ‘What do Objects Want?’, Journal of Archaeological Method and Theory, 12/​
3: 193–​211.
Grahame, M. (2000). Reading Space: Social Interaction and Identity in the Houses of Roman
Pompeii:  A  Syntactical Approach to the Analysis and Interpretation of Built Space. BAR
International Series 886. Oxford: Archaeopress.
Greene, K. (2008). ‘Learning to Consume: Consumption and Consumerism in the Roman
Empire’, Journal of Roman Archaeology, 21/​1: 64–​82.
Harding, A. (2008). ‘A Conversation with Colin Renfrew (Professor Lord Renfrew of
Kaimsthorn)’, European Journal of Archaeology, 11/​2–​3: 143–​170.
Haverfield, F. (1912). The Romanization of Roman Britain. 2nd edn. Oxford: Clarendon Press.
Hill, J. D. (1995). ‘The Pre-​Roman Iron Age in Britain and Ireland (ca. 800 bc to ad 100): An
Overview’, Journal of World Prehistory, 9/​1: 47–​98.
Hill, J. D. (2001). ‘Romanisation, Gender and Class:  Recent Approaches to Identity in
Britain and their Possible Consequences’, in S. James and M. Millett (eds), Britons and
Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for
British Archaeology, 12–​18.
Hills, C. (2003). The Origins of the English. London: Duckworth.
Hingley, R. (1989). Rural Settlement in Roman Britain. London: Seaby.
Hingley, R. (1996). ‘The “Legacy” of Rome:  The Rise, Decline, and Fall of the Theory of
Romanization’, in J. Webster and N. J. Cooper (eds), Roman Imperialism:  Post-​Colonial
Perspectives. Leicester: University of Leicester, School of Archaeological Studies, 35–​48.
Hingley, R. (2000). Roman Officers and English Gentlemen: The Imperial Origins of Roman
Archaeology. London: Routledge.
Hingley, R. (2001). ‘A Comment on Ray Laurence’s Roman Narratives:  The Writing
of Archaeological Discourses—​a View from Britain?’, Archaeological Dialogues, 8/​
2: 111–​114.
504

504   Andrew Gardner

Hingley, R. (2008). ‘Not so Romanized? Tradition, Reinvention or Discovery in the Study of

Roman Britain’, World Archaeology, 40/​3: 427–​443.
Hingley, R. (2015). ‘Post-​Colonial and Global Rome: The Genealogy of “Empire”’, in M. Pitts
and M. J. Versluys (eds), Globalisation and the Roman World: Perspectives and Opportunities.
Cambridge: Cambridge University Press, 32–46.
Hingley, R., and Willis, S. (2007) (eds). Roman Finds: Context and Theory. Oxford: Oxbow Books.
Hitchner, R. B. (2008). ‘Globalization avant la lettre:  Globalization and the History of the
Roman Empire’, New Global Studies, 2/​2: 1–​12.
Hodder, I. (1985). ‘Post-​ Processual Archaeology’, in M. Schiffer (ed.), Advances in
Archaeological Method and Theory 8. New York: Academic Press, 1–​26.
Hodder, I. (1991). Reading the Past: Current Approaches to Interpretation in Archaeology. 2nd
edn. Cambridge: Cambridge University Press.
Hodder, I. (1993). ‘Bridging the Divide: A Commentary on Theoretical Roman Archaeology’,
in E. Scott (ed.), Theoretical Roman Archaeology:  First Conference Proceedings.
Aldershot: Avebury, pp. xiii–​xix.
Hodder, I. (1999). The Archaeological Process: An Introduction. Oxford: Blackwell.
Hodder, I. (2011). ‘Human–​ Thing Entanglement:  Towards an Integrated Archaeological
Perspective’, Journal of the Royal Anthropological Institute, ns 17: 154–​177.
Hodder, I. (2012). Entangled: An Archaeology of the Relationships between Humans and Things.
Chichester: Wiley-​Blackwell.
Hodder, I., and Hassall, M. (1971). ‘The Non-​Random Spacing of Romano-​British Walled
Towns’, Man, ns 6: 391–​407.
Hodson, F. R. (1964). ‘Cultural Groupings within the British Pre-​Roman Iron Age’, Proceedings
of the Prehistoric Society, 30: 99–​110.
Hurcombe, L. (2007). ‘A Sense of Materials and Sensory Perception in Concepts of Materiality’,
World Archaeology, 39/​4: 532–​545.
Ingold, T. (1993). ‘The Temporality of the Landscape’, World Archaeology, 25/​2: 152–​174.
Ingold, T. (2007). ‘Materials against Materiality’, Archaeological Dialogues, 14/​1: 1–​16.
James, S. (2003). ‘Roman Archaeology: Crisis and Revolution’, Antiquity, 77/​295: 178–​184.
Johnson, M. H. (1989). ‘Conceptions of Agency in Archaeological Interpretation’, Journal of
Anthropological Archaeology, 8/2: 189–​211.
Johnson, M. H. (2004). ‘Agency, Structure and Archaeological Practice’, in A. Gardner (ed.),
Agency Uncovered: Archaeological Perspectives on Social Agency, Power and Being Human.
London: UCL Press, 241–​247.
Johnson, M. H. (2006). ‘On the Nature of Theoretical Archaeology and Archaeological Theory’,
Archaeological Dialogues, 13/​2: 117–​132.
Johnson, M. H. (2011). ‘A Visit to Down House: Some Interpretive Comments on Evolutionary
Archaeology’, in E. E. Cochrane and A. Gardner (eds), Evolutionary and Interpretive
Archaeologies: A Dialogue. Walnut Creek, CA: Left Coast Press, 307–​324.
Jones, S. (1997). The Archaeology of Ethnicity. London: Routledge.
Jundi, S., and Hill, J. D. (1998). ‘Brooches and Identities in First Century ad Britain: More than
Meets the Eye?’, in C. Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 97: Proceedings of
the Seventh Annual Theoretical Roman Archaeology Conference, Nottingham 1997. Oxford:
Oxbow Books, 125–​137.
Knappett, C. (2005). Thinking through Material Culture:  An Interdisciplinary Perspective.
Philadelphia: University of Pennsylvania Press.
 Changing Materialities   505

Knappett, C. (2012). ‘Materiality’, in I. Hodder (ed.), Archaeological Theory Today.

London: Routledge, 188–​207.
LaMotta, V. M. (2012). ‘Behavioural Archaeology’, in I. Hodder (ed.), Archaeological Theory
Today. London: Routledge, 62–​92.
Latour, B. (1993). We Have Never Been Modern, trans. C. Porter. Cambridge, MA: Harvard
University Press.
Latour, B. (1999). Pandora’s Hope:  Essays on the Reality of Science Studies. Cambridge,
MA: Harvard University Press.
Laurence, R. (1999). ‘Theoretical Roman Archaeology’, Britannia, 30: 387–​390.
Laurence, R. (2001). ‘Roman Narratives: The Writing of Archaeological Discourse—​a View
from Britain? (With Comments and Reply)’, Archaeological Dialogues, 8/​2: 90–​122.
Lavan, L., Swift, E., and Putzeys, T. (2007) (eds). Objects in Context, Objects in Use: Material
Spatiality in Late Antiquity. Leiden: Brill.
Leone, M. (1984). ‘Interpreting Ideology in Historical Archaeology:  Using the Rules of
Perspective in the William Paca Garden in Annapolis, Maryland’, in D. Miller and C. Tilley
(eds), Ideology, Power and Prehistory. Cambridge: Cambridge University Press, 25–​35.
Liversidge, J. (1968). Britain in the Roman Empire. London: Routledge.
Lucas, G. (2001). Critical Approaches to Fieldwork. London: Routledge.
McAnany, P. A., and Yoffee, N. (2010). ‘Why we Question Collapse and Study Human
Resilience, Ecological Vulnerability and the Aftermath of Empire’, in P. A. McAnany and
N. Yoffee (eds), Questioning Collapse:  Human Resilience, Ecological Vulnerability and the
Aftermath of Empire. Cambridge: Cambridge University Press, 1–​17.
McCarthy, E. D. (1984). ‘Towards a Sociology of the Physical World: George Herbert Mead on
Physical Objects’, Studies in Symbolic Interaction, 5: 105–​121.
Malafouris, L., and Renfrew, C. (2010) (eds). The Cognitive Life of Things. Cambridge: McDonald
Institute for Archaeological Research.
Martin, A. (2005). ‘Agents in Inter-​Action: Bruno Latour and Agency’, Journal of Archaeological
Method and Theory, 12/​4: 283–​311.
Mattingly, D. (1997). ‘Dialogues of Power and Experience in the Roman Empire’, in D. Mattingly
(ed.), Dialogues in Roman Imperialism. Portsmouth, RI:  Journal of Roman Archaeology
Supplementary Series 23, 1–​24.
Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of
Roman Archaeology, 17: 5–​25.
Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Penguin/​
Allen Lane.
Mattingly, D. (2011). Imperialism, Power, and Identity:  Experiencing the Roman Empire.
Princeton: Princeton University Press.
Meskell, L. (2005). ‘Objects in the Mirror Appear Closer than they Are’, in D. Miller (ed.),
Materiality. Durham, NC: Duke University Press, 51–​7 1.
Miller, D. (1994). ‘Artefacts and the Meaning of Things’, in T. Ingold (ed.), Companion
Encyclopedia of Anthropology. London: Routledge, 396–​419.
Miller, D. (2008). The Comfort of Things. Cambridge: Polity.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Moore, T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and
Roman Studies’, Journal of Social Archaeology, 11/​3: 334–​360.
506

506   Andrew Gardner

Morphy, H. (2009). ‘Art as a Mode of Action:  Some Problems with Gell’s Art and Agency’,
Journal of Material Culture, 14/​1: 5–​27.
Naerebout, F. G. (2007). ‘Global Romans? Is Globalisation a concept that is Going to Help us
Understand the Roman Empire?’, Talanta, 38–​39: 149–​170.
Nesbitt, C., and Tolia-​Kelly, D. (2009). ‘Hadrian’s Wall: Embodied Archaeologies of a Linear
Monument’, Journal of Social Archaeology, 9/​3: 368–​390.
O’Hanlon, R., and Washbrook, D. (1992). ‘After Orientalism: Culture, Criticism, and Politics in
the Third World’, Comparative Studies in Society and History, 34/​1: 141–​167.
Olsen, B. (2006). ‘Scenes from a Troubled Engagement:  Post-​Structuralism and Material
Culture Studies’, in C. Tilley, W. Keane, S. Küchler, M. Rowlands, and P. Spyer (eds),
Handbook of Material Culture. London: Sage, 85–​103.
Olsen, B. (2010). In Defense of Things:  Archaeology and the Ontology of Objects. Lanham,
MD: AltaMira Press.
Olsen, B. (2012). ‘Symmetrical Archaeology’, in I. Hodder (ed.), Archaeological Theory Today.
2nd edn. Cambridge: Polity, 208–​228.
Osborne, R., and Tanner, J. (2007) (eds). Art’s Agency and Art History. Oxford: Blackwell.
Pauketat, T. (2001). ‘A New Tradition in Archaeology’, in T. Pauketat (ed.), The Archaeology of
Traditions: Agency and History before and after Columbus. Gainesville, FL: University Press
of Florida, 1–​16.
Peacock, D. P.  S. (1982). Pottery in the Roman World:  An Ethnoarchaeological Approach.
London: Longman.
Peña, J. T. (2007). Roman Pottery in the Archaeological Record. Cambridge:  Cambridge
University Press.
Pitts, M. (2007). ‘The Emperor’s New Clothes? The Utility of Identity in Roman Archaeology’,
American Journal of Archaeology, 111: 693–​7 13.
Pitts, M. (2008). ‘Globalizing the Local in Roman Britain: An Anthropological Approach to
Social Change’, Journal of Anthropological Archaeology, 27: 493–​506.
Pitts, M. (2010). ‘Artefact Suites and Social Practice: An Integrated Approach to Roman
Provincial Finds Assemblages’, Facta, 4: 125–​152.
Preucel, R. W., and Mrozowski, S. A. (2010). ‘The New Pragmatism’, in R. W. Preucel and S.
A. Mrozowski (eds), Contemporary Archaeology in Theory: The New Pragmatism. 2nd edn.
Chichester: Wiley-​Blackwell, 3–​49.
Reece, R. (1988). My Roman Britain. Cirencester: Cotswold Studies.
Reece, R. (1993a). ‘Theory and Roman Archaeology’, in E. Scott (ed.), Theoretical Roman
Archaeology: First Conference Proceedings. Aldershot: Avebury, 29–​38.
Reece, R. (1993b). ‘British Sites and their Roman Coins’, Antiquity, 67/​257: 863–​869.
Reece, R. (1997). ‘Writing Roman Britain:  Past Indicative, Future Perfect?’, Britannia,
28: 473–​478.
Reece, R. (2002). The Coinage of Roman Britain. Stroud: Tempus.
Renfrew, C. (1982). ‘Explanation Revisited’, in C. Renfrew, M. Rowlands, and B. A. Segraves-​
Whallon (eds), Theory and Explanation in Archaeology. New York: Academic Press, 5–​23.
Renfrew, C. (2012). ‘Towards a Cognitive Archaeology:  Material Engagement and
the Early Development of Society’, in I. Hodder (ed.), Archaeological Theory Today.
London: Routledge, 124–​145.
Revell, L. (2000). ‘Creation of Multiple Identities in Roman Italica’, in G. Fincham, G. Harrison,
R. R. Holland, and L. Revell (eds), TRAC 99: Proceedings of the Ninth Annual Theoretical
Roman Archaeology Conference, Durham 1999. Oxford: Oxbow Books, 1–​7.
 Changing Materialities   507

Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge

University Press.
Richmond, I. A. (1955). Roman Britain. Harmondsworth: Penguin.
Rippengal, R. (1993). ‘Villas as a Key to Social Structure? Some Comments on Recent
Approaches to the Romano-​British Villa, and some Suggestions towards an Alternative’,
in E. Scott (ed.), Theoretical Roman Archaeology:  First Conference Proceedings.
Aldershot: Avebury, 79–​101.
Rogers, A. (2011). Late Roman Towns in Britain:  Rethinking Change and Decline.
Cambridge: Cambridge University Press.
Salway, P. (1993). The Oxford Illustrated History of Roman Britain. Oxford:  Oxford
University Press.
Scott, E. (1993a). ‘Writing the Roman Empire’, in E. Scott (ed.), Theoretical Roman
Archaeology: First Conference Proceedings. Aldershot: Avebury, 5–​22.
Scott, E. (1993b). ‘Introduction:  TRAC (Theoretical Roman Archaeology Conference)
1991’, in E. Scott (ed.), Theoretical Roman Archaeology:  First Conference Proceedings.
Aldershot: Avebury, 1–​4.
Scullard, H. H. (1979). Roman Britain: Outpost of the Empire. London: Thames and Hudson.
Selkirk, A. (2006). ‘Review of An Imperial Possession:  Britain in the Roman Empire (D.
Mattingly, 2006)’, Current Archaeology, 205(18.1): 30.
Shennan, S. J. (2002). Genes, Memes and Human History: Darwinian Archaeology and Cultural
Evolution. London: Thames and Hudson.
Shennan, S. (2011). ‘An Evolutionary Perspective on the Goals of Archaeology’, in E. Cochrane
and A. Gardner (eds), Evolutionary and Interpretive Archaeologies:  A  Dialogue. Walnut
Creek, CA: Left Coast Press, 325–​344.
Shennan, S. (2012). ‘Darwinian Cultural Evolution’, in I. Hodder (ed.), Archaeological Theory
Today. London: Routledge, 15–​36.
Sillar, B. (2009). ‘The Social Agency of Things? Animism and Materiality in the Andes’,
Cambridge Archaeological Journal, 19/​3: 367–​377.
Silva, F. C. da, and Baert, P. (forthcoming). ‘Evolution, Agency and Objects: Re-​Discovering
Classical Pragmatism’, in A. Gardner, M. Lake, and U. Sommer (eds), The Oxford Handbook
of Archaeological Theory. Oxford: Oxford University Press.
Skibo, J. M., and Schiffer, M. B. (2008). People and Things: A Behavioural Approach to Material
Culture. New York: Springer.
Smith, A. (2007). ‘The Late Iron Age and Roman Landscape’, in D. Miles, S. Palmer, A.
Smith, and G. Perpetua Jones, Iron Age and Roman Settlement in the Upper Thames
Valley:  Excavations at Claydon Pike and Other Sites within the Cotswold Water Park.
Oxford: Oxford Archaeology, 391–​403.
Smith, A., and Powell, K. (2010). ‘In the Shadow of Corinium—​the Later Iron Age to Late
Roman Period’, in K. Powell, A. Smith, and G. Laws, Evolution of a Farming Community in the
Upper Thames Valley. Volume 1: Site Narrative and Overview. Oxford: Oxford Archaeology,
96–​166.
Stein, G. J. (2005) (ed.). The Archaeology of Colonial Encounters:  Comparative Perspectives.
Santa Fe: School of American Research Press; Oxford: James Currey.
Storey, G. R. (1999). ‘Archaeology and Roman Society: Integrating Textual and Archaeological
Data’, Journal of Archaeological Research, 7/​3: 203–​248.
Strathern, M. (1988). The Gender of the Gift: Problems with Women and Problems with Society in
Melanesia. Berkeley and Los Angeles: University of California Press.
508

508   Andrew Gardner

Swift, E. (2000). The End of the Western Roman Empire:  An Archaeological Investigation.
Stroud: Tempus.
Taylor, J. (2001). ‘Rural Society in Roman Britain’, in S. James and M. Millett (eds), Britons and
Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for
British Archaeology, 46–​59.
Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151.
York: Council for British Archaeology.
Thomas, C. (1990). ‘Review of Rural Settlement in Roman Britain (R. Hingley, 1989) and The
Economy of Romano-​British Villas (K. Branigan and D. Miles (eds), 1989)’, Antiquity, 64/​
242: 183–​184.
Thomas, J. (1996). Time, Culture and Identity:  An Interpretative Archaeology.
London: Routledge.
Thomas, J. (2004). Archaeology and Modernity. London: Routledge.
Thomas, J. (2007). ‘The Trouble with Material Culture’, in V. O. Jorge and J. Thomas (eds),
Overcoming the Modern Invention of Material Culture. Journal of Iberian Archaeology, 9–​10
Special Issue, 11–​24.
Thomas, J. (2012). ‘Archaeologies of Place and Landscape’, in I. Hodder (ed.), Archaeological
Theory Today. 2nd edn. Cambridge: Polity, 167–​187.
Tilley, C. (1984). ‘Ideology and the Legitimation of Power in the Middle Neolithic of
Southern Sweden’, in D. Miller and C. Tilley (eds), Ideology, Power and Prehistory.
Cambridge: Cambridge University Press, 111–​145.
Tilley, C. (1989). ‘Interpreting Material Culture’, in I. Hodder (ed.), The Meanings of Things.
London: Harper Collins, 185–​194.
Tilley, C. (1999). Metaphor and Material Culture. Oxford: Blackwell.
Trigger, B. G. (2006). A History of Archaeological Thought. 2nd edn. Cambridge: Cambridge
University Press.
Tyers, P. A. (1996). Roman Pottery in Britain. London: Batsford.
Van Oyen, A. (2012). ‘Knowledge Systems in the Production of terra sigillata:  Moving
beyond the Local/​Global Paradox’, in M. Duggan, F. McIntosh, and D. J. Rohl (eds), TRAC
2011:  Proceedings of the Twenty-​First Annual Theoretical Roman Archaeology Conference,
Newcastle 2011. Oxford: Oxbow Books, 48–​59.
Webmoor, T., and Witmore, C. L. (2008). ‘Things Are Us! A Commentary on Human/​Thing
Relations under the Banner of a “Social” Archaeology’, Norwegian Archaeological Review,
41/​1: 53–​70.
Webster, J. (1996). ‘Roman Imperialism and the Post-​Imperial Age’, in J. Webster and N. Cooper
(eds), Roman Imperialism:  Post-​Colonial Perspectives. Leicester:  University of Leicester,
School of Archaeological Studies, 1–​17.
Webster, J. (1999). ‘A New Roman Britain? Recent Developments in Romano-​ British
Archaeology’, American Journal of Archaeology, 103/​1: 122–​125.
Webster, J. (2001). ‘Creolizing the Roman Provinces’, American Journal of Archaeology,
105: 209–​225.
White, R. (2007). Britannia Prima: Britain’s Last Roman Province. Stroud: Tempus.
White, R., and Barker, P. (1998). Wroxeter: Life and Death of a Roman City. Stroud: Tempus.
Wilkinson, P. (2000). What the Romans Did for Us. London: Boxtree.
Witcher, R. (1998). ‘Roman Roads: Phenomenological Perspectives on Roads in the Landscape’,
in C. Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 97:  Proceedings of the Seventh
 Changing Materialities   509

Annual Theoretical Roman Archaeology Conference, Nottingham 1997. Oxford:  Oxbow

Books, 60–​70.
Witcher, R. (2000). ‘Globalisation and Roman Imperialism: Perspectives on Identities in
Roman Italy’, in E. Herring and K. Lomas (eds), The Emergence of State Identities in Italy in
the First Millennium bc. London: Accordia Research Centre, 213–​225.
Witcher, R. (2015). ‘Globalisation and Roman Cultural Heritage’, in M. Pitts and M. J. Versluys
(eds), Globalisation and the Roman World: Perspectives and Opportunities. Cambridge:
Cambridge University Press, 198–222.
Wobst, H. M. (1977). ‘Stylistic Behavior and Information Exchange’, University of Michigan
Museum of Anthropology Anthropological Papers, 61: 317–​342.
Woodward, I. (2007). Understanding Material Culture. London: Sage.
Woolf, G. (1991). ‘Review of The Romanization of Britain (M. Millett, 1990)’, Britannia,
22: 341–​342.
Woolf, G. (2004). ‘The Present State and Future Scope of Roman Archaeology: A Comment’,
American Journal of Archaeology, 108: 417–​428.
510

Chapter 25

F orms of Know l e d g e
Changing Technologies of Romano-​British Pottery

Jeremy Evans

Introduction

The field covered in this chapter is a very broad one and subject to considerable scholar-
ship and publications, so this can be no more than a rapid scan through parts of the field,
trying to address the questions asked of this author with a few examples. It is also very
much a personal view and does not attempt to reflect a consensus position. The editors
kindly informed the author that: ‘We envision this chapter addressing the social expla-
nations for and consequences of the changing forms of ceramic technologies and vessel
types, both in the immediate post-​conquest period, but also through to the late Roman
period. You may wish to consider the continuity of indigenous technologies, as well as
the adoption of new forms.’ The author will attempt to do this.

Prolegoma: From Caesar to Nero

To discuss Roman impact on ceramic use, technology, and form it is necessary to look at
these from the mid-​first century bc onwards. Some of the most significant transforma-
tions of previous Iron Age ceramic traditions take place in this period and what happens
in this period seems largely to determine subsequent developments.
Creighton (2000) points out that later ‘Iron Age’ coinage in the south-​east becomes
Romanized in the post-​Caesarian period in much the same way as coinages in other
Roman client-​kingdoms (Braund 1984). He points to strong parallels within the pool of
Roman client-​kingdom coinages and demonstrates that their imagery is clearly from a
direct knowledge of Rome itself. He speculates that this is the direct result of the Roman
education of prominent hostages and their subsequent return to their kingdoms. (He
 Forms of Knowledge    511

draws a further series of strong parallels between material from elite late pre-​Roman
Iron Age (LPRIA) graves and Roman ritual practice.) The fundamental point is that the
Caesarian conquest of the south-​east really was of significance and that, subsequent
to this, a series of client-​kingdoms existed here that provided Roman hostages, from
at least the Augustan period, who returned with a strongly ‘Roman’ outlook. This, of
course, parallels the histories of client-​kingdoms elsewhere in the Roman world.
Hill (2002) discusses in some detail the introduction of wheel-​made, predominantly
grog-​tempered pottery into East Anglia. He points to the association of this with new
cremation burial rites of the Aylesford–​Swarling tradition, oppida, and a close associa-
tion with north-​east Gaul. He also notes evidence of a knowledge of Gallic wheel-​made
vessels from the fifth century bc and highlights the question of why wheel-​made ceram-
ics were adopted in the first century bc /​ad rather than earlier (cf. Mepham 1997). He
points out (Hill 2002: 144) that, ‘despite the rejection of the notion of immigrants caus-
ing all change in late Iron Age south-​east England, the assumption that foreign potters
are necessary to start wheel-​turned pottery production remains to this day—​on those
rare occasions people have considered the problem.’ As Freestone and Rigby (1988)
argue:

These first wheel-​thrown ceramics are very similar to the pottery being produced
at this time in the Belgic areas of Gaul. Although the British pots are made in local
fabrics, typologically and technically they are identical to assemblages in northern
France… These exceptionally strong parallels seem to imply that the first of these
new wheel-​thrown wares were produced by immigrant or itinerant potters from
Gaul, working in Britain and using local clays.

Hill goes on to analyse the new pottery forms that appear with wheel-​made pottery,
to show that these represent a considerable widening of functional types available in
ceramics and that many appear to be associated with the consumption of alcohol. He
demonstrated this by comparing the height/​diameter ratios of mid and later Iron Age
pottery. The former have a linear relationship, as, essentially, most vessels were just
larger and smaller versions of the same basic form, while there is no relationship with
the latter, as many different functional types are represented. Hill suggests that it is only
at this time that wheel-​made pottery can be introduced, because a demand has appeared
at this time for drinking vessels and tablewares. ‘The cultural and social conditions (and
not just the technical and ‘economic’ conditions) needed to be right for the rapid adop-
tion of the technology and of its products by others’ (Hill 2002: 153). This is probably the
case and certainly suggests the technologically deterministic views of Peacock (1982);
linking modes of production to ceramic technology should be thoroughly questioned.
However, it is only because both the demand for new pottery types and the Gallo-​
Roman pottery to express them came from a series of classically educated Roman client-​
kings and their followers.
Creighton’s client-​kingdom model (2000) provides the explanation for the lack of
samian ware across most of the south-​east and the preponderance of ‘Belgic’ wares.
512

512   Jeremy Evans

The ‘Belgic’ horizon across the south-​east of England ought to be better regarded for
what it is: a Gallo-​Roman horizon. Sites such as oppida are deeply ‘Romanized’ in the
pre-​conquest period, in the Gallo-​Roman tradition of north Gaul (Evans 2001b: fig-
ure 9; 2005). In many ways these sites appear much more ‘Romanized’ than the early
Romano-​British sites such as the Claudian forts and London. They are, however, in a
different cultural tradition and one that was still probably economically ‘embedded’ (cf.
Reece 1979).
The appearance of personal toilet instruments (Hill 1997) offers a further matching
distribution, largely concentrated in the ‘eastern’ (Creighton 2000: figure A.1) coin-​
using area, where Creighton’s arguments (2000) make a very strong case for locating a
Roman client-​kingdom (Braund 1984). Figure 25.1 shows these from Hill’s study (1997)
superimposed onto the coin series area maps (Creighton 2000: figure A.1) and the dis-
tribution of Dressel 1 amphorae (Fitzpatrick 1985). The coincidence of the distribu-
tions is striking and eloquent. These toilet instruments need to be reconsidered in their
‘Gallo-​Roman’ context. The ‘growth of the individual’ and the ‘changing lifestyles’ rep-
resented by these must now be seen in the context of social emulation of a Gallo-​Roman
elite within a Roman client-​kingdom (cf. Hill 1997; Jundi and Hill 1998). If they repre-
sent growing individualization then they reflect it as a growing ‘Roman’ individualism.

Kiln
Contour Bar
80%
70%
60%
50%
40%
30%
20%
10%
Western IA Coin Core Distribution 0%

Figure 25.1  Map showing quantitative evidence for the distribution of Severn Valley wares, in
comparison with the core distribution of the western Iron Age coinage group.
Note: Columns represent the percentage at a site; dashes represent sites without Severn Valley wares. Known kiln sites
producing Severn Valley wares are shown. Both the contour map and the bar plots show that there is a very sharp fall-​off at
the edge of the distribution of the fabrics.
Source: © J. Evans.
 Forms of Knowledge    513

One of the great strengths of Creighton’s model (2000) is that it is based on observa-
tion of trends in the material evidence (in this case coinage) and is not simply an abstract
model or one applied fashionably from a different period or discipline, as is so prevalent
in much of academic archaeology.
Another point not much remarked upon is the very widespread use of a single tem-
per type, namely grog, in this ‘Gallo-​Roman’ horizon. Although it is occasionally used
as a temper type later in the Roman period, it is far from common, yet at this period it
is near universal. At a basic functional level the grog temper will have provided good
thermal shock resistance, as it has a low coefficient of thermal expansion (Rye 1976; A.
Woods 1983–​4), and thus would help the vessels survive in bonfire and simple kiln fir-
ings. The very widespread use of this temper type, rare in following periods, like the
common koine of ‘Gallo-​Roman’ forms, points to a single potting tradition with a tradi-
tional clay preparation recipe, clearly imported from north Gaul and employed by pot-
ters who were probably itinerant (P. J. Woods 1974) and operating in an ‘embedded’ Iron
Age economy.
It is interesting to look at the pottery tradition in England outside Hill’s study area
(2002) in this period. Here, pottery was used either not at all or, very rarely, in Wales,
Shropshire, the north-​west, and the Pennines, or handmade traditions, often three-​
quarters of a millennium old, continued largely unchanged, as in East Yorkshire, or
with the Cornish Gabbroic tempered wares, and in Dorset Durotrigian handmade
wares (Brailsford 1958) change little to become BB1 (Black Burnished ware category 1
(Williams 1977)). It is very noticeable that most areas that were not pre-​Claudian client-​
kingdoms do not seem to contain individuals who feel the need for the material expres-
sion of their ‘individualism’ or the need for a kit of drinking vessels that just happen to
match those used in north Gaul!

The Development of Romano-​British

Traditions: ad 61+

In comparison with the ‘Gallo-​Roman’ pottery being used on oppida and in the country-
side in the Claudio-​Neronian periods, very different assemblages are seen on military
sites and on the truly ‘colonial’ towns at Colchester, London, and Verulamium (‘colonial’
here being used as a synonym for ‘planted towns’, not for centres with the Roman status
of a colonia). Here the ceramics looked very different. Rather than Terra Nigra and Terra
Rubra as principal finewares, as in the countryside (Evans et al., forthcoming; Geoff
Dannell 2011, Roman Pottery Study Group conference) or in north Gaul (Déru 1996:
225), samian ware is the principal fineware (Evans 2005; Pitts and Perring 2006; Evans
et al., forthcoming), and, whereas most ‘Gallo-​Roman’ assemblages are predominantly
reduced, these are substantially or predominantly oxidized (Evans 2005: figure 9.9).
It can be objected that, as a municipium, Verulamium ought to be different, but the
514

514   Jeremy Evans

ceramic evidence does not suggest so; the town appears to have predominantly immi-
grant settlers, although the grandees controlling it from Gorhambury and the like were
probably indigenous (Evans et al., forthcoming).
Quantities of ‘Gallo-​Roman’ forms are relatively low on these forts and ‘colonial’
towns compared with oppida assemblages (cf. Willis 1997: table 5). This author (Evans
2001b: figure 9) suggested how oppida assemblages, in particular in Britain and north
Gaul, are dominated by drinking vessels, which are much more common there than
on major Romano-​British ‘colonial’ town groups of Boudiccan date (data from Millett
1983), and Pitts and Perring (2006) have provided further confirmation. The common
use of samian ware in the ‘colonial’ towns and forts reflects similar usage in military
assemblages on the Rhine frontier. This is part of a core Mediterranean tradition, then
already over 350 years old, deriving from late Hellenistic types of red gloss vessels, which
are skeumorphs of bronze ones, and thus often heavily carinated and sometimes stamp
decorated. The technology of samian-​ware production was highly sophisticated, using
muffle kilns and with vessels thrown in moulds or over templates (Webster 1996), but it
was almost absent from the countryside until the major revolution in ceramic use in the
south-​east in the decade ad 60–​70.
The ‘Gallo-​Roman’ horizon described thus far, extending back to a little before the
turn of the millennium and including strong evidence of immigrant Gallic potters from
sites such as Rushden (Woods and Hastings 1984), comes to an end in the decade ad
60–​70 when grog-​tempered pottery production largely ceases (Pollard 1988: 40; Evans
et al., forthcoming). New permanently based kiln sites seem to appear from this period
onwards; for example, Horningsea (Evans et al., forthcoming), the north Kent industries
(Pollard 1988), and Alice Holt (Lyne and Jefferies 1979), while samian ware, previously
restricted to military and immigrant Roman ‘colonial’ sites, starts to reach rural sites
(Millett 1980; Evans 2005; Pitts and Perring 2006; Evans et al., forthcoming). The peri-
patetic potters of the earlier period seem to have disappeared, and so do the imported
Terra Nigra fineware vessels from the south-​east, although small quantities are still
imported among the military supply to the forts of the west and north (Rigby 1977).
The reason for this seems reasonably clear, if unfashionable—​namely, the suppression
of the client-​kingdoms and the imposition of regular Roman provincial administration
in the aftermath of the Boudiccan Revolt. Alongside this, in the south-​east, probably
came direct money taxes and the breakdown of an embedded Iron Age economy (cf.
Reece 1979).
It is of note that the tendency for pottery kilns in the early Roman period to have
urban locations largely develops from the Flavian period, in contrast to the scattered
rural locations of ‘Belgic’ pottery kilns.
Elsewhere things are rather different. In the ‘highland zone’ there is very little change,
except in forts, their vici, and settler towns. Gabbroic pottery continues in the far south-​
west, ‘Romanizing’ a little in form (Peacock 1988). Wales (excluding the south Welsh
plain), Shropshire, and the north-​west, along with County Durham, remain largely
aceramic. In the pre-​Hadrianic period, north Welsh rural sites are mainly restricted to a
very few pieces of, often decorated, south Gaulish samian ware and, after that, to samian
 Forms of Knowledge    515

ware and BB1, which make up most of the less than 500 sherds that generally come from
such sites (Evans, forthcoming a). These would appear to be the only types of Roman pot-
tery, apart from mortaria, that were in demand on these sites, since a much wider range
of types was available at local fort vici. County Durham remains aceramic throughout
the Roman period and demonstrates continuity in that throughout the following pagan
Anglo-​Saxon period (Haselgrove et al. 1988). East Yorkshire continues its strong hand-
made, largely calcite gritted, Iron Age ceramic tradition in the first and second centuries
ad, with very little impact from ‘Roman’ forms or fabrics outside the forts and vici (Evans
1995). In the neighbouring Vale of York, pottery is little used on most rural sites, and most
of what is used is in handmade fabrics of Iron Age tradition in the first and second cen-
turies—​the exception being the sites of indigenous morphology at Lingcroft Farm, just 3
miles from York, where oxidized fabrics from York actually penetrate.
Similarly in West Yorkshire, Roman pottery does reach rural sites in low levels, with
generally less than 500 sherds from a site. What pottery does reach these sites is mainly
Roman greywares, chiefly from the South Yorkshire industry (Buckland et al. 1980).
Between the ‘highland zone’ and the south-​east there are some interesting traditions
in the south-​west. Pottery traditions in the territory usually assigned to the Dobunni
are curiously divergent. In the southern part of the area, Severn Valley wares are largely
absent, although they do reach the oppidum at Bagendon, along with grog-​tempered
wares (Trow 1988). However, in the Severn Valley, from as far south as the Frocester villa
and as far west as Ariconium, wheel-​made oxidized Severn Valley wares appear as the
Transitional horizon. Timby (1990: 251) observed that:

the Severn Valley ware industry appears to owe its origins to a pre-​conquest late Iron
Age tradition marked by the introduction of wheelmade wares… Many of the vessel
types produced can be recognized in other regions of Southern Britain in local pre-​
conquest assemblages, notably the butt-​beakers, platters and necked cordoned jars.
The tankards/​handled mugs, however, are less common and can only be paralleled
by Durotrigian types.

Severn Valley wares would appear to be part of a ‘Romanizing’ tradition but one sepa-
rate from the south-​eastern grog-​tempered wares, against which their distribution butts
up in Warwickshire. This context also helps to explain another feature of Severn Valley
wares, which can be seen particularly in Warwickshire and to the south—​namely, that
the tradition seem to have a very ‘hard boundary’. Figure 25.2 shows a quantified distri-
bution map of Severn Valley wares in the second to fourth centuries (while this broad
time period masks some trends in frequency with time, it does give a fair representation
of the general frequency of the fabrics).
There is no good evidence of urban marketing of Severn Valley wares or the
Malvernian Metamorphic tempered ware that usually accompanies them (Evans 2005,
contra Hodder 1974) or of most Romano-​British coarsewares. As Figures 25.1 and 25.2
show (leaving aside military sites further north), there seems to be a sharp fall-​off of
Severn Valley wares at the edges of their distribution, a phenomenon that is generally
516

516   Jeremy Evans

Dressel 1
Toilet Instruments

North-Eastern
East Anglian

Western

Eastern
South -Eastern

Southern
South-Western

Figure 25.2  Map showing the core areas of the seven regional coin series of Iron Age Britain,
after Creighton (2000: figure A.1) compared with the distribution of Dressel 1 amphorae finds, after
Fitzpatrick (1985) and the distribution of pre-​conquest Romanized toilet instruments (Hill 1997).
Source: © J. Evans.

interpreted (Hodder and Orton 1976: figure 5.81; cf. Evans 1988) as reflecting social con-
straints at crossing group boundaries.
If we take Creighton’s work (2000) as a starting point, the appearance of Severn Valley
wares is part of the horizon of ‘Gallo-​Roman’ material that seems to be closely associated
with the presence of ‘friendly kings’. The form repertoire is similar to that employed in
other areas in the south-​east, except for the addition of tankards, which clearly reflect a
local cultural trait, but the fabric is very different. This is clearly a matter of local choice;
knowledge of the form repertoire in the rest of lowland Britain clearly implies that
information was equally available on the nature of the fabrics used there. It may well be
significant that the fabric used was an oxidized ware, the predominant fabric tradition
in the Mediterranean. Severn Valley wares clearly start as a local variant of the south-​
eastern ‘Gallo-​Roman’ horizon with a strong cultural knowledge of the ‘Roman’ types
they are copying. Yet, while within the same geographical area covered by the Dobunnic
coinage, the south-​eastern part of this area uses grog-​tempered wares like south-​eastern
England. If Severn Valley wares were just a distinctive type of grog-​tempered pot-
tery that could be traced to a common origin, then it could reasonably be argued that
their distribution was determined by distribution constraints and marketing factors.
 Forms of Knowledge    517

However, Severn Valley wares were actually making the same culturally ‘Gallo-​Roman’
statements in terms of form as the pottery in the rest of the Dobunnic coinage area,
while strongly differentiating the ceramics in the northern area from that in the south-​
east of the territory and using the colour of most Mediterranean pottery. It hardly seems
probable that this is meaningless, and it is very difficult to see this suggesting a lack of
deliberate differentiation from the southern area. Given a largely common form rep-
ertoire, why should the northern group utilize a fabric so different from the southern
group and the rest of the south-​east if it was not also defining a social group? Similarly
why should Severn Valley wares map onto a major difference in rural site drinking hab-
its, or have what appears to be a socially constrained distribution throughout the Roman
period, if their distribution had no significance in terms of social identity?
In contrast, in Dorset there is considerable continuity of Iron Age pottery traditions.
Brailsford (1958) identified the range of types in Durotrigian ware, known as BB1 in the
Roman period (Williams 1977; see also Holbrook and Bidwell 1991: 90–​91). As Cunliffe
(1974: figure A30) points out, handled tankards are also among the repertoire. ‘Gallo-​
Belgic’ imports are largely absent and do not seem to have had much influence on the
forms of Durotrigian vessels, although Gallic imports, in the form of Hengistbury
class B wares and graphite coated ware, are present, and Cunliffe (1974: 98) suggests the
forms of ‘the necked bowls and tazze may well have been derived’ from these imports.
There is no evidence from the coinage that this area was ever a Roman client-​kingdom.
Durotrigian pottery seems to have been supplied across the area of the Durotrigian
coinage distribution from Poole Harbour from the late Iron Age onwards (Holbrook
and Bidwell 1991: 90). It continued to be until the end of the Roman period. Allen and
Fulford (1996: figures 1 and 8) have provided quantified distribution maps for BB1 in
general for the second to fourth centuries in the south-​west. It is clear the fabric was
the dominant type throughout the Durotrigian area, but that it travelled much further
afield. As was noted, the army developed an early relationship with the BB1 industry;
it supplied military installations in the south-​west from the first century (Holbrook
and Bidwell 1991: 91), as well as having an offshoot industry, South Western BB1 to the
west, which played a major role in the supply of Exeter and other south-​western forts.
It is interesting to note that the early fort at Lake Farm, Dorset (Darling 1977), was sup-
plied by kilns at Corfe Mullen, ‘which produced not only the well-​known flagons and
mortaria but also wheel-​thrown grey ware jars in Durotrigian-​inspired forms (Calkin
1935, classes C, D, and E). It seems probable that the kiln was set up specifically to serve
the fortress at Lake Farm… and closed down when the army moved on c. ad 55–​60’
(Holbrook and Bidwell 1991: 91). Although wheel technology was introduced early in
the Roman period to potters working in the Durotrigian tradition, here this had no
effect on the technology of Durotrigian/​BB1 vessels.
Allen and Fulford (1996: 266) note that:

one puzzling aspect of the level of supply in the region is that it was possible for a sec-
ond major industry to emerge in the New Forest less than 45km from Poole Harbour
in the mid-​to late third century. Although the latter produced fine wares which did
518

518   Jeremy Evans

not compete with SEDBB1, a major part of its production was the domestic, grey
wares whose distribution was almost mutually exclusive with that of SEDBB1. Thus,
paradoxically, while there was an apparent inability to supply demand several kilom-
eters to the east of Poole Harbour, there was no difficulty in maintaining a supply to
the Somerset coast.

This might be restated to say that there is a more rapid fall-​off than expected at the
eastern boundary of Durotrigian territory, which would seem to imply a fairly ‘hard
boundary’ here.
BB1 producers developed a relationship in army supply from the mid-​first century,
which was maintained until the mid-​fourth century, and also links with the negotiatores
who supplied urban markets in the south of the province.
What is very clear in discussing continuing Iron Age traditions is that they generally
take place in those areas that used ceramics fairly heavily in the Iron Age but did not
have a ‘Belgic’ pottery horizon. A semi-​exception to this was the Severn Valley, where
Severn Valley wares could, in some ways, be regarded as such a horizon. Other excep-
tions may be the Harrold shell-​tempered ware industry and the Milton Keynes pink-​
grogged ware in the south-​east.
It seems notable that all the western and northern industries maintained fairly ‘hard
boundaries’—​that is, the local fabric dominates local supply and adjacent fabric groups
tend to fall off heavily at the boundary of the region—​although at some period the local
fabric may also travel (considerable) distances beyond its core territory. Most of these
areas, in contrast to those that were largely aceramic, seem to have been integrated into a
Roman market economy and saw villa development in the later Roman period.

The Development of Romano-​British

Civilian Traditions

The non-​military pottery traditions that developed in England from the Flavian period
onwards were predominantly of the ubiquitous sandy greywares. These do seem to
become the model for Romano-​British pottery, particularly in the south-​east. As has
been commonly observed, pottery kilns tend to be frequent in the early Roman period
and tend to have an urban or near urban location, with kilns becoming fewer and larger
in scale in the later Roman period and having rural locations, often near civitas bounda-
ries (Swan 1984: maps 4–​7; Millett 1990: figure 52, after Saunders 1986). It seems to be
assumed that the near urban location of early kilns is related to marketing, although
there is scant evidence for urban marketing for coarsewares (Evans 2005) and pre-​
Flavian kilns are almost universally rural. One question for pottery production in the
south-​east might be how many of the new Flavian industries were actually founded by
immigrant potters, which might be what their urban location is implying, or how far
 Forms of Knowledge    519

does their location reflect the growing importance of towns as a market economy starts
to develop. Greywares become the predominant Romano-​British pottery type, and it
would appear clear that purchasers wanted this, but greywares are not predominant in
the military pottery traditions of the first century and are not a common fabric class in
most parts of the wider Empire, although they are fairly common in the Low Countries.
Evidence of migrant continental civilian potters can be adduced for a number of
industries. As noted later, the glazed ware potters were clearly immigrants. Equally,
among the Verulamium region wares, the form range included Gauloise-​type ampho-
rae and ‘honey pots’, and a continental origin for the potters would be unsurpris-
ing. Similarly, the Colchester mortaria-​makers starting production in the pre-​Flavian
period were also probably immigrants. Tyers (1996: table 19) lists some of the migrant
potters operating in the pre-​Flavian period, both military and civilian, noting that they
divide into a Rhenish tradition and north Gaulish ones—​in the long term, the former
tradition being the more important in the development of Romano-​British forms. Tyers
(1996: 63) also points to the insular development of continental types by the Flavian
period and the separate development of Romano-​British traditions, particularly among
flagon types.
As Millett (1980) demonstrated in West Sussex, and as other studies have clarified
(Evans et al., forthcoming), samian ware does not effectively reach rural sites until the
Flavian period. However, it rapidly becomes the most important fineware on most sites
until there is a crisis of supply in the late Flavian and Trajanic periods (Marsh 1981),
when Les Martres-​de-​Veyre was producing in central Gaul but La Graufesenque had
ceased production in south Gaul and Lezoux had not started mass production. The
quantity of Les Martres production was not great, and a samian shortage resulted. This
led to a flourishing of a series of other fineware industries (Marsh 1978), particularly of
‘London-​type’ black gloss finewares and mica-​dusted wares, all largely copying samian
forms, although there were also small quantities of glazed wares.
One side note that should be made here is about glazed wares. These were definitively
studied by Arthur (1978). They appear from the later Flavian period. The potters were
pretty certainly immigrants to the province, and a series of production centres have been
isolated (Arthur 1978). Forms in these glazed wares tend to be decorated bowls, beak-
ers, and flasks. Arthur (1978: 295) notes: ‘Apart from two undoubted military industries,
the other British centres largely marketed their glazed pottery to the civilian population
in southern England and the Midlands. By the second quarter of the second century
ad production appears to have ceased.’ The forms of these vessels indicate their use as
finewares, and they clearly were not regarded as technologically superior to other table-
ware or drinking vessels, despite modern perceptions. As Arthur (1978: 296) suggests,
they seem to have been put out of business by the massive expansion of samian-​ware
production at Lezoux in the Hadrianic period, like many other small Flavian−Trajanic
fineware producers. Glazed wares do not appear again in Roman Britain—​and do not
become common as tablewares in the Roman world—​until the seventh and eighth cen-
turies ad, in Constantinople (Hayes 1997: 64).
520

520   Jeremy Evans

Mortaria supply in south-​east England and in Britain in general is of interest.

Mortaria are assumed to be a ‘Roman’ type par excellence by those who have not looked
at Mediterranean Roman period pottery assemblages. As Hartley’s map (1998: figure 8)
of the distribution of stamped mortaria neatly illustrates, however, apart from on the
west coast of Italy, they are largely very few or absent in most of the Mediterranean and,
apart from Britain, which seems to be their chief home, they are common in Switzerland
and in Gallia Belgica. Mortaria are present, but very rare, in assemblages from pre-con-
quest oppida and are even rarer on pre-​Flavian rural sites. In Flavian–​Trajanic groups
in Cambridgeshire and Bedfordshire ‘mortaria reach the urban sites in small numbers
but are generally absent from the rural sites… More widespread mortaria usage does
not take place before the Hadrianic period… The phenomenon seen here is part of a
wider pattern across southern England (Rush 1997: figure 1) of the non-​use of mortaria
on rural sites before the second or third centuries’ (Evans et al., forthcoming: ch. 5). In
the south of England mortaria supply in the early Roman period is also remarkably
dominated by continental sources, with few indigenous industries setting up to compete
(Rush 1997). This, he suggests, is economically irrational because of transport costs, but
this does not allow for the huge subsidy to civilian transport costs from the large-​scale
movement of military supplies (Carreras Montfort 2002; Evans, forthcoming b).
In contrast, mortaria were popular in the north from the conquest, and not just on
military sites. Cool (2004b) made some interesting observations on northern mortaria
(although the contrast with mortaria use in the south is not particularly recognized),
noting that mortaria were often in high demand on basic level rural sites (many of these
being ones that barely used Roman ceramics (Evans, forthcoming a)) and, rightly, sug-
gesting that ‘this must surely cast further doubt on the assumption that they were all
used as utensils in a Romanised kitchen’ (Cool 2004b: 31). She also highlighted the fre-
quency with which mortaria are sooted, implying they have been used in cooking rather
than just grinding, something that subsequent evidence largely confirms (e.g. Evans and
Mills 2008). Unlike in the south, it would appear that, in the north, there was a military
demand for mortaria, which reflected that of military forces on the Rhine frontier, but
there was also a civilian demand, even on ‘highland-​zone’ rural sites that barely used
Roman pottery, and the latter would appear to be to fulfil an existing need, which clearly
was absent in the south. A possible explanation for this demand might be porridge,
especially as Iron Age type vessels with heavy internal carbonization seem to disappear
rapidly from the region in the areas where they were found (Evans 1995; Evans and Mills
2008; Evans, in Mills, forthcoming).
Essentially, later Roman fineware industries—​except Nene Valley and Crambeck —​
produced, in large part, samian-​type copies once samian wares had become sufficiently
infrequent for there to be a demand for copies. That the quantity of samian wares reach-
ing Britain after the end of the second century drops massively is well known (Marsh
1981). The reason for it is less clear. It is also widely reported that East Gaulish samian
wares tend to have an east coast distribution (Willis 2006), something not seen with
South and Central Gaulish material. There is good evidence that the interprovincial
supply of olive oil (and wine) for the army ceases in the first half of the third century
 Forms of Knowledge    521

(Carreras Montfort and Funari 1998; Carreras Montfort 2002; Evans, forthcoming b),
and strong indications that the end of this trade massively reduced the level of inter-
provincial trade with Britain, as it took away a large indirect subsidy to freight costs.
However, this seems to take place in the second quarter of the third century, rather
than at the end of the second when Central Gaulish ware apparently ceases to be sent to
British markets.
Among large-​scale coarseware producers, what does seem clear is that most major
kiln sites that were established by the second century succeeded in surviving until the
end of the Roman period. There do not seem to be common reasons as to why those kiln
sites that came to an end did so and, indeed, generally there is not much clarity on why
industries did come to an end. There does not seem to be any parallel extinction horizon
comparable with the ending of many pre-​Flavian kilns in the south-​east.

Regional Types

One of the more interesting areas in Romano-​British pottery studies is the strongly
emerging evidence of differences, not simply between pottery use on different types of
site and the changes in these with time, but also in different regional traditions that are
not apparently related to sources of supply.
One vessel type that is well known is the tankard. In the Severn Valley tankards are the
commonest drinking vessel on all types of site. These tankards seem to be quite as com-
mon on rural sites as on urban ones (Evans 1999). This is a phenomenon that appears
unique in the province; in other regions, apart from on oppida and high-​status transi-
tional period sites, drinking vessels are almost invariably found in much lower quan-
tities on basic level rural sites than on urban ones. However, in the Severn Valley this
is not the case (Evans 2001b: figure 7), with drinking vessels, usually tankards, being
remarkably common on basic level rural sites and occurring at levels usually associated
with an urban assemblage.
Evans (1999: 118) suggested: 

Speculatively one might suggest a society where the consumption of drink, presum-
ably alcoholic (?cider/​?beer) was deeply embedded in social relations and that the
conversion of agricultural surplus into such and the provision of hospitality with this
was a means of acquiring and/​or maintaining social standing. Either way this unique
regional feature in the composition of pottery assemblages would appear here to go
back to the Iron Age and not to have been a result of ‘Romanization’ in ceramic use.

In Warwickshire and Oxfordshire, with outliers in Worcestershire and possibly

Cambridgeshire, there are a series of fired clay discs that seem to indicate a regional food
tradition. A reasonably large assemblage of these has been recovered recently from a series
of sites excavated for the Transco pipeline in northern Warwickshire (Evans 2009: figure 46).
522

522   Jeremy Evans

The discs all have squared edges and seem to be either smoothed on both surfaces or
smoothed just on their apparent upper surface. Some are burnt, and burning tends to
be on the underside, while some others show evidence of carbonized deposits on their
upper surfaces, near their circumference. The most obvious explanation to this author is
that they are ‘chapatti discs’, bakestones for making some form of bread or pizza.
As Booth (2001) points out, there is no evidence in Oxfordshire for any of these discs
on Iron Age sites in the area as in Warwickshire. The earliest dated examples at Alchester
and Tiddington might suggest a late-​first-​century date. It may be that earlier occur-
rences for them will be found in Oxfordshire, but currently there is no published evi-
dence known to this author to substantiate this. Thus, they appear to be an innovation
of the Roman period. However, they do not seem to originate elsewhere, and the strong,
but erratic, rural distribution of the type, and comparatively weak urban distribution
(common in the village at Tiddington but with only a single example from Alcester),
suggests a fairly plebeian fashion.
If this interpretation is accurate, then the appearance of the discs would seem to sug-
gest the development of a possibly novel cuisine in the west Midlands in the Roman
period, probably mainly among the peasantry.
Another class of Romano-​British pottery vessel that seems to have distinct regional
variations as to presence and frequency are wide-​mouthed jars, often similar to what
in the Mediterranean would be called basins. These seem to be common in the Severn
Valley and around the Humber, but are virtually absent in many other areas.
Cool (2006) has produced a stimulating study that makes a first attempt to reintegrate
the evidence from pottery studies, glass studies, artefact studies, documents, and envi-
ronmental evidence actually to look at what may be fashionably described as food ways
in the late Iron Age and Roman Britain. She looks at the different strands of evidence
individually and then takes a series of groups to illustrate change and the contempo-
rary differences between sites. The principal evidence actually drawn on are functional
analyses of pottery (and glass) assemblages and species proportions in the animal bone.
Given the evidence from functional analyses of pottery assemblages (Evans 1993, 2005),
it is perhaps unsurprising that she concludes: ‘The cases presented in this chapter have
shown that deep-​seated regional differences persisted well into the Roman period and
that even in the third century Britain was far from sharing a common cuisine or attitude
to food’ (Cool 2006: 219). Other types that seem to exhibit regionality include face pots
(Braithwaite 1984, 2007) and triple vases.

The Late Roman Period

In the late Roman period pottery use, or non-​use, continued largely as it had in the
early Roman period. Areas that eschewed pottery use in the early Roman period did
not ‘catch-​up’; indeed, it seems probable that pottery use on rural sites in north Wales
actually ceased by the mid-​fourth century (Evans, forthcoming a), this perhaps relating
 Forms of Knowledge    523

to the apparent end of coin use on these sites at much the same period (Davies 1983). The
infrequency of late-​fourth-​century pottery on rural sites in West Yorkshire might also
give occasion to speculate on whether pottery use here was also declining. Equally, areas
that predominantly used handmade pottery in the first century, in East Yorkshire and
the south-​west, are still doing that in the fourth.
Within the pottery-​using areas of the diocese the fourth century, and particularly its
latter half, appears to see a major divide in the ways pottery was used. On rural sites in
the north and south during the earlier Roman period, the secular trend in the functional
composition of assemblages had been gradually to diversify from a jar-​based (mid) Iron
Age-​style assemblage to one where other functions were represented and more table-
wares were present (Evans 1993). Even in London this can be seen to be operating to
some extent in the early Roman period (Davies et al. 1994: figure 148). Millett (1979) pro-
duced functional sequences from three south-​eastern towns, Chichester, Verulamium,
and Neatham. All the sequences show a clear trend from high early jar levels to low late
ones and rising tableware levels with time. That these are typical of the pattern in the
south-​east is confirmed by the material from the A421 Alchester suburb (Evans 2001a)
and by the clear trends from many sites in Cambridgeshire and southern Bedfordshire
(Evans et al., forthcoming). Particularly of note is the jump to high tableware and low
jar levels on sites of the very latest fourth–​early fifth-​century date (Evans et al., forth-
coming). (Cool (2006: 230) unfortunately misses this south-​eastern trend and takes the
northern pattern to be much more widespread than it is.)
In the north, however, jar levels, having fallen in the first to third centuries, start to
rise again in the early fourth, and rise considerably in the later fourth/​early fifth centu-
ries (Evans 1985, 1993). Sites such as Ilchester, Cirencester, and Alcester may be part of
an intermediate zone where jar levels neither rise not fall, but further work is required
on this.
A similar dichotomy between north and south can be seen in the levels of finewares
on sites. In the north, fineware levels are generally fairly low and tend to decline in the
earlier fourth century (although they may rise very slightly in the later fourth), but even
on forts, at the top of the settlement hierarchy, they range only from 9 per cent to 14
per cent (Evans 1985; 1993: 111). In the south, however, although fineware levels fluc-
tuate across the area in part, at least, probably depending on the status of the site and
the proximity of fineware kiln sites, the trend of rising overall fineware levels from the
later third century onwards, which generally peak in the very latest ‘Roman’ deposits, is
pretty clear.
Evans (2001a: 372) noted: 

This author (Evans 1996) has interpreted this contrast as principally owing to dis-
tribution mechanisms operating in the later fourth century with Oxfordshire and
South Western Brown slipped ware fineware tablewares being lighter and more
easily distributed than Southern Shell tempered coarsewares from Harrold across
much of the Midlands, in particular. However, at Alchester local sources of both
tablewares and coarsewares are available and the site still conforms to this pattern,
524

524   Jeremy Evans

perhaps suggesting some more genuine cultural factors influencing the divergence
between northern and southern patterns of ceramic use in this period. The de-​
Romanization of much of the northern material perhaps being the most obvious fac-
tor (cf. Evans 1985).

A further point about fourth-​century finewares in the north that might be noted is
the near absence of red finewares. As noted earlier, these are the ‘Roman’ tradition par
excellence, dating back to the Hellenistic period and continuing in the Mediterranean
for another 300 years. Essentially, northern pottery assemblages are becoming more and
more similar to those of the East Yorkshire Iron Age tradition as the fourth century goes
on. From the mid-​fourth century onwards, the vast majority of these assemblages are
all supplied from kiln sites in East Yorkshire, although this appears to have been as a
result of a military contract arrangement for the supply of northern forts (Evans 1989).
The majority of this pottery was of handmade calcite gritted ware jars, drawing heavily
on East Yorkshire traditions stretching back into the Iron Age. Nonetheless, although
pottery supply for the frontier may have been localized within Britannia Secunda, it was
still as organized as it had been when it came from the south of the diocese. Dobney
(2001: 44) points to a similar level of organization in meat supply at the signal station at
Carr Naze, Filey, Yorkshire, and that ‘the evidence from Filey indicates that this central-
ised administrative network still possessed much of the “cultural baggage” of the Roman
tradition as late as the very end of the fourth century’.
In the south-​east the latest horizon is a logical development of previous trends and,
at one level, would seem to imply a strongly ‘Roman’ identity to them. However, the
occurrence of the groups—​which are relatively few and seem to be found only on some
urban and villa sites—​and their composition suggest a level of economic stress. In most
of the locations of the late pottery groups they seem to have been supplied by road over
quite considerable distances by major centralized industries. Local industries seem
to have ceased or, at least, seem to be poorly or not represented in these groups (for
example, the lack of Horningsea ware at Teversham villa (Evans et al., forthcoming);
the lack of Severn Valley wares in the late group from Alcester (Evans 1996)). Woodfield
(1983: 79) first observed this phenomenon in discussing the later Roman pottery supply
to Towcester, noting that it ‘looks like the survival of large production centres outside
modern Northamptonshire servicing an area which had lost its local industries’. A fairly
similar picture would seem to be indicated in Essex:

the assemblages from these late contexts show the greatest fabric changes, suggesting
a dynamism in terms of marketing which is not expressed in typological variety or
stylistic innovation. This vitality can indicate either a healthy economy or increasing
competition for a shrinking market. It has been suggested on the evidence of typo-
logical and stylistic fossilization that the major pottery industries were in decline
from c. ad 350 (Fulford 1979: 129); a view supported by apparently declining amounts
of pottery at Chelmsford and, perhaps, by the evidence of cessation of production at
Inworth, Chelmsford and, possibly, Rettendon. Elsewhere in the county small pro-
duction sites appear to be in decline. (Going 1987: 118)
 Forms of Knowledge    525

The overall picture seems to reflect stress rather than dynamism, and also indicates
that the last surviving industries in most of the pottery-​using part of the province were
the major producers.

Conclusions

Peacock (1982) examined Romano-​British pottery using extensive modern ethno-

graphic parallels with a clear determination to classify potteries as the result of house-
hold industries, workshop industries, or manufactories. There is a clear element of
technological determinism applied, which results in handmade fabrics being seen in
terms of household industries, or, at a push, workshop industries, while wheel-​made
ones are all seen as workshop industries or manufactories. As is clear, the evidence
clearly fails to match the implications of this hierarchy of modes of production and fails
to match modern assumptions about technologically superior methods. Glazed wares
were known, but rejected after a short spell for samian ware. Stonewares were known,
but rarely used. BB1 was actually the most successful industry in Britain, despite its
being handmade. It remained handmade, despite the evidence from Corfe Mullen of
BB1 tradition potters being educated in wheel-​throwing techniques at an early date. It is
also relatively standardized. Hill (2002: 153) is undoubtedly right that ‘the cultural and
social conditions (and not just the technical and “economic” conditions) needed to be
right for the rapid adoption of the technology [wheel-​throwing] and of its products by
others’.
What does become very clear from this brief survey of Romano-​British ceramics is
just how strongly the pattern of pottery use prior to ad 43 conditioned pottery use in
the Roman period. There are three quite distinct zones of Romano-​British pottery use:
first, most of the south-​east including Lincolnshire for these purposes, secondly, East
Yorkshire, the Severn Valley, and Dorset. and, thirdly, the remainder of England and
Wales. In most of the south-​east area ‘Belgic’-​style material is introduced prior to the
conquest and usually ends in the Flavian period; thereafter Romano-​British pottery tra-
ditions develop and pottery is widely used. In the second group Iron Age tradition mate-
rial continues to dominate the pottery use of the area throughout the Roman period.
Among the third group little or no pottery is used in the Iron Age and little or no pot-
tery is used in the Romano-​British period. There is a strong parallel between the first
and second areas and those that see the development of Romano-​British villas (Sargent
2002: figure 3) or non-​military sites that have temples or shrines (Sargent 2002: figure
4). What is particularly notable is the lack of ‘catch-​up’. Aceramic areas do not particu-
larly move towards using pottery later on. There are some marginal changes; in north
Wales samian ware is used from the Flavian period, but coarse pottery tends to wait
until the advent of BB1, and then this fabric is selectively used on rural sites to the exclu-
sion of most other Roman coarse pottery except mortaria (Evans, forthcoming a), so
that assemblages typically can contain 45–​93 per cent of BB1. Similarly, in the Vale of
526

526   Jeremy Evans

York Iron Age-​style pottery, which is largely absent in the Iron Age, is used on Roman
sites in small quantities in the first and second centuries at Mourie Farm and Sike Spa,
Crayke, along with an increasing but minority use of ‘Romanized’ pottery. Despite the
rather weak material evidence for ‘resistance’ in many of the claimed examples of such
in the pottery of the south-​east (cf. Fincham 2002), across the north, Wales, and the far
south-​west, there is very strong and very clear evidence of it. However, a failure to adopt
Roman cultural norms does not mean that the populations of these areas were politi-
cally opposed to Roman government. There were no obligations on subject populations
to ‘Romanize’, only to display political loyalty to the emperor and pay their taxes. There
is generally little evidence that the population of the ‘highland zone’ failed to do this.
However, in this ‘highland-​zone’ region there is precious little evidence of economic
interaction in the Romano-​British period; local societies apparently continued to func-
tion much as they had in the Iron Age. There were no towns as such—​that is centres
whose economy was articulated with that of the local countryside. Most nucleated set-
tlements were vici, dependent on their adjacent military communities, including the
larger military ‘R and R’ centres at Corbridge and Carlisle; others in the north-​west are
military supply production centres, probably with a largely servile population (Evans,
forthcoming b). Even apparent towns such as Wroxeter seem to be autarchic centres of
an alien population group, quite possibly Latin-​speaking, working their hinterland but
not interacting much with the surrounding rural economy.
It can be argued that ‘highland-​zone’ rural sites would have to participate in the
wider provincial economy in order to raise cash to pay their taxes. Given the scarcity
of coinage, the prevalence of local garrisons, and the only partially hypothetical view
that stock loomed large in indigenous societies, direct taxation in kind in this region
might, perhaps, be envisioned. It was possible in the early Empire, although cash taxa-
tion was more usual. The direct rendering of local cattle to the military, cattle that might
have been very important in assessing status and identity in local communities, might
also have led to festering resentments that helped maintain local communal material
identities. The absence of villas from the ‘highland zone’ might well point to the absence
of a market in land in these areas (cf. Gregson 1988); were we to see the land as commu-
nally owned in a predominantly stock-​raising area, then this might be a serious obstacle
to developing ‘Roman’ forms of individual status display, which were largely based on
ownership. It is notable that it was these largely aceramic areas that did not become sub-
ject to Saxon takeover in the initial colonization, perhaps, particularly in an area like the
kingdom of Elmet, because they were not very integrated into a Romano-​British provin-
cial economy and so suffered comparatively little disruption when it ended.
We have seen that there is developing evidence of regional variations in the functional
classes of vessels used and, therefore, probably of food ways. Further evidence of this
from other material evidence has been collected by Cool (2006), such as the northern
used of barley griddle scones rather than wheat (Cool 2006: 77–​78). It is likely that much
more evidence of regional patterning may be teased out yet, especially by examining evi-
dence from different finds classes together. As discussed, in the later fourth century (and
beyond), there does seem to be a regional north versus south–​central English division in
 Forms of Knowledge    527

the use of pottery that may represent a growing cultural divide between the areas, with
the northern army developing usages more similar to those of the East Yorkshire Iron
Age. In the interests of relative brevity, this chapter has not considered military pottery
supply, but there is growing evidence in this field of pottery reflecting particular cultural
groups (e.g. Swan 1992, 2009; Swan and Bidwell 1998; Cool 2004a).

References
Allen, J., and Fulford, M. G. (1996). ‘The Distribution of South-​East Dorset Black Burnished
Category I Pottery in South-​West Britain’, Britannia, 27: 223–​281.
Arthur, P. (1978). ‘The Lead Glazed Wares of Roman Britain’, in P. Arthur and G. Marsh (eds),
Early Fine Wares in Roman Britain. BAR British Series 57. Oxford: Archaeopress, 293–​356.
Booth, P. (2001). ‘Fired Clay’, in P. Booth, J. Evans, and J. Hiller (eds), Excavations in the Extra-​
Mural Settlement of Roman Alchester, Oxfordshire, 1991. Oxford Archaeology Monograph
1. Oxford: Oxford Archaeology, 260–​261.
Brailsford, J. W. (1958). ‘Early Iron Age “C” in Wessex’, Proceedings of the Prehistoric Society,
24: 101–​119.
Braithwaite, G. (1984). ‘Romano-​British Face Pots and Head Pots’, Britannia, 15: 99–​131.
Braithwaite, G. (2007). Faces from the Past: A Study of Roman Face Pots from Italy and the Western
Provinces of the Roman Empire. BAR International Series 1651. Oxford: Archaeopress.
Braund, D. (1984). Rome and the Friendly King:  The Character of the Client Kingship.
Beckenham: Croom Helm.
Buckland, P., Magilton, J. R., and Dolby, M. J. (1980). ‘The Roman Pottery Industries of South
Yorkshire: A Review’, Britannia, 11: 145–​164.
Calkin, J. B. (1935). ‘An Early Romano-​British Kiln at Corfe Mullen, Dorset’, Antiquities Journal,
15: 42–​55.
Carreras Montfort, C. (2002). ‘The Roman Military Supply during the Principate: Transportation
and Staples’, in P. Erdkamp (ed.), The Roman Army and the Economy. Amsterdam:  J.
C. Gieben, 70–​89.
Carreras Montfort, C., and Funari, P. P. A. (1998). Britannia y el Mediterráneo. Estudios sobre
el abastecimiento de aceite bélico y Africano en Britannia. Col. Instrumento 5.  Barcelona
<http://​ceipac.gh.ub.es/​biblio/​Data/​A/​0021.pdf>.
Cool, H. E. M. (2004a). The Roman Cemetery at Brougham, Cumbria: Excavations 1966-​67.
Britannia Monograph Series 21. London: Society for the Promotion of Roman Studies.
Cool, H. E. M. (2004b). ‘Some Notes on Spoons and Mortaria’, in B. Croxford, H. Eckardt, J.
Meake, and J. Weekes (eds), TRAC 2003: Proceedings of the Thirteenth Annual Theoretical
Roman Archaeology Conference, Leicester 2003. Oxford: Oxbow Books, 28–​35.
Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge
University Press.
Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge
University Press.
Cunliffe, B. (1974). Iron Age Communities in Britain: An Account of England, Scotland and Wales
from the Seventh Century BC until the Roman Conquest. London: Routledge.
Darling, M. J. (1977). ‘Pottery from Early Military Sites in Western Britain’, in J. Dore and K.
Greene (eds), Roman Pottery Studies in Britain and Beyond. BAR International Series 30.
Oxford: Archaeopress, 57–​100.
528

528   Jeremy Evans

Davies, B., Richardson, B., and Tomber, R. (1994). A Dated Corpus of Early Roman Pottery from
the City of London. The Archaeology of Roman London Volume 5/​CBA Research Report 98.
York: Council for British Archaeology.
Davies, J. L. (1983). ‘Coinage and Settlement in Roman Wales and the Marches: Some
Observations’, Archaeologia Cambrensis, 132: 78–​94.
Déru, X. (1996). La Céramique Belge dans le Nord de la Gaule. Caractérisation, chronologie,
phénomènes culturels et économiques. Publications d’Histoire de l’Art et d’Archéologie de
l’Université Catholique de Louvain, 89.
Dobney, K. (2001). ‘A Place at the Table:  The Role of Vertebrate Zooarchaeology within a
Roman Research Agenda’, in S. James and M. Millett (eds), Britons and Romans: Advancing
an Archaeological Agenda. CBA Research Report 125. York:  Council for British
Archaeology, 36–​45.
Evans, J. (1985). ‘Aspects of Later Roman Pottery Assemblages in Northern England’.
Unpublished Ph.D. thesis, University of Bradford.
Evans, J. (1988). ‘All Yorkshire is Divided into Three Parts:  Social Aspects of Later Roman
Pottery Distributions in Yorkshire’, in J. Price and P. R. Wilson (eds), Recent Research in
Roman Yorkshire. BAR British Series 193. Oxford: Archaeopress, 323–​337.
Evans, J. (1989). ‘Crambeck: The Development of a Major Northern Pottery Industry’, in P.
R. Wilson (ed.), The Crambeck Roman Pottery Industry. Yorkshire Archaeological Society
Roman Section Monograph. York: Yorkshire Archaeological Society, 43–​90.
Evans, J. (1993). ‘Function and Finewares in the Roman North’, Journal of Roman Pottery
Studies, 6: 95–​118.
Evans, J. (1995). ‘Reflections on Later Iron Age and “Native” Romano-​British Pottery in North-​
Eastern England’, in B. Vyner (ed.), Moorland Monuments. CBA Research Report 101.
York: Council for British Archaeology, 46–​68.
Evans, J. (1996). ‘The Gas House Lane (AL23) Roman Pottery’, in S. Cracknell (ed.), Roman
Alcester. Volume 2: Defences and Defended Area. CBA Research Report 106. York: Council
for British Archaeology, 58–​97.
Evans, J. (1999). ‘Roman Pottery’, in S. C. Palmer, ‘Archaeological Excavations in the Arrow
Valley, Warwickshire’, Transactions of the Birmingham and Warwickshire Archaeological
Society, 103: 101–​125.
Evans, J. (2001a). ‘Iron Age, Roman and Anglo-​Saxon Pottery’, in P. Booth, J. Evans, and J.
Hiller (eds), Excavations in the Extra-​Mural Settlement of Roman Alchester, Oxfordshire,
1991. Oxford Archaeology Monograph 1. Oxford: Oxford Archaeology, 263–​383.
Evans, J. (2001b). ‘Material Approaches to the Identification of Different Romano-​British Site
Types’, in S. James and M. Millett (eds), Britons and Romans: Advancing an Archaeological
Agenda. CBA Research Report 125. York: Council for British Archaeology, 26–​35.
Evans, J. (2005). ‘Pottery in Urban Romano-​British Life’, in A. MacMahon and J. Price (eds),
Roman Working Lives and Urban Living. Oxford: Oxbow Books, 145–​166.
Evans, J. (2009). ‘The Ceramic Objects’, in S.  C. Palmer, ‘Neolithic, Bronze Age, Iron Age,
Romano-​British and Anglo-​Saxon Sites Excavated on the Transco Churchover to Newbold
Pacey Gas Pipeline in 1999’, Transactions of the Birmingham and Warwickshire Archaeological
Society, 113: 129–​133.
Evans, J. (forthcoming a). ‘Some Small Harvest, Pottery from “Highland Zone” Sites in North
Wales and the North-​West’, Journal of Roman Pottery Studies.
Evans, J. (forthcoming b). ‘King Street, the Roman Frontier in the North-​West and Roman
Military Supply’, Journal of Roman Pottery Studies.
 Forms of Knowledge    529

Evans, J., and Mills P. (2008). ‘The Mortaria’, in H. E.  M. Cool and D. J.  P. Mason (eds),
Roman Piercebridge:  Excavations by D.  W. Harding and Peter Scott 1969–​1981. AASDN
Research Report 7.  Durham:  Architectural and Archaeological Society of Durham and
Northumberland, 197–​207.
Evans, J., Macaulay, S., and Mills, P. J. (forthcoming). The Horningsea Roman Pottery Industry
in Context: An Area Study of Ceramic Supply in the Cambridgeshire Region. East Anglian
Archaeology.
Fincham, G. (2002). Landscapes of Imperialism:  Roman and Native Interaction in the East
Anglian Fenland. BAR British Series 338. Oxford: Archaeopress.
Fitzpatrick, A. (1985). ‘The Distribution of Dressel 1 Amphorae in North-​West Europe’, Oxford
Journal of Archaeology, 4: 305–​340.
Freestone, I. C., and Rigby, V. (1988). ‘The Introduction of Roman Ceramic Styles and
Techniques into Roman Britain: A Case Study from King Harry Lane Cemetery, St Albans,
Hertfordshire’, in E. Sayre, P. Vandicer, J. Druzik, and C. Stevenson (eds), Material Issues
in Art and Archaeology. MRS Proceedings, volume 123, number 1. Cambridge: Cambridge
University Press, 109–​115.
Fulford, M. G. (1979). ‘Pottery Production and Trade at the End of Roman Britain: The Case
against Continuity’, in P. J. Casey (ed.) The End of Roman Britain. BAR British Series 71.
Oxford: Archaeopress, 120–​132.
Going, C. (1987). The Mansio and Other Sites in the South-​Eastern Sector of Caesaromagus: The
Roman Pottery. CBA Research Report 62. London: Council for British Archaeology.
Gregson, M. (1988). ‘The Villa as Private Property’, in K. Branigan and D. Miles (eds), The
Economies of Romano-​British Villas. Sheffield: University of Sheffield, 21–​33.
Hartley, K. F. (1998). ‘The Incidence of Stamped Mortaria in the Roman Empire, with Special
Reference to Imports to Britain’, in J. Bird (ed.) Form and Fabric: Studies in Rome’s Material
Past in Honour of B. R. Hartley. Oxbow Monograph 80. Oxford: Oxbow Books, 199–​217.
Haselgrove, C. C., Ferrel, G., and Turnbull, P. (1988). The Durham Archaeological Survey 1983–
87. Durham: University of Durham, Department of Archaeology.
Hayes, J. W. (1997). Handbook of Mediterranean Roman Pottery. London: British Museum Press.
Hill, J. D. (1997). ‘The End of one Kind of Body and the Beginning of Another Kind of Body?
Toilet Instruments and ‘Romanization’ in Southern England during the First Century ad’, in
A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies. Oxford: Oxbow Books,
96–​107.
Hill, J. D. (2002). ‘Just about the Potter’s Wheel? Using, Making and Depositing Middle
and Later Iron Age Pots in East Anglia’, in A. Woodward and J. D. Hill, Prehistoric
Britain:  The Ceramic Basis. Prehistoric Ceramic Research Group Occasional Publication
3. Oxford: Oxbow Books, 143–​160.
Hodder, I. R. (1974). ‘Some Marketing Models for Romano-​British Coarse Pottery’, Britannia,
5: 340–​359.
Hodder, I., and Orton C. (1976). Spatial Analysis in Archaeology. Cambridge:  Cambridge
University Press.
Holbrook, N., and Bidwell, P. T. (1991). Roman Finds from Exeter. Exeter Archaeology Report 4,
Exeter: Exeter City Council.
Jundi, S., and Hill, J. D. (1998). ‘Brooches and Identities in First Century ad Britain: More than
Meets the Eye?’, in C. Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 97: Proceedings
of the Seventh Annual Theoretical Roman Archaeology Conference, Nottingham 1997.
Oxford: Oxbow Books, 125–​137.
530

530   Jeremy Evans

Lyne, M. A. B., and Jefferies, R. S. (1979). The Alice Holt/​Farnham Roman Pottery Industry. CBA
Research Report 30. London: Council for British Archaeology.
Marsh, G. (1978). ‘Early Second Century Fine Wares in the London Area’, in P. Arthur and G.
Marsh, Early Fine Wares in Roman Britain. BAR British Series 57. Oxford: Archaeopress,
119–​224.
Marsh, G. (1981). ‘London’s Samian Supply and its Relationship to the Development of the Gallic
Samian Industry’, in A. C. Anderson and A. S. Anderson (eds), Roman Pottery Research
in Britain and North-​West Europe. BAR International Series 123. Oxford:  Archaeopress,
173–​238.
Mepham, L. N. (1997). ‘Pottery’, in A. Fitzpatrick (ed.), Archaeological Excavations on the
Route of the A27 Westhampnett Bypass, West Sussex, 1992. Volume 2: The Cemeteries. Wessex
Archaeology Report 12. Salisbury: Wessex Archaeology.
Millett, M. (1979). ‘An Approach to the Functional Interpretation of Pottery’, in M. Millett (ed.),
Pottery and the Archaeologist. University of London Institute of Archaeology Occasional
Publications 4. London: University of London, 35–​48.
Millett, M. (1980). ‘The Archaeology of Sussex Pottery: Aspects of Romano-​British Pottery in
West Sussex’, Sussex Archaeological Collections, 118: 57–​68.
Millett, M. (1983). ‘A Comparative Study of some Contemporaneous Pottery Assemblages from
Roman Britain’. Unpublished D. Phil. thesis, University of Oxford.
Millett, M. (1990). The Romanization of Britain. Cambridge: Cambridge University Press.
Mills, P. J. (2015). ‘The Iron Age and Roman Pottery’, in P. Halkon, M. Millett, and H.
Woodhouse (eds), Hayton, East Yorkshire: Archaeological Studies of the Iron Age and Roman
Landscapes. Leeds: Yorkshire Archaeological Society.
Peacock, D. P.  S. (1982). Pottery in the Roman World:  An Ethnoarchaeological Approach.
Harlow: Longman.
Peacock, D. P. S. (1988). ‘The Gabbroic Pottery of Cornwall’, Antiquity, 62/​235: 302–​304.
Pitts, M., and Perring, D. (2006). ‘The Making of Britain’s First Urban Landscapes: The Case of
Late Iron Age and Roman Essex’, Britannia, 27: 189–​212.
Pollard, R. (1988). The Roman Pottery of Kent. Kent Archaeological Society Monograph 5.
Maidstone: Kent Archaeological Society.
Reece, R. (1979). ‘Roman Military Impact:  Addendum’, in B. C. Burnham and H. B.
Johnson (eds), Invasion and Response: The Case of Roman Britain. BAR British Series 73.
Oxford: Archaeopress, 216–​217.
Rigby, V. (1977). ‘The Gallo-​Belgic Pottery from Cirencester’, in J. Dore and K. Greene
(eds), Roman Pottery Studies in Britain and Beyond. BAR International Series 30.
Oxford: Archaeopress, 37–​43.
Rush, P. (1997). ‘Symbols, Pottery and Trade’, in K. Meadows, C. Lemke, and J. Heron (eds),
TRAC 96:  Proceedings of the Sixth Annual Theoretical Roman Archaeology Conference,
Sheffield 1996. Oxford: Oxbow Books, 55–​64.
Rye, O. S. (1976). ‘Keeping your Temper under Control:  Material and the Manufacture of
Papuan Pottery’, Archaeology and Physical Anthropology in Oceania, 11/​2: 106–​137.
Sargent, A. (2002). ‘The North–​South Divide Revisited: Thoughts on the Character of Roman
Britain’, Britannia, 33: 219–​226.
Saunders, L. (1986). ‘A Study of the Distribution of Romano-​British Pottery Kilns with Respect
to the Major Settlements of the Province’. Unpublished B.Sc. dissertation, University of
Durham.
Swan, V. G. (1984). The Pottery Kilns of Roman Britain. London: RCHM/​HM.
 Forms of Knowledge    531

Swan, V. G. (1992). ‘Legio VI and its Men: African Legionaries in Britain’, Journal of Roman
Pottery Studies, 5: 1–​33.
Swan, V. G. (2009). Ethnicity, Conquest and Recruitment: Two Case Studies from the Northern
Military Provinces. JRA Supplementary Series 72. Portsmouth, RI:  Journal of Roman
Archaeology.
Swan, V. G., and Bidwell, P. T. (1998). ‘Camelon and Flavian Troop-​Movements on Southern
Britain: Some Ceramic Evidence’, in J. Bird (ed.), Form and Fabric: Studies in Rome’s Material
Past in Honour of B. R. Hartley. Oxford: Oxbow Books, 21–​30.
Timby, J. R. (1990). ‘Severn Valley Wares: A Reassessment’, Britannia, 21: 243–​251.
Trow, S. D. (1988). ‘Excavations at Ditches Hillfort, North Cerney, Gloucestershire, 1982–​3’,
Transactions of the Bristol and Gloucestershire Archaeological Society, 106: 19–​85.
Tyers, P. (1996). Roman Pottery in Britain. London: B. T. Batsford.
Webster, P. V. (1996). Roman Samian Pottery in Britain: Background Notes. London: Council for
British Archaeology.
Williams, D. F. (1977). ‘The Romano-​ British Black-​Burnished Industry:  An Essay in
Characterization by Heavy Mineral Analysis’, in D. P. S. Peacock (ed.), Pottery and Early
Commerce, Characterization and Trade in Roman and Later Ceramics. London: Academic
Press, 163–​220.
Willis, S. (1997). ‘Samian: Beyond Dating’, in K. Meadows, C. Lemke, and J. Heron (eds), TRAC
96:  Proceedings of the Sixth Annual Theoretical Roman Archaeology Conference, Sheffield
1996. Oxford: Oxbow Books, 38–​54.
Willis, S. H. (2006). ‘Samian Pottery—​A Resource for the Study of Roman Britain and Beyond:
The Results of the English Heritage Funded Samian Project: An e-​monograph’, Internet
Archaeology, 17/​1 <http://​intarch.ac.uk/​journal/​issue17/​1/​toc.html/​>.
Woodfield, C. (1983). ‘The Pottery’, in A. E. Brown and C. Woodfield, ‘Excavations at Towcester,
Northamptonshire: The Alchester Road Suburb’, Northamptonshire Archaeology, 18: 74–​100.
Woods, A. (1983–​4). ‘The Old Pot Boiler’, Bulletin of the Experimental Firing Group, 2: 25–​31.
Woods, P. J. (1974). ‘Types of Late Belgic and Early Romano-​British Pottery Kilns in the Nene
Valley’, Britannia, 5: 262–​281.
Woods, P. J., and Hastings, B. C. (1984). Rushden:  The Early Fine Wares.
Northampton: Northamptonshire County Council.
532

Chapter 26

M etals a nd
M etalwork i ng

David Dungworth

Introduction

There is hardly any aspect of life in Roman Britain that did not rely on metals. Metals
were used to make the practical tools, weapons, and implements that formed the eco-
nomic basis of the province. Metals also reached into less obvious realms such as per-
sonal grooming (copper alloy toilet sets) and adornment (brooches used to express
aspects of identity).
The metals that were exploited can be considered in four groups: precious metals,
copper alloys, tin and pewter, and iron. The two precious metals used were gold and
silver—​these were often alloyed with each other and also with copper. The precious
metals were used for the production of a limited range of coins and jewellery. A vari-
ety of copper alloys were used, including pure copper, bronze (an alloy of copper and
tin), and brass (an alloy of copper and zinc). In addition, many copper alloys also con-
tained some lead. Copper alloys had limited use for the production of tools, but were
most extensively used for the manufacture of jewellery, such as brooches. Lead and tin
were used both separately and in combination (as solder and pewter). Lead was used in
many different ways, such as for fishing weights and for the pipes for delivering water to
bathhouses. Pewter (an alloy of tin and lead) was extensively used to produce tableware
(imitating silverware). Iron (and its alloys) was used for the manufacture of practical
tools, which were absolutely essential to most other craft activities, such as carpentry.

Gold

Gold was used in Roman Britain for the manufacture of objects of personal adornment
(although the scarcity of gold means more items were made from more base metals).
Gold was also used for the production of high value coins.
 Metals and Metalworking    533

Gold ores are found in very few parts of the Britain (Figure 26.1), but at Dolaucothi,
Dyfed, there are very important traces of Roman mining. At Dolaucothi there are
remains of numerous leets and reservoirs (Lewis and Jones 1969), which were used for
hushing (or hydraulic mining). The leets gathered water and stored this in the reservoirs,

Malton

Dolaucothi

London

Chichester
Exeter

Figure 26.1  Map of Roman Britain showing selected sites with evidence for goldworking.
Source: © David Dungworth.
534

534   David Dungworth

from where it could be released suddenly to scour the soil and reveal the bedrock
(including any gold-​bearing rock). The rock was dug out to leave a substantial opencast
mine. Water continued to be gathered and used to process the ore. The rock would be
crushed and the lighter, non-​gold-​bearing portion could be washed away. The crushing
appears to have made use of water power and trip hammers, and the scale of the Roman
gold-​working is attested by the discovery that the apparent medieval motte is actually a
spoil heap comprising both coarse and fine waste rock (Burnham and Burnham 2004).
Some gold-​mining also took place underground, as fragments of a wheel used to raise
water from the mine were discovered in the 1930s (Boon and Williams 1966).
It is likely that the working of gold used techniques that are better known from
copper-​alloy working, in particular casting using moulds and hammering. Pure gold
melts at 1064°C (almost the same as pure copper) and so requires the use of carefully
made crucibles, which will retain their strength at high temperature. Evidence for cast-
ing is very limited; while crucibles and moulds are routinely analysed to determine the
nature of any metal traces left, gold is rarely detected in this way. Remains of more base
metals (especially copper) are regularly found, but this may be because these metals are
more reactive and so more likely to become trapped in a crucible or mould. Those cru-
cibles, in which significant traces of gold have been detected, are usually quite small—​in
some cases with the capacity for only 10–​20 g of gold (Bayley and Rehren 2007).
Most Roman gold objects are actually gold–​silver alloys, or gold–​silver–​copper alloys
(Cowell et al. 1983; Cowell and Hook 2010). Pure gold is a very soft metal ,and the pres-
ence of small amounts of other metals can improve its hardness and may alter its colour;
for example, silver will produce a ‘white’ gold, while copper can produce a ‘red’ gold; but
the careful balance of both copper and silver can yield a gold-​coloured alloy, even when
the proportion of gold is quite low. Hammering has always been a suitable method for
working gold, as, in its pure state, it does not work-​harden—​most metals will become
harder the more they are hammered, but gold remains soft. This allows gold to be ham-
mered into very thin sheets or even foil.
Gold was used to coat (or gild) the surface of many objects, thus giving them the
appearance of gold without the expense. Two techniques were widely used: leaf-​gilding
and amalgam-​gilding (Oddy 1981). For leaf-​gilding, thin gold foil sheets were applied
to the surface of the object and then burnished. For amalgam-​gilding, a mixture (amal-
gam) of gold and mercury was painted onto the surface; when heated, the mercury
would be driven off as a vapour, leaving the gold behind. Evidence for gilding in Roman
Britain is almost entirely restricted to surviving gilded objects (analysis of some of these
has detected traces of mercury, which confirms the use of amalgam-​gilding).
Certain evidence for a gold workshop in Norton, near Malton in Yorkshire, is pro-
vided by a building inscription (‘feliciter sit genio loci servvle utere felix
tabern am aurefi cinam’ Prosperity to the genius of this place! O Servulus, enjoy
thy goldsmith’s shop in happiness), which identifies the workshop as that of a gold-
smith and that it was run by a slave (RIB 712). This may be an example of what Love
(1991: 133) describes as apophora: semi-​independent, small-​scale production by
slaves.
 Metals and Metalworking    535

Archaeological evidence has been found that provides a fascinating insight into the
technologies used to recover and recycle gold. Gold alloys were widely used for cur-
rency and for the manufacture of jewellery, but there were clearly occasions when
people wanted their gold in a pure form. Removing copper from a gold alloy could be
achieved using cupellation, but separating silver from gold required a technique known
as ‘parting’ (Ramage and Craddock 2000). The gold–​silver alloy would be hammered
into sheets and placed in a sealed pot with salt; as the pot was heated, the salt would react
with the silver and leave the gold behind. This process is attested by the survival of part-
ing vessels from mid–​late first century ad contexts in London, Chichester, and Exeter
(Bayley 1991).

Silver and Lead

Silver was most commonly used in Roman Britain for the manufacture of tableware
such as plates, cups, and spoons, and is represented in several spectacular late Roman
hoards (Painter 1977a, b; Johns and Potter 1983; Bland and Johns 1993; Johns 2010). Silver
was also used for the production of a variety of coins (see Walton and Moorhead, this
volume, for a discussion on coins and the economy).
Silver ores are almost unknown in Britain, and most silver has instead been obtained
from lead ores, which often contain several hundred parts per million of silver. Lead
ores occur in the Mendips, Shropshire, the Peak District (Derbyshire), the Pennines
(Yorkshire and Northumberland) and Flint (Wales) (Figure 26.2). The ore is usually
found as a lead sulphide and is much easier to smelt than the ores of other metals. While
most metal-​smelting would require a reducing atmosphere, lead-​smelting can be car-
ried out under relatively oxidizing conditions and may even be achieved in a bonfire
rather than a furnace.
Lead ores routinely contain several hundred parts per million (ppm) of silver, and this
was undoubtedly a significant factor in the Roman lead industry. Over six tonnes of lead
ingots are known from Roman Britain, and some of these bear inscriptions (Tylecote
1986: 58–​69). The EX ARG inscription (e.g. RIB 2404.4; Frere et al. 1990: 40–4​1) on some
ingots has traditionally been taken as evidence that the lead derived from silver-​lead
works and that the lead had been de-​silvered by cupellation. Some inscriptions include
the name of an emperor (e.g. RIB 2404.6; Frere et al. 1990: 42), suggesting a state inter-
est in production and some name military units (e.g. Legio II Augustus, RIB 2404.25;
Frere et al. 1990: 50), which has prompted some to see the lead as having been produced
under military supervision (e.g. Elkington 1976). However, only two out of just over fifty
inscribed ingots refer to military units; more inscriptions refer to individuals or associa-
tions (socio) and suggest that most production was not directly controlled by the mili-
tary. Both the find-spots of the ingots and the inscriptions they bear have encouraged
some to develop detailed models for the organization of the Roman lead–​silver industry
(e.g. Elkington 1976; Tylecote 1986: 61–​68).
536

North Pennines

Flint Peak District

Brompton
Shropshire

Mendips
St Algar’s Farm

Figure 26.2  Principal lead–​silver orefields and sites of primary silver extraction.
Source: © David Dungworth.
 Metals and Metalworking    537

Although lead ingots imply lead-​mining and -smelting, there is relatively little direct
archaeological evidence for these activities. Lead-​working is more often assumed than
demonstrated, and there are relatively few published excavations (Ashworth and Palmer
1957–​8; Anon. 1971; Rahtz and Greenfield 1977; Todd 2007). Research by the Early Mines
Research Group has identified a lead smelting hearth at Cwmystwyth, Dyfed, which has
been dated by radiocarbon to the early Roman period (Timberlake 2002). A similar date
has been obtained for the site of a lead smelter at Llangynfelyn, on the edge of Borth
Bog, Dyfed (Page 2005).
Although the de-​silvering of lead is demonstrated by the presence at several Roman
sites of primary litharge, the relationship between the known lead ingots and the sil-
ver industry remains unclear. Several of the ingots with an EX ARG inscription have
been shown to contain residual fragments of lead ore (Smythe 1939–​40). The presence
of ore is not reconcilable with the idea of re-​smelted lead, and it must be concluded that
these ingots had not been de-​silvered, despite their EX ARG inscriptions. Indeed, the
silver content of most Roman ingots is less than 100ppm, and it is doubtful whether de-​
silvering such lead would ever have been economically sound.
When the lead ore is smelted, all of the silver will be concentrated in the metallic lead
(rather than any associated waste slag). The lead can then be processed to extract the
silver by cupellation (Bayley and Eckstein 2006). The lead would be melted and exposed
to air so that it would form lead oxide (litharge). The litharge would be absorbed into
the lining of the hearth, leaving a puddle of silver behind. Litharge is occasionally
found as fragments of substantial disks (1 m in diameter) typical of primary cupella-
tion. Unequivocal evidence for primary cupellation is known from a fort at Brompton,
Shropshire (Bayley and Eckstein 1998), and the small civil settlement at St Algar’s Farm,
Somerset (Dunster and Dungworth 2013). The litharge could then be re-​smelted to
obtain metallic lead.
Cupellation could also be used to extract the silver from a silver–​copper alloy (sec-
ondary cupellation). The silver–​copper alloy would be melted with sufficient lead and
exposed to the air. The lead and the copper would both oxidize and be absorbed into
the hearth lining. Litharge from secondary cupellation is always found as fragments of
smaller disks (0.1–​0.2 m in diameter) and is often stained green from the copper it con-
tains. Secondary cupellation is most frequently recovered from urban settlements—​for
example, London, Exeter, Leicester, Silchester, Wroxeter, and Winchester (Bayley and
Eckstein 2006).
Silver was rarely used in its completely pure state: in most cases it was alloyed with
other metals. Most analysed Roman silver has a composition fairly close to Sterling
silver: 90–​5 per cent silver with copper and traces of other metals (Cowell et al. 1983;
Cowell and Hook 2010). Denarii of the early empire usually had a high silver content
(>90 per cent), but this declined until, by the middle of the third century, ‘silver’ coins
were being produced with as little as 2 per cent silver (Walker 1976–​8; Butcher and
Ponting 2012). Subsequently, the silver content was increased as part of various emper-
ors’ coinage reforms; however, most reforms were separated by periods of further
debasement.
538

538   David Dungworth

Silver can be melted and cast into shape, or it can be hammered. Pure silver melts
at 962°C and so requires the use of carefully made crucibles. The analysis of crucibles
and moulds occasionally reveals the presence of silver, but it is often unclear whether
this indicates the melting of a gold–​silver alloy or a silver–​copper alloy. Most of the late
Roman silver tableware would have been produced by a combination of casting and
wrought work. Much of the initial shaping, especially of small simple objects such as
spoons, would have been achieved by melting and casting silver into suitable moulds.
The shaping and decoration of tableware was mostly carried out by hammering; some-
times from the back to produce raised decoration.
Metallic lead was widely used in combination with other metals. Lead was added to
many copper alloys and was used with tin to produce pewter. Lead was also used on its
own for water pipes, cisterns, and coffins.

Copper and its Alloys

Copper and its various alloys were widely used in Roman Britain. Copper alloys are
most prominent in the archaeological record as brooches and other items of personal
adornment, but they were also used for toilet implements, cutlery, tablewares, cooking
vessels, box and furniture fittings, horse fittings, armour, statuettes, and coins.
The principal copper alloys were bronze and brass. Tin bronze could be produced
using specific proportions of copper and tin. The volatility of zinc meant that brass could
be made only using a cementation process. Widespread recycling gave rise to many cop-
per alloys, which contain significant proportion of both tin and zinc (often referred to as
gunmetal). All of these alloys could also contain varying levels of lead.
While there is abundant evidence for the use of copper and its various alloys in Britain
from the second millennium bc onwards, there is relatively little evidence for its pro-
duction. Copper ores are mostly found in Cornwall and Wales, and it is likely that a great
deal of the evidence for Roman mining and smelting has been lost as a result of later
large-​scale mining. Until quite recently, many of the mines with apparent pre-​modern
activity were assumed to be Roman in date (Davies 1935), although more recent field-
work and radiocarbon dating indicate that many of these early mines are prehistoric
(Timberlake 2003). Fieldwork at Alderley Edge has recently identified a rock-​cut shaft
(1.5–​1.9 m wide), which was excavated with iron picks (Timberlake and Prag 2005). The
shaft descended 11 m, and radiocarbon dating of oak planks from the base gave a late
Iron Age or early Roman date (the maturity of the wood and style of the shaft are both
consistent with an early Roman date).
No Roman period copper-​smelting sites have yet been positively identified in Britain.
Despite the lack of any furnaces or slag, copper-​smelting is implied by the existence of at
least thirty plano-​convex ingots (typically 0.3 m in diameter and weighing 15kg). Most of
these have been found on Anglesey, with the majority of the others from neighbouring
Gwynedd. The circumstances of their recovery do not usually provide information on
 Metals and Metalworking    539

the context of the deposition (most were recovered before the beginning of the twen-
tieth century). Ten of the ingots have Latin inscriptions and so are certainly of Roman
date, but the others are uncertain. The Latin inscriptions on the copper ingots are much
shorter than those on lead ingots; none refers to a military unit or an emperor (RIB
2403.1–​2403.11; Frere et al. 1990: 34–​37). Most inscriptions appear to be personal names,
although one (RIB 2403.3; Frere et al. 1990: 34–35) has socio romae natsol (Tylecote
1986: 20–​21), which suggests production by an association.
Evidence for the casting of copper in the form of crucibles, moulds, and slag is rou-
tinely found on Roman sites. The earliest Roman crucibles are handmade and hemi-
spherical in form, but simple wheel-​thrown forms became common from the end of
the first century. The wheel-​thrown crucibles were often provided with an additional
outer layer of clay to protect the crucible itself (Bayley 1988). The crucibles are typically
100 mm in diameter and 100 mm high and had a capacity for melting 2–​3kg of copper
alloy (Bayley and Rehren 2007).
Producing a casting requires a ‘pattern’—​an original that gives the desired shape to
the mould. Two moulding techniques were commonly used: lost-​wax casting and piece-​
moulding. In the former case a unique pattern was prepared in wax, which would then
be covered in clay. The clay would be heated to melt the wax and create the mould. Lost-​
wax casting was widely used in the pre-​Roman period to produce horse-​riding equip-
ment and vehicle fittings (Wainwright 1979) and, after the conquest, may have been used
to manufacture small figurines and other complex shapes (Draper 1985).
In piece-​moulding, a more durable pattern was produced in wood, bone, or lead. The
clay mould was formed from two pieces of clay: the pattern would be pressed into the
first piece of clay and then covered with a second lump of clay. The two pieces of clay
would be separated, the pattern removed and the clay joined back together to form a
mould (Bayley and Butcher 2004: 26–​40). Clay moulds do occasionally survive in sub-
stantial quantities that will allow the detailed reconstruction of the objects being cast;
however, they are usually friable and they had to be broken in order to recover the cast-
ing and so often survive as tiny fragments of fired clay. While some casting took place on
a small scale, a few sites have provided evidence for the production of large numbers of
identical moulds, which would all have been filled with molten metal at the same time.
Thousands of fragments of moulds were found at Roman Castleford for the manufac-
ture of spoons, and the painstaking reconstruction of these showed that the individual
moulds were stuck together to form a cone, allowing the simultaneous casting of sixteen
spoons (Bayley and Budd 1998). Metal moulds for the manufacture of brooches have
also recently been identified (Bayley and Butcher 2004: 28–​29).
Some copper alloy artefacts were largely formed by wrought techniques, especially
hammering. The manufacture of copper alloy sheet would have required prolonged
hammering and, as copper alloys tend to harden, the metal would have had to be peri-
odically annealed (heated to soften it again). A similar approach would have been used
for the manufacture of many artefacts such as toilet implements.
The chemical analysis of copper alloy artefacts shows a complex relationship between
artefact type and alloy type (Dungworth 1997; Bayley and Butcher 2004). Some artefacts
540

540   David Dungworth

display a limited range of alloy compositions (and these can, on occasion, be distin-
guished from other alloys by quite subtle differences), while other artefacts are made from
a wide range of alloy types. In some cases, the function of the object prescribed a par-
ticular alloy type—​for example, Roman copper alloy mirrors were made from a high tin
bronze (typically 20–​25 per cent tin), which gave a silvery metal. In some cases, the alloy
differences can be relatively easily linked to technical aspects of artefact production (the
chaîne opératoire). Colchester-​type brooches are almost all made of brass (the exception
being a single variant type, which is probably pre-​conquest), while almost all Colchester-​
derivative brooches are made from leaded bronze (Figure 26.3). The Colchester-​type
brooches were made from a single piece of metal and required extensive working to
shape the spring and pin, and brass is well suited to this sort of forming technique. The
Colchester-​derivative brooches comprise several different components, and the main
part of the brooch required no significant working after casting—​a leaded bronze was
the technically appropriate alloy for this manufacturing technique. The Colchester-​type
brooches and Langton Down/​Thistle-​type brooches are both usually made of brass; how-
ever, the Colchester-​type brooches usually contain slightly more zinc, while the Langton
Down/​Thistle-​type brooches contain more tin and lead (cf. Dungworth 1997: table 1).
Nevertheless, these differences are small, and the exact mechanism responsible is uncer-
tain. It is possible that the two types of brooches were manufactured in different work-
shops and that each was supplied with metal from different sources.
While bronze could easily have been made by adding the desired amount of tin to
copper (and indeed leaded alloys could be made in the same way), the manufacture

30
Colchester-type
Colchester derivative
25

20
Zinc (wt%)

15

10

0
0 5 10 15 20 25 30
Lead (wt%)

Figure 26.3  Zinc and lead content of Colchester-​type brooches (single piece construction and
made of brass) and Colchester derivative brooches (multiple component construction and made
of leaded bronze).
Source: © David Dungworth.
 Metals and Metalworking    541

of brass required a different technology (Bayley 1984; Rehren 1996). Zinc boils at a
relatively low temperature (903°C), and so, during smelting, it will usually be lost as a
vapour. In pre-​modern Britain, brass would be manufactured by ‘cementation’, a tech-
nique used widely from the first century bc. Metallic copper, zinc ore, and charcoal were
heated in a sealed crucible; the zinc ore would be reduced to metal but could not escape,
so diffused into the copper to form brass. Brass was used for sestertii and dupondii from
23 bc (Dungworth 1996), and was used to make the lorica segmentata armour used by
legionaries from the Claudian period (Bishop and Coulston 1993). Virtually unknown
in Britain prior to the conquest, brass was soon used in the manufacture of many every-
day artefacts (Dungworth 1997), and there is evidence for small-​scale cementation from
several Roman towns (Bayley 1984).
The use of different copper alloys undergoes considerable change during the Roman
period. By the end of the first century very few brooches were being made using brass
(Bayley 1998; Bayley and Butcher 2004). After the middle of the second century little
brass was used in Roman military equipment (Dungworth 1997). The analysis of a wide
selection of artefact types shows that the proportion of brass used (Dungworth 1997:
figure 26.6) and the average zinc content decreases with time. The proportion of zinc in
sestertii and dupondii also declines, and this was originally attributed to a loss of brass-​
making technology (e.g. Caley 1964). More recently (Dungworth 1996), it has been
suggested that the change in the composition of the base metal coinage was actually an
attempt to maintain some parity with the silver coinage (denarius), which was being
deliberately debased with copper (Figure 26.4).
Copper-​alloy workshops are known from a number of urban and military sites. In
most cases the evidence consists of timber strip buildings with hearths and associ-
ated dumps of crucibles and moulds. In many cases moulds and crucibles have been

100
Silver content (wt%) of contemporary denarii

95

90

85

80

75

70
0 5 10 15 20 25
Zinc content (wt%) of contemporary sestertii and dupondii

Figure 26.4  Silver content of denarii and zinc content of contemporary sestertii and dupondii
(data from Walker 1976–​8; Dungworth 1996).
Source: © David Dungworth.
542

542   David Dungworth

recovered from secondary contexts, and the workshops are inferred (Bayley and Budd
1998). Some workshops have been uncovered with evidence for the secondary work-
ing of copper alloys, especially the final finishing and polishing (Frere 1972; Zienkiewicz
1993; Wilson 2002).

Tin and Pewter

Tin was used in Roman Britain as an important component of many copper alloys and
for the production of soft solder and pewter (Lee 2009). Tin is rare in Europe, and his-
torically one of the most significant sources has been Cornwall. There are persistent
(if vague) references to islands (Cassiterides, Islands of Tin) off the north-​west coast
of Europe as sources of tin (Hawkes 1984). These begin with Herodotus (The Histories,
Book 1) and are repeated (in the first century) by Posidonius, Diodorus Siculus, and
Strabo (see Hawkes 1984 and references). Although it is generally accepted that Cornish
tin has been exploited since prehistoric times, there is very little concrete evidence for
mining or smelting the metal (Penhallurick 1986; Gerrard 2000). Plano-​convex tin
ingots are often thought to be prehistoric or Roman; however, they have not been recov-
ered from stratified archaeological contexts, and most have no inscription (Fox 1996;
Penhallurick 1986: 225–2​36; Fox 1996), although an ingot from Carnanton, Cornwall,
found in 1819, bears a helmeted head and several IENN stamps (RIB 2405; Penhallurick
1986: 204–​207; Frere et al. 1990: 67). Despite the paucity of direct evidence, the palaeo-​
environmental record provides strong support for exploitation from the end of the first
century ad (Meharg et al. 2012).
Tin ores occur as veins within Cornish granite and as eroded fragments within valley
sediments. There is no surviving evidence for any early mining of vein tin, but there is
evidence for Roman tin streaming (using water to erode and sort valley sediments that
contain tin ore) at Boscarne, in the Camel valley (Penhallurick 1986). The early Roman
fort of Nanstallon (occupied c. 60–​80 ad) overlooks the tin-​streaming area at Boscarne
and may be associated with a state interest in tin (Fox and Ravenhill 1972). The recov-
ery of a tin ingot and tourmaline minerals often associated with tin ores from a fourth-​
century context at Trethurgy implies tin-​smelting, although no furnace or slag was
recovered (Quinnell 2004: 72–7​6). The excavation of a pipeline near Botallack Head,
Cornwall, recently recovered tin-​smelting slag from a Roman context (Lawson-​Jones
2013).
Pewter, an alloy of tin and lead, occurs in some earlier contexts, but most belongs to
the late third century and fourth century, when it is largely used for the production of
tableware (Beagrie 1989; Lee 2009). The use and deposition of pewter tableware follows
some of the trends seen in contemporary silverware. Pewter will melt at temperatures
in the region of 300°C, so does not require refractory crucibles and would not usually
give rise to slaggy waste products. Pewter could be melted in domestic vessels and cast in
stone moulds (Lee 2009).
 Metals and Metalworking    543

Iron

Iron was widely used in the Roman period for a great variety of weapons, tools, and
implements (e.g. Manning 1985), and most craft activities were reliant on the availability
of iron for tools. While non-​ferrous ores were found in only some parts of Britain, iron
ores can be found almost everywhere (Young 1993). While some areas (for example, the
Weald and the Forest of Dean) with well-​known medieval iron production have long
attracted attention, iron production has been found throughout most of the province
(Figure 26.5).
The iron ores of the Weald (to the south of London) are commonly found as nodules
of iron carbonate within the Weald Clay (Worssam and Gibson-​Hill 1976), although
there was some exploitation of iron concretions in post-​Cetaceous deposits (Paynter
2008: 283). The iron ores of the Forest of Dean are found as deposits of goethite or limo-
nite within dolomite and limestone rocks (Young and Thomas 1999). Iron carbonates
are known along the Jurassic scarp from Cleveland to Oxfordshire (Young 1993). Little
attention has been paid to bog iron ores, but these are likely to have been an impor-
tant resource for some iron manufacture. Bog iron ores will readily form in waterlogged
deposits where the environment is poor in oxygen (Tylecote and Clough 1983). Bog iron
ores are easy to exploit, as they are found near to the surface; however, they have rarely
been mapped, and nineteenth-​century land improvements will have removed many
deposits. Iron ores would undergo some sort of processing before they were smelted
to produce iron. ‘Roasting’ would often be used to drive off water contained in the ore
and convert carbonates and hydrated oxides to hematite. Roasting hearths consisting
of simple structures are known from a few sites—​for example, Bardown, Sussex (Cleere
1970: 13–​15, figure 26.5). The ores would be broken to obtain an optimum size; deposits
of small fragments of ore (‘fines’) are occasionally found on smelting sites (Cleere 1970:
6, figure 26.3).
The technique of iron-​smelting in the Roman period was the direct or bloomery pro-
cess: the iron was smelted but not melted, and variations of this technique were used
throughout the Iron Age and into the medieval period. Iron ore and charcoal were
charged into a clay-​built furnace, with the charcoal serving as fuel to achieve the tem-
peratures required and as a reducing agent to transform the iron ore to metallic iron. All
ores contain at least some impurities (especially silica), which form waste slag, and most
iron production sites are characterized by heaps of waste slag. Some Roman slag heaps
may contain in excess of 10 tonnes of slag; however, many of the larger examples have
been quarried as material for later blast furnaces or for road-​metalling (Cleere 1974).
The solid-​state reduction of iron ore to metallic iron produced a spongy mass (or
bloom) of iron. The character of the bloom would vary depending on the nature of the
ore used and the smelting procedure. The only element in the ore (other than iron) that
was routinely reduced is phosphorus. Ores that contained significant proportions of
phosphorus (for example, some bog ores) would usually give rise to phosphoric iron
544

East
Midlands

Forest
of Dean

Weald

Figure 26.5  Principal iron production sites.

Source: © David Dungworth.
 Metals and Metalworking    545

blooms. Phosphoric iron is generally tougher than plain iron. It was also possible to con-
duct a smelt such that a proportion of the carbon in the fuel entered the iron, yielding a
steel bloom. It was difficult to produce steel from a phosphoric ore.
Oak was the most common species of wood used for the production of charcoal used
in iron-​smelting, although some sites have shown that a wide variety of tree species
could be used—​for example, Bardown (Cleere 1970: 16). The study of the charcoal from
the iron-​smelting site at Woolaston (Fulford and Allen 1992) showed that the majority
of the charcoal derived from branches between 7 and 18 years old and that many may
have come from managed woodland (coppice or pollard). The charcoal from Bardown
(Cleere 1970: 16) is described as coming from branch wood rather than trunk wood, and
it is suggested that the wood was collected by ‘lopping’.
Clay sources are even more abundant than iron ores, and iron workers are unlikely
to have travelled far to obtain the clay they used to build furnaces or hearths. The fur-
naces at Woolaston used clay from several possible sources all within 1.5 kilometres
(Fulford and Allen 1992: 193). In most cases, the clays used are very rich in silica, but
it is not clear whether the smelters exploited naturally silica-​rich clays or deliberately
added sand. Excavated furnaces walls are typically ‘100mm thick and grading in col-
our and texture from a soft orange-​red on the outside, to a hard, glassy, vesicular grey
on the inside’ (Fulford and Allen 1992: 169). In many cases, only the lowest parts of
furnaces survive and it is difficult fully to reconstruct the size and shape of the fur-
nace. Many Roman examples had an internal diameter of c. 0.3 metres and may have
had a height of c. 1 metre.
Over 100 Roman iron-​smelting sites are known from the Weald (Cleere and Crossley
1995: 57; Hodgkinson 2008), although relatively few sites have benefited from detailed
modern excavation. Those sites that have provided dating evidence (Figure 26.6)

16

14

12
Number of sites

10

0
0 100 200 300 400
Date (AD)

Figure 26.6  Dated Roman iron production sites in the Weald, after Cleere (1974).
Source: © David Dungworth.
546

546   David Dungworth

indicate that production peaked in the second century and that, by the fourth century,
there was little production (if any). The site at Beauport Park (just north of Hastings)
had a slag heap that was largely quarried in the nineteenth century but has been esti-
mated as having originally comprised a staggering 100,000 tonnes of slag (Cleere 1976).
The site was provided with a bathhouse (Cleere and Crossley 1995: figure 21), but there
has been no excavation of any iron production features. Cleere’s important excava-
tions at Bardown and Holbeanwood (Sussex) included the recording of furnaces, ore
roasting pits and other structures; however, to date only an interim report is available
(Cleere 1970). Most of the furnaces were shafts 0.3 m in diameter and probably 1 m high.
Furnaces were frequently built into banks at the edge of hollows; some of them may
originally have been formed by excavating clay or iron ore (Cleere 1970; Cartwright
1992). Some Wealden furnaces were much larger: 1 m or more in diameter at the base but
with a narrower shaft (Hodgkinson 2008: figure 13, plate 6).
Cleere (1974) has suggested (largely on the absence of villas or urban sites) that the
Weald may have been an imperial estate specifically for the exploitation of iron, which
was, in its latest phases, directly organized by the Channel fleet (Classis Britannica).
Military control of metal production (apart from precious metals) is largely unknown
in the empire, and most of the industry was probably independent (Millett 2007: 178–17​
9). Smaller-​scale production is attested at many other sites, such as Broadfields, Crawley
(Cartwright 1992), and Westhawk Farm, Ashford (Paynter 2008).
Over 100 sites have been linked with Roman iron production in the Forest of Dean
(e.g. Jones and Mattingly 1990: 193–​195); however, many of these sites are known only as
earthworks (or ploughed fields), and dating is often limited to a few surface finds. One
of the most important sites was the small town of Ariconium (Weston-​under-​Penyard,
Herefordshire), which was surrounded by numerous furnaces and extensive slag heaps
(Jackson 2012). The excavated furnaces are usually severely truncated; however, they can
all be identified as shaft furnaces (0.3–​0.5 m in diameter) and are associated with tap
slag. The settlement and associated iron manufacture appear to have pre-​Roman ori-
gins, and both continued into the fourth century. The recent synthesis concludes that
no military or other state control of the iron industry can be detected and suggests that
production was controlled by local elites (Jackson 2012). The extensive evidence for
iron-​smelting at Worcester (Darlington and Evans 1992; Dalwood and Edwards 2004;
Cuttler and Buteaux 2011) shows that this was mainly a late Roman phenomenon, which
was based (at least in part) on goethite ores brought from the Forest of Dean. Roman
iron-​smelting is also known from the small town at Monmouth (Marvell 2001). Smaller
production sites are represented by Chesters villa, Woolaston (Fulford and Allen 1992),
where furnaces were recorded within a large timber building dating to the late Roman
period perhaps forming part of a ‘mixed estate economy’ (Fulford and Allen 1992:
200). Estimates of the scale of production are fraught with uncertainties; however, it
is clear that the Forest of Dean was producing more iron than was required for local
consumption.
The ironstones of the Jurassic ridge that runs through the east Midlands were exten-
sively exploited in the Roman period (Condron 1997; Schrüfer-​Kolb 2004). Some
 Metals and Metalworking    547

iron-​smelting was carried out at isolated rural settlements such as Harringworth

(Jackson 1981), probably following the pre-​Roman pattern of iron production. Even
sites with more intensive iron-​smelting, such as Wakerley (Jackson and Ambrose 1978),
appear to have combined industry and agriculture. Most of the furnaces in this region
appear to be shaft furnaces with an internal diameter of approximately 0.3 m and an
assumed height in the region of 1 m. Some were built on the contemporary land sur-
face, and slag would be tapped from them; others were built in shallow, wide pits, and
slag-​tapping would have been limited (Bellamy et al. 2001). Evidence for iron-​smelting
on a much larger scale comes from nearby Laxton (Northamptonshire), where excava-
tions uncovered large furnaces 1.3–​1.4 m in diameter, with multiple tuyère (air) holes
(Jackson and Tylecote 1988; Crew 1998). The large diameter of the Laxton furnaces and
the multiple tuyère holes suggest that each furnace would have produced perhaps half
a dozen separate blooms. Associated pottery suggests that the iron-​smelting took place
in the late first century (or early second century). The production in this region has
been characterized as displaying a hierarchy from subsistence to large-​scale produc-
tion at specialized sites for export beyond the region (Condron 1997). For the most
part, iron-​smelting occurs in rural settings, and urban sites in the region provide evi-
dence for smithing only.
While the three traditional iron-​producing regions (Weald, Forest of Dean, and
the east Midlands) have attracted considerable attention, ironsmelting is known from
many other regions. The widespread occurrence of iron-​smelting reflects the fact that
in many cases a small industry could be sustained from quite small ore deposits that
largely escaped the interests of later geologists and industrialists. The investigation of a
probable villa at Mantles Green (Buckinghamshire Chilterns) identified the remains of
a mid-​to late-​second-​century slag heap (Yeoman and Stewart 1992). While no furnace
survived, study of the iron slag confirmed that smelting had taken place. Iron-​smelt-
ing at Mantles Green was probably part of a mixed economy and served the immediate
needs of the estate and, possibly, the surrounding settlements. The identification of iron-​
smelting in the Chilterns is surprising, as there is no historically attested iron-​smelt-
ing industry in the region, and geological maps indicate no iron ores. It is likely that
the early iron-​smelters exploited small deposits within Quaternary deposits. This could
have included bog ore, which would have been much more abundant before the agricul-
tural improvements of more recent times. Other iron-​smelting sites are known in the
area—​for example Cow Roast, Hertfordshire (Morris and Wainwright 1995).
Other regions, such as Norfolk, whose iron-​ore deposits were of limited economic
value in recent times, may have been significant sources of Roman iron production.
Excavations at Ashwicken revealed well-​preserved furnaces and extensive slag depos-
its dated to the second century (Tylecote and Owles 1960). The furnaces, which were
built into the edges of hollows or shallow quarries, were 0.3m in diameter and perhaps 1
m high. Comparable evidence for eight furnaces has been recorded at North Wootton,
where ironworking seems to have occupied an area of 5 hectares in the late Roman
period (Smallwood 1989). Comparable data (although without any in situ furnace struc-
tures) have been recovered from Snettisham (Lyons 2004).
548

548   David Dungworth

Blacksmithing was carried out on almost every Roman period site in Britain. In most
cases this would have comprised the occasional repair of objects, but many sites have
provided evidence for intensive iron-​working. Most urban sites have yielded dedicated
buildings or tools or at least dumps of smithing waste; however, the evidence is most
prominent at ‘small’ towns—​for example, Ashton (Northamptonshire), Bourton-​on-​
the-​Water (Gloucestershire), Braintree (Essex), Brampton (Norfolk), Godmanchester
(Cambridgeshire), and Wilderspool (Cheshire) (Burnham and Wacher 1990). Extensive
evidence for smithing in Southwark, on the south side of the Thames, comprised sev-
eral workshops in an area of four hectares, which specialized in iron-​working (Hammer
2003). The workshops appear to have been rather simple wooden buildings, with hearths
that were set at ground level, showing that smithing took place on the floor rather than
using raised hearths and anvils.
The work of Roman blacksmiths can also be seen in the nature of the iron objects
they made (Manning 1985). Scientific study of the microstructure of Roman iron-
work shows a wide range of materials were used, with varying degrees of blacksmith-
ing skill (Tylecote and Gilmour 1986; Lang 1988, 1995). Some Roman blacksmiths were
highly proficient and could weld hard, but somewhat brittle, steel to soften plain iron
to produce a highly effective composite cutting tool; however, most everyday iron-
work appears to have been prepared with relatively little skill (Tylecote and Gilmour
1986: 99).

Conclusion

Although the Roman conquest had some significant impacts on the ways in which met-
als were made and used in Britain, in other ways the conquest is not immediately obvi-
ous. A great deal is known about how, where, and when metals were produced, but there
still remain large gaps in the evidence, and the interpretation of much of the available
evidence is often problematic.
Classical accounts have been taken to indicate that Britain was seen as a source of
metals and that the invasion was motivated—​at least in part—​by a desire to control and
exploit this resource: Tacitus (Agricola 12) says that ‘Britain yields gold, silver, and other
metals, to make it worth conquering’. The existence of lead ingots with imperial inscrip-
tions dateable to the Flavian period are most frequently cited as evidence that the lead–​
silver industry was established quickly and under military control for the benefit of the
empire (Elkington 1976). Nevertheless, most lead ingots are not inscribed, and the rea-
sons why some were inscribed are not clear. While it seems logical that exploitation of
lead took place primarily to extract silver, the scientific examination of the lead does not
provide unequivocal support for this. Tin and copper ingots are known, but most lack
inscription and so cannot be dated or any real inference made about the organization
of their production. Very little detailed evidence is currently available for the mining
and smelting of non-​ferrous metals beyond the ingots. If anything, even less is known
 Metals and Metalworking    549

about the pre-​conquest industry, and so no useful suggestions can be advanced as to the
changes that the Roman conquest may have brought about. The casting and working of
copper alloys for the manufacture of small items (mostly jewellery) appear to have taken
place in small workshops that can be found on a high proportion of Roman settlements,
including both rural and urban examples. Archaeological evidence for the refining of
gold has been recovered from a number of urban sites and mostly of first-​century date. It
is tempting to see this activity as a Romanization and demonetization of the native pre-
cious metal stock (for example, coins). Similarly, it might be suggested that secondary
cupellation (the recovery of silver from debased silver–​copper alloys), which is often
found in urban centres and dated to the later Roman period, may have been a response
to programmes of coinage debasement.
Substantial evidence is available for the nature of iron manufacture in prehistoric
and Roman Britain. Many of the sites that are listed in the literature were investigated
many years ago, and in some cases the information is not reliable: ‘the report on the
metalworking debris demonstrates such a breathtaking ignorance of iron-​working
technology that it cannot be relied upon’ (Crew 1990: 154). A feature of the Roman
period is a limited number of sites where production took place on an enormous scale
and left slag heaps to be measured in hectares or thousands of tonnes. Sadly, many of
these sites were treated as sources of material for later blast furnaces or road-​metalling
and, in some cases, the evidence has been completely lost. Like many other aspects
of Roman archaeology, the publication research on ironworking sites has often been
delayed, and much evidence is known in summary form only. Nevertheless there
were three main areas of iron production: the Weald, the Forest of Dean, and the east
Midlands (although surplus production beyond local subsistence needs clearly took
place in many other regions). Much of the iron production took place in a rural context
and in some cases was probably integrated into the annual farming cycle. Iron-​smith-
ing was ubiquitous, but the most intensive, specialized smithing appears to have taken
place in urban contexts, although this is most clearly seen in ‘small’ towns. There is vir-
tually no direct evidence for the organization of Roman iron production, but this has
rarely restrained interpretations and modelling building. Condron (1997: 5) suggests
that ‘villa-​based smiths may have partly replaced the services formerly carried out by
itinerant smiths’ without addressing the lack of any evidence of itinerant smiths in the
Roman period or before (cf. Rowlands 1971). Tylecote and Owles (1960) draw attention
to the fact that iron production at Ashwicken was only 3 miles from the villa at Gayton
Thorpe and, by implication, assume that the resident/​owner of the latter controlled the
iron production at the former—​even though a substantial proportion of England is
within 3 miles of a Roman villa. Cleere’s model (1974) of state control for iron produc-
tion in the Weald is tempting for many, but it is driven by presumptions derived from
assumptions about the Mediterranean world with little or no direct archaeological evi-
dence. A better understanding of the role and importance of metallurgy within Roman
Britain needs to be based on the nature of the archaeological evidence. This is a continu-
ing challenge, as the archaeological record is by its very nature partial and at times open
to multiple interpretations.
550

550   David Dungworth

Abbreviation
RIB  R
 . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I:
Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Primary Sources
Diodorus Siculus, Histories, trans. C. H. Oldfather. Loeb Classical Library.
London: Heinemann, 1946.
Herodotus, The Histories, ed. A. De Sélincourt and J. Marincola. London: Penguin, 2003.
Posidonius, Fragments. Volume II: The Commentary, ed. I. G. Kidd. Cambridge: Cambridge
University Press, 2004.
Strabo, Geography, trans. H. L. Jones. Loeb Classical Library. London: Heinemann, 1917–​32.
Tacitus, Agricola and the Germania, trans. H. Mattingly, rev. S.  Handford.
London: Penguin, 2003.

Secondary Sources
Anon. (1971). ‘Romano-​British Lead Working at Scarcliffe Park, near Chesterfield’, Bulletin of
the Historical Metallurgy Group, 5: 38.
Ashworth, H. W. W., and Palmer, L. S. (1957–​8). ‘Romano-​British Metallurgical Workings at
Green Ore, Somerset’, Annual Report of the Wells Natural History and Archaeological Society,
69/​70: 10–​15.
Bayley, J. (1984). ‘Roman Brass-​Making in Britain’, Historical Metallurgy, 18: 42–​43.
Bayley, J. (1988). ‘Non-​Ferrous Metal Working:  Continuity and Change’, in E. A. Slater
and J. O. Tate (eds), Science and Archaeology:  Glasgow 1987. BAR International 196.
Oxford: Archaeopress, 93–​208.
Bayley, J. (1991). ‘Archaeological Evidence for Parting’, in E. Pernicka and G. A. Wagner (eds),
Archaeometry ’90. Heidelberg: Birkhäuser Verlag, 19–​28.
Bayley, J. (1998). ‘The Production of Brass in Antiquity with Particular Reference to Roman Britain’,
in P. T. Craddock (ed.), 2000 Years of Zinc and Brass. London: British Museum Press, 7–​26.
Bayley, J., and Budd, P. (1998). ‘The Clay Moulds’, in H. E. M. Cool and C. Philco (eds), Roman
Castleford Excavations 1974–​85. Volume I:  The Small Finds. Wakefield:  West Yorkshire
Archaeology Service, 195–​222.
Bayley, J., and Butcher, S. (2004). Roman Brooches in Britain: A Technological and Typological
Study Based on the Richborough Collection. London: Society of Antiquaries.
Bayley, J., and Eckstein, K. (1998). ‘Metalworking Debris from Pentrehyling Fort, Brompton,
Shropshire’, Ancient Monuments Laboratory Report 13/​98. London: English Heritage.
Bayley, J., and Eckstein, K. (2006). ‘Roman and Medieval Litharge Cakes: Structure and
Composition’, in J. Pérez-​Arantegui (ed.), Proceedings of the 34th International Symposium
on Archaeometry. Zaragoza: Institución Fernando el Católico, 145–​153.
Bayley, J., and Rehren, T. (2007). ‘Towards a Functional and Typological Classification
of Crucibles’, in S. La Niece and P. Craddock (eds), Metals and Mines:  Studies in
Archaeometallurgy. London: Archetype Books, 46–​55.
 Metals and Metalworking    551

Beagrie, N. (1989). ‘The Romano-​British Pewter Industry’, Britannia, 20: 169–​191.

Bellamy, B., Jackson, D., and Johnston, G. (2001). ‘Early Iron Age Smelting in the Rockingham
Forest Area’, Northamptonshire Archaeology, 29: 103–​128.
Bishop, M. C., and Coulston, J. C. N. (1993). Roman Military Equipment. London: Batsford.
Bland, R., and Johns, C. (1993). The Hoxne Treasure: An Illustrated Introduction. London: British
Museum.
Boon, G. C., and Williams, C. (1966). ‘The Dolaucothi Drainage Wheel’, Journal of Roman
Studies, 56: 122–​127.
Burnham, B. C., and Burnham, H. (2004). Dolaucothi-​Pumsaint: Survey and Excavation at a
Roman Gold-​Mining Complex (1987–​1999). Oxford: Oxbow Books.
Burnham, B. C., and Wacher, J. (1990). The Small Towns of Roman Britain. London: Batsford.
Butcher, K., and Ponting, M. (2012). ‘The Beginning of the End? The Denarius in the Second
Century’, Numismatic Chronicle, 172: 63–​83.
Caley, E. R. (1964). Orichalcum and Related Ancient Alloys. New York: American Numismatic
Society.
Cartwright, C. (1992). ‘The Excavation of a Romano-​British Iron-​Working Site at Broadfield,
Crawley, West Sussex’, Sussex Archaeological Collections, 130: 22–​59.
Cleere, H. F. (1970). The Romano-​British Industrial Site at Bardown, Wadhurst:  An Interim
Report on Excavations, 1960–​1968. Chichester: Sussex Archaeological Society.
Cleere, H. F. (1974). ‘The Roman Iron Industry of the Weald and its Connections with the
Classis Britannica’, Archaeological Journal, 131: 171–​199.
Cleere, H. (1976). ‘Some Operating Parameters for Roman Ironworks’, Bulletin of the Institute of
Archaeology, 13: 233–​246.
Cleere, H., and Crossley, D. (1995). The Iron Industry of the Weald. 2nd edn. Cardiff: Merton
Priory.
Condron, F. (1997). ‘Iron Production in Leicestershire, Rutland and Northamptonshire in
Antiquity’, Transactions of the Leicestershire Archaeological and Historical Society, 71: 1–​20.
Cowell, M. R., and Hook, D. R. (2010). ‘The Analysis of the Metal Artefacts’, in C. Johns, The
Hoxne Late Roman Treasure:  Gold Jewellery and Silver Plate. London:  British Museum,
175–​183.
Cowell, M. R., La Niece, S., Meeks, N. D., Oddy, W. A., and Holmes, R. (1983). ‘The
Scientific Examination of the Thetford Treasure’, in C. Johns and T. Potter, The Thetford
Treasure: Roman Jewellery and Silver. London: British Museum, 56–​67.
Crew, P. (1990). ‘Late Iron-​Age and Roman Iron Production in North-​West Wales’, in B. C.
Burnham and J. L. Davies (eds), Conquest, Co-​existence and Change: Recent Work in Roman
Wales. Lampeter: University College, 150–​160.
Crew, P. (1998). ‘Laxton Revisited: A First Report on the 1998 Excavation’, Historical Metallurgy,
32: 49–​53.
Cuttler, R., and Buteaux, S. (2011). Life and Industry in the Suburbs of Roman Worcester. BAR
British Series 533. Oxford: Archaeopress.
Dalwood, C. H., and Edwards, R. (2004). Excavations at Deansway, Worcester 1988–​89.
Romano-​British Small Town to Late Medieval City. York: Council for British Archaeology.
Darlington, J., and Evans, J. (1992). ‘Roman Sidbury, Worcester:  Excavations 1959–​1989’,
Transactions of the Worcestershire Archaeological Society, 13: 5–​104.
Davies, O. (1935). Roman Mines in Europe. Oxford: Clarendon Press.
Draper, J. (1985). Excavations at Hill Farm, Gestingthorpe, Essex. Gressenhall:  East Anglian
Archaeology.
552

552   David Dungworth

Dungworth, D. (1996). ‘Caley’s “Zinc Decline” Reconsidered’, Numismatic Chronicle,

156: 228–​234.
Dungworth, D. B. (1997). ‘Roman Copper Alloys: Analysis of Artefacts from Northern Britain’,
Journal of Archaeological Science, 24: 901–​910.
Dunster, J., and Dungworth, D. (2013). St Algar’s Farm, Frome, Somerset: The Analysis of Lead-​
working Waste. Research Report 15/​2012. Portsmouth: English Heritage.
Elkington, D. (1976). ‘The Mendip Lead Industry’, in K. Branigan and P. J. Fowler (eds), The
Roman West Country. London: David and Charles, 183–​198.
Fox, A. (1996). ‘Tin Ingots from Bigbury Bay, South Devon’, Bulletin of the Peak District Mines
Historical Society, 13: 150–​151.
Fox, A., and Ravenhill, W. (1972). ‘The Roman Fort at Nanstallon, Cornwall’, Britannia,
3: 56–​111.
Frere, S. S. (1972). Verulamium Excavations: Volume I. London: Society of Antiquaries.
Frere, S. S., Roxan, M., and Tomlin, R. S. O (1990). The Roman Inscriptions of Britain. Volume II.
Fasicule 1: Instrumentum Domesticum. Stroud: Alan Sutton.
Fulford, M. G., and Allen, J. R.  L. (1992). ‘Iron-​Making at the Chesters villa, Woolaston,
Gloucestershire: Survey and Excavations 1987–​91’, Britannia, 23: 159–​215.
Gerrard, S. (2000). The Early British Tin Industry. Stroud: Tempus.
Hammer, F. (2003). Industry in North-​West Roman Southwark. London: Museum of London.
Hawkes, C. F.  C. (1984). ‘Ictis Disentangled and the British Tin Trade’, Oxford Journal of
Archaeology, 3: 211–​233.
Hodgkinson, J. (2008). The Wealden Iron Industry. Stroud: Tempus.
Jackson, D. A. (1981). ‘Archaeology at an Ironstone Quarry in the Harringworth–​Wakerley
Area 1968–​79’, Northamptonshire Archaeology, 16: 14–​33.
Jackson, D. A., and Ambrose, T. M. (1978). ‘Excavations at Wakerley, Northants, 1972-​75’,
Britannia, 9, 115–​242.
Jackson, D. A., and Tylecote, R. F. (1988). ‘Two New Romano-​British Iron Working Sites in
Northamptonshire: A New Type of Furnace?’, Britannia, 19: 275–​298.
Jackson, R. (2012). Ariconium, Herefordshire: An Iron Age Settlement and Romano-​British Small
Town. Oxford: Oxbow Books.
Johns, C. (2010). The Hoxne Late Roman Treasure:  Gold Jewellery and Silver Plate.
London: British Museum.
Johns, C., and Potter, T. (1983). The Thetford Treasure:  Roman Jewellery and Silver.
London: British Museum.
Jones, G. D. B., and Mattingly, D. (1990). An Atlas of Roman Britain. Oxford: Blackwell.
Lang, J. (1988). ‘Study of the Metallography of Some Roman Swords’, Britannia, 19: 199–​216.
Lang, J. (1995). ‘A Metallographic Examination of Eight Roman Daggers from Britain’, in B.
Raftery, with V. Megaw and V. Rigby (eds), Sites and Sights of the Iron Age. Oxford: Oxbow
Books, 119–​132.
Lawson-​ Jones, A. (2013). ‘Discoveries along the North Land’s End Pipeline’, Cornish
Archaeology, 52.
Lee, R. (2009). Production, Use and Disposal of Romano-​ British Pewter Tableware.
Oxford: Oxbow Books.
Lewis, P. R., and Jones, G. D. B. (1969). ‘The Dolaucothi Gold Mines, I: The Surface Evidence’,
Antiquaries Journal, 49: 244–​272.
Love, J. R. (1991). Antiquity and Capitalism. London: Routledge.
 Metals and Metalworking    553

Lyons, A. (2004). Romano-​British Industrial Activity at Snettisham, Norfolk. Dereham: Norfolk

Museum and Archaeology Service.
Manning, W. H. (1985). Catalogue of the Romano-​British Iron Tools, Fittings and Weapons in the
British Museum. London: British Museum.
Marvell, A. (2001) (ed.). Investigations along Monnow Street, Monmouth. BAR British Series
320. Oxford: Archaeopress.
Meharg, A. A., Edwards, K. J., Schofield, J. E., Raab, A., Feldmann, J., Moran, A., Bryant, C. L.,
Thornton, B., and Dawson, J. J. C. (2012). ‘First Comprehensive Peat Depositional Records
for Tin, Lead and Copper Associated with the Antiquity of Europe’s Largest Cassiterite
Deposits’, Journal of Archaeological Science, 39: 717–​727.
Millett, M. (2007). ‘Roman Kent’, in J. H. Williams (ed.), The Archaeology of Kent to ad 800.
Woodbridge: Boydell and Kent County Council, 135–​184.
Morris, M., and Wainwright, A. (1995). ‘Iron Age and Romano-​British Settlement, Agriculture
and Industry in the Upper Bulbourne Valley, Hertfordshire: An Interim Interpretation’, in R.
Holgate (ed.), Chiltern Archaeology: Recent Work. Dunstable: The Book Castle, 68–​75.
Oddy, A. (1981). ‘Gilding through the Ages’, Gold Bulletin, 14: 75–​79.
Page, N. (2005). ‘Industrial Complex and Timber Trackway at Llangynfelin’, Archaeology in
Wales, 45: 103–​104.
Painter, K. S. (1977a). The Mildenhall Treasure: Roman Silver from East Anglia. London: British
Museum.
Painter, K. S. (1977b). The Water Newton Early Christian Silver. London: British Museum.
Paynter, S. (2008). ‘Metalworking Remains’, in P. Booth, A.-​M. Bingham, and S. Lawrence,
The Roman Roadside Settlement at Westhawk Farm, Ashford, Kent:  Excavations 1998–​9.
Oxford: Oxford Archaeological Unit, 267–​301.
Penhallurick, R. D. (1986). Tin in Antiquity. London: Institute of Metals.
Quinnell, H. (2004). Trethurgy. Excavations at Trethurgy Round, St Austell: Community and
Status in Roman and Post-​Roman Cornwall. Truro: Cornwall County Council.
Rahtz, P. A., and Greenfield, E. (1977). Excavations at Chew Valley Lake, Somerset. London: HMSO.
Ramage, A., and Craddock, P. T. (2000). King Croesus’ Gold: Excavations at Sardis and the
History of Gold Refining. London: British Museum.
Rehren, T. (1996). ‘Large Scale-​Small Size Production: Roman Brass Making Crucibles’, Journal
of Archaeological Science, 26: 1083–​1087.
Rowlands, M. J. (1971). ‘The Archaeological Interpretation of Prehistoric Metalworking’, World
Archaeology, 3: 210–​223.
Schrüfer-​Kolb, I. E. (2004). Roman Iron Production in Britain: Technological and Socio-​
Economic Landscape Development along the Jurassic Ridge. BAR British Series 380. Oxford:
Archaeopress.
Smallwood, J. P. (1989). ‘Romano-​British Iron Working at North Woottom, King’s Lynn,
Norfolk’, Britannia, 20: 243–​245.
Smythe, J. A. (1939–​40). ‘Roman Pigs of Lead from Brough’, Transactions of the Newcomen
Society, 20: 139–​145.
Timberlake, S. (2002). ‘Medieval Lead Smelting Boles near Penguelan, Cwmystwyth,
Ceredigion’, Archaeology in Wales, 42: 45–​59.
Timberlake, S. (2003). ‘Early Mining Research in Britain: The Developments of the Last Ten
Years’, in P. Craddock and J. Lang (eds), Mining and Metal Production through the Ages.
London: British Museum, 21–​42.
554

554   David Dungworth

Timberlake, S., and Prag, A. J.  N. W. (2005). The Archaeology of Alderley Edge:  Survey,
Excavation and Experiment in an Ancient Mining Landscape. BAR British Series 396.
Oxford: Archaeopress.
Todd, M. (2007). Roman Mining in Somerset: Excavations at Charterhouse-​on-​Mendip, 1993-​
1995. Exeter: Mint Press.
Tylecote, R. F. (1986). The Prehistory of Metallurgy in the British Isles. London:  Institute of
Metals.
Tylecote, R. F., and Clough, R. E. (1983). ‘Recent Bog Iron Ore Analyses and the Smelting of
Pyrite Nodules’, Offa, 40: 115–​118.
Tylecote, R. F., and Gilmour, B. J. J. (1986). The Metallography of Early Ferrous Edge Tools and
Edged Weapons. BAR British Series 155. Oxford: Archaeopress.
Tylecote, R. F., and Owles, E. (1960). ‘A Second-​Century Iron Smelting Site at Ashwicken,
Norfolk’, Norfolk Archaeology, 32: 142–​162.
Wainwright, G. J. (1979). Gussage All Saints: An Iron Age Settlement in Dorset. London: HMSO.
Walker, D. R. (1976–​8). The Metrology of Roman Silver Coinage: Volumes 1–​3. BAR International
Series 5, 22, and 40. Oxford: Archaeopress.
Whittick, G. C. (1982). ‘The Earliest Roman Lead-​Mining on Mendip and in North Wales: A
Reappraisal’, Britannia, 13: 113–​123.
Wilson, P. R. (2002). Cataractonium:  Roman Catterick and its Hinterland. Excavations and
Research, 1958–​1997: Part II. York: Council for British Archaeology.
Worssam, B. C., and Gibson-​Hill, J. (1976). ‘Analyses of Wealden Iron Ores’, Historical
Metallurgy, 10: 77–​82.
Yeoman, P. A., and Stewart, I. J. (1992). ‘A Romano-​British Villa Estate at Mantles Green,
Amersham, Buckinghamshire’, Records of Buckinghamshire, 34: 107–​182.
Young, T. (1993). ‘Sedimentary Iron Ores’, in R. A.  D. Pattrick and D. A. Polya (eds),
Mineralization in the British Isles. London: Chapman and Hall, 446–​489.
Young, T. P., and Thomas, G. R. (1999). ‘Provenancing Iron Ore from the Bristol
Channel Orefield:  The Cargo of the Magor Pill Boat’, in A. M. Pollard (ed.),
Geoarchaeology:  Exploration, Environments, Resources. London:  Geological Society of
London, 103–​121.
Zienkiewicz, J. D. (1993). ‘Excavations in the Scamnum Tribunorum at Caerleon: The Legionary
Museum Site 1983–​5’, Britannia, 24: 27–​140.
 Chapter 27

M edic i ne

Patricia Baker

Introduction

Historical and anthropological examinations of medicine have shown that variant

cultural perceptions of the body, healthcare, and medical practices existed among dif-
ferent societies in the past, as in the present. Even Greco-​Roman medical writers com-
mented on regional differences in medical treatments and philosophies (e.g. Celsus, de
Med proem. 30–​1; Soranus, Gyn. 2. 12). Anthropologists have found that beliefs about
the body are often deeply embedded into a society’s manner of thinking—​so much so
that, when new traditions are introduced by one group of people to another, there can
be strong reactions towards the new ideas. In some instances they are rejected, in others
they are adapted to fit the beliefs of the society receiving the information, and in other
cases they can be adopted outright (e.g. Kleinman 1980; Nijhuis 1995). In regards to this
research and the comments made by the ancient writers, it should be expected that stud-
ies of Roman provincial medicine would point to variations in practice, especially since
the Roman Empire covered a large geographical area and was populated by different
societies, which held their own beliefs, rituals, and customs. However, this is generally
not the case in scholarship on the subject of ancient medicine.
Until roughly the early 1990s, and in some cases even now, the idea was promoted
that when the Romans occupied a province their medical practices were brought with
them and adopted by the local inhabitants; thus medical procedures were deemed to be
homogenous across the empire. The consistent use of the term ‘Roman medicine’ in ref-
erence to provincial healthcare contributes to these scholarly assertions. It was implied,
and in some instances argued, that the native populations would have welcomed Roman
medicine and they would have viewed it to be superior to their own indigenous tra-
ditions (e.g. Simpson 1856; Richmond 1952; Nutton 1972; Jackson 1988; Allason-​Jones
1993, 1999; Breitwieser 1998; Künzl 2002). The use of Roman-​style medical objects, the
construction and design of places used for treatment, and the meanings of the Latin
or Greek terms for ‘doctor’ found on inscriptions, papyri, and wooden tablets, and in
556

556   Patricia Baker

medical texts, are all described as if their functions and meanings were universally
understood across the empire.
Yet, these ideas have been called into question for Roman Britain and elsewhere (e.g.
Jackson 1997, 2007; Baker 2001, 2002a, 2004a, b, 2011; Draycott 2011). Both the influ-
ence of the anthropological studies mentioned above and developments in text-​based
scholarship on ancient medicine have informed a more critical approach to provincial
Roman medicine. Scholars who focus on text have found there were multi-​variant phi-
losophies about the body, illness, and treatment among the ancient medical writers who
lived contemporaneously in Greece and Rome (e.g. Smith 1979; Nutton 1995). They have
also argued that that there is little evidence to support the traditional idea that Greco-​
Roman medicine was ‘rational’ (Longrigg 1993)—​that which was empirical in regards to
modern scientific standards of assessment and divorced from divine elements. To put
this simply, although the ancient medical writers and philosophers were attempting to
find natural causes for how the cosmos and the body worked, they still saw nature as
divine. Thus, ancient philosophical questions concerning the nature of the body and
illness were posed within the cultural frameworks of the particular place and time in
which the philosophers lived (e.g. Lloyd 1983; van der Eijk 2005).
The questions that these studies force us to ask about medicine in the provinces, par-
ticularly Britannia, are: (1) if writers of ancient medicine living within the same period
and region disagree with one another about the functions of the body and care, is it pos-
sible to argue for homogeneity elsewhere in the empire?; (2) if medical concepts were
introduced from one region to another, which and/​or whose ideas were they?; and (3)
since the idea of Greco-​Roman ‘rational’ medicine has now been shown to be culturally
constructed, can it be argued that those living in the provinces found it superior to their
own practices and, therefore, adopted it?
In the light of these questions, it is the aim of this chapter to present recent evalu-
ations of the archaeological and literary excerpts that survive from Roman Britain to
show that there was likely to have been a hybridity of medical beliefs and practices that
developed through trade with other provinces; contact with the Roman army that con-
sisted of units from different regions of the empire; and contact with Roman and provin-
cial diasporas through governance and occupation. It is argued that it is more accurate
to refer to ‘Romano-​British medicine’ or ‘Roman-​period medicine’ than ‘Roman medi-
cine’ when describing healthcare in the provinces, because it conveys the multi-​variant
understandings of the body, health, and healing in the ancient world.

Historical Overview

Before the recent scholarship on Romano-​British medicine is presented, a brief back-

ground of the subject of medical history is given here to show how early studies of
ancient medicine influenced the interpretations of provincial healthcare.
 Medicine   557

From the mid-​nineteenth to mid-​twentieth centuries, medical history was taught

as part of the curriculum in medical schools in Europe and North America. This
aspect of research and teaching was mainly undertaken by medical practitioners who
also had a traditional classical education. They taught and studied the extant medical
literature, such as the works of the Hippocratic writers and Galen, for two main rea-
sons. The first was to show the development of medicine from antiquity to the present
and the second was to use these ‘great men of medicine’ as role models for doctors
to emulate, since they had social responsibilities and interacted with people on an
intimate basis. Translations of the ancient medical treatises were made to facilitate
the study of the subject (e.g. Fee and Brown 2004; Nutton 2004). Even the archae-
ological remains of medical implements were examined by physicians to see how
the field of surgery evolved. One of the most notable contributions at this time was
Milne’s Surgical Instruments in Greek and Roman Times (1907). In the book, he com-
pared medical objects with medical texts to determine their function. Yet, there are few
descriptions of the actual tools given in the extant literature, so in some places it is pos-
sible to question whether Milne had correctly identified the tools he was describing.
Surgical instruments found in Roman Britain were often published by medical doctors,
rather than archaeologists (e.g. Francis 1926; Thomas 1963). Along with this, archaeolo-
gists who did discuss ancient medicine used their surroundings and familiarity with
the medicine of the time in which they lived to describe Roman civilian and military
practices (e.g. Koenen 1904; Richmond 1952).
Towards the second half of the twentieth century, medical history went from
being a subject studied by physicians to one that was mainly examined by histori-
ans, classicists, and archaeologists. The focus was on ancient literature, although
archaeologists and paleopathologists also added to our awareness of the subject
(e.g. Scarborough 1968, 1976; Davies 1969a, b, 1970, 1972, 1989; Nutton 1969, 1970,
2004; Bliquez 1981, 1988; Künzl 1983, 1996; Jackson 1988, 1990, 1993, 1994a, b, 1995,
1996, 2002, 2011). At this time there was a move away from studying biographies of
medics towards examining the social role of doctors and medicine in society. It became
apparent in the textual examinations that not all of the ancient writers agreed with
one another on their theories of the body, as mentioned above (e.g. Nutton 2004). The
archaeological studies, however, continued to be interpreted in terms of rational and
homogenous practices. Ultimately these arguments are founded in traditional views of
Romanization.
Since the mid-​1990s interpretations of the archaeological and epigraphic evidence
in the Roman provinces have been affected not only by the text-​based examinations,
but also by theoretical developments in archaeological methods of interpretation. It is
becoming more apparent that some of the long-​held assertions about the Romanization
of medicine in Britain and elsewhere cannot be supported (Jackson 1997, 2011; Baker
2001, 2002a, 2004a, 2011; Crummy 2002). In Britain, variant practices are evident
in surviving medical tools, inscriptions, structures, and osteological remains, as
discussed below.
558

558   Patricia Baker

Medical Instruments

The archaeological remains of surgical instruments found throughout the Roman

Empire tend to be similar in design, so are described in an analogous manner in site
reports and catalogues. This type of description contributes to the idea that the tools rep-
resent the spread of Roman medicine. However, these objects are multi-​functional and
could have been used for mixing and applying ointments, cosmetics, and paints. Like
any object, when it was introduced to a new society, local customs could also have been
applied to them. A closer examination of the instruments’ context and associated finds
is a more appropriate comparison to make to determine how they may have been used
and thought about in Roman Britain. Unfortunately, many of the remains of Roman-​
style medical instruments found in the province are single finds without a clear context
(e.g. Bosanquet 1904: 289; Wheeler 1928: 169, no. 4, figure 33.1; Nash-​Williams 1932: 90,
no. 5, figure 38; Fox 1940: 127, no. 1, figure 5; Thomas 1963; Gilson 1978, 1981, 1982, 1983;
Allason-​Jones and Miket 1984; McCarthy 1990: 137–​138; Wilmott 1997: 287–​288).
Some of the tools that have a recorded provenance indicate that the artefacts might
not have had a straightforward medical function and show a mixing of Roman and Iron
Age traditions in how they were used and understood (Baker 2004b). For instance,
a spatula probe was found deliberately placed at the bottom of a post-​hole in what is
believed to be the centurion’s quarters of a barrack in the legionary fortress of Caerleon
(Baker 2004a: 182–​183, no. 18). Probes are common multi-​functional objects that, in a
votive context, might have represented a person’s body and/​or work. The location of the
artefact suggests the possibility that it was used as a foundation offering to protect the
health of the person or people who lived and worked within the structure (Baker 2004a:
21).
Another example of a medical tool being used as a possible votive object was a
stray find of surgical forceps found on the bank of the River Trent in Littleborough,
Nottinghamshire—​an area with evidence for Roman occupation, including an auxiliary
fort. The object was found by a metal detectorist, so there remains some doubt about its
exact context (Jackson and Leahy 1989). Although the artefact might have been lost or
washed down the river, medical instruments were found in rivers and bodies of water in
Roman Gaul and Germany and, sometimes, in places known for Iron Age votive deposi-
tion. The practice of placing votive offerings in water continued in the Roman period,
but with the use of newly introduced Roman-​style medical objects (Baker 2004b: 16–​20;
2011: 167–​170). It is, therefore, possible that people living in Roman Britain who wished
to make an offering in relation to themselves or their health could have placed an object
associated with their body in areas deemed appropriate for votive deposition.
A further indication that the medical objects were utilized in regional customs is
noted by some being placed in burials. The practice of burying medical objects with
the dead is not common, but, archaeologically, more evidence for this has been found
in Gallia Belgica and its surrounding areas. Gallia Belgica has eleven burials, the border
 Medicine   559

area with Germania has seven burials, and Raetia has four (Künzl 1983: 2, 59–​74). The
number of burials with medical tools in Italy is seven; yet of these, four were found in
the Alpine region near Raetia (Künzl 1983: 2, 103–​108). The number of burials in Britain
with instruments is three. Two have single instruments in them, but the third has a set
of thirteen tools. This cremation burial was found in 1996 in Stanway near Colchester,
Essex, which was an area in contact with Gaul and Rome (Crummy 2002, 2007; Jackson
1997, 2007).
The Stanway find is also exceptional because the design of the medical tools is not
entirely like those found in Italy and elsewhere. The ‘doctor’s burial’ (CF47) is so named
because of the medical tools found within it. It was unearthed in the fifth burial enclo-
sure located at the site and dates to ad 40–​50/​55. The cremated remains are of indeter-
minate age and sex (Jackson 1997; 2007: 236–252; Crummy 2002; Crummy et al. 2007:
201–​253; Mays 2007: 380–​381; Wiltshire 2007: 394–​388). It is suggested that the remains
might be those of a male on account of two brooch types found within the burial that
are generally associated with men (Crummy 2007: 333–​345). However, these might also
have been gifts for the dead.
With one exception, all of the instruments found in the burial were manufactured dif-
ferently from those found elsewhere in the Roman Empire. There exist some rare Iron
Age medical tools from continental Europe, and the Stanway set is somewhat compara-
ble to them. Roman-​period instruments are commonly made of copper alloy, but in this
instance the tools were crafted out of both iron and copper alloy, suggesting localized
technological practices in style and preference for fabric (Crummy 2002: 51; Jackson
2007: 245). This mixing of traditions is also supported by the associated artefacts found
within the burial.
The selection of thirteen instruments consists of one Roman-​style copper-​alloy scoop
probe. There were two one-​piece scalpels made of iron. In comparison, Roman-​style
scalpels were made of two parts with a bronze handle and detachable steel blade. There
were two combined hook instruments that had a sharp and blunt hook on each of their
ends. One of the hooks was made of copper alloy and the other was fabricated out of
iron. Also included were another double-​hooked instrument and a smooth-​jawed cop-
per-​alloy forceps that was similar in style to Roman ones, but longer than those found
elsewhere. The remaining instruments consisted of a set of three iron-​handled needles
graded in size, a small iron saw with a copper-​alloy and bone handle, a possible retrac-
tor, and an unidentified instrument (Crummy 2002: 51; Jackson 1997; 2007: 236–​245,
250–25​2).
When the tools were excavated, ten of them were found spread out on the top of
a wooden gaming board with the other three placed near to it. The surgical saw was
deliberately broken into five pieces (like pottery in some of the other burials found at
the site) and arranged in the centre of the board. The gaming counters were laid out on
the board and the cremated remains were also placed on it. It was originally thought
that they represented a game in progress, but this has now been disputed (Schädler
2007: 359–3​75). One of the more perplexing finds from the burial was a set of eight rods;
four of these were made of copper alloy and four of iron. Each group consisted of two
560

560   Patricia Baker

small and two large rods. Eight copper-​alloy rings were found close by the rods. Some
of the rings show signs of having been attached to leather straps. It may be that these
objects were used in some form of divination, but their use(s) are indeterminate. Three
of the rods were also placed end-​to-​end to overlap the board; the rest were placed in a
bundle to the side of it.
Aside from the unusual finds placed on or near the gaming board, the burial also con-
tained eleven Gallo-​Belgic vessels: an amphora, a ceramic flagon, a samian bowl, a cop-
per-​alloy pan, a copper-​alloy strainer bowl, three brooches, a glass bead, and various
fragments of wood and textiles. All of the vessels were placed in the burial in an upright
position, and they were divided into two groups according to type. The pottery had been
placed together at one end of the burial and the metal vessels were placed at the other
end along with the flagon and samian bowl. The gaming board was open and was laid
next to the metal vessels (Crummy 2002; 2007: 201–2​36).
The strainer bowl might have been used in the preparation of medicines. Pollen anal-
ysis showed that it contained Artemisia (mugwort or wormwood). This was as an ingre-
dient that was used to flavour wines, and it also had a medicinal function. There is no
evidence of wine in the strainer bowl, which suggests that the ingredient was more likely
to be used in a medical remedy (Wiltshire 2007: 398).
Overall, this burial is a significant archaeological find in Britain because it shows
signs of a mixing of medical traditions that possibly came from Rome, Iron Age Gaul,
and Britannia. Besides the design of the objects and their context, further support for
hybrid practices is found with the Gallic pottery in the burial (Dannell 2007: 213–​215).
Its date and proximity to the site of Colchester are argued by Jackson (1997) to be a sign
of interaction with the invading Romans.
Since there is an obvious link to Gaul, Jackson (2007: 250) has recently suggested the
possibility that the person buried in the grave might have been a druid on account of an
argument put forth by J. Webster on the abolition of the Druids from Gaul. Crummy
(2007: 445), however, disputes Jackson on the grounds that when Caesar (de Bello
Gallico 6. 13–​14) mentions Druids he does not state that they were healers. Although it
is impossible to say who this person actually was, the objects are of interest because they
show links with Gaul in relation to medicine.
Thus, the few medical instruments found in Britain that have a context do not indi-
cate a full adoption of Roman practices. They seem to have been used and held mean-
ings that were local to the region.

Collyrium Stamps

Closer examination of the material properties of objects can also be used to deter-
mine how people might have understood their functions and meanings. Additional
support for the above-​mentioned observation that there might have been an influ-
ence of Gallic medical practices in Britain is realized in the distribution of collyrium
 Medicine   561

stamps in the southern part of the province. Collyrium stamps are stone objects that
were used in the Roman period to mark medicines, particularly eye remedies. Of the
314 or so known stamps, roughly 202 are from the provinces of Gaul, mainly Belgica
(Voinot 1984, 1999). The majority of the stamps found outside Gaul were from the
nearby provinces of Germania Inferior, Germania Superior, and Britannia (Jackson
1996: 2241; Voinot 1999: 12–​14); much like the distribution of burials with medical
instruments. Twenty-​eight stamps have so far been found in Britain, the majority
from the south.
The objects tend to be quadrilateral and measure about 4–​5 centimetres in width and
a few millimetres in depth. The inscriptions located on them are always in retrograde,
upper-​case letters. Three elements are usually inscribed: a personal name in the genitive
case, the name of the medical ingredients, and the name of the ailment for which the
medicine might have been used. There are sometimes different remedies and ailments
inscribed on the various sides of the same object. There is no certainty who the person
named on the stamp was. However, since it was inscribed in the genitive case, it could
mean that the stamp, the medical recipe, or both, belonged to the named person. It is
possible that the object belonged to a healer who might have mixed, bought, sold, and/​
or used the remedies. The names are not of a single origin and are Roman and Greek
(Voinot 1999: 23–2​8).
There are very few stamps from Italy, so they seem to be Gallic objects created after
Roman occupation. Although there have been a number of scholarly discussions about
their Gallic ubiquity (e.g. Nutton 1972; Boon 1983; Jackson 1996), it was only recently
that a consideration of the properties and deposition of the objects was made (Baker
2011). It was found that a majority of the objects were crafted of two types of stone: schist
and steatite. Particular stones are mentioned in the ancient texts for their uses in treat-
ments, and it is possible that these stones were thought to have had special properties
necessary for healing vision. It was also noted that many were green in colour. Green is
frequently mentioned in ancient texts for its use in restoring vision. Pliny, for example,
suggested that gem carvers stare at emeralds and green beetles to rest their eyes (Natural
History 29, 132; 37, 62–​64). Thus, it is possible that some Roman ideas about the health of
the eyes were spread, in this instance, beliefs about the colour green, but incorporated
into provincial objects and practices.
When their contexts were noted, many of the stamps were found in places known for
votive deposition, such as water and Gallic sanctuaries, so it is possible that these objects
might also have been used as votives or amulets—​much like some of the medical tools
discussed above. Again there seems to be a continuation of indigenous practices with
the use of newly introduced Roman-​period objects (Baker 2011).
There has not been a thorough study of collyrium stamps found in Britain, but
research for this chapter has brought about the opportunity to present a brief survey
of the properties of those listed in the Roman Inscriptions of Britain II Fascicule 4. The
stamps are comparable to those found in Gaul. There are thirty stamps listed, includ-
ing one from Ireland and two stamp impressions found on pottery vessels. Eighteen
of the twenty-​five stones that are described are green (RIB II. 2446.1–​4, 6, 8–​9, 11–​12,
562

562   Patricia Baker

14–​16, 19, 26–​27, 30–​32), three are grey (RIB II. 2446.21–​22, 28), and four are greyish
or greenish in colour (RIB II. 2446.10, 13, 23–2​4). One even has an inscription of a
medicine called a ‘green’ (Cholron) salve (RIB II. 2446.3), but there is no illness men-
tioned on this stamp for which it might have been used to cure. Another mentions
a ‘sea green’ (Th[a]‌la[s](s)er(os)) salve for clear sight (RIB II. 2446.10), which relates
to Pliny’s observation mentioned above. The types of stones used are also like those
found in Gaul: seven are of schist (RIB II. 2446.2, 4, 19, 30–3​2) and six are made of
steatite (RIB II. 2446.6, 9, 11–​12, 15, 27). Hence, it is possible that these objects repre-
sent a Gallic, or Gallo-​Roman, medical tradition that was passed onto the popula-
tion of southern Britain.

Vindolanda Tablets and Inscriptions

There are no direct ancient literary references to medical practices in Roman Britain,
but the Vindolanda Tablets and medical inscriptions shed some light onto medicine in
the province. Some of these remains also show links to Gallia Belgica.
The Vindolanda Tablets are an important resource for understanding life in a Roman
military situation. Yet, because of their fragmentary condition, it is often impossible to
extract precise and detailed information from them. Although they do mention some
aspects related to medicine, it is argued here that they should be used with care. A dis-
cussion of the other information surviving on the tablets should be presented when
interpretations are made of their meaning in regards to medical practices at the fort.
Therefore, each tablet that has a reference to medicine will be discussed below to explain
what can be derived from it about the possible healthcare available to the soldiers sta-
tioned at the site.
One of the most complete tablets is a list of soldiers from the First Cohort of
Tungrians, a unit from Gallia Belgica, who were absent from duty (Tab. Vindol. II.
154)  (these can be found online at <http://​vto2.classics.ox.ac.uk/​>). Those that are
missing from duty because of a medical problem are divided into three groups: fif-
teen were sick (aegri), six were wounded (uulnerati), and ten suffered from some-
thing called lippientes (an inflammation of the eyes). Some papyri found in Egypt lists
absent and ill soldiers, but they are not divided into particular categories like those
listed on this tablet. There are a couple of possibilities for why these classifications
might have been made: they might indicate that the soldiers were cared for by differ-
ent specialists or placed in different areas of the fort for treatment (Baker 2004a: 42).
Another suggestion could be based on the cultural background of the unit. A high
number of collyrium stamps, some archaeological evidence of medical tools for cata-
ract surgery, and a relief sculpture with a representation of eye treatment were found
in Gallia Belgica (CIL XIII. 4668; Baker 2011: 166). From this evidence it seems as if
the care of the eyes was of particular importance for the indigenous population of
Gaul (Baker 2011). The soldiers in the unit might have preferred to bring some of their
 Medicine   563

own medical philosophies with them, explaining the division of those absent because
of medical problems.
Another fragmentary tablet (Tab. Vindol. II. 155) lists 343 people associated with the
workshop; for example, shoemakers and builders for the bathhouse. The word valetu-
dinarium (hospital) also appears on it, but the context is uncertain. Above it is listed a
fragment of the possible word for ‘wagons’ ([a]d.ar[]) and below it survives ‘to the kilns’
(ad furnaces). From this fragment it is possible to say only that something called a hos-
pital probably existed at the site. We do not know if this document indicates people who
worked in it, built it, or made objects for it. It might even be interpreted as the hospital
being part of the workshop.
A doctor (medicus) named Marcus (Tab. Vindol. II. 156) is described on another doc-
ument as going with people to build the hospitium (residence), which provides support
for the idea that a doctor was stationed at the fort. However, we do not know what medi-
cal tradition(s) he practised, how long he was stationed at the site, or if he worked with a
team of medics.
A fragmentary letter to Lepidina (Tab. Vindol. II. 294) might have a medical reference
in it; however, there are difficulties with the translation. It seems to speak of remedies for
a fever that will be brought to Lepidina. If the translation is correct, it suggests that med-
ical treatments were not only available from doctors in the fort, but also shared between
laypeople.
Vitalius the seplasiarius (ungent seller or ‘pharmacist’) is listed on two fragments
(Tab. Vindol. II. 586 and 877; see Bowman et al. 2011: 125–​127). In number 586, there is
reference to him receiving five modii of gruel (halicae), while in 877 his name is listed on
the back of a document. It is possible that he made remedies for the soldiers without the
need for doctors or that he made them for the medicus(i) stationed at the fort, but fur-
ther than this little more can be said about him.
Finally, a list of ingredients was found on one of the tablets (Tab. Vindol. II. 591) that
might have been needed for producing medicines. All of the materials mentioned on
it have known uses in surviving medical recipes, but they could also have been spices,
foods, or even writing materials (given reference to wax and blacking) (see Van der
Veen, this volume, for some uses of herbs in medicine). It could simply be a list of sup-
plies for many purposes, although the suggestion for medicines is also a possibility,
especially since there was a seplasiarius stationed at the fort.
The lists and letters from Vindolanda provide a small amount of evidence for the
availability of medical treatments in that fort, but for other sites we have only a few epi-
graphic remains that have a medical association. These tend to be altars to Asclepius, the
healing god. Asclepius was worshipped throughout the Greco-​Roman world, and there
are some statuettes of him found in Roman Britain (Jackson 2011). Two of the altars (RIB
461 and 1028) were dedicated by doctors, and four of them were set up by soldiers or
civilians (RIB 609, 808, 1052, 1072). All but two of the altars are inscribed in Latin. Two
were inscribed in Greek (RIB 461, 808) and a third has the same dedication given in
both Greek and Latin (RIB 1072). The fact that doctors, particularly those who wrote
in Greek, were worshipping the god can be used to dispute the argument for Greek
564

564   Patricia Baker

‘rational’ medicine. It also shows that the worship of Asclepius, along with Hygieia, the
daughter of Asclepius, who is sometimes worshipped with him (RIB 609), was wide-
spread in the Roman world, possibly because people might have placed more faith in
healing deities than medical treatments or because the doctors felt they needed divine
support to do their job properly.
There is also a tomb of a doctor (medicus ordinarius) from the fort at Housesteads
(RIB 1618). Again, we see a link to Gallia Belgica, because he was in the First Cohort of
Tungrians. One of the altars dedicated to Asclepius (RIB 1028) was set up by a doctor
who served in a cohort of Vettonians, who were from Lusitania. The Greek inscription
to Asclepius at Chester (RIB 461) does not mention a unit, although it was likely that the
doctor was associated with the legion based there.

Military Medicine and Hospitals

Since many of the inscriptions were found in military contexts, this provides an
appropriate lead into a discussion on Roman military medicine. Traditionally, the
topic of military medicine was written as if there was a medical unit organized in a
homogenous manner, which included a hierarchy of doctors and healers in each unit
(e.g. Richmond 1952: 195; Penn 1964; Davies 1969b, 1970, 1972, 1989; Wilmanns 1995).
There was a debate about how this system was organized and who received treatment,
but there was little doubt that a uniform system existed in the army (Scarborough
1968; Nutton 1969), even though the basis for these claims was slight. In order to
see if there was evidence to support the idea of a uniform system of treatment in the
Roman army, a study was undertaken that examined inscriptions, medical tools, and
structures thought to be hospitals on the Rhine, Danube, and British frontiers (Baker
2004a). It was found that there was not enough evidence to support the underlying
supposition that there was a homogenous medical unit. Through the examination, a
pattern developed in the extant material that showed it was auxiliary units from Gaul
and Spain that had more evidence for Roman-​style medical treatment than units from
elsewhere. In contrast, units from Britain, for example, stationed in other provinces
had little evidence in terms or instruments or inscriptions for Roman-​style medical
treatment.
For the military units stationed in Britain, 121 medical objects were found in 12 differ-
ent fortifications: the legionary bases at Caerleon, York, Chester, and at nine auxiliary
forts on Hadrian’s Wall. The forts that had objects were all occupied by units originally
raised in Gaul and Spain, with the exception of Halton Chesters. The forts that had med-
ical implements and were occupied by units originally from Gaul were at South Shields,
Wallsend, Housesteads, Birdoswald, and Carlisle; those raised in Spain were based at
Chesters and Corbridge. All of the medical tools were common Roman-​style objects
(Baker 2004a: 73–​81).
 Medicine   565

The question that arose was why was there evidence for Roman-​style medi-
cal treatment only in certain units? Based on the available material, it is likely that
medicine in the Roman army was varied and depended on a number of aspects: the
culture of the unit, military virtues, and even personal choice (Baker 2002b,
2004a, 2009).
A re-​evaluation of the identification of hospitals was also made in this study (Baker
2004a: 83–​114). Something called a valetudinarium definitely existed in the Roman
world, which is evidenced by inscriptions and scant excerpts in ancient literature.
However, the inscriptions, literature, and archaeological remains of structures iden-
tified as hospitals in terms of their location, layout, and artefacts indicate that the
buildings might have served other or multiple functions or may not even have been
hospitals. The early descriptions of Roman hospitals made by archaeologists were
based on modern perceptions of the design and function of hospitals that are seem-
ingly inconsistent with the function of hospitals prior to the early modern period
(Baker 2002b, 2004a).
There are no monumental inscriptions from Roman Britain that mention either the
word valetudinarium or the military position of the optio valetudinarii (someone in
charge of the hospital). The word valetudinarium derives from the Latin adjective vale-
tudinus/​a/​um meaning sick. In its substantive masculine or feminine form, the adjective
is translated as a noun meaning ‘invalid’ and in its substantive neuter form it translates
to ‘a sick room, infirmary or hospital’ (Lewis and Short 1980: 1955).
The references to these buildings in the ancient literature provide little detail of
the structures, their layout, their organization, or the treatments available in them.
According to Hygenius’ description of marching camps (Liber de Munitionibus
Castrorum 4), the valetudinarium was to be constructed beyond the praetorium and 70
Roman feet away from the veterinarium and the fabrica—​which was a quiet location. If
the marching camp design described is compared to permanent fortifications, the so-​
called hospital buildings that have been identified were not always located in the place
mentioned by Hygenius.
Archaeologically, the first Roman military valetudinarium identified was that at the
legionary fortress of Neuss, located on the lower Rhine (Koenen 1904: 180–​182). The
courtyard-​style building had small rooms divided by hallways and contained a room
with ten probes. Four scalpels were found in other rooms of the structure. Eighty-​one
other instruments were found in different structures of the fortress, some of which were
courtyard buildings. The courtyard plan was comparable to German civilian and mili-
tary hospitals in use at the time this structure was excavated and identified. Once the
structure’s classification as a hospital was made, buildings of a similar design found in
other fortifications were labelled as hospitals—​even when no medical tools were found
in them.
In Britain ‘hospitals’ have been recognized at Benwell (Simpson and Richmond 1941:
22), Caerleon (Murray-​Threipland 1969: 86–​123; Boon 1972: 75–​77), Chester (Frere 1983:
297–29​8; Mason 2000: 410), Fendoch (Richmond and McIntyre 1938–​9: 132–13​4), Hod
566

566   Patricia Baker

Hill (Richmond 1968), Housesteads (Bosanquet 1904; Wilson 1972: 306–30​8; Charlesworth
1976; Crow 1995: 50–​1; Rushworth 2009), Inchtuthil (Pitts and St Joseph 1985: 91–​101), and
Wallsend (Frere 1983: 289; 1984: 277). The structure originally identified as a hospital at
Corbridge has since been argued to be a workshop (Richmond and McIntyre 1938–​9: 132–​
134; Daniels 1969: 97–​101; Davies 1969–​70: 177; Gillam and Tate 1971: 8; Watermann 1980:
30).
Comparisons of the so-​called hospitals in Britain have shown that their designs
were similar but they vary in their sizes and structural details. Many were only par-
tially excavated, so some of the plans are based on comparisons with complete court-
yard buildings rather than the actual remains—​particularly at Benwell, Caerleon,
Chester, Fendoch, and Hod Hill. Artefacts and hearths found in some of the ‘hospi-
tals’ suggest the buildings, or parts of them at least, were used as workshops. Hearths
found in the ‘hospital’ at Housesteads showed evidence for intense burning, suggest-
ing some form of metalworking (Crow 1995: 50–​51). Inchtuthil had a large room that
was argued to have been an operating theatre, based on the fact that one of the small
rooms next to it had a number of hearths thought to have been used to sterilize the
instruments; yet, the Romans had no concept of sterilization. Moreover, there are no
descriptions of operating theatres in ancient literature. We know only that doctors
need a place with a good source of light so that they could see the part of the body they
were treating (e.g. Celsus, de Med. 7.7. 14C). Some ‘hospitals’ seem to have been used
as places for storage, as large numbers of amphora fragments were found in Inchtuthil
(Pitts and St Joseph 1985: 91–​101) and at Caerleon (Murray-​Threipland 1969: 86–​123;
Boon 1972: 75–​7 7).
Drains, latrines, and evidence for water supplies were found in some of the structures,
but these are only in a limited few and do not evince medical treatment alone. A drain
was uncovered at Caerleon and tanks were found in the courtyard of the structure that
may have been used to collect rain from the roofs of the building (Murray-​Threipland
1969: 86–​123). It is believed that the rainwater would have been used for medical pur-
poses, but, since many Roman houses had impluvia to catch rainwater for other pur-
poses, it is difficult to justify the argument that the water was medicinal alone. At
Wallsend, a latrine was found in the southern end of the structure (Frere 1984: 277) and
one was discovered in the lower south-​west corner of the building in Housesteads
(Bosanquet 1904: 239).
Most so-​called hospitals in Britain have no medical artefacts in them. A compara-
tive examination of the distribution of medical tools from fortifications that yielded
these artefacts showed that medical objects were found in numerous structures,
including baths, barracks, and headquarters buildings, indicating that treatments
were available in many areas not one. The baths at Caerleon have evidence indicating
the possibility that dentistry was performed in them. Some milk teeth were found in
the drains, which showed indentations that might have been caused by dental for-
ceps (Zienkiewicz 1986: 223). These also indicate that children were receiving medical
treatment in the military bathhouse. Hence, the evidence from the military shows the
 Medicine   567

possibility for a variety of practices and places for treatment, rather than homogenous
medical units.

Skeletal and Environmental Remains

It should not go without mention that there has also been a growing interest in taking
a bioarchaeological approach, which compares skeletal remains with the environment
in which the person might have lived to determine health and lifestyles in the past
(Roberts and Cox 2003: 107–​163). The condition of bones is a good indicator of health,
because they can show both general and specific stress indicators. General stress indi-
cators show that someone might have suffered a childhood disease or malnutrition
when their bones were developing, while specific stress indicators can show signs of
trauma or ailments.
Redfern (2003) found that, when comparing the condition of bones with the envi-
ronment, women and children living in Roman Britain suffered more from disease in
comparison to men. She argued that urban lifestyles could be grim, with evidence for
unhygienic living conditions that included head-​lice, poor diets, contaminated drink-
ing water, and animal/​human waste in the yards of houses, which would have affected
people’s health. On a wider level, Redfern has shown that the growth of urbanism after
Roman occupation might have led to poorer health in certain cases, rather than to the
traditional view of a healthier Romanized population.

Conclusion

In the introduction to this chapter a few questions were raised about the transferral
of Roman medicine onto the provinces, because it was traditionally assumed that the
Romans brought with them better medical practices and a healthier living environment.
Lifestyles in Britain changed with Roman occupation, but, as indicated by Redfern, not
necessarily for the better in terms of a healthier lifestyle. Moreover, in relation to the
practice of medicine, the idea that the indigenous populations adopted Roman medical
practices cannot be supported.
Ultimately, the evidence shows a mingling of traditions most likely from Gaul, Britain,
and Rome. Although there were probable regional differences in Britain, in terms of
treatment and understandings of the body, the evidence at the moment is too limited
to make a strong case for what the practices might have been. Therefore, the use of the
terms ‘Romano-​British medicine’ or ‘Roman-​period medicine’ will suffice in discussing
medical practices in the province and help to articulate the variant medical traditions
across the empire.
568

568   Patricia Baker

Abbreviations
CIL 
Corpus Inscriptorum Latinorum. Consilio et Ductoritate Academie Litterarum
Regiae Borussical Edition. Berlin: Academieder Wissenschaften, 1862–​to date.
RIB 
R. G.  Collingwood and R.  P. Wright, The Roman Inscriptions of Britain.
Stroud: Alan Sutton, 1992–​to date.

References
Primary Sources
Caesar [C. Iulii Caesaris], Commentariorum Libri VI De Bello Gallico, ed. R. du Pontèt.
Oxford: Clarendon Press, 1900–​1908.
Celsus, De Medicina, trans. W. G. Spencer. Loeb Classical Library. Cambridge, MA and
London: Harvard University Press and William Heinemann Ltd, 1971.
Hyginus Gromaticus, Hygini Gromatici:  Liber de Munitionibus Castrorum, ed. W. Gemoll.
Leipzig: B. G. Teubner, 1897.
Pliny the Elder, Naturalis Historiae, ed. L. Ianus. Leipzig: B. G. Teubner, 1880.
Soranus, Gynecology, trans. O. Temkin. Baltimore: Johns Hopkins University Press, 1956.

Secondary Sources
Allason-​Jones, L. (1993). ‘Health Care in the Roman North’, in D. Gardner-​Medwin, A.
Hargreaves, and E. Lazenby (eds), Medicine in Northumbria:  Essays on the History of
Medicine in the North-​East of England. Newcastle upon Tyne: The Pybus Society for the
History and Bibliography of Medicine, 24–​39.
Allason-​Jones, L. (1999). ‘Health Care in the Roman North’, Britannia, 31: 131–​146.
Allason-​Jones, L., and Miket, R. (1984). The Catalogue of Small Finds from South Shields Roman
Fort. Newcastle upon Tyne: The Society of Antiquaries.
Baker, P. (2001). ‘Medicine, Culture and Military Identity’, in G. Davies, A. Gardner,
and K. Lockyear (eds), Theoretical Roman Archaeology Conference Proceedings 2000.
Oxford: Oxbow Press, 51–​70.
Baker, P. (2002a). ‘Diagnosing Some Ills: The History and Archaeology of Roman Medicine’, in
P. Baker and G. Carr (eds), Practitioners, Practices and Patients New Approaches to Medical
Archaeology and Anthropology: Conference Proceedings. Oxford: Oxbow Press, 16–​29.
Baker, P. (2002b). ‘The Roman Military valetudinaria:  Fact or Fiction?’ in R. Arnott (ed.),
The Archaeology of Medicine Proceedings of the Theoretical Archaeology Group 1998. BAR
International Series 1046. Oxford: Archaeopress, 69–​80.
Baker, P. (2004a). Medical Care for the Roman Army on the Rhine, Danube and British Frontiers from
the First through Third Centuries ad. BAR International Series 1286. Oxford: Hadrian Books.
Baker, P. (2004b). ‘Roman Medical Instruments:  Archaeological Interpretations of their
Possible “Non-​Functional” Uses’, Social History of Medicine, 17: 3–​21.
Baker, P. (2009). ‘Medicine, Death and Military Virtues’, in F. Marco Simón, F. Pino Polo, and J.
Remesal Rodríguez (eds), Formae Mortis: El Tránsito de la Vida a la Muerte en las Sociedades
Antiguas. Barcelona: Universidad de Barcelona, 25–​37.
 Medicine   569

Baker, P. (2011). ‘Collyrium Stamps:  An Indicator of Regional Medical Practices in Roman

Gaul’, European Journal of Archaeology, 14/​1–​2: 158–​189.
Bliquez, L. (1981). ‘An Unidentified Roman Surgical Instrument in Bingen’, Journal of History of
Medicine, 36/​2: 219–​220.
Bliquez, L. (1988). Roman Surgical Instruments and Minor Objects found in the University of
Mississippi. Göteborg: Paul Åstroms Fõrlag.
Boon, G. C. (1972). Isca:  The Roman Legionary Fortress at Caerleon, Monmouthshire.
Cardiff: National Museum of Wales.
Boon, G. C. (1983). ‘Potters, Oculists and Eye Troubles’, Britannia, 14: 1–​12.
Bosanquet, R. C. (1904). ‘Excavations on the Line of the Roman Wall in Northumberland’,
Archaeologia Aeliana, 25 [Series 3], 193–​300.
Bowman, A. K., Thomas, J. D., and Tomlin, R. S. O. (2011). ‘The Vindolanda Writing Tablets
(Tabulae Vindolandenses IV, Part 2)’, Britannia, 42: 113–​144.
Breitwieser, R. (1998). Medizin im römischen Österreich. Linz:  Linzer Archäologische
Forschungen.
Charlesworth, D. (1976). ‘The Hospital: Housesteads’, Archaeologica Aeliana [Series 5], 4: 17–​30.
Crow, J. (1995). Housesteads. Bath: B. T. Batsford/​English Heritage.
Crummy, P. (2002). ‘A Preliminary Account of the Doctor’s Grave at Stanway, Colchester,
England’, in P. Baker and G. Carr (eds), Practitioners, Practices and Patients: New Approaches
to Medical Archaeology and Anthropology. Oxford: Oxbow Books, 47–​57.
Crummy, N. (2007). ‘The Identities of the “Doctor” and the “Warrior”’, in P. Crummy, S. Benfield,
N. Crummy, V. Rigby, and D. Shimmin (eds), Stanway: An Élite Burial Site at Camulodunum.
Britannia Monograph Series 24. London: Society for the Promotion of Roman Studies, 444–​447.
Crummy, P., Benfield, S., Crummy, N., Rigby, V., and Shimmin, D. (2007) (eds), Stanway: An
Élite Burial Site at Camulodunum. Britannia Monograph Series 24. London: Society for the
Promotion of Roman Studies.
Daniels, C. (1969). ‘Corbridge’, Current Archaeology, 15: 97–​101.
Dannell, G. B. (2007). ‘Decorated Samian’, in P. Crummy, S. Benfield, N. Crummy, V. Rigby, and
D. Shimmin (eds), Stanway: An Élite Burial Site at Camulodunum. Britannia Monograph
Series 24. London: Society for the Promotion of Roman Studies, 213–​215.
Davies, R. (1969a). ‘Joining the Roman Army’, Bonner Jahrbücher, 169: 208–​232.
Davies, R. (1969b). ‘The medici of the Roman Armed Forces’, Epigraphische Studien, 8: 83–​99.
Davies, R. W. (1969–​70). ‘A Note on the Hoard of Roman Equipment Buried at Corbridge’,
Durham University Journal, 62: 177.
Davies, R. (1970). ‘Some Roman Medicine’, Medical History, 14/​1: 101–​106.
Davies, R. (1972). ‘Some More Military medici’, Epigraphische Studien, 9: 3–​11.
Davies, R. (1989) [original 1970]. ‘The Roman Military Medical Service’, in D. Breeze and V.
Maxfield (eds), Service in the Roman Army. Edinburgh: Edinburgh University Press, 209–​236.
Draycott, J. (2011). ‘Approaches to Healing in Roman Egypt’, Ph.D.  thesis, University of
Nottingham.
Fee, E., and Brown, T. M. (2004). ‘Using Medical History to Shape a Profession’, in F. Huisman
and J. H. Warner (eds), Locating Medical History: The Stories and their Meanings. Baltimore:
Johns Hopkins University Press, 139–​164.
Fox, A. (1940). ‘The Legionary Fortress at Caerleon in Monmouthshire: Excavations at Myrtle
Cottage Orchard, 1939’, Archaeologia Cambriensis, 95: 101–​152.
Francis, A. (1926). ‘On a Romano-​British Castration Clamp Used in the Rites of Cybele’,
Proceedings of the Royal Society of Medicine (Section Medical History), 19: 95–​110.
570

570   Patricia Baker

Frere, S. S. (1983). ‘Sites Explored 1982’, Britannia, 14: 279–​335.

Frere, S. S. (1984). ‘Sites Explored 1983’, Britannia, 15: 265–​332.
Gillam, J. P., and Tate, J. (1971). ‘The Investigation of the “Commander’s House” Area, on Site
XI, Corbridge, 1958 to 1970: The Structures’, Archaeologia Aeliana [Series 4], 49: 1–​28.
Gilson, A. (1978). ‘A Doctor at Housesteads’, Archaeologia Aeliana, 6 [Series 5], 162–​165.
Gilson, A. (1981). ‘A Group of Roman Surgical and Medical Instruments from Corbridge’,
Saalburg Jahrbuch, 37: 5–​9.
Gilson, A. (1982). ‘A Roman Iron Cautery from Verulamium’, Britannia, 13: 303–​304.
Gilson, A. (1983). ‘A Group of Roman Surgical and Medical Instruments from Crammond,
Scotland’, Medizin Historisches Journal, 18/​4: 384–​393.
Jackson, R. (1988). Doctors and Diseases in the Roman Empire. Norman, OK, and
London: University of Oklahoma Press.
Jackson, R. (1990). ‘Roman Doctors and their Instruments: Recent Research into Ancient
Practice’, Journal of Roman Archaeology, 3: 5–​27.
Jackson, R. (1993). ‘Roman Medicine: Practitioners and their Practices’, Aufstieg und Niedergang
der Romischen Welt II, 37/​1: 79–​100.
Jackson, R (1994a). ‘Styphylagra, Staphylocaustes, Uvulectomy and Haemorrhoidectomy: The
Roman Instruments and Operations’, in A. Krug (ed.), From Epidauros to
Salerno: Symposium Held at the European University Centre for Cultural Heritage, Ravello,
April 1990. Rixensart: Pact Belgium, 167–​185.
Jackson, R. (1994b). ‘The Surgical Instruments, Appliances and Equipment in Celsus’
De Medicina’, in G. Sabbah and J. Mundry (eds), La Médecine de Celse. Saint-​
Étienne: Publications de l’Université Saint-​Étienne, 167–​209.
Jackson, R. (1995). ‘The Composition of Roman Medical instrumentaria as an Indicator of
Medical Practice: A Provisional Assessment’, in P. J. van der Eijk, H. F. J, Horstmannshoff,
and P. H. Schrijvers (eds), Ancient Medicine in its Socio-​Cultural Context. Amsterdam and
Atlanta: Rodopi Press, 189–​208.
Jackson, R. (1996). ‘Eye Medicine in the Roman Empire’, Aufstieg und Niedergang der romischen
Welt II, 37/​3: 2228–​2251.
Jackson, R. (1997). ‘An Ancient British Medical Kit from Stanway, Essex’, Lancet, 350/​
9089: 1471–​1473.
Jackson, R. (2002). ‘Roman Surgery: The Evidence of the Instruments’, in R. Arnott (ed.), The
Archaeology of Medicine. BAR International Series 1046. Oxford: Archaeopress, 87–​94.
Jackson, R. (2007). ‘The Surgical Instruments’, in P. Crummy, S. Benfield, N. Crummy, V.
Rigby, and D. Shimmin (eds), Stanway:  An Élite Burial Site at Camulodunum. Britannia
Monograph Series 24. London: Society for the Promotion of Roman Studies, 236–​252.
Jackson, R. (2011). ‘Medicine and Hygiene’, in L. Allason-​Jones (ed.), Artefacts in Roman
Britain: Their Purpose and Use. Cambridge: Cambridge University Press, 243–​268.
Jackson, R., and Leahy, K. (1989). ‘A Roman Surgical Forceps from near Littleborough and a
Note of the Type’, Britannia, 20: 211–​214.
Kleinman, A. (1980). Patients and Healers in the Context of Culture. Berkeley and Los
Angeles: University of California Press.
Koenen, C. (1904). ‘Beschreibung von Novaesium’, Bonner Jahrbucher, 111–​112: 97–​242.
Künzl, E. (1983). Medizinische Instrumente aus Sepulkralfunden der römischen Kaiserzeit.
Cologne: Rheinland Verlag GmbH.
Künzl, E. (1996). ‘Forschungsbericht zu den antiken medizinischen Instrumenten’, Aufstieg
und Niedergang der romischen Welt II, 37/​3: 2433–​2639.
 Medicine   571

Künzl, E. (2002). Medizin in der Antike:  Aus einer Welt ohne Narkose und Aspirin.
Stuttgart: Konrad Theiss Verlag GmBh.
Lewis, C. T., and Short, C. (1980). A Latin Dictionary. Oxford: Clarendon Press.
Lloyd, G. E. R. (1983). Science Folklore and Ideology. Cambridge: Cambridge University Press.
Longrigg, J. (1993). Greek Rational Medicine: Philosophy and Medicine from Alcmaeon to the
Alexandrians. London and New York: Routledge.
McCarthy, M. R. (1990). The Roman, Anglican and Medieval Site at Blackfriars Street. Carlisle:
Cumberland and Westmorland Antiquarian and Archaeological Society Research Series 4.
Mason, D. (2000). Chester: The Elliptical Building’, Current Archaeology, 167: 404–​416.
Mays, S. A. (2007). ‘The Cremated Remains’, in P. Crummy, S. Benfield, N. Crummy, V. Rigby,
and D. Shimmin (eds), Stanway: An Élite Burial Site at Camulodunum. Britannia Monograph
Series 24. London: Society for the Promotion of Roman Studies, 377–​381.
Milne, J. (1907). Surgical Instruments in Greek and Roman Times. Oxford: Clarendon Press.
Murray-​Threipland, L. (1969). ‘The Hall: Caerleon Excavations on the Site of the Legionary
Hospital’, Archaeologia Cambriensis, 118: 86–​123.
Nash-​ Williams, V. E. (1932). ‘The Roman Legionary Fortress at Caerleon in
Monmouthshire: Report on the Excavations Carried out at Prysg Field, 1927–​9. Part II: The
Finds’, Archaeologia Cambriensis, 87: 48–​104.
Nijhuis, K. (1995). ‘Greek Doctors and Roman Patients: A Medical Anthropological Approach’,
in P. J. van der Eijk, H. F. J. Horstmannshoff, and P. H. Schrijvers (eds), Ancient Medicine in
its Socio-​Cultural Context. Amsterdam and Atlanta: Rodopi Press, 49–​66.
Nutton, V. (1969). ‘Medicine and the Roman Army: A Further Reconsideration’, Medical
History, 13: 260–​270.
Nutton, V. (1970). ‘A Greek Doctor at Chester’, Journal of Chester Archaeological Society, 55: 7–​13.
Nutton, V. (1972). ‘Roman Oculists’, Epigraphica, 34: 16–​29.
Nutton, V. (1995). ‘The Medical Meeting Place’, in P. J. van der Eijk, H. F. J. Horstmannshoff,
and P. H. Schrijvers (eds), Ancient Medicine in its Socio-​cultural Context. Amsterdam and
Atlanta: Rodopi Press, 3–​26.
Nutton, V. (2004). ‘Ancient Medicine from Berlin to Baltimore’, in F. Huisman and J. H. Warner
(eds), Locating Medical History: The Stories and their Meanings. Baltimore: Johns Hopkins
University Press, 115–​138.
Penn, R. G. (1964). ‘Medical Service in the Roman Army’, Journal of the Royal Army Medical
Corps, 110: 253–​258.
Pitts, L., and St Joseph, J. K. (1985). Inchtuthil:  The Roman Legionary Fortress. Britannia
Monograph Series 6. London: Society for the Promotion of Roman Studies.
Redfern, R. (2003). ‘Sex and the City: A Biocultural Investigation into Female Health in Roman
Britain’, in G. Carr, E. Swift, and J. Weekes (eds), TRAC 2002:  Proceedings of the Twelfth
Annual Theoretical Roman Archaeology Conference, Canterbury 2002. Oxford:  Oxbow
Books, 146–​170.
Richmond, I. A. (1952). ‘The Roman Army Medical Service’, University Medical Gazette
(June), 2–​7.
Richmond, I. A. (1968). Hod Hill, Volume 2: Excavations Carried out between 1951 and 1958.
London: The Trustees of the British Museum.
Richmond, I. A., and McIntyre, J. (1938–​9). ‘The Agricolan Fort at Fendoch’, Proceedings of the
Society of Antiquaries of Scotland, 73: 110–​154.
Roberts, C., and Cox, M. (2003). Health and Disease in Britain: From Prehistory to the Present
Day. Stroud: Sutton Publishing.
572

572   Patricia Baker

Rushworth, A. (2009) (ed.). Housesteads Roman Fort—​The Grandest Station: Excavation and
Survey at Housesteads, 1954–​95, by Charles Daniels, John Gillam, James Crow and Others.
Volume 1: Structural Report and Discussion. Volume 2: The Material Assemblages. Swindon:
English Heritage.
Scarborough, J. (1968). ‘Roman Medicine and the Legions: A Reconsideration’, Medical History,
12: 254–​261.
Scarborough. J. (1976). Roman Medicine. Ithaca, NY: Cornell University Press.
Schädler, U. (2007). ‘The Doctor’s Game: New Light on the History of Ancient Board Games’,
in P. Crummy, S. Benfield, N. Crummy, V. Rigby, and D. Shimmin (eds), Stanway: An Élite
Burial Site at Camulodunum. Britannia Monograph Series 24. London:  Society for the
Promotion of Roman Studies, 359–​375.
Simpson, J. Y. (1856). Was the Roman Army Provided with Medical Officers?
Edinburgh: Sutherland and Knox.
Simpson, F. G., and Richmond, I. A. (1941). ‘The Roman Fort on Hadrian’s Wall at Benwell’,
Archaeologia Aeliana [Series 4], 19: 1–​43.
Smith, W. D. (1979). The Hippocratic Tradition. Ithaca, NY: Cornell University Press.
Thomas, P. H. (1963). ‘Graeco-​Roman Medical and Surgical Instruments with Special Reference
to Wales and the Borders’, Journal of the Royal College of General Practitioners, 6: 495–​502.
van der Eijk, P. (2005). Medicine and Philosophy in Classical Antiquity: Doctors and Philosophers
on Nature, Soul, Health and Disease. Cambridge: Cambridge University Press.
Voinot, J. (1984). ‘Inventaire des cachets d’oculistes gallo-​romains’, Conférences Lyonnaises
d’Optalmologie, 150: 1–​578.
Voinot, J. (1999). Les Cachets à Collyres dans le Monde Romain. Instrumentum 7. Montagnac:
Monique Mergoil.
Watermann, R. A. (1980). Mensch und Medizin Zwischen Macht und Militär der römischen
Kaiserzeit. Frankfurt am Main: Gerhard Michter Verlag.
Wheeler, R. E.  M. (1928). ‘The Roman Amphitheatre at Caerleon, Monmouthshire’,
Archaeologia, 78: 111–​128.
Wilmanns, J. C. (1995). Der Sanitätsdienst im römischen Reich. Medizin der Antike 2.
Hildesheim, Zürich and New York: Olms Weidmann.
Wilmott, C. T. (1997). ‘A New Building Type in an Auxiliary Fort: The Biroswald Basilica and
its Discovery’, in W. Groenman-​van Waateringe, B. L. van Beek, W. J. H. Willems, and S. L.
Wynia (eds), Roman Frontier Studies 1995. Oxbow Monograph 91. Oxford: Oxbow Books,
581–​586.
Wilson, D. R. (1972). ‘Roman Britain in 1971: Sites Explored’, Britannia 4, 298–​351.
Wiltshire, P. E. J. (2007). ‘Palynological Analysis of the Organic Material Lodged in the Spout
of the Strainer Bowl’, in P. Crummy, S. Benfield, N. Crummy, V. Rigby, and D. Shimmin
(eds), Stanway:  An Élite Burial Site at Camulodunum. Britannia Monograph Series 24.
London: Society for the Promotion of Roman Studies, 394–​397.
Zienkiewicz, J. D. (1986). The Legionary Fortress Baths at Caerleon II: The Finds. Cardiff: Cadw/​
Welsh Historical Monuments.
 Chapter 28

S o cioling u i st i c s

Alex Mullen

Introduction

Roman Britain has not traditionally been on the agenda for discussing ancient sociolin-
guistics. Language rarely makes an appearance in the standard handbooks, general his-
tories, or archaeological syntheses, even though everyone knows that the arrival of the
Romans made a significant impact linguistically through the introduction of the Roman
Reichsprache and its cultural handmaiden, Greek. The linguistic content of the result-
ant Latin (and, to a much smaller extent, Greek) epigraphy is generally deemed to be
standardized, formulaic, and not really worthy of comment beyond specialist works on
language. Furthermore, Celticists studying the languages of Britain have largely skipped
straight to the Middle Ages, since virtually no pre-​Roman or Roman period Celtic epig-
raphy exists, in contrast to the continent, where Celtic languages are attested in Gaul (in
Greek and Roman script, see RIG I, II.1, II.2, III, IV; Woolf 1994; Lambert 2003; Mullen
2013a: 95–​121); Italy (in the alphabet of Lugano, a form of the Etruscan alphabet, see RIG
II.1; Lejeune 1971) and Spain (in Iberian script, see MLH IV; Simkin 2012). However,
while it is fair to admit that direct evidence is relatively scarce, numerous indirect routes
to reconstructing linguistic contours are available, and, used with care, linguistic infor-
mation can prove surprisingly fertile in understanding the nature of identities and cul-
tural interactions more broadly.
In this chapter I briefly confront some of the key questions that we might ask of the
evidence and present a view of the state of play. How does the linguistic situation evolve
from the Iron Age, through Roman occupation, and thereafter? How widespread was
bilingualism? Who spoke which language, to whom, and when? How similar or dissimi-
lar linguistically were different parts of Roman Britain? Are regional or social dialectal
variations reconstructible? Where is our British Romance language? But we should first
ask a more basic question: why should anyone other than sociolinguists care?
Simply put, languages express identity. As a result, language use should be rele-
vant to everyone interested in understanding Roman Britain. As soon as we speak, we
574

574   Alex Mullen

communicate, not only the intended message but also numerous other subtle, intended or
unintended, clues about our multi-​layered identities. It is often assumed that the uptake
of Latin in the provinces indicates a move towards being more Roman. ‘Latinization’,
indeed, is used almost interchangeably with ‘Romanization’. In a similar vein, evidence
of resistance to Latinization and the persistence of local languages have been taken as a
sign of rejection of both Roman culture and integration into the empire. Consider, for
example, the entry for Britannia in the second edition of the Oxford Classical Dictionary
(Hammond and Scullard 1970: 181): ‘The class distinction, however, manifest in the con-
trast between the Celtic spoken by the common man and the stilted Latin of his superiors,
had the seeds of trouble.’ Employing concepts such as language and culture interchange-
ably brings a tacit appreciation that languages might be very closely identified with
cultures themselves: Latin is Roman culture, Celtic is indigenous culture. Herein lies a
grain of truth that needs to be explored, but the full picture will be much more intricate.
Languages always express identity, but we know that the relationships between language
and culture and language and ethnicity are not so straightforward.

The Evidence

The evidence at our disposal proves surprisingly varied. The most obvious, main source
of linguistic material derives from Roman epigraphy. Taken in its widest sense, this
encompasses the large numbers of lapidary inscriptions (RIB I, III), right through to
the writing tablets from some twenty sites including Vindolanda, Carlisle, and London
(Pearce 2004: 47–​48; for the texts, see RIB II.4; Carlisle; Tab. Vindol. I, II, III, IV.1, IV.2;
Tomlin 2003). ‘Curse tablets’, which in Britain typically ask for justice after a theft, have
been turning up with regularity, and well over thirty sites now offer examples, though
Bath, and to a lesser extent Uley, lay claim to the largest share. Also featuring in British
epigraphy are words, often simply names, found on a range of types of ceramic, metal,
glass, gems, plaster, mosaic (Ling 2007), leather (Rhodes 1987), stone, bone, and wood,
which are subsumed under the grand title instrumenta domestica and have been pub-
lished in the eight fascicules of RIB II. Inscriptions on coins constitute a closely related
set of direct evidence, but, as the preserve of a distinct discipline, are published sepa-
rately (for a discussion of coins, see Walton and Moorhead, this volume). Other non-​
epigraphic evidence consists of occasional, and sometimes opaque, discussions of the
linguistic situation in Britain found in authors such as Tacitus, the Roman grammarians,
and, later, Bede (e.g. Historia Ecclesiastica 1.1). Literary texts by Roman authors of British
origin might also have provided us with dialectal words used without comment, but,
again, their absence—​or perhaps our inability to identify them—​is striking. Additional
important material includes borrowings between languages: both Celtic borrowings in
Latin and Latin borrowings that can be uncovered in the subsequently attested Celtic
languages. A comprehensive discussion of all this material is not possible here; rather
key material will be presented to illustrate salient points.
 Sociolinguistics   575

Language in Pre-​Roman Britain

The point of departure for assessing continuity and change in the linguistic composition
of Roman Britain must be to establish the language(s) of pre-​Roman Britain. Scholars
generally concur that the main language of the island was Celtic—​one of the distinctive
branches of the Indo-​European family of languages—​and yet are equally convinced that
we have no epigraphic remains. So how can we be so sure?
The broader linguistic and historical context of the British Isles indicates that Celtic
languages were spoken over a large part of the area in the Iron Age.1 The earliest written
evidence from or about Britain in Latin and Greek repeatedly presents us with personal,
deity, and place names of Celtic origin and, very occasionally, non-​onomastic Celtic fea-
tures. Tacitus states that the local languages of northern Gaul and Britain hardly differed
(sermo haud multum diversus, Agricola 11) and we have direct evidence for the linguistic
composition of the former, the Celtic language, Gaulish. In the absence of any evidence
for a significant linguistic shift in the late pre-​Roman Iron Age, it is hard to argue against
the view that Celtic was widely spoken in Iron Age Britain. For how many generations,
by how many, in which communities and alongside which other languages cannot be
firmly ascertained.
A major consideration in defining the linguistic contours of pre-​Roman Britain will
be the advent of Latin. Britannia, technically speaking, comes to life in ad 43, but this is
by no means the first contact with Romans, nor with Latin. Caesar had investigated the
island in the mid-​first century bc and cross-​channel diplomatic and economic links had
long been established (Gosden 2004: 104–​110; Cunliffe 2005: 446–​484) so, by the first
century bc, a growing percentage of the communications in trade and other matters
must have been undertaken in Latin, at least in southerly areas. Unfortunately, the vast
majority of these early communications have not been transmitted across the centuries.
This is no doubt partly a result of the nature of some of the transactions, but perhaps also
because, even when the communications may have been suitable for permanent record,
there was no local tradition of writing.
Virtually no pre-​conquest inscriptional evidence exists from Britain, though finds of
writing equipment from contexts dated to before ad 43 indicate that writing was taking
place, though probably to a limited extent (see Hanson and Conolly 2002: 156, 159, for
the early styli). The inscribed material itself is restricted to a handful of graffiti on pots
that are mostly imported (Partridge 1982; Hanson and Conolly 2002: 156), and we can-
not exclude the possibility that they may have been incised on the continent or by visi-
tors to Britain. Apart from this meagre evidence, we are reliant on writing on coins (see
Nash 1987: 118–​142; Mays 1992; Williams 2007), some of which Williams (2001: 10, 14, 7)
cogently suggests may betray knowledge of other non-​numismatic inscriptional types
and which demonstrate knowledge of ‘non-​Roman’, local writing in Gaul.2 Several pre-
conquest British coin legends show Celtic names, some with local rather than Latinate
endings (e.g. VOLISIOS, CARTIVELLAVNOS), and also, possibly, the Celtic word for
576

576   Alex Mullen

‘king’ in the accusative RICON—​as in TASCIO RICON—​where RIX might be equivalent

to Latin REX (de Bernardo Stempel 1991; Creighton 2000: 168–​172). If we do interpret
RICON as a Celtic noun, it would demonstrate that British communities were willing to
write British, even though in a restricted and specific usage. Intriguingly, writing of the
local language never seems to extend beyond British coin legends, despite the relatively
widespread writing of Celtic on the continent. Although the strength of the oral tradi-
tion in the country may have been a factor against the uptake of literacy, the most plau-
sible explanation for this restricted usage must be that literacy arrives later in Britannia
than elsewhere and that, by the time it does, Latin has proved its worth as the written
language.

Latinization and the Extent

of Bilingualism

In the early twentieth century the orthodoxy stated that no written Celtic was to be
found in Roman Britain because Latin had become ubiquitous. Too much was made of
Tacitus’ comment that Agricola ‘principum filios liberalibus artibus erudire … ut qui
modo linguam Romanam abnuebant, eloquentiam concupiscerent’ (‘educated the sons
of the leading men in the liberal arts … so that those who just lately had been reject-
ing the Roman language now conceived a desire for eloquence’ (Agricola 21)). The Latin
inscriptions were also over-​optimistically interpreted; inscriptions scratched on tiles,
such as Primus fecit x, were described by Haverfield (1923: 32) as showing that literacy
in Latin (and therefore the Latin language) had permeated to workmen. However, these
formulaic phrases do not necessarily imply anything but low-​level literacy, and there
is no reason to assume that the authors were not speaking Celtic, either exclusively or
in tandem with Latin. We must not fall into the trap of viewing the textual record as a
direct reflection of spoken language and must be aware of the distancing effect not only
of formulaic language but also of scribes, stonemasons, and manuals (Pulgram 1950; see
Hope, this volume, for a discussion of inscriptions).
The Celticist Jackson, in his magisterial work Language and History in Early Britain
(1953), argued that the attitude of earlier scholars towards the linguistic situation ‘was
fortified by the then very general ignorance about Celtic linguistics and history, and by
the prejudice rife among some English historians against everything Celtic’ (Jackson
1953: 94). Indeed, anti-​Celtic sentiment and Classical cultural arrogance, conscious
or subconscious, do seem to have been factors in the development of scholarship on
Britannia. Jackson (1953: 105) formulated a more sophisticated approach, concluding
that:

Latin was the language of the governing classes, of civil administration and of the
army, of trade, of the Christian religion, and very largely (but perhaps not entirely)
 Sociolinguistics   577

of the people of the towns. The rural upper classes were bilingual; the peasantry of
the Lowland Zone, who constituted the great bulk of the population, spoke British
and probably knew little Latin; and the language of the Highland Zone (apart from
the army and its native camp-​followers) was to all intents and purposes exclusively
British.

Jackson’s illustration of a variegated linguistic landscape for Britannia resonates with

recent research into ancient multilingualism, which has reached beyond Latin and
Greek literary and other high-​status remains and has used modern linguistic theory to
help interpret the complexity of language contact (Adams 2003; Mullen and James 2012;
Mullen 2013a). One important aspect of this cross-​disciplinary research is an apprecia-
tion of the value of sociolinguistics in understanding the large numbers of interlinking
factors that can determine the nature, extent, and vitality of both societal and individual
bilingualism. There are no hard and fast rules for linguists; rather an appreciation of ten-
dencies built up through numerous detailed case studies. Sociolinguistic models do not
support the Jacksonian view that bilingualism in Roman Britain would have effectively
been restricted to the rural upper class. We would instead expect varying levels of bilin-
gualism dependent on social background, occupation, education, area of birth/​work,
age, gender (a factor missing from Jackson’s account), and so on. Levels of Latin would
have been disparate: from the flawless Latin of highly educated local and immigrant
elites, to muddled scraps of Latin spoken, of necessity and rarely, by isolated commu-
nities. Most of the Latin in Britain would fit somewhere on the continuum in between
and would have displayed regional and social variation (Adams 2007, 2013), particularly
given the lack of formalized, mass education. We would expect to find contact phenom-
ena that occur in situations of bilingualism: the transfer of idioms, morpho-​syntax, pho-
nology, and loanwords. Some of these phenomena would have been restricted to the
language of individuals, specific communities, or regions, though others may have been
more pervasive. In some areas reduced languages, which are created to facilitate com-
munication where there is little education, may have formed.
Can these hypotheses based on models and plausibility find evidential support? That
there must have been widespread bilingualism in Roman Britain is witnessed by the
large number (c. 1,000) of Latin loanwords that can be found in the Insular Celtic lan-
guages: Welsh was so affected that it has been described as ‘akin to the Romance lan-
guages descended from Latin’ (Charles-​Edwards 2013: 76). Many Latin loanwords are
likely to have been borrowed in the Roman period, and, in principle, it should be pos-
sible to try to describe cultural contacts through borrowings. However, their dating is
fraught with difficulties, and we are not certain of the exact linguistic origin of several
forms. Scholarship on Latin loanwords in Brittonic languages (Breton, Cornish, and
Welsh) has a long history (for an overview of scholarship dating back to the Renaissance,
see D. E. Evans 1983: 960–​963), and the need for an up-​to-​date, precise, and well-​docu-
mented study was highlighted by Gratwick in 1982, but the gap still remains to be filled
satisfactorily. Even when we do have a firm idea of etymology and date, the interpreta-
tion can be challenging. Although some borrowings are easy to interpret—​for instance,
578

578   Alex Mullen

Latin scribendum ‘writing’ underlies Welsh ysgrifennu, and Latin grammatica ‘grammar’
gives Welsh grammadeg—​others are much harder given native forms already existed, for
example, the borrowing of Latin pontem ‘bridge’ (Welsh pont), piscis ‘fish’ (Welsh pysg),
and bracchium ‘arm’ (Welsh braich). The large numbers and wide range of lexical bor-
rowings and also the morphological and syntactical features that may be influenced by
Latin3 strongly suggest that a substantial group of bilinguals has to be assumed as well as
prolonged and intimate contact (Mac Cana 1976: 195).
Bilingualism between Latin and Celtic entailed a two-​way traffic of loanwords. Latin
itself contains around 150 Celtic loanwords, which might be informative about lin-
guistic and cultural contacts (Schmidt 1967; Wild 1976; Porzio Gernia 1981; Lambert
2003: 204–​207). Loans concerning wheeled vehicles are common (for example, esse-
dum ‘war-​chariot’; petorritum ‘four-​wheeled wagon’), as are clothing terms (e.g. birrus
‘hooded cloak’; sagus/​m ‘tunic’ (see Tab. Vindol. II. 192)), but these were almost cer-
tainly borrowed through contact between Latin and Celtic speakers on the continent
rather than in Britain. The Celtic words attested in the Vindolanda Tablets may already
have been in military Latin created on the continent or borrowed from the speech of
the continental auxiliaries stationed at Vindolanda, rather than through contact with
the local Celtic-​speaking Britons. The linguistic make-​up of the auxiliaries, who were
composed of a mixture of Germanic-​and Celtic-​speaking communities from Batavia
(southern Netherlands) and Tungria (Belgium), makes tracing the origins of Celtic fea-
tures in the texts complex. The tablets do demonstrate links with the world outside the
fort: one writer moans about the Brittunculi (Tab. Vindol. II. 164), and a civilian com-
plains about his treatment at the hands of the soldiers, but it is difficult to assert specifi-
cally local authorship. Indeed, the complaining civilian describes himself as a hominem
trasmarinum (sic) (Tab. Vindol. II. 344; l. 15), and it is often impossible to tell whether
the Celtic features are British or continental. The word souxtum (Tab. Vindol. II. 301)
illustrates this point: originally interpreted by Adams (1996) as a ‘Celticized’ form of
Latin sumptum, we are now confident that it is a Celtic word for a type of vessel, since it
is attested in a second-​century potter’s account from Vayres (Gironde) and occurs later
in Insular Celtic (Early Irish suacht, Scottish suacan, Old Cornish seit) (Lambert 2000,
2004; Adams 2007: 597–​598; Jørgensen 2008). The problem arises in trying to work
out how and when the loanword entered the Latin used at Vindolanda, whether from
continental or local British Celtic (if, in fact, it is a borrowing and not a code-​switch or
interference).
Whatever the precise origins of the Celtic loanwords, contact between military
and civilian populations would naturally be a route for their dissemination through
British Latin (Adams 2007: 581–​582). Indeed, while research into the distribution of
classes of settlement type has, at the highest level of abstraction at least, supported
the traditional division of Britannia into north and west ‘military’ and south and east
‘civil’ zones (e.g. Sargent 2002: 224; Taylor 2007: 109–​118), Jackson’s view (1953: 106)
that the Highland was almost ‘exclusively British’ in language must be questioned.
The linguistic realities will never follow closely such invented boundaries, as Jackson
 Sociolinguistics   579

(1953: 106) himself admits, and the military sites may have constituted influential cen-
tres of Latinization.
Hanson and Conolly (2002: 156) have demonstrated that ‘many lower-​status set-
tlements across the country-​side have produced examples of stili’, and, since these
would presumably have been used to write Latin, this may indicate a more widespread
use of Latin and Latin literacy than usually conceived. Indeed data from the Portable
Antiquities Scheme (PAS) might help us to explore this proposition concerning rural
literacy in more detail. Approximately 100 styli have been identified and assigned the
date range ‘broad period Roman’ on the database. These, along with the whole range
of instrumenta scriptoria, need to be thoroughly assessed, as they allow us to explore
the dynamics of Latinization. This is a tall order: instrumenta scriptoria are not always
easy to identify; we are well aware of the broader problems of the biases in metal detect-
ing, reporting, and analysis that affect the PAS; and, for the distribution of instrumenta
scriptoria to be of value, the PAS data must be supplemented with information from
the Historic Environment Record, published and unpublished excavation reports,
museum and private collections, and so on. While we wait for more detailed research to
be completed, Figure 28.1 provides a rough indication—​caveat lector—​of the distribu-
tion of styli: some possible constraints on metal detecting are shown (Robbins 2014),
and a kernel density plot indicates the finds of styli. The map presents us with a rela-
tively unsurprising distribution that mirrors relatively closely the total distribution
of PAS Roman finds (see Richards et al. 2009: figures 40, 41), though there are some
areas of interest that will require further exploration. For example, Mattingly has stated,
based on Hanson and Conolly’s survey of excavation reports and information from
museum and local authority archaeologists, that ‘a number of interesting blanks in the
distribution [of styli] can be noted (Cornwall, Devon, Leicestershire, Nottinghamshire,
Derbyshire, Shropshire), corresponding to areas with comparatively few villas’
(Mattingly 2006: 461). This statement is based on the partial information collected by
Hanson and Conolly and requires modification, given that the densest area of PAS finds
lies in the east Midlands. As we explore further and think about increased Latinization
in Britannia, we must not be binary: more Latin does not necessarily mean less Celtic,
but it almost certainly means widespread bilingualism.
Since Latin–​Celtic bilingualism was not unusual in Roman Britain, we might legiti-
mately ask where the direct evidence for written Celtic is, especially given our less
pessimistic views of levels of civilian literacy. Jackson’s dictum that there is no written
Celtic to be found whatsoever has only recently been reconsidered. He stated (Jackson
1953: 99–​100) in no uncertain terms:

It should always be borne in mind that British was not a written language, and
that the only language of writing was Latin; it would not occur to anyone to write
in British, nor would they know how to do so … In Roman Britain those who had
enough education to know the alphabet had enough to know some Latin and those
who had none did not write at all.
580

580   Alex Mullen

Kernel density plot of PAS

Roman styli
Land above 200 m
Constraints on PAS finds on a nation-
wide scale (Robbins 2014, fig. 4)

North Sea

T he Channel

0 50 200 km

Figure 28.1  Kernel density plot of the distributions of ‘broad period Roman’ styli recorded by
the Portable Antiquities Scheme as of 17 January 2014, plotted against selected constraints.
Source: Robbins 2014: figure 4 for constraints, with kind permission of the author; background map data from Ordnance
Survey/​EDINA supplied service, Crown Copyright/​database right 2013. Produced by L. Wallace.

The material available to him in the early twentieth century merited that kind of
response. Unfortunately, the effect has been that few Celticists have focused on Roman
Britain, and Classicists have been left to deal with the evidence. In 1987, Tomlin
announced that two tablets in the Bath collection (Tab. Sulis 14, 18) might be Celtic and
asked Celticists to ‘enlighten their classical colleagues’ (Tomlin 1987: 19). The tablets are
difficult to interpret, but have been reasonably analysed as Celtic (for a consideration of
both tablets, see RIG II.2 *L-​107–​08; Mees 2005; Mullen 2007b; for a discussion of Tab.
Sulis 18 only, see Schrijver 2004: 16–​17; 2005, 57–​60). Tab. Sulis 18 (Figure 28.2) is more
 Sociolinguistics   581

Figure 28.2  Tomlin’s line drawing of Tab. Sulis 18.

Source: reproduced with the kind permission of the author.

conducive to analysis as the text is complete, the transcription more certain, and the
word boundaries clear. The transcription can be given as:

1 adixoui
2 deịạna/​deṿina
3 deịeda/​deṿeda
4 andagin
5 vindiorix
6 cuamịịn
7 ạị

One interpretation of the text yields the following suggested translation: ‘I, Vindiorix, O
divine Deveda(?), shall fix an evil (?fate vel sim.) on Cuamiina’ (Mullen 2007b: 41).
An important consideration is whether these Celtic tablets are rare attestations
of written British Celtic or traces of continental visitors to the shrine. The exact
582

582   Alex Mullen

relationships between the forms of Celtic attested on the continent and in Britain are
still under examination, making firm linguistic conclusions elusive. Although we know
that British Celtic is closely related to Gaulish, it is unclear for this period precisely
which linguistic features would have distinguished the two varieties, so we are left argu-
ing from other evidence. We know that Aquae Sulis was a cosmopolitan sanctuary and
visitors were frequent from the continent, where Sulis and the related Sulevia group of
goddesses were also worshipped (see RIB I for inscriptions by continental visitors). On
the continent, curse tablets were written in Gaulish until at least the third century ad,
so it would have been conceivable for a Gaul to write to the goddess in his vernacular.
However, we have no certain evidence that Celtic-​speaking Britons could not do the
same, and we await future publications and finds to decide whether Tab. Sulis 14 and 18
are more likely to be British or Gaulish (RIB IV will contain all the known curse tablets
and writing tablets from Roman Britain).
Curse tablets often provide an insight into lower-​level society (for the status of the authors,
see Tab. Sulis 95–​98; Tomlin 2002: 171–​172, 174; for the issues in assigning authorship, see Tab.
Sulis 98–​101; Tomlin 2002: 170–​171) and are a more fruitful source of Celtic than much of
the epigraphic record (Russell 2006). Although the full edition of the Uley tablets remains
unpublished, Tomlin has commented that ‘three or four puzzling tablets from Bath and Uley
… may be British Celtic transliterated’ (Uley 114). One published tablet (Uley 33), provides a
possible parallel to Tab. Sulis 18. Although it is damaged, the best reading of line three contains
the word aexsieumo, which is not obviously Latin and may be a Celtic verb. The naming con-
text also has a non-​Latin flavour with three of the four names analysable as Celtic: Minu(v)
assus, Senebel[l]‌ena, and (possibly) Lucilia, which Hassall and Tomlin (1995: 378 n. 1) suggest
is not the Latin nomen but conceals a Celtic name element (see CPNRB; Mullen 2007a). The
addition of Latin filia and filius to the transcription is modern; these names almost certainly
show the non-​Roman use of plain genitive names as patronymics.
Celticists have largely overlooked the naming evidence from Roman Britain, concen-
trating instead on continental Celtic onomastics. However, the evidence is not meagre
and can help to reveal the indigènes within the Latin discourse: around 500 Celtic names
can be gleaned from the epigraphic corpora, plus a further 48 in numismatic and 22 in
literary sources (see the annually updated database CPNRB). Celtic names rarely appear
in either duo or tria nomina formulae (see Mullen 2007a). In fact, there are no instances
of Celtic names in these formulae in any of the hundreds of published curse tablets
from Roman Britain. A mere 6 per cent of Celtic names appear in duo nomina, almost
all from tombstones and militarized areas, while fewer than 2 per cent appear in tria
nomina, again all from highly Roman contexts (e.g. the inscription of Tiberius Claudius
Togidubnus (RIB I 91; Bogaers 1979)).
Across Britannia we see that filiation markers show a similar pattern. British Celtic
would probably have simply used the genitive of the father’s name to mark filiation. Of the
naming formulae attested in Britain that include Celtic names and filiation, 70 per cent
of the instances from RIB I show the Latinate filiation marker filius, fil or f, compared to
only 20 per cent in the curse tablets. The latter amounts to only three examples, two of
which are found in the elegant Tab. Sulis 30 (Figure 28.3) and are very likely the work of a
 Sociolinguistics   583

Figure 28.3  Tomlin’s line drawing of Tab. Sulis 30.

Source: reproduced with the kind permission of the author.

scribe. We should caution against the modern tendency to add Latin filiation markers to
the transcriptions of British inscriptions. Clearly, it is understandable in certain circum-
stances to reconstruct a filiation marker for the inscriptions on stone where there is an
obvious space for it and traces of letters. However, where there is definitely no space (e.g.
Sule(u)is Sulinus Bruceti VSLM (RIB I 105)) and/​or the filiation marker is not necessarily
expected (e.g. Lucilia Mellossi … Minu(v)assus Senebel[l]‌enae (Uley 33)), it is misleading to
make this addition, as it implies assimilation to the Latin formula, which had not necessar-
ily occurred.
Celtic names are interesting in themselves as they are composed of lexical elements
(e.g. Vindiorix in Tab. Sulis 18 is formed from lexical elements denoting ‘white/​shining’
and ‘king’), and, when they are compared with the Latin names used alongside them,
tentative assumptions might be made concerning linguistic and cultural contact. There
seem to be four main ways in which Britons adopt Latin nomenclature:

1. adoption of ‘colourless’ Latin cognomina such as Latinus, Civilis, or Maximus;

2. adoption of Latin names wholly or partly homophonous with Celtic names, per-
haps because they expressed an awareness of Celtic–​Latin interaction and/​or
584

584   Alex Mullen

facilitated comprehension. The three most common British Celtic name elements,
Sen, Luc, and Bel, all have common parallels in Latin;
3. translation of the Celtic name into Latin: for example, Primus is the translation
name for Cintusmus; this requires a comprehension of the lexical meaning of the
names in both Celtic and Latin and perhaps expresses duality of identity;
4. adoption of names of patrons and other important figures.

These different methods of adoption, which, arguably, could roughly indicate a scale
of increasing ‘Roman-​ness’, have been deemed useful in assessing the level of integra-
tion into new lifestyles and familiarity with the Reichssprache (Mullen 2007a). However,
in many cases names may have been adopted for personal and particular reasons that
are now irrecoverable, and the hierarchy for adoption is a blunt tool. Approaches to life
under Roman imperial power were multi-​faceted and complex, sometimes unpredict-
able, and it would be a mistake to assume too much about the identity of people on the
evidence of naming choice and naming formulae.4 Nevertheless, setting the names in
their context (Raybould 1999; Mullen 2007a), and in the context of other evidence for
bilingualism and linguistic varieties, may aid us in our quest.

Latinitas Britannica: The Nature

of British Latin

Trying to define ‘British Latin’ is as ridiculous as trying to describe ‘British English’.

Languages are never monolithic and display overwhelming variation. British Latin
would have been no different, although, in contrast to British English, we have little
chance of describing the complexity given our meagre resources. Traditional opinions
on British Latin are again largely attributable to Jackson (1953). He stated that, on the
one hand, the language shares features with the western empire, especially Gaul, but, on
the other, there is ‘a peculiar look to the picture of Vulgar Latin in Britain’—​namely, that
‘the sound-​system of Latin in Britain was very archaic by ordinary continental stand-
ards’ (Jackson 1953: 94, 107). Partly as a result of an occasional lack of clarity in the text
as to whether he refers specifically to the language of the ‘squirearchy’ or to the Latin of
Britain more generally, his view that elements of the British Latin were conservative and
old-​fashioned was uncritically adopted outside linguistic circles and extended to foster
the still relatively widely held belief that British Latin was archaic (see Gratwick 1982: 6,
69–​70).
But, even if we understand Jackson’s archaic Latinitas Britannica to be the preserve of
the Romano-​British elite, his vision of the Romano-​British linguistic landscape should
be reconsidered, not least because we now have much better evidence. Jackson was pri-
marily interested in the phonology of the Brittonic languages from the first to twelfth
centuries ad, and his conclusion was formulated using the evidence of Latin loanwords
 Sociolinguistics   585

into the Insular Celtic languages as transmitted by mainly post-​Roman sources. The
material on stone and instrumenta domestica from Roman Britain had not yet been
gathered into an easily accessible corpus (RIB) and, more importantly, there were hardly
any known curse tablets—​now our best source for British Latin. A great deal hinged on
the reconstruction of the form of Latin loanwords at the point at which they were bor-
rowed into Celtic—​that is, the spoken form of Latin with which Celtic speakers came
into contact: a subject of no little controversy.
In the 1980s, the Classicist Gratwick launched a scathing attack on Jackson’s Latinitas
Britannica, followed by a rigorous defence by the Celticist, McManus (1984). Smith
(1983) took the debate further by presenting a non-​partisan, meticulous study of the lin-
guistic features of the material published in RIB I and other ‘similar sources’ (1983: 896),
this time also considering morphology, lexis and semantics.5 In distilling the responses
to Jackson’s British Latin, several important points emerge:

1. Dating and nature of evidence. Borrowings may retain traits from the donor lan-
guage long after those traits have been lost or modified in the donor; that is, just
because early loans into British Celtic show features of early Latin does not mean
that the Latin of Britain continued to be of early type throughout the Roman
period. Conversely, much of the loanword evidence is potentially late, and some
liable to be that of the learned Latin of the Church. Furthermore, Roman lapidary
evidence is generally standardized and formulaic, and a proportion, sometimes
not identifiable, will have been put up by visitors to the province.6
2. Broader picture. Jackson underestimates the impact of British Celtic and ‘insu-
larity’ on British Latin and overlooks the evidence of Latin loanwords into Irish
(McManus 1983).
3. Linguistic detail. Jackson’s ‘Twelve Conservative Features’ of British Latin partly
dissolve under scrutiny.7
4. Interpretation. Jackson’s assumption that the ‘squirearchy’ is the main route of
transfer for Latin into British Celtic is an oversimplification, as is his ‘apartheid’
construction of ‘Roman’ and ‘Briton’ (Gratwick 1982).
5. Conclusion. British Latin can ‘no longer be accepted uncritically as archaic or con-
servative’ (Russell 1985: 29, see also D. E. Evans 1983: 979).

After considering the claims and counter-​claims of the scholarship on Latin in Roman
Britain, we are left with a pervasive scepticism as to whether the evidence will ever
allow us to describe the linguistic variation.8 However, these scholars were working in a
largely pre-​curse-​tablet era. The curses are often not subject to quite as many barriers to
reaching ‘British Latin’ and are now allowing insights into the possible range and variety
of non-​standard British Latin. Linguistic analysis of the Bath curse tablets suggests, for
example, that British Latin contains some possible regionalisms (such as, perhaps, the
preference for involo instead of furo for the verb ‘to steal’) and that Gallic and British
Latin share some ‘north-​western’ characteristics caused by the common Celtic back-
ground, cross-​Channel contacts, and distance from Italy (Adams 1992; Schrijver 2002).
586

586   Alex Mullen

Curse tablets also allow us to examine more closely the only one of Jackson’s ‘Twelve
Conservative Features’ of British Latin whose existence in Britain (though not interpre-
tation) was not questioned by the scholarship of the 1980s. The feature is B/​V alterna-
tion in spelling (Jackson’s point number seven), which indicates a merger of standard
Latin [b]‌and [w] and shows regional variation across provincial Latin. It occurs rarely
in the evidence from Britannia and the Balkans, but very commonly in Italy and North
Africa and commonly elsewhere. For Britain, the inscriptional material and the evi-
dence of Latin loanwords into the Brittonic languages (Welsh, Cornish, and Breton),
where [b] and [w] appear to be kept distinct, implies that the merger did not take
place. A revised interpretation (Mullen 2013b) of a curse tablet from Ratcliffe-​on-​Soar
(Nottinghamshire) suggests that we must scrutinize even this feature.
Initially edited with some degree of uncertainty (Hassall and Tomlin 1993: 310–​314),
the revised analysis of the inscription found in 1990 at Red Hill, Ratcliffe-​on-​Soar
(Figure 28.4) argues that it concerns the theft of a mule and contains two examples of B/​
V alternation: vissacio for bisaccio ‘saddle-​bags’ (l. 7) and pavlatoriam for pabulatoriam
‘fodder-​bag’ (l. 5). The original text of the 1990 find, perhaps dating to the third or fourth
century, is written from right to left and almost entirely in mirror image capitals.
Here is a revised transcription and tentative translation.

nomine Camụlorigi et Titọcune molam quam perdeḍerunt

in fanum dei ḍ[..]v[.]‌ c̣uicumquẹ ṇ[…]n involasit
mola illam ut samguin suum [m]‌ittat usque die‘m’ q[..]
moriatur [quic]ụmque iṇṿollṿa[…] hụrta mor‘i’atur
et pavlatoriam quicumque [……] involạsit
et ipse moriato mo[ri]atur quicumqui illam
involasit et vertogn ḍe ospitio uel vissacio
uicumque illam involasit a devo moriotur

Figure 28.4  Tomlin’s line drawing of the lead curse tablet at Red Hill, Ratcliffe-​on-​Soar.
Source: Hassall and Tomlin (1993: 311), reproduced with the kind permission of the author.
 Sociolinguistics   587

In the name of Camulorix and Titocuna I have dedicated in the temple of the god the
mule which they have lost. Whoever stole that mule, whatever his name, may he let
his blood until the day he die. Whoever stole the objects of theft(?) /​fencing(?), may
he die, and the fodder-​bag, whoever stole it, may he also die. Whoever stole it and the
trinket(?) from the stable or the saddle-​bag, whoever stole it, may he die by the god.

The alternation of B/​V occurs twice in this short text and the conclusion seems irre-
sistible: the merger of /​b/​and /​w/​is likely to have occurred in the idiolect of the author.
The problem with extrapolating beyond this meagre evidence to say that the merger
may have been more widespread is that, so far, no other published curse tablet offers a
clear example of a B/​V alternation (Adams 2007: 653). Indeed, in his survey of British
Christian and pagan lapidary inscriptions spanning the first to the ninth centuries ad,
Barbarino (1978: 30–​39; his sources are RIB I and Macalister 1945) found 80 correct
usages of B in initial position and no examples of V and 354 correct usages of initial V
and only 2 examples of B for V, and, in intervocalic position, only 1 example of V for
expected B in 107 tokens and no examples of B for V in 139. However, we have to con-
sider correctness in spelling as a possible factor for the lack of B/​V alternation in certain
material from Britain, rather than necessarily assuming that the merger has not taken
place in the spoken language. The curse tablet from Ratcliffe-​on-​Soar provides us with
fertile linguistic material: a context of potentially low-​level education, a local author,9
and hints of a spoken variety (Adams, in press). It looks as if, on the basis of this tab-
let, the [b]‌–​[w] merger may have been a feature of British Latin in some idiolects, peri-
ods, and areas, though the B/​V alternation was avoided in writing through adherence to
orthographic norms. The other possible examples of B/​V alternation in the inscriptional
record (e.g. bagis bitam (RIB I 1); Vivio (17); Betto (2144); Vrocatae (RIB II 2503.160);
parbo (RIB II 2503.444)), which have often been brushed aside as masons’ errors or the
work of continental visitors, should now be re-​examined.
We might consider that the traditional interpretation of distinct [b]‌and [w] sounds in
Latin loanwords into British Celtic also needs to be revisited (Jackson 1953: 88–​90, 363–​
365, 413). If, after careful examination, we can show that the sounds are kept apart, then
we need to establish securely the dating and nature of these loanwords. Some may, of
course, be early (before the merger in spoken Latin of the empire) and others late (either
from learned ecclesiastical Latin (Thomas 1981: 61–​79), or from spoken Latin after the
differentiation of the sounds in initial and post-​liquid positions, as we see in some of
the Romance languages). If the loanwords showing no merger are, at least in part, impe-
rial in date, and we now have evidence that the B/​V alternation (indicating a merger)
may have been attested in some varieties of British Latin in the imperial period, then
this may be useful evidence for the variegated linguistic landscape of the province along
social, temporal, and geographical lines (Hamp 1975; Smith 1983: 935–​938).
Jackson’s view (1953) that the countryside was Celtic speaking, with the exception of
the rural elites, has been undermined by the finds of Latin curse tablets in rural con-
texts, whose authors complain, often in low-​level Latin, about relatively minor thefts.
As we have seen, current research is encouraging us to be more generous about levels
588

588   Alex Mullen

of literacy (Pearce 2004: 44), and therefore Latin and bilingualism, among the civilian
population. Bilingualism theory teaches us that a restriction of Celtic–​Latin bilingual-
ism to the rural elite is unlikely and that the authors of curse tablets, such as the one from
Ratcliffe, were very likely to have been bilingual to an extent. The rural elite are clearly
not the almost exclusive purveyors of Latin loanwords into British Celtic, as Jackson
(1953) proposed. We need to attempt to identify, describe, and explain the differences
and similarities between the Latin(s) of the loanwords and the Latin(s) on the ground,
which may demonstrate mixtures of different types of Latin from a range of time depths
and contexts. The vigorous and complex discussion of the nature of Latin loanwords in
the Brittonic languages and Latinitas Britannica finds itself with more fuel for debate.

The End of Britannia: Post-​Roman

Inscriptions and the Germanic
Languages

As we creep past the end of Roman rule, Ogam inscriptions, often under-​appreciated
by Classicists, are important evidence for assessing the linguistic composition of
Britain. These texts date from the fourth century ad onwards and are attested in south-
ern Ireland, Scotland, the Isle of Man, Wales, and south-​western England (for the texts,
see Macalister 1945; Nash-​Williams 1950; Thomas 1991–​2; Okasha 1993; Tedeschi 2005;
Edwards 2007, 2013; Redknap and Lewis 2007; for discussions, see McManus 1991; Sims-​
Williams 2003; Charles-​Edwards 2013:  esp.  96–​173; for dating, see Charles-​Edwards
2013: 119). They are the earliest attestations of Insular Celtic (with the possible exception
of one or two curse tablets) and are inscribed using an alphabet comprised of linear inci-
sions and loosely influenced by the Roman alphabet, according to a majority of scholars
(Harvey 1987, 1989; Lehmann 1989). The earliest examples, incised on standing stones,
mainly consist of personal names and formulae. In Britain, the inscriptions appear in
Celtic in Ogam script, and also, much more commonly, bilingually in Celtic in Ogam
script and in Latin (Figure 28.5). In addition, related texts occur written in Latin using
only the Roman script (conversely, the inscriptions in Ireland are almost always only
in Irish; see Charles-​Edwards 2013: 119). This is not the place for a full discussion of the
details of the Ogam stones and their Latin congeners, but it is worth touching on the
debate that has sprung up around the Latin versions.
Charles-​Edwards (1995: 715–​7 18) has argued that the presence of Latin on British
Ogam stones suggests that the inscribing communities were bilingual and that their
British Latin shows some distinctive ‘spoken’ developments, not least the collapse of the
case system, which was also happening simultaneously in British Celtic (Koch: 1982–​
3). The fifth-​century ‘Corbalengi stone’ from Penbryn, Ceredigion, for example, reads
CORBALENGI IACIT ORDOVS, and offers what seems to be a genitive singular per-
sonal name where standard Latin would require a nominative. Adams (2007: 616–​620)
 Sociolinguistics   589

(a) (b)

Figure  28.5 Latin and Ogam-​inscribed stone, St Dogmaels/​Llandudoch, Pembrokeshire,

Wales, (a) photograph, (b) drawing and transcription.
Source: © Crown copyright: Royal Commission on the Ancient and Historical Monuments of Wales.

presents a different interpretation of the same material. For him, the mistakes in case
agreement suggest that Latin is not known by the inscribing communities; rather they
recognize that they might want to use Latin but, in the absence of competence, incor-
rectly copy features from other Ogam inscriptions into their scrappy, written Latin.
Given our poor knowledge of the inscribing communities (in many cases it is hard to be
sure whether we are dealing with Irish speakers in Wales or local communities)10 and in
the absence of lengthy texts, it is difficult to make any firm decisions between what seem
590

590   Alex Mullen

to be plausible views of the same material. Nevertheless, other evidence—​for example,

the contemporary non-​ecclesiastical loanwords into Old Irish, which come via British
Latin rather than British Celtic, and the fact that apparently ‘Ogam-​influenced’ Latin
texts occur in areas with no Ogam inscriptions—​suggests that a Celtic-​influenced Latin
may have been spoken beyond the context of the church in the post-​Roman Highland
zone (Charles-​Edwards 2013: 109–​110). What we can state with certainty is that written
Latin had some currency in communities in the far west of what had been Britannia,
even after the decline of Rome.11
One other major clue for reconstructing what might have been happening to the lin-
guistic composition of Britain during the Roman period could be the languages that the
Germanic-​speaking incomers to the island encountered from the fifth century onwards.
Unfortunately, as with many aspects of the history of post-​imperial Britain, there are
far more questions than answers. In recent decades, the crux of the debate concerning
the adventus Saxonum has moved away from the issues of precisely when, where, and
who, and has involved a reconsideration of the mass invasion of Germanic-​speaking
peoples altogether. This change was motivated by a move in the twentieth century away
from Germanic origins for England, reinterpretation of the historical sources, and, most
importantly, increasingly sophisticated analysis of the growing archaeological evidence.
The revisionist stance argued that there had been no large-​scale change in population
and that cultural change had been brought about by a small number of elite incomers.
The current consensus still seems to support immigration of relatively significant num-
bers into Britain (Hamerow 1997; Ward-​Perkins 2000: 519–​523), though one of the revi-
sionists, Higham (2007: 12), has stated that ‘there is a real danger that an elite dominance
interpretation of cultural change is becoming the new orthodoxy’. This particular debate
seems set to continue.
Linguistic evidence has increasingly been deemed an essential element in mak-
ing sense of the ambiguous remains from the post-​Roman period. The clear success
of the English (Germanic) language and the apparent lack of elements of Latin and,
especially, British Celtic in Old English have led some linguists to view a massive
invasion and change of population or ‘cultural annihilation’ as indispensable (e.g.
Coates 2007; Padel 2007).12 But we should hesitate before assigning a victory to the
‘mass-​migrationists’:  the evidence from linguistics needs to be handled more cau-
tiously. First, language change may be brought about without the condition of mas-
sive population replacement. Not only that, but when speakers of one language shift
to another with little access to education, the contact-​induced change in the target
language may tend to involve transfer of phonology, morphology, and syntax rather
than loanwords. In other words, when speakers of British Celtic/​Latin shifted to Old
English, the impact on Old English may have been grammatical rather than lexical.
This type of change is much harder to recover, especially for typologically similar lan-
guages (though Schrijver (2002) and Tristram (2007) claim to have identified some
possible examples), and will be unlikely to have affected the standardized Old English
of the elite, which was the main variety of language committed to writing until the
 Sociolinguistics   591

twelfth century (Tristram 2007: 201). So the possibility of language change without

population replacement remains open, though we need to try to work out how and
why, if they stayed put, the local communities of England steadily shifted away from
both their British Celtic and their Latin in favour of the Germanic tongues of the
incomers.13
The element often missing from the immediately post-​Roman story has been a seri-
ous consideration of Latin, though Schrijver (2002, 2007) has recently been putting
spoken Latin back in the picture (see Russell (2012: 219–​224) for a series of associated
methodological issues). The question often posed by scholars of post-​Roman Britain
is: ‘why don’t the English speak Welsh?’ (Tristram 2007: 192). But we might legiti-
mately wonder what has happened to spoken Latin too, if we think that Latinization
had been successful—​at least in certain areas and social groups—​and given the lapi-
dary evidence just discussed and the continued use of spoken Latin forms on the
continent, despite incursions from outside the old Roman Empire (Ward-​Perkins
(2000) and Halsall (2013) compare the post-​Roman situations in northern Gaul and
Britain). Indeed, considering how important the Roman Empire must have been for
at least some Germanic-​speaking mercenaries and traders originating outside the
official boundaries of the empire, we might even wonder whether the lingua franca
between local and incomer in post-​Roman Britain might have been Latin. But there
is no consensus on how much Latin the members of the adventus encountered in
Britain. We can be sure that the linguistic situation varied along geographical and
social lines, and the new arrivals would have encountered different varieties of lan-
guage, depending on where and when they landed and with whom they came into
contact. It is perhaps the fact that the end of Roman rule upset the social framework
for this complex Latin–​Celtic societal bilingualism that provided the ideal condi-
tions for the uptake of a new language, whose speakers may have had a strong sense
of ethnolinguistic identity.

Conclusion

Though the indigenous communities of Roman Britain did not have a writing system
or literature of their own and barely feature in Classical texts, they can be materialized
in the material remains of the province, including its epigraphy. Sociolinguistic analysis
can provide information for multiple identities and cultural contacts through naming
choices and linguistic features. The oft-​neglected linguistic evidence, which can even
be enshrined in later languages, is essential in reaching nuanced conclusions on social
interactions in the province and is constantly renewed with new finds and revisions of
the old. This evidence should not be used to supplant the other archaeological material,
or its literary and historical handmaidens, but rather to complement it, and is a vital
component in enabling us to understand Roman Britain.
592

592   Alex Mullen

Notes
1. The difficult issues of possible ‘pre-​Celtic’ and non-​Indo-​European/​Indo-​European (e.g.
Germanic) languages in contact with Celtic (D. E. Evans 1983: 952–​954) and of Pictish in
Scotland (Charles-​Edwards 2013: 89–​92) will not be tackled here.
2. This is due to their use of theta and barred d, which are employed on the continent to
represent the so-​called tau gallicum sound of Gaulish language; see Mullen (2013a:
102–​103).
3. For example, the definite article from the demonstrative pronoun, compounding of two
or more prepositions, syntax of subordinate clauses, and the pluperfect, for which see Mac
Cana (1976). Russell (2012: 222–​223) argues that, while the pluperfect may have been created
under influence from Latin, the model presented by Mac Cana cannot work. See D. E. Evans
(1983: 973) for citation of other possible non-​lexical features of contact-​induced change.
4. Toponyms (place names) are another source of linguistic and cultural information,
though their interpretation raises just as many problems as for personal names (see Rivet
and Smith 1979; Parsons and Sims-​Williams 2000; de Hoz et al. 2005). Parsons (2011: 122)
notes that no more than 7% of place names in Roman Britain are Latin and many post-​
conquest names, even for Roman creations, are Celtic (Rivet 1980).
5. Mann (1971) produced an incomplete list of features found in the inscriptional material,
which he purported to be evidence of spoken Latin in Britain, supplemented by features
in the coinage a few years later by Shiel (1975). These attempts to identify spoken language
through the testimony of lapidary and numismatic material were met with distain by
Smith (1983: e.g. 895–​896, 935–​936).
6. See Mann (1985) for the view that the epigraphic habit never really catches on among local
Britons, even recruits into the army (see also Hope 1997; Mattingly 2008; Noy 2010). J.
Evans (1987) notes that the graffiti on pottery are more useful than lapidary inscriptions
for charting literacy in Britain, since they are not biased towards the military and for-
eign authors. For literacy in Roman Britain, see also Raybould (1999); Ingemark (2000);
Williams (2001, 2002, 2005, 2007); Hanson and Conolly (2002); Pearce (2004).
7. According to Smith (1983), Jackson’s numbers 3, 5, 10 (1953:  87–​88, 93–​94) need to be
withdrawn altogether, numbers 1, 2 (1953: 86–​87) modified, and numbers 4, 6, 8–​9, 11–​12
(1953: 88, 90–​93, 94) retained but with a warning that perhaps the developments were pos-
sibly rather later in Vulgar Latin itself than Jackson allowed.
8. If only we had better evidence, Hamp’s study (1975) of ‘social gradience’ in British Latin could be
a useful approach. Adams (2007: 591–​596), however, notes that Hamp’s article is inaccurate and
expresses frustration that it ‘goes on being quoted with approval’.
9. The penultimate word of the Ratcliffe curse, devo, allows identification of a potentially
‘local’ feature; see Mullen (2013b: 269).
10. Sims-​Williams (2002) has pointed out that a significant percentage (70%) of the pre-​
Norman inscriptions from Wales are ‘Irish’ in some sense. Indeed, in the stone inscription
cited in the main text, CORBALENGI is a name of Irish form (with an -​a-​composi-
tion vowel) and yet appears to belong to a member of the Ordovices (see Sims-​Williams
2002: 26; Russell 2012: 216 n. 74; Charles-​Edwards 2013: 176).
11. Latin also continues to be used in elite contexts: parts of the Llandaff charters date to the
early seventh century and Gildas is well educated in Latin (see Lapidge 1984).
 Sociolinguistics   593

12. There are many more early Latin loanwords than British Celtic in Old English, even once
the later ecclesiastical Latin borrowings have been removed; some of these may be from
continental Latin, but surely not all (see Parsons 2011: 120–​121).
13. See Ward-​Perkins (2000: 523–​524) for the suggestion that the Law of Ine, which gave pref-
erential treatment to Saxons, may have been a motivating factor.

References
Major British and Continental Corpora
Carlisle R. S.  O. Tomlin, ‘Roman Manuscripts from Carlisle:  The Ink-​Written Tablets’,
Britannia, 29 (1998), 31–​84.
CPNRB A. Mullen and P. Russell, Celtic Personal Names of Roman Britain <www.asnc.cam.
ac.uk/​personalnames>, updated annually (accessed 12 March 2013).
MLH IV J. Untermann, Monumenta linguarum hispanicarum: IV. Die tartessischen, keltiber-
ischen und lusitanischen Inschriften. Wiesbaden: L. Reichert, 1997.
RIB I R. G. Collingwood et al., The Roman Inscriptions of Britain. Volume I: Inscriptions
on Stone. Oxford: Oxford University Press, 1965; rev. edn. 1995.
RIB II S. S. Frere et  al., The Roman Inscriptions of Britain. Volume II:  Fascicules 1–​
8: Instrumentum domesticum. Gloucester: Sutton, 1990–​5.
RIB III R. S. O. Tomlin, R. P. Wright, and M. W. C. Hassall (2009). The Roman Inscriptions
of Britain. Volume III: Inscriptions on Stone. Oxford: Oxbow Books.

[New inscriptions post-​RIB published in the Journal of Roman Studies and Britannia ‘Roman
Britain in [calendar year]’]
RIG J.-​
B. Colbert de Beaulieu, P.-​ M. Duval, B. Fischer, M. Lejeune, P.-​ Y.
Lambert, and G. Pinault (1985). Recueil des inscriptions gauloises, I, II.1, II.2,
III. Paris: Éditions du C.N.R.S.
Tab. Sulis R. S.  O. Tomlin, ‘The Curse Tablets’, in B. Cunliffe (ed.), The Temple of
Sulis Minerva at Bath, ii. The Finds from the Sacred Spring. Oxford: Oxford
Committee for Archaeology, 1988, 59–​269.
Tab. Vindol. I A. K. Bowman and J. D. Thomas, Vindolanda:  The Latin Writing-​Tablets
(Tabulae Vindolandenses I). London: British Museum Press, 1983.
Tab. Vindol. II A. K. Bowman and J. D. Thomas, The Vindolanda Writing-​Tablets (Tabulae
Vindolandenses II). London: British Museum Press, 1994.
Tab. Vindol. III A. K. Bowman and J. D. Thomas, The Vindolanda Writing-​Tablets (Tabulae
Vindolandenses III). London: British Museum Press, 2003.
Tab. Vindol. IV.1 A. K. Bowman, J. D. Thomas, and R. S. O. Tomlin, ‘The Vindolanda Writing-​
Tablets (Tabulae Vindolandenses IV, Part 1)’, Britannia, 41 (2010), 87–​224.
Tab. Vindol. IV.2 A. K. Bowman, J. D. Thomas, and R. S. O. Tomlin, ‘The Vindolanda Writing-​
Tablets (Tabulae Vindolandenses IV, Part 2)’, Britannia, 42 (2011), 113–​144.
Uley ‘The Inscribed Lead Tablets: An Interim Report’, in A. Woodward and P.
Leach (eds), The Uley Shrines: Excavation of a Ritual Complex on West Hill,
Uley, Gloucestershire, 1977–​79. London: English Heritage, 1993, 113–​130.
594

594   Alex Mullen

Primary Sources
Bede, Historia Ecclesiastica (Bede’s Ecclesiastical History of the English People), ed. B. Colgrave
and R. A. B. Mynors. Oxford: Clarendon Press, 1969.
Tacitus, Agricola, ed. R. M. Ogilvie and I. A. Richmond. Oxford: Clarendon Press, 1969.

Secondary Sources
Adams, J. N. (1992). ‘British Latin: The Text, Interpretation and Language of the Bath Curse
Tablets’, Britannia, 23: 1–​26.
Adams, J. N. (1996). ‘The Interpretation of souxtum at Tab. Vindol. II.301.3’, Zeitschrift für
Papyrologie und Epigraphik, 110: 238.
Adams, J. N. (2003). Bilingualism and the Latin Language. Cambridge: Cambridge University Press.
Adams, J. N. (2007). The Regional Diversification of Latin 200 bc–​ad 600. Cambridge:
Cambridge University Press.
Adams, J. N. (2013). Social Variation and the Latin Language. Cambridge:  Cambridge
University Press.
Adams, J. N. (in press). Informal Varieties of Latin: Selected Passages from the Second Century
bc to the Eighth Century ad, with Commentary. Cambridge: Cambridge University Press.
Barbarino, J. L. (1978). The Evolution of the Latin /​b/​—​/​ṷ/​Merger:  A  Quantitative and
Comparative Analysis of the B-​V Alternation in Latin Inscriptions. Chapel Hill, NC: UNC
Department of Romance Languages.
Bogaers, J. E. (1979). ‘King Cogidubnus in Chichester: Another Reading of RIB 91’, Britannia,
10: 243–​254.
Charles-​Edwards, T. (1995). ‘Language and Society among the Insular Celts ad 400–​1000’, in
M. J. Green (ed.), The Celtic World. London: Routledge, 703–​736.
Charles-​Edwards, T. (2013). Wales and the Britons, 350–​1064. Oxford: Oxford University Press.
Coates, R. (2007). ‘Invisible Britons: The View from Linguistics’, in N. J. Higham (ed.), Britons
in Anglo-​Saxon England. Woodbridge: Boydell Press, 172–​191.
Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge
University Press.
Cunliffe, B. (2005). Iron Age Communities in Britain:  An Account of England, Scotland
and Wales from the Seventh Century bc until the Roman Conquest. 4th edn. London and
New York: Routledge.
de Bernardo Stempel, P. (1991). ‘Die Sprache altbritannischer Münzlegenden’, Zeitschrift für
celtische Philologie, 44: 36–​55.
de Hoz, J., Luján, E. R., and Sims-​Williams, P. (2005) (eds). New Approaches to Celtic Place
Names in Ptolemy’s Geography. Madrid: Ediciones Clásicas.
Edwards, N. (2007). A Corpus of Early Medieval Inscribed Stones and Stone Sculpture in Wales,
ii. South-​West Wales. Cardiff: University of Wales Press.
Edwards, N. (2013). A Corpus of Early Medieval Inscribed Stones and Stone Sculpture in Wales,
iii. North Wales. Cardiff: University of Wales Press.
Evans, D. E. (1983). ‘Language Contact in Pre-​Roman and Roman Britain’, Aufstieg und
Niedergang der römischen Welt, 29/​2: 949–​987.
Evans, J. (1987). ‘Graffiti and the Evidence of Literacy and Pottery Use in Roman Britain’,
Archaeological Journal, 144: 191–​204.
 Sociolinguistics   595

Gosden, C. (2004). Archaeology and Colonialism: Cultural Contact from 5000 bc to the Present.
Cambridge: Cambridge University Press.
Gratwick, A. S. (1982). ‘Latinitas Britannica: Was British Latin Archaic?’, in N. P. Brooks (ed.),
Latin and the Vernacular Languages in Early Medieval Britain. Leicester: Leicester University
Press, 1–​79.
Halsall, G. (2013). Worlds of Arthur:  Facts and Fictions of the Dark Ages. Oxford:  Oxford
University Press.
Hamerow, H. (1997). ‘Migration Theory and the Anglo-​Saxon “Identity Crisis”’, in J. Chapman
and H. Hamerow (eds), Migrations and Invasions in Archaeological Explanation. BAR
International Series 664. Oxford: Archaeopress, 33–​44.
Hammond, N. G. L., and Scullard, H. H. (1970) (eds). The Oxford Classical Dictionary. 2nd edn.
Oxford: Oxford University Press.
Hamp, E. P. (1975). ‘Social Gradience in British Spoken Latin’, Britannia, 6: 150–​162.
Hanson, W. S., and Conolly, R. (2002). ‘Language and Literacy in Roman Britain: Some
Archaeological Considerations’, in A. E. Cooley (ed.), Becoming Roman, Writing Latin? Literacy
and Epigraphy in the Roman West. Journal of Roman Archaeology Supplementary Series 48.
Portsmouth, RI: Journal of Roman Archaeology, 151–​164.
Harvey, A. (1987). ‘Early Literacy in Ireland: The Evidence from Ogam’, Cambridge Medieval
Celtic Studies, 14: 1–​15.
Harvey, A. (1989). ‘Some Significant Points of Early Insular Celtic Orthography’, in D.
O’Corráin, L. Breatnach, and K. McCone (eds), Sages, Saints and Storytellers, Celtic Studies
in Honour of Professor James Carney. Maynooth: An Sagart, 56–​66.
Hassall, M. W.  C., and Tomlin, R. S.  O. (1993). ‘Roman Britain in 1992. II. Inscriptions’,
Britannia, 24: 310–​322.
Hassall, M. W.  C., and Tomlin, R. S.  O. (1995). ‘Roman Britain in 1994. II. Inscriptions’,
Britannia, 26: 371–​390.
Haverfield, F. (1923). The Romanization of Roman Britain. 4th edn. Oxford: Clarendon Press.
Higham, N. J. (2007). ‘Britons in Anglo-​Saxon England: An Introduction’, in N. J. Higham
(ed.), Britons in Anglo-​Saxon England. Woodbridge: Boydell Press, 1–​15.
Hope, V. M. (1997). ‘Words and Pictures: The Interpretation of Romano-​British Tombstones’,
Britannia, 28: 245–​258.
Ingemark, D. (2000). ‘Literacy in Roman Britain:  The Epigraphical Evidence’, Opuscula
Romana, 25–​26: 19–​30.
Jackson, K. H. (1953). Language and History in Early Britain. Edinburgh: Edinburgh University Press.
Jørgensen, A. R. (2008). ‘Varia III an Additional Cognate of Gaulish souxtu and Irish
suacht: Old Cornish seit’, Ériu, 58: 183–​185.
Koch, J. T. (1982–​3). ‘The Loss of Final Syllables and Loss of Declension in Brittonic’, Bulletin of
the Board of Celtic Studies, 30: 201–​233.
Lambert, P.-​Y. (2000). ‘Varia III Gaulish souxtu: Early Irish suacht’, Ériu, 51: 189–​192.
Lambert, P.-​Y. (2003). La Langue gauloise. 3rd edn. Paris: Editions Errance.
Lambert, P.-​Y. (2004). ‘Varia V Gaulish souxtu: Addendum’, Ériu, 54: 263–​264.
Lapidge, M. (1984). ‘Gildas’s Education and the Latin Culture of Sub-​Roman Britain’,
in M. Lapidge and D. Dumville (eds), Gildas:  New Approaches. Woodbridge:  Boydell
Press, 27–​50.
Lehmann, R. P. M. (1989). ‘Ogham: The Ancient Script of the Celts’, in W. Senner (ed.), The
Origins of Writing. Lincoln, NE: University of Nebraska Press, 59–​170.
596

596   Alex Mullen

Lejeune, M. (1971). ‘Lepontica’, Études celtiques, 12/​2: 357–​500.

Ling, R. (2007). ‘Inscriptions on Romano-​British Mosaics and Wall-​Paintings’, Britannia
38: 63–​91.
Mac Cana, P. (1976). ‘Latin Interference on British:  The Pluperfect’, in J. J. O’Meara and B.
Naumann (eds), Latin Script and Letters ad 400–​900: Festschrift presented to Ludwig Bieler
on the Occasion of his 70th Birthday. Leiden: Brill, 194–​203.
Macalister, R. A.  S. (1945). Corpus Inscriptionum Insularum Celticarum. Dublin:  Four
Courts Press.
McManus, D. (1983). ‘A Chronology of the Latin Loan-​Words in Early Irish’, Ériu, 34: 21–​7 1.
McManus, D. (1984). ‘Linguarum diversitas: Latin and the Vernaculars in Early Medieval
Britain’, Peritia, 3: 151–​188.
McManus, D. (1991). A Guide to Ogam. Maynooth: An Sagart.
Mann, J. (1971). ‘Spoken Latin in Britain as Evidenced in the Inscriptions’, Britannia, 2: 218–​224.
Mann, J. (1985). ‘Epigraphic Consciousness’, Journal of Roman Studies, 75: 204–​206.
Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Penguin.
Mattingly, D. (2008). ‘Urbanism, Epigraphy and Identity in the Towns of Britain under
Roman Rule’, in H. M. Schellenberg, V. E. Hirschmann, and A. Krieckhaus (eds), A
Roman Miscellany:  Essays in Honour of Anthony R.  Birley on his Seventieth Birthday.
Gdańsk: Foundation for the Development of Gdańsk University, 53–​7 1.
Mays, M. (1992). ‘Inscriptions on British Celtic Coins’, Numismatic Chronicle, 152: 57–​82.
Mees, B. (2005). ‘The Celtic Inscriptions of Bath’, Studia Celtica, 39: 176–​181.
Mullen, A. (2007a). ‘Linguistic Evidence for Romanization:  Continuity and Change in
Romano-​British Onomastics’, Britannia, 38: 35–​61.
Mullen, A. (2007b). ‘Evidence for Written Celtic from Roman Britain: A Linguistic Analysis of
Tabellae Sulis 14 and 18’, Studia Celtica, 41: 31–​45.
Mullen, A. (2013a). Southern Gaul and the Mediterranean:  Multilingualism and Multiple
Identities in the Iron Age and Roman Periods. Cambridge: Cambridge University Press.
Mullen, A. (2013b). ‘New Thoughts on British Latin: A Curse Tablet from Red Hill, Ratcliffe-​
on-​Soar (Nottinghamshire)’, Zeitschrift für Papyrologie und Epigraphik, 187: 266–​272.
Mullen, A., and James, P. (2012) (eds). Multilingualism in the Graeco-​ Roman Worlds.
Cambridge: Cambridge University Press.
Nash, D. (1987). Coinage in the Celtic World. London: Seaby.
Nash-​Williams, V. E. (1950). The Early Christian Monuments of Wales. Cardiff: University of
Wales Press.
Noy, D. (2010). ‘Epigraphic Evidence for Immigrants at Rome and in Roman Britain’, in H.
Eckardt (ed.), Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the
Roman Empire. Journal of Roman Archaeology Supplementary Series 78. Portsmouth, RI:
Journal of Roman Archaeology, 13–​26.
Okasha, E. (1993). Corpus of Early Christian Inscribed Stones of South-​ West Britain.
Leicester: Leicester University Press.
Padel, O. J. (2007). ‘Place-​Names and the Saxon Conquest of Devon and Cornwall’, in N. J.
Higham (ed.), Britons in Anglo-​Saxon England. Woodbridge: Boydell Press, 215–​230.
Parsons, D. N. (2011). ‘Sabrina in the Thorns: Place-​Names as Evidence for British and Latin in
Roman Britain’, Transactions of the Philological Society, 109: 113–​137.
Parsons, D. N., and Sims-​Williams, P. (2000) (eds). Ptolemy: Towards a Linguistic Atlas of the
Earliest Celtic Place-​Names of Europe. Aberystwyth: CMCS.
Partridge, C. (1982). ‘Graffiti from Skeleton Green’, Britannia, 13: 325–​326.
 Sociolinguistics   597

Pearce, J. (2004). ‘Archaeology, Writing Tablets and Literacy in Roman Britain’, Gallia,
61: 43–​51.
Porzio Gernia, M. L. (1981). ‘Gli elementi celtici del latino’, in E. Campanile (ed.), I Celti d’Italia.
Pisa: Giardini, 97–​122.
Pulgram, E. (1950). ‘Spoken and Written Latin’, Language, 26/​4: 458–​466.
Raybould, M. (1999). A Study of Inscribed Material from Roman Britain. An Inquiry into Some
Aspects of Literacy in Romano-​British Society. BAR British Series 281. Oxford: Archaeopress.
Redknap, M., and Lewis, J. M. (2007). A Corpus of Early Medieval Inscribed Stones and
Stone Sculpture in Wales, i. Glamorgan, Brecknockshire, Monmouthshire, Radnorshire and
Geographically Contiguous Areas of Herefordshire and Shropshire. Cardiff:  University of
Wales Press.
Rhodes, M. (1987). ‘Inscriptions on Leather Waste from Roman London’, Britannia, 18: 173–​181.
Richards, J. D., Naylor, J., and Holas-​Clark, C. (2009). ‘Anglo-​Saxon Landscape and Economy:
Using Portable Antiquities to Study Anglo-​ Saxon and Viking Age England’, Internet
Archaeology, 25. <http://​dx.doi.org/​10.11141/​ia.25.2> (accessed 15 January 2015).
Rivet, A. L. F. (1980). ‘Celtic Names and Roman Places’, Britannia, 11: 1–​19.
Rivet, A. L. F., and Smith, C. (1979). The Place-​Names of Roman Britain. London: Batsford.
Robbins, K. (2014). Portable Antiquities Scheme: A Guide for Researchers. <http://​finds.org.uk/​
documents/​guideforresearchers.pdf> (accessed 17 March 2015).
Russell, P. (1985). ‘Recent Work in British Latin’, Cambridge Medieval Celtic Studies, 9: 19–​29.
Russell, P. (2006). ‘VILBIAM (RIB 154): Kidnap or Robbery?’, Britannia, 37: 363–​367.
Russell, P. (2012). ‘An habes linguam Latinam? Non tam bene sapio: Views of Multilingualism
from the Early Medieval West’, in A. Mullen and P. James (eds), Multilingualism in the
Graeco-​Roman Worlds. Cambridge: Cambridge University Press, 193–​224.
Sargent, A. (2002). ‘The North–​South Divide Revisited: Thoughts on the Character of Roman
Britain’, Britannia, 33: 219–​226.
Schmidt, K. H. (1967). ‘Keltisches Wortgut im Lateinischen’, Glotta, 44: 151–​174.
Schrijver, P. (2002). ‘The Rise and Fall of British Latin:  Evidence from English and
Brittonic’, in M. Filppula, J. Klemola, and H. Pitkänen (eds), The Celtic Roots of English.
Joensuu: University of Joensuu, Faculty of Humanities, 87–​110.
Schrijver, P. (2004). ‘Der Tod des Festlandkeltischen und die Geburt des Französischen,
Niederländischen und Hochdeutschen’, in P. Schrijver and P.-​A. Mumm (eds), Sprachtod
und Sprachgeburt. Bremen: Hempen, 1–​20.
Schrijver, P. (2005). ‘Early Celtic Diphthongization and the Celtic–​Latin Interface’, in J. de Hoz,
E. R. Luján, and P. Sims-​Williams (eds), New Approaches to Celtic Place Names in Ptolemy’s
Geography. Madrid: Ediciones Clásicas, 55–​67.
Schrijver, P. (2007). ‘What Britons Spoke around 400 ad’, in N. J. Higham (ed.), Britons in
Anglo-​Saxon England. Woodbridge: Boydell Press, 165–​171.
Shiel, N. (1975). ‘The Coinage of Carausius as a Source of Vulgar Latin Forms’, Britannia,
6: 147–​149.
Simkin, O. B. (2012). ‘Language Contact in the Pre-​ Roman and Roman Iberian
Peninsula: Direct and Indirect Evidence’, in A. Mullen and P. James (eds), Multilingualism in
the Graeco-​Roman Worlds. Cambridge: Cambridge University Press, 77–​105.
Sims-​Williams, P. (2002). ‘The Five Languages of Wales in the Pre-​Norman Inscriptions’,
Cambrian Medieval Celtic Studies, 44: 1–​36.
Sims-​Williams, P. (2003). The Celtic Inscriptions of Britain: Phonology and Chronology, c.400–​
1200. Oxford: Blackwell.
598

598   Alex Mullen

Smith, C. (1983). ‘Vulgar Latin in Roman Britain: Epigraphic and Other Evidence’, Aufstieg und
Niedergang der romischen Welt 29/​2: 893–​948.
Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. York: Council for British
Archaeology.
Tedeschi, C. (2005). Congeries lapidum:  Iscrizioni Britanniche dei Secoli V–​VII. 2  vols.
Pisa: Scuola normale superiore di Pisa.
Thomas, C. (1981). Christianity in Roman Britain to ad 500. London: Batsford.
Thomas, C. (1991–​ 2). ‘The Early Christian Inscriptions of Southern Scotland’, Scottish
Archaeological Journal, 17: 1–​10.
Tomlin, R. S. O. (1987). ‘Was Ancient British Celtic Ever a Written Language? Two Texts from
Bath’, Bulletin of the Board of Celtic Studies, 34: 18–​25.
Tomlin, R. S. O. (2002). ‘Writing to the Gods in Britain’, in A. E. Cooley (ed.), Becoming Roman,
Writing Latin? Literacy and Epigraphy in the Roman West. Journal of Roman Archaeology
Supplementary Series 48. Portsmouth, RI: Journal of Roman Archaeology, 165–​176.
Tomlin, R. S. O. (2003). ‘“The Girl in Question”: A New Text from Roman London’, Britannia,
34: 41–​51.
Tristram, H. (2007). ‘Why Don’t the English Speak Welsh?’, in N. J. Higham (ed.), Britons in
Anglo-​Saxon England. Woodbridge: Boydell Press, 192–​214.
Ward-​Perkins, B. (2000). ‘Why Did the Anglo-​Saxons not Become More British?’, English
Historical Review, 115: 513–​533.
Wild, J. P. (1976). ‘Loanwords and Roman Expansion in North-​ West Europe’, World
Archaeology, 8/​1: 57–​64.
Williams, J. (2001). ‘Coin Inscriptions and the Origins of Writing in Pre-​Roman Britain’, British
Numismatic Journal, 71: 1–​17.
Williams, J. (2002). ‘Pottery Stamps, Coin Designs, and Writing in Late Iron Age Britain’, in A.
E. Cooley (ed.), Becoming Roman, Writing Latin? Literacy and Epigraphy in the Roman West.
Journal of Roman Archaeology Supplementary Series 48. Portsmouth, RI: Journal of Roman
Archaeology, 135–​149.
Williams, J. (2005). ‘Coinage and Identity in Pre-​Conquest Britain:  50 bc–​ad 50’, in C.
J. Howgego, V. Heuchert, and A. M. Burnett (eds), Coinage and Identity in the Roman
Provinces. Oxford: Oxford University Press, 69–​78.
Williams, J. (2007). ‘New Light on Latin in Pre-​Conquest Britain’, Britannia, 38: 1–​11.
Woolf, G. (1994). ‘Power and the Spread of Writing in the West’, in A. K. Bowman and G.
Woolf (eds), Literacy and Power in the Ancient World. Cambridge: Cambridge University
Press, 84–​98.
 Chapter 29

Art in Rom a n Bri ta i n

Ben Croxford

Introduction

Art in Roman Britain occupies an unfortunate position. It is at once on the edge of

empire and an outlier in terms of production and, as is more usually the focus, quality. In
relation to Roman Britain, art is typically defined as mosaics, wall-​paintings, and sculp-
ture. The artistic output of craftspeople producing utilitarian objects, such as pottery
or items of personal adornment, though often featuring anthropomorphic or zoomor-
phic motifs, is somewhat marginalized. This appears to spring from a division between
functional and non-​functional; objects versus art objects. This, of course, assumes that
art objects have no function, beyond being decorative. This is a fundamental issue that
follows through all later consideration of the pieces in question. Art in Roman Britain
has received, and continues to receive, attention, though often from a purely art histori-
cal perspective. This has certain implications for the way in which the art is portrayed,
understood, and discussed. To this day, the often overriding concern relating to art in
Roman Britain is that of its quality, or rather the lack thereof.

What is Art?

The standard types of art considered in relation to Roman Britain, the mosaics, wall-​
paintings, and sculpture, are all, to a greater or lesser extent, elite products consumed in
small quantities and in highly restricted locations. This is particularly true of mosaics
and wall-​paintings, which have issues of access that will have made their contemporary
visibility low, if not near invisible, for much of the population. Sculpture, therefore, pre-
sents a more appealing prospect, if only because it was more widely available, and can
demonstrably be shown to have had a wider prevalence and impact, occurring across the
province and in a variety of forms. What was probably the most widely distributed and
600

600   Ben Croxford

frequently encountered art object is not often considered in these terms—​namely, coin-
age. Coins were amongst the first objects decorated with Roman artistic tropes to reach
late Iron Age Britain, influencing earliest coinage there (but, importantly, seemingly
only coinage). This process continued within the Roman period as coins were widely
produced and circulated (though by no means universally adopted). Whether delib-
erate choices were made in terms of reverse type has a long history of study, concern-
ing imperial decisions and relations to the provinces (e.g. Reece 1979: 231). Questions
remain concerning how much attention might have been paid to the imagery on the
reverse (anecdotal parallels from modern examples suggest perhaps little). Whether a
connection was made at all between the imagery on coins and other forms of Roman art
is perhaps unknowable, but in any event they feature classical images and were acces-
sible to much of the population.
Beyond coinage and the elite forms already outlined, other art items from Britain
are often niche products, again with elite-focused circulation. Intaglios frequently fea-
ture classical motifs, but, owing to material and function alone, will have had limited
availability to the general population. Decorated pottery, whether imported samian or
similar or locally produced, can feature anthropomorphic scenes and was more widely
available. It, and coinage, therefore probably represent the mainstream of individual
access to art. There are also lost mediums to consider. Archaeological survival of soft
furnishings is limited, but these would probably have featured images drawn from the
classical repertoire—​see, for example, the leather upholstery from Bloomberg Place,
London (Booth 2013: 325–​326). Wooden products are perhaps the most surprising in
their absence. Nothing on the scale of sites found in Gaul, such as those at the source of
the Seine and at Chamalières, has been located yet in Britain (see Aldhouse-​Green 2004
for numerous examples of Gallic wooden pieces). Indeed, there are only a handful of
wooden artistic products, mostly sculpture and some of those rather crude in form and
of uncertain function—​see, for example, the piece from Ickham (Clark 2010). This raises
issues of adoption of the artistic habit in toto.

Distribution of Art

Art is not uniformly employed within Roman Britain (see Figure 29.1). The occurrence
of mosaics and wall-​paintings is essentially determined by the distribution of towns and
villas (see Jones and Mattingly 1990: 241, map 7:6). Likewise, the distribution of coins
and other small personal objects is often subject to modern factors of investigation, par-
ticularly metal detecting. The sculpture of Roman Britain offers a more interesting and
nuanced means of considering where art was distributed. The objects were not restricted
to single site types (though did occur more regularly in towns and villas) and were not as
portable, so have often remained close to their point of use.
Definite concentrations of material are focused on the Cotswolds region and the
areas around the Antonine Wall and Hadrian’s Wall. The impression given is that these
 1
2–3
4–10
11–30
31–50
51–108

0 50 100 200 km

Figure 29.1  Finds per site of anthropomorphic sculpture.

602

602   Ben Croxford

are clearly the main areas for the production and use of sculpture in Roman Britain. In
many respects this comes as little surprise. The two exceptionally large Corpus Signorum
Imperii Romani (CSIR) volumes covering just Hadrian’s Wall (Phillips 1977a; Coulston
and Phillips 1988) obviously indicate that the region was particularly productive where
sculpture was concerned. In a similar fashion, it has long been recognized that the
Cotswolds seemed to be a focus in terms of production, or indeed skill (Henig 1996).
This distribution encompasses the entire period, concealing variations by time. The
dating of sculpture is problematic in many instances, but the general impression is that
the peak of production was spread over the second and third centuries, and apparently
had all but ceased by the early fourth century. There are, however, some concerns relat-
ing to these dating assumptions, particularly given the fourth-​century boom seen in
relation to other art forms—​namely, in mosaic production (Ling 1997; Scott 2000) but
also in silverware (Johns 1990; Hunter and Painter 2013). Other problems with the dat-
ing of art in general relate to stylistic details, which can occasionally apparently conflict
with stratigraphic dating methods. At Kingscote a late-​third-​or early fourth-​century
building has a second-​century wall-​painting, if some interpretations of stylistic ele-
ments are to be believed (Ling 1998: 86). The possibility exists that the fourth century
saw greater production of art objects in general, not just mosaics, but that they have
been incorrectly assigned to earlier centuries.
As with coins in general, it must be remembered that not all art from Roman Britain
was made in or for Britain. A small body of material was of continental production and
only later imported into the island. Furthermore, some, or perhaps much, of the art pro-
duced within Roman Britain was made not by Britons (Romano-​or otherwise) but by
immigrants from elsewhere in the empire. It has essentially been assumed that the ori-
gin of the craftsperson can be inferred from the quality of the piece (Toynbee 1964: 5–​8).
There are distinct problems inherent in such assumptions, especially given the absence
of external confirmation via inscriptions and so on. More importantly, it is problem-
atic simply to assume that anything of quality must have been externally produced. This
leads us from art to questions of its achievements.

A Question of Quality or Taste

Much of the way in which the art of Roman Britain has been considered, and continues
to be considered, can be traced back to the oft-​quoted damning opinion of Collingwood
(Collingwood and Myres 1937: 249–​250). Featuring words such as blundering, stupid,
ugliness, and vulgar, this opinion was so strongly worded that it has resonated ever since,
usually appearing, as here, at the outset of any consideration (e.g. Reece 1977: 1; Millett
1990: 117; Johns 2003a: 29; 2003b: 10; Scott 2003: 1; Aldhouse-​Green 2004: 195). In part
this could be because many struggle to reject this reading, and interest in the subject
matter often assumes a defensive stance, attempting to justify the material (or at least
the application of serious thought to such a collection). Toynbee’s copious work (1962,
 Art in Roman Britain    603

1964, 1965) on Roman art often features examples from Roman Britain or returns to it
despite an empire-​wide scope. This trend continues today, most commonly when the
writer in question is based in Britain (e.g. Stewart 2003, 2008). This is not to suggest that
such interest springs simply from a nationalistic drive, but Henig (1995: 11) defends what
are elsewhere described as shortcomings, highlighting key examples of British produce
to demonstrate that the damning verdict from the 1930s was short-​sighted, and empha-
sizing that the British material is good after all. Importantly, it is on the same aesthetic
terms as the original judgement that this defence is mounted.
More recently, the views of Collingwood are still repeated but new considerations
are put forward, in express opposition to the remarks concerning sticky fingers and
the ghastliness of Romano-​British art; the original insult is simply repeated to demon-
strate intellectual progress and distance from such antiquated views. The rejection of
the Victorian preferences and contemporary influences that coloured Collingwood’s
outlook (Johns 2003a: 29) has enabled new appreciations of the material in question to
be developed, with the modern visual experience no longer of primary and overriding
concern. This trend reached its extreme in the early 2000s, when the entire structure
was inverted to hail what had been failings as successes, as expressions of pre-​Roman
or non-​Roman sensibilities via a very Roman medium as an act of purposeful resist-
ance (Aldhouse-​Green 2003, 2004; Scott 2003; Webster 2003). The greatest pitfall on
any such endeavour, though, remains the simple fact that the ‘distortions’ held to be the
product of ‘Celtic’ sensibilities are more likely to be the result of varying competence
among those doing the carving (Johns 2003a, b). Most recently, the issue of quality has
been redefined as one of provincialism in an attempt to account for variations in artis-
tic output or norms prevalent in the Roman imperial provinces, and with the realiza-
tion that Roman Britain is not exceptional in terms of its art or its quality (e.g. Stewart
2010). What remains apparent, then, is that the most overriding concern is still why art
looks the way it does in Roman Britain. Implicit within that is the worry that it is as
Collingwood described. Indeed, this may be the reason that it is still openly described as
a problem and even as “disappointing’ (Stewart 2010: 155–​156), perhaps supporting the
accusation that Romanists still do not like the art of Britain (Aldhouse-​Green 2004: 13).

What is Quality Then?

Defining quality often requires a departure into aesthetics, the field from which the initial
damnation sprang (Reece 1977). Attempting to avoid the modern sensibilities that inevi-
tably manifest themselves when assessing artistic output is difficult but can be achieved
via a focus upon competence in terms of physical achievement (see Johns 2003a, b). In
this sense, each piece of artistic production known from Roman Britain is successful as
it represents the adoption of the practice and the completion of a piece executed in what
are essentially alien techniques and often with alien materials (when compared with the
late Iron Age). Qualifying and quantifying the level of this success are more difficult, but,
604

604   Ben Croxford

accepting classical norms as the intended result, it is possible to create a series of gradu-
ated classes of artistic achievement in relation to sculpture. This process is fraught with
issues, not least of which is the assumption that a classical form was the intended result
(cf. Lindgren 1980: 11), but provides a way into the question of quality, allowing variations
to be defined and explored. Material from Britain may be divided into five classes, based
upon the execution and the composition of the image, rising from the simplest examples
to the well-​finished and classical pieces. Divisions within the system are typically based
upon whether there are any items of classical iconography such as paraphernalia of par-
ticular deities—​the caduceus, for example—​but also take account of basic carving tech-
nique. Class 1 sculpture is that which lacks any classical imagery and Class 5 pieces fully
conform to the norms of Roman art and are well executed. This system has the advan-
tage of being largely self-​contained, acknowledging the classical origin or intention but
not necessarily evaluating each piece against something on the level of the Ara Pacis, for
instance, against which nearly all would ‘fail’—​perhaps the origin of Collingwood’s dis-
like. This class system outlined here is an attempt to evaluate the material on its own mer-
its within its context as a provincial product.
The sculpture of Roman Britain falls into the five classes in a standard population
curve, centred around Class 3. There is a certain logic to the distribution at the upper
ends of the scale, as these images would have required additional capital and rarer skills,
thus making them less common generally. Were pieces at the upper end of the scale
to be the most common, this would have had interesting implications for the produc-
tion and use of art. What is less expected is the apparent rarity of the simplest image
forms. There may be a problem springing from assuming that Class 3 is a norm, and
thus evaluating all pieces against this, creating small sets of outliers above and below. It
is possible, then, that the extremes of Class 3 are actually grey areas and that more pieces
should fall into the adjacent classes. If this is the case, however, it makes no substan-
tive difference to the general pattern. In particular, it would not increase the number of
Class 1 sculptures. It would appear, based upon the frequency of recognizable classical
motifs, that a classical image (or at least subject) was the intended result in the major-
ity of instances. This is interesting for two principal reasons. Firstly, it suggests that the
art form is not being adopted and re-​purposed, in that it is overwhelmingly not used to
display non-​or pre-​Roman images. Secondly, it suggests that the adoption of sculpting
as a habit was limited. The more basic sculpture, particularly the Class 1 type, which
requires considerably less capital and skill to create, should have no impediment to pro-
duction. Individuals with minimal skill or experience and limited time may produce the
more simple carvings relatively easily. A modern attempt to carve an imitation of a fairly
typical Romano-​British style of carving produced a good Class 1 piece in about an hour
(Guy de la Bédoyère, pers. comm.). The availability of raw material and access to tools
may have been a contributing factor, but the rarity of such pieces cannot be completely
accounted for by this means.
As shown in Figure 29.1, the distribution of sculpture is not even across Britain. There
are clear zones of use, and within these there is variation in relation to the distribution of
the quality classes (Figure 29.2). Class 1 and 2 sculptures are largely concentrated around
 Class 1

Class 2

Class 3

Class 4

Class 5

0 40 80 160 km

Figure 29.2  Zoning of sculpture quality.

606

606   Ben Croxford

the greatest areas of general production. It would appear, therefore, that, in these areas,
a greater use of sculpture does transfer to a greater incidence of lower-​quality pieces
in general. The low numbers of such pieces generally is therefore a product of the low
uptake of the sculptural habit in the rest of Britain. At the other extreme, Class 4 and
5 sculptures are largely restricted to the core of the Cotswold area, London, and a few
isolated cases across the south-​east, particularly so with Class 5 pieces. Outside of these
areas the driving force of the military presence is apparent. Indeed, comparing mili-
tary tombstones with civilian ones shows a greater frequency of Class 3, 4, and 5 pieces
among military contexts (though the subject is not always directly from the military
themselves). The presence of Class 1 and 2 objects within the military areas in the north,
and the restriction of production of any class of sculpture to this immediate zone, sug-
gests that poor quality does not equate to a non-​Roman context of production, or at least
display. It would appear instead that production is stimulated by the presence of other
pieces, a general sculptural habit, but all operating within a Roman setting and not sig-
nificantly spreading beyond this.
Suitable stone is not uniformly available across Britain, and it has long been noted
that the Jurassic limestone ridge that includes the Cotswolds was a focus for produc-
tion, with the implication being that it is the raw material itself that promotes manu-
facture (Henig 1993: p. xiii; 1996). It has, however, more recently become apparent that
material was moved, often considerable distances (Hayward 2006a, b). Furthermore,
though limited in number, the occurrence of sculpture in other materials still repro-
duces the same patterning observed for all sculpture. This is particularly noticeable for
bronze sculpture, which, despite limited overall numbers (Croxford, in press), still sees
greatest use in those centres already identified as key users of sculpture in general. Such
materialistic explanations also fail to account for the concentrations along the military
installations of the north, none of which is located on the Jurassic ridge. It is, therefore,
too simplistic to suggest that material alone is the driving force behind the use or non-​
use of sculpture.
The conformity evident in the general artistic output raises interesting questions
concerning apparent restrictions on the use of sculpture. On the whole, only classical
subjects are depicted, and there is little innovation or re-​purposing of the medium. If
resistance was being expressed through the use of a very Roman medium but conform-
ing to native sensibilities, this never extended to a general use to depict non-​Roman sub-
jects and was common to other western provinces. Instead, it suggests that the artistic
habit was restricted in practice, being employed only in certain circumstances and by
particular people.

Who was Making Art?

The question of who was actually producing the art has a simplistic answer—​namely,
that it must have been artists. This belies the complexity of the artists as individuals,
 Art in Roman Britain    607

specifically whether they were skilled dedicated craftspeople or jobbing amateurs pro-
ducing pieces as and when time or demand permitted. It also fails to acknowledge that
we have a dearth of evidence relating to the producers of art, from evidence of where
they worked to more basic details such as who they were. There are few names that can
confidently be advanced as artists, mostly because of the rarity of inscriptions associated
with art in general, let alone claims of authorship. The problem then is that for the most
part we have only their product to work from. It is unwise to make assumptions con-
cerning the origin of the individuals who were making art based purely on the quality
of their output (cf. Toynbee 1964: 8; Phillips 1977b), especially given the general rarity of
sculpture in Britain. The presence of unique pieces (well-​executed examples in particu-
lar) does not automatically indicate the work of an itinerant craftsperson arriving from
elsewhere in the empire and producing a piece for a specific commission, before disap-
pearing back again. The frequency of such unique pieces means it is not easily possible
to identify workshops or individuals working across areas, where sculpture is concerned
(cf. Phillips 1976; Henig 1996, 2012; Pearson 2006: 68). Schools of mosaicists have been
identified for Britain, operating within geographic and chronological limits (Ling 1997:
264–​265; Scott 2000). How these were established and how significant hiatuses in activ-
ity were bridged are less clearly understood. A reliance upon pattern or copy books is
sometimes supposed, even with the suggestion that the later resumption of production
entailed selection from such books by patrons not best equipped to know what they were
selecting and artists not fully able to deliver (Ling 1997: 281–​282). The complexity of the
transmission of the image ideal is often overlooked, usually at the expense of the success
or failure of the individual artist to have faithfully reproduced the classical norms from a
thousand miles away. Pattern books are commonly suggested as a likely explanation for
image types repeated (with varying degrees of success) across the provinces (Toynbee
1964: 10–​11). Some familiarity with the norms of display, for Mercury for example, evi-
dently must have existed, though was perhaps more likely to have been provided by the
existence of small portable figurines (see Durham 2012). The exact thought processes
behind any image’s origin are unclear in most instances. Rare cases, such as the temple at
Uley (Woodward and Leach 1993), do allow some speculation on this front. The repeti-
tion of the same image type (Mercury with attendant animals) across what is presumed to
be the main cult image and a number of altars suggests that later images were based upon
the first, producing a uniform collection. Determining which was the first and what con-
trols were in place to ensure this conformity (or whether it was a natural progression) is
impossible. Ultimately all that can be said is that the same image type was reproduced by
an unknown number of artists and dedicants over an unknown period.
There are three principal possible structures under which art may have been pro-
duced: a patron and artist dynamic, an artist producing stock images, and self-​made
products. The first of these, that of a direct patron and artist relationship (or the buyer
and the maker as relabelled by Johns (2003a)), is the most commonly imagined struc-
ture for the production of art in Britain (Henig 1995). The patron approaches an artist
and commissions a specific piece, dictating requirements and expectations of the out-
put. The artist influences the outcome due to their predilections and competence, but it
608

608   Ben Croxford

is the patron who is thought to have the greatest input. There are a number of problems
with such a scenario, mostly because it assumes the existence of dedicated artists open to
commission, despite the general rarity of sculpture from Britain. It also fails to account
for the quality issue, most often having to fall back onto ideas of the patron being una-
ble to appreciate art sufficiently to reject the ‘substandard’ product. The counter to this
argument is that the patron settles for a poor product because they are aware of their
limited options and makes do with what can be had, as, ‘unlike cheese, art cannot be had
on demand’ (Henig 1995: 107). This also leads to ideas that the artists were less skilled
than perhaps they should have been, but were trying to fill a market demand as best they
could. This raises issues of how much of a market economy Roman Britain truly was,
and also only really works best in the military north, where it is imagined soldiers had a
pressing need for art and drew upon their own pool of ability to supply it.
The second scenario, that artists worked to produce a stock of images, is less commonly
advanced but is actually well supported by a few examples that suggest it may have been
more common. Identifying this particular dynamic is largely reliant upon tombstones.
There are tombstones with images of adults raised over the graves of children—​for exam-
ple, the tombstone of Vacia from Carlisle, CSIR 1.6 495, RIB 961. Such instances suggest
the existence of workshops with stock ready prepared that could be purchased in a near
complete form, requiring only an inscription to personalize it. It also suggests that the
range of stock was limited, as an ‘appropriate’ image was not selected. It is impossible to
know the precise chain of events or decisions that led to the mismatch between subject
and image. It cannot be ruled out that the desire was to show the child as the adult they
could have become. If artists were working to produce a range of images from which cus-
tomers could select, the patron is reduced to being simply a buyer, faced with a restricted
range of ready-​made pieces from which the best-​fit selection must be made. In these cir-
cumstances, it is the artist, and perhaps market forces, that determine the number and the
types of image made, as well as the quality. Related to this possible mode of production
is the importation of finished works. In this scenario the patron probably has little input,
if any, into the form of the image. It is unclear how imports were handled, whether they
were commissioned remotely, made and moved on spec in the hope of finding a buyer,
or moved as private possessions. Collections such as the material from the Walbrook
Mithraeum in London (Toynbee 1986) or the Lullingstone busts (Toynbee 1951) could
have been personal possessions brought to Britain by an individual or group. The iso-
lated occurrence of Class 5 pieces such as the Lullingstone busts fits a possible situation
where the market for art is extremely limited and only exceptional pieces are acquired.
These then are displayed only in closed settings and do not trigger a wider adoption of the
habit, a situation that appears to fit well in Kent, where there is an abundance of villas but
extremely little in the way of sculpture or indeed art of any form.
The third scenario is one of self-​manufacture. Here the artist is also the patron, making
the piece for their own needs. This particular means of production probably accounts for
the Class 1 and possibly Class 2 pieces, though is again assuming that only an incompetent
sculptor working for themselves would settle for the products in question. There are only a
small number of inscriptions mentioning a name in relation to the production (as opposed
 Art in Roman Britain    609

to the dedication) of a monument or sculpture (Johns 2003a: 33, 35). The problem in part
arises from the Latin used to indicate construction. F(ecit) and F(aciendum) C(urauit) are
the two most common elements appearing in inscriptions that relate to the actual manu-
facture of the piece. The distinction between fecit (made) and faciendum curauit (had this
made) is subtle, and there are clear instances where, though fecit is stated, it is unlikely
to mean ‘made’ in the sense of personal use of hammer and chisel. RIB 233, for instance,
appears to indicate that the deceased made the memorial during his lifetime. It is unlikely
in the extreme, however, that the strator consularis would carve something personally,
let alone the monumental tomb to which this slab appears to have belonged, discount-
ing the possibility of a very unusual hobby. The abbreviation and instances of apparent
confusion (RIB 796 with FE, perhaps intended to be F C or a less common abbreviation of
fecit) serve to further this problem. It is only in instances in which two names are recorded,
or explicit mention of a sculptor is made, that we may say with confidence that a sculp-
tor is employed (e.g. RIB 132, 151, 274). This is not to presume that personal manufacture
was therefore the common means of production, only that the epigraphic habit—​though
sharing a similar distribution to that of sculpture (see Jones and Mattingly 1990: 152 map
5:10)—​was not automatically linked to the production or use of sculpture.
However the art was produced, the consistency of output strongly suggests that the
work must have been acceptable to whomever was immediately responsible for its crea-
tion and consumption. Furthermore, this acceptance must have included the subsequent
public exposure of the piece. Given the social role of art, there must have been external
pressures on the patron or dedicant where the quality of the piece was concerned. Much
can be made of the contemporary account of an informed observer from the centre of
the empire encountering such provincial output: Arrian’s reaction to poorly executed
imagery was to have it replaced (Periplus Ponti Euxini 1–​2). What is important, though,
is that the sculpture had been erected in the first place, evidently meeting the local stand-
ards. The problem comes when attempting to distinguish between acceptance because
of an inability to know better or because of limited resources dictating what could be
achieved (Millett 1990: 117; Ling 1997: 281–​282). It has been stated that there was a gen-
eral shortage of professionally trained sculptors in the military zone of Britain, with the
partial exception of the legionary bases, but that, even though the resulting product was
‘inept’, their very production is evidence of an aspirational motivation (Mattingly 2006:
210–​211): the need to erect sculpture outweighed the problems of its quality. What this
need was and how strong it was, given the rarity of sculpture for the rest of Britain, is
uncertain. This moves us from issues of who was physically producing the art to ques-
tions of what function it served and why the art was made in the first place.

Who was Using Art and How?

The ‘use of art’ has two principal meanings: there is the functional application of
pieces (where they were displayed, what they were used for) and there are the social
610

610   Ben Croxford

implications of their use (the message that was transmitted by them and by their very
use). The second element is often inextricably linked with the first, determining why
somebody would employ art in the first place. Leaving aside the why and who initially,
we shall instead focus on the how. A great deal of art in Roman Britain can be simplisti-
cally described as decorative. At the most basic level the function of a mosaic is as a floor,
but deliberate choices are made in selecting a mosaic floor over a plain (and cheaper)
alternative. The mosaic serves to decorate the space it is installed in. The same is true of
wall-​paintings. Sculpture is somehow different, in that it alone is commonly perceived
as having some form of function beyond simply decorating a space (though see Perring
2003 for meaning in mosaics). Outside of considerations of art, sculpture appears most
frequently in considerations of Romano-​British religion, illustrating deities and types
of dedication. Though only c. 9 per cent of material comes from identifiably religious
sites (Croxford 2008: 77), the subjects, excepting tombstones, are predominantly divine
in nature. As such, religious sculpture forms the largest identified functional type and
principal use of statues, reliefs, and altars in Britain. Tombstones may be added to this
group, equating a religious or devotional meaning to funerary commemoration. This
overly religious nature of sculpture is perhaps unusual for Britain, as there are far fewer
decorative or portrait pieces compared with other provinces. Among what little there is,
by and large, the majority are of emperors, presumably public art pieces that once graced
official settings such as the fora of Britain. It is an unusual quirk of preservation that
means particular emperors are better represented among surviving material, specifi-
cally Nero (see Croxford, in press). There is a problem in attempting to classify imperial
images as simply portraits, as they are also public monuments and potential religious
items at the same time. They remain uncommon though and it is interesting that, even
with this particular type of monument, the sculptural habit does not appear to be have
been strongly embraced in Britain. This has been linked to a reluctance on the part of
the British elite to engage in euergetism, also explaining the limited numbers of public
buildings in much of Britain (Stewart 2010: 24).
To return to the religious use then, the assumption that sculpture is in the most part
intended for religious use is usually made just on the basis of the image portrayed (cf.
Henig 1993: p. xvi). This is obviously ill-​advised, and it is highly probable that a greater
number of pieces served decorative functions but that their iconography and the con-
straints of archaeological interpretation render them less obvious. The material from
Woodchester villa (Clarke 1982) is one such example that should perhaps be thought of
as decorative. It is perhaps closer to modern popular understandings of what constitutes
‘art’ given the quality of the pieces and some of the subjects. CSIR 1.7 2, for instance,
the statue of Cupid and Psyche, is less likely to have attracted ‘straightforward’ religious
devotion because of its mythical and allegorical nature; it may have been an art piece.
In this respect, the classical imagery embodied by sculpture such as the Woodchester
marbles may have played a role similar to such pieces from houses of the nobility in
Britain from very much later. A more direct comparison may be made with continental
sites where the term villa is used more confidently. The Villa of the Papyri demonstrates
the ideal version of a villa with sculpture—​a large assemblage featuring many pieces
 Art in Roman Britain    611

clearly of decorative function (Mattusch 2005). The rarity of such instances in Britain
raises questions concerning the definition of villa and the role of sculpture in relation
to these sites. Were decorative sculpture to be regarded as a defining character of a villa,
the distribution of such sites in Britain would be dramatically reduced. Attempting to
detect art objects among the assemblage is falling into a single-​purpose mode of thought
where the contemporary use or interpretation of sculpture is concerned. The fragments
of marble and bronze from the temple site at Maiden Castle (CSIR 1.2 98 and Wheeler
1943: 75) may relate to art pieces initially displayed in a well-​appointed town house in
nearby Dorchester. They were clearly late arrivals on the site (the buildings being con-
structed only in the fourth century) and so were perhaps converted into religious objects
late in their life. The marble Bacchus from Spoonley Wood (CSIR 1.7 1) was found in a
grave and may likewise have been initially an art object before becoming a unique grave
good (Hutchinson 1986: 157). This raises the issue of change over time, and, more impor-
tantly, the longevity of art objects.
The questions of who was using art and why are even more complex. The interplays
of desires, abilities, and capital are a fascinating aspect of the lives of sculpture, deter-
mining the very form and existence of the material. They are, however, more easily
explored in other more sculpture-​rich areas, particularly in the Mediterranean area,
where the wealth of available sources and evidence enables more detailed understand-
ings of the role that art fulfilled as well as the factors that shaped it (Stewart 2003,
2010). For Roman Britain, we have fewer examples and it was less readily employed
in general. It has been suggested that it was a conscious decision to produce sculp-
ture as a means of displaying Romanitas (Millett 1990: 116–​117). More recent consid-
eration suggests that it was being employed by particular social groups in order to
demonstrate their status, or rather to project an image of confidence and acceptance
perhaps beyond the reality. This is particularly noted with tombstones, which appear
to have been employed by the non-​citizen soldiery and their associated communi-
ties in greater numbers and to have drawn upon particular motifs at certain times
(Hope 1997: 255–​257). The use in religious settings certainly suggests attempts at emu-
lation of other provincial settings, given the presumed aniconic nature of late Iron
Age Britain. More generally, the use of art, particularly in elite settings, could well
have been a simple message of conspicuous consumption of an expensive commod-
ity. It could also have been more nuanced, intended to be read by other members of
the educated elite, to demonstrate learning. The use of literary references (Ovid and
Virgil in mosaics and wall-​paintings, from Lullingstone (RIB 2448.6) and Otford vil-
las (British Museum. 1928,1011.1) respectively) has often been thought of as a very
deliberate display of learning intended to signal status of the owner (Ling 2007: 64–​65,
76–7​9, 84). They are, though, rare exceptions, in their use of Latin, their execution,
and their arena of display. The rarity of examples to draw upon, and a modern famili-
arity with the media employed, can lead to confidence in readings that might not have
been common or possible at the time. More complex readings of artistic motifs or
inscriptions in mosaics, for example, are perhaps over-​reaching, attempting to find
hidden meanings that might well not be there.
612

612   Ben Croxford

Understanding Art

Recognizing the subject of images is an area fraught with problems, not aided by the
habit of some scholars, when presented with an unknown piece, confidently to name any
given deity or historical personage. For instance, many of the pieces discovered prior to
the nineteenth century have had a great many different interpretations applied to them
as the study of the subject developed over the years. This is a process still occurring
today: the head of ‘Claudius’ from the River Alde (CSIR 1.8 23) has recently undergone
just such a transformation in interpretation and is now thought to be of Nero (Russell
2006: 44), though opinion is still divided (see correspondence in British Archaeology 90
and 91 from Henig, Plouviez, and Twinch). Some pieces have had an eventful life since
discovery, with numerous interpretations applied to them over the years (in the case of
the head from the Alde, one of the first considerations accused the Vikings of involve-
ment (Athenaeum, 1908: 766)). Aside from these issues, which are common to much of
academia, consisting of constant reappraisal of past identifications and the development
of new views concerning old pieces, there are more complex issues relating to the iden-
tification of subject.
The repetition of characteristics and conventions of display permit narrow typolo-
gies of image type to be generated (see Lindgren 1980: 3–​4); particularly obvious are the
features of Mercury, for example, in his usual garb and with his well-​known equipment
and/​or companions. There is an over-​reliance on, and false confidence in, these sym-
bols and the ability of modern observers to read them correctly in every instance. The
problem in part arises because of a lack of available confirmation in the form of inscrip-
tions, but there has been a failure in the past to observe the conflict between modern
identifications and those key pieces that do feature inscriptions. The knowledge that
Minerva is depicted with aegis, shield, spear, and helmet leads to an automatic assump-
tion that all depictions of a female subject with these accoutrements (or some of them
at least) must therefore be Minerva. CSIR 1.4 12, however, is identified by an inscrip-
tion (RIB 2091) as Brigantia. The depicted female figure possesses attributes of Minerva,
Minerva–​Victoria or Victory, and a city Tyche. A more recent example is the Senuna
treasure from near Ashwell in Hertfordshire. Among the images there are pieces that
are ‘clearly’ of Minerva, and yet the subject is fortunately named as Senuna, an other-
wise unattested deity (Tomlin and Hassall 2005). In these instances it is a British deity
depicted using essentially unmodified image types. This phenomenon is not restricted
to simple instances of borrowing image types to depict new deities.
CSIR 1.7 40 and CSIR 1.7 60 are a pair of similar reliefs from Bisley, both depicting
standing male figures, one in a long garment with a prominent belt buckle and the other
armed and armoured. The first figure would, under other circumstances, be labelled as
a Genius owing to its garb, attitude, and the large double-​cornucopia that it holds. The
armoured figure conforms to the Mars-​type image and indeed has been identified as
such (Lindgren 1980: 112). There are, however, inscriptions on both: the first figure is
 Art in Roman Britain    613

identified as Mars Olludius (RIB 131) while the second is Romulus (RIB 132). The pres-
ence of these two deities at Bisley is unusual in its own right (see Stewart 2000) but is
secondary to the importance of the discrepancy between the interpretation based upon
image type and that provided by the inscriptions. The armed and armoured figure (CSIR
1.7 60) would under no circumstances be described as Romulus in the absence of the
inscription, despite the fact that he shares the iconography of his father. This point is
amply demonstrated, as in no other instance is Romulus advanced as the subject of any
of the many Mars-​type sculptures in Roman Britain. With the other image (CSIR 1.7 40)
this message is reiterated. It is unlikely that this Genius-​type image would, were it not
for the inscription, be identified as Mars, let alone Mars Olludius. In the absence of an
inscription or a genuine confidence encouraged by the abundant presence of identifying
objects specific to one deity and one deity alone, and even in such cases, images should
be hesitantly identified.
This issue of the modern ability to identify intended subjects largely relates to the
specific use made of a particular type of art—​that is, religious art used to inform about
Romano-​British religious practices. It has wider implications though and potentially
also occurred in the past. There are some examples where it would appear that the
iconographic message was either ignored or illegible to some contemporary viewers.
Again, the temple at Uley offers an excellent insight into this possibility. There, despite
the apparent abundance of images of Mercury, there are offerings that fall outside the
expected range for that deity and, more explicitly, seemingly some confusion on the part
of dedicants in written prayers as to whom they should be addressing at this apparently
unequivocal temple of Mercury (Tomlin 1993).
Contemporary viewers will have had a number of additional clues to aid an identifi-
cation that are no longer available. Foremost among these must have been verbal com-
munication; simply another person telling the viewer what was intended. This would
rely upon the informer being correctly informed to begin with, a problem that will no
doubt have increased with time. It also assumes an equality between the two parties, and
between these individuals and the original artist/​patron. Discrepant experience of art is
a somewhat overlooked phenomenon. The restricted spaces in which it was displayed or
circulated are probably the main cause of discrepant experience, but it must be stressed
that not all observers will have approached art from the same position. Levels of classi-
cal education will have determined familiarity with the subject. The literary references
associated with the art of the Lullingstone and Otford villas must only have been appar-
ent to a small subset of observers. More generally, there is the issue of basic familiarity
with art in general, particularly large-​scale anthropomorphic pieces.
There is a tendency to compress the timescale in relation to the Roman period in
Britain. Situations that may have existed at the moment of conquest (unfamiliarity with
Roman cultural norms, practices, or products) are unthinkingly allowed to apply at all
periods of the occupation, some 360 years in reality (not counting the century of contact
before conquest). The first-​century rural population would have differed greatly (if not
always in material terms) from the late Roman population (cf. Taylor 2013). Exposure
to art forms and products will only have been truly alien at the initial period of contact.
614

614   Ben Croxford

Familiarity will have been a norm from perhaps the second generation onwards, thanks
to publicly displayed pieces and the availability of mass-​produced items displaying such
imagery. The shock of the new engendered by first contact with an unfamiliar medium
can have occurred only at the earliest stages of the Roman occupation or on an individ-
ual basis, probably only early on in life, perhaps on the occasion of a first visit to a town
or temple. Despite some indirect assertions that the late Roman population would have
been unable to ‘read’ a sculpture and discern its essential function or nature (see Roach
Smith 1859: 71), we must assume a passing familiarity with art in one form or another for
much of the population of Roman Britain. How frequently they could access it, and the
likelihood of them owning, let alone commissioning such, is another issue. Divergent
readings and interactions with art, along with discrepant experience, are likely only to
have increased with the passage of time. The chronological distance from the moment
of creation introduces the same problems already outlined in relation to interpretation
and, for contemporary observers, issues of relevance. Just as the timescale in general is
compressed, so too is it often forgotten that art, particularly hardy forms such as sculp-
ture, continue long after the moment of creation.
Sculpture experienced an initially slow take-​off, with few pieces produced in the first
century. In any event, deposition of sculpture, though starting from the first century, by
no means kept pace with production. This resulted in a growing population of sculp-
ture, one that would have begun noticeably to decrease only towards the very end of the
Roman period. As such, material produced generations before would have remained
on display. All parties to its creation would have long since passed, and there is no guar-
antee that the ‘correct’ interpretation would have survived this process. The life of art
is very much more than the complex interplay of factors surrounding its creation. It is
becoming apparent that art objects were physical objects and that they enjoyed complex
lives. These lives were often highly dependent upon the material of their production but
were ultimately connected with their physicality. The late Roman and late antique inter-
action with sculpture in particular has received a great deal of interest in recent years
(see Croxford 2003, 2008; Myrup Kristensen and Stirling, forthcoming). How much of
this interaction, particularly the destructive elements, is as a result of the longevity of
the pieces in question is not always apparent. The relevance of art objects to later observ-
ers cannot be determined and response can often be informed by new and unexpected
motivations.

Art in Roman Britain

There is a great deal more to art in Roman Britain that meets the eye. The product is
varied and the use more so. This variety has in the past been somewhat obscured by the
focus being fixed upon the quality of the vast majority of pieces. More recently this has
switched to an emphasis upon particular examples at either end of the spectrum of ‘bad-
ness’ for similar but opposed reasons. The truth probably lies somewhere between these
 Art in Roman Britain    615

extremes and is best helped by the realization that the art of Roman Britain is not excep-
tional in terms of quality but is actually ‘normal’ for provincial products. Quality is often
in the eye of the beholder—​and that person is unthinkingly too frequently allowed to be
the modern observer. Art must have passed the standards of the day, as it was displayed,
seemingly without consequence (though a ‘more-​educated’ viewer might have reacted to
it quite differently from the ‘provincials’). The greater interest of the material is actually
to be found beyond the initial modern aesthetic reaction to the pieces. Seemingly simple
questions such as who was making it are surprisingly difficult to answer and more com-
plex questions concerning why they were doing so are often unanswerable. It is possible
to make some assumptions concerning the function of some pieces and the probable ori-
gins of those responsible, most notably in relation to military zone products, especially
the tombstones. Elsewhere it is less apparent, but what can be said is that quality alone is
not sufficient evidence to attempt to guess the likely origin of the individual who made
the piece, or the one who commissioned or used it. The uses to which art objects were put
are complex, and it must be remembered that the use-​life of a piece extended well beyond
the initial stage of creation. The dynamics and intentions of that moment, along with the
skills and materials available, are fundamental to the form of the object but are not the be-​
all and end-​all of the piece. Once created, the object almost takes on a life of its own. Its
display, assuming public access, will have permitted discrepant readings. The longevity
of some pieces must have afforded ample opportunities for the loss of the initial intention
and plenty of scope for reinterpretation. Indeed, the modern experience of such pieces is
simply another stage on this continuum; the pieces being found today are recognized as
art and displayed as such, often at the expense of the social object that the piece must have
been. They are also frequently used as illustration, as images to flesh out our presenta-
tion of the Roman past. This is most prevalent in relation to divine subjects, which it is
perhaps too commonly assumed most are. There needs to be greater caution in modern
attempts to read such pieces, particularly where the application of firm identities to indi-
vidual pieces is concerned. We have enough examples now to show that classical ‘normal’
image types were being appropriated and used in an unmodified form but to represent a
range of otherwise unattested deities. A Minerva-​type or Mercury-​type sculpture is just
that—​the image may be recognizable to the modern observer but the loss of the associ-
ated name makes us unable fully to understand the subject in too many instances. This
is not to say there are no images of Minerva or Mercury from Roman Britain—​simply
that among them there are now anonymized examples of other deities dressed in seem-
ingly easily read image types. Overall, it is most important to realize that the adoption of
Roman art techniques or habits are not universal: there are clear regions of production
and/​or use. Some of these do relate to military zones of occupation, with little or no pen-
etration beyond these bubbles into the wider population around them. Others seemingly
occur in relation to available resources. But this alone cannot account for the wholesale
adoption of the custom. It is likely that particular social forces were absent or stronger
in some parts of the province, and it is these that drove production. The material from
Roman Britain has much more to offer than illustrative examples of various deities wor-
shipped, and is considerably more than just bad.
616

616   Ben Croxford

Abbreviations
CSIR  Corpus Signorum Imperii Romani
RIB   . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I:
R
Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date

References
Aldhouse-​ Green, M. J. (2003). ‘Alternative Iconographies:  Metaphors of Resistance in
Romano-​British Cult-​Imagery’, in P. Noelke, F. Naumann-​Steckner, and B. Schneider
(eds), Romanisation und Resistenz in Plastik, Architektur und Inschriften der Provinzen des
Imperium Romanum: Neue Funde und Forschungen. Mainz: Zabern, 39–​48.
Aldhouse-​Green, M. J. (2004). An Archaeology of Images: Iconology and Cosmology in Iron Age
and Roman Europe. London: Routledge.
Athenaeum No. 4223, 12 December 1908.
Booth, P. (2013). ‘Roman Britain in 2012: 7. Greater London’, Britannia, 44: 324–​333.
British Museum. Object No. 1928,1011.1. <http://​www.britishmuseum.org/​research/​collec-
tion_​online/​collection_​object_​details.aspx?objectId=1363000&partId=1&searchText=BCB
70859&page=1> (accessed 16 August 2015).
Clark, P. (2010). ‘Figurine’, in P. Bennett, I. Riddler, and C. Sparey-​Green (eds), The Roman
Watermills and Settlement at Ickham, Kent. The Archaeology of Canterbury New Series 5.
Canterbury: Canterbury Archaeological Trust, 215–​217.
Clarke, G. (1982). ‘The Roman Villa at Woodchester’, Britannia, 13: 197–​228.
Collingwood, R. G., and Myres, J. N. L. (1937). Roman Britain and the English Settlements. 2nd
edn. Oxford: Clarendon Press.
Coulston, J. C., and Phillips, E. J. (1988). Corpus Signorum Imperii Romani, Great Britain Vol. 1
fasc. 6: Hadrian’s Wall West of the North Tyne, and Carlisle. Oxford: Published for the British
Academy by Oxford University Press.
Croxford, B. (2003). ‘Iconoclasm in Roman Britain?’, Britannia, 34: 81–​95.
Croxford, B. (2008). ‘Sculpture from Roman Britain:  Aspects of Origin, Use, Disuse and
Deposition’. Unpublished Ph.D thesis, University of Cambridge.
Croxford, B. (in press). ‘Metal Sculpture from Roman Britain: Scraps but not Always Scrap’,
in T. Myrup Kristensen and L. Stirling (eds), The Afterlife of Roman Sculpture: Late Antique
Responses and Practices.
Durham, E. (2012). ‘Depicting the Gods: Metal Figurines in Roman Britain’, Internet
Archaeology, 31 <http://​dx.doi.org/​10.11141/​ia.31.2> (accessed 16 August 2015).
Hayward, K. M. J. (2006a). ‘The Early Development of the Roman Freestone Industry in South-​
Central England: A Geological Characterisation Study of Roman Funerary Monuments and
Monumental Architecture’. Unpublished Ph.D. thesis, University of Reading.
Hayward, K. M. J. (2006b). ‘A Geological Link between the Facilis Monument at Colchester
and First Century Army Tombstones from the Rhineland Frontier’, Britannia, 37: 359–​363.
Henig, M. (1993). Corpus Signorum Imperii Romani, Great Britain Vol.1 fasc. 7: The Cotswold
Region, with Devon and Cornwall. Oxford: Published for the British Academy by the Oxford
University Press.
Henig, M. (1995). The Art of Roman Britain. London: Batsford.
 Art in Roman Britain    617

Henig, M. (1996). ‘Sculptors from the West in Roman London’, in J. Bird, M. Hassall, and
H. Sheldon (eds), Interpreting Roman London:  Papers in Memory of Hugh Chapman.
Oxford: Oxbow Books, 97–​103.
Henig, M. (2012). ‘Workshops, Artists and Patrons in Roman Britain’, in T. Myrup Kristensen
and B. Poulson (eds), Ateliers and Artisans in Roman Archaeology. Journal of Roman
Archaeology Supplementary Series, 92: 113–​128.
Hope, V. M. (1997). ‘Words and Pictures: The Interpretation of Romano-​British Tombstones’,
Britannia, 28: 245–​258.
Hunter, F., and Painter, K. (2013). Late Roman Silver:  The Traprain Treasure in Context.
Edinburgh: Society of Antiquaries of Scotland.
Hutchinson, V. J. (1986). Bacchus in Roman Britain:  The Evidence for his Cult vol. (i), BAR
British Series 151 (i). Oxford: British Archaeological Reports.
Johns, C. (1990). ‘Research on Roman Silver Plate’, Journal of Roman Archaeology, 3: 28–​43.
Johns, C. (2003a). ‘Romano-​ British Sculpture:  Intention and Execution’, in P. Noelke, F.
Naumann-​Steckner, and B. Schneider (eds), Romanisation und Resistenz in Plastik, Architektur
und Inschriften der Provinzen des Imperium Romanum. Neue Funde und Forschungen.
Mainz: Zabern, 27–​38.
Johns, C. (2003b). ‘Art, Romanisation, and Competence’, in S. Scott and J. Webster (eds),
Roman Imperialism and Provincial Art. Cambridge: Cambridge University Press, 9–​23.
Jones, B., and Mattingly, D. (1990). An Atlas of Roman Britain. Oxford: Basil Blackwell.
Lindgren, C. (1980). Classical Art Forms and Celtic Mutations: Figural Art in Roman Britain.
Park Ridge, NJ: Noyes.
Ling, R. (1997). ‘Mosaics in Roman Britain: Discoveries and Research since 1945’, Britannia,
28: 259–​295.
Ling, R. (1998). ‘The Wall Plaster’, in J. R. Timby, Excavations at Kingscote and Wycomb,
Gloucestershire:  A  Roman Estate Centre and Small Town in the Cotswolds with Notes on
Related Settlements. Cirencester: Cotswold Archaeological Trust, 77–​87.
Ling, R. (2007). ‘Inscriptions on Romano-​British Mosaics and Wall-​Paintings’, Britannia, 38: 63–​91.
Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409.
London: Allen Lane, Penguin Books.
Mattusch, C. C. (2005). The Villa dei Papiri at Herculaneum: Life and Afterlife of a Sculpture
Collection. Los Angeles: J. Paul Getty Museum.
Millett, M. (1990). The Romanization of Britain: An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Myrup Kristensen, T., and Stirling, L. (forthcoming) (eds). The Afterlife of Roman
Sculpture: Late Antique Responses and Practices.
Pearson, A. (2006). The Work of Giants: Stone and Quarrying in Roman Britain. Stroud: Tempus.
Perring, D. (2003). ‘“Gnosticism” in Fourth-​ Century Britain:  The Frampton Mosaic
Reconsidered’, Britannia, 34: 97–​127.
Phillips, E. J. (1976). ‘A Workshop of Roman Sculptors at Carlisle’, Britannia, 7: 101–​108.
Phillips, E. J. (1977a). Corpus Signorum Imperii Romani, Great Britain Vol. 1 fasc. 1: Corbridge,
Hadrian’s Wall East of the North Tyne. Oxford: Published for the British Academy by Oxford
University Press.
Phillips, E. J. (1977b). ‘The Classical Tradition in the Popular Sculpture of Roman Britain’, in J.
Munby and M. Henig (eds), Roman Life and Art in Britain. BAR 41/​1: 35–​49.
Reece, R. (1977 with later alterations, unpublished/​forthcoming). ‘The Badness of British Art
under the Romans’.
618

618   Ben Croxford

Reece, R. (1979). ‘Romano-​British Interaction’, in B. C. Burnham and J. Kingsbury (eds),

Space, Hierarchy and Society: Interdisciplinary Studies in Social Area Analysis. BAR int. ser.,
59: 229–​240.
Roach Smith, C. (1859). Illustrations of Roman London. London: Printed for subscribers, and
not published.
Russell, M. (2006). ‘Forgotten Hero’, British Archaeology, 89: 43–​45.
Scott, S. (2000). Art and Society in Fourth-​Century Britain: Villa Mosaics in Context. Oxford
University School of Archaeology Monograph 53. Oxford: Oxford University School of
Archaeology.
Scott, S. (2003). ‘Provincial Art and Roman Imperialism:  An Overview’, in S. Scott and J.
Webster (eds), Roman Imperialism and Provincial Art. Cambridge: Cambridge University
Press, 1–​7.
Stewart, P. (2000). ‘How Romulus came to Bisley: Sculpture in Roman Britain’, Apollo, 461:
15–​19.
Stewart, P. (2003). Statues in Roman Society:  Representation and Response. Oxford:  Oxford
University Press.
Stewart, P. (2008). The Social History of Roman Art. Cambridge: Cambridge University Press.
Stewart, P. (2010). Geographies of Provincialism in Roman Sculpture. RIHA Journal 0005
(27 July 2010) <http://​www.riha-​journal.org/​articles/​2010/​stewart-​geographies-​of-​
provincialism> (accessed 8 April 2014).
Taylor, J. (2013). ‘Encountering Romanitas:  Characterising the Role of Agricultural
Communities in Roman Britain’. Britannia, 44: 171–​190.
Tomlin, R. S. O. (1993). ‘The Inscribed Lead Tablets: An Interim Report’, in A. Woodward and
P. Leach, The Uley Shrines, Excavation of a Ritual Complex on West Hill, Uley, Gloucestershire:
1977–​9. English Heritage Archaeological Report 17. London: English Heritage, 113–​130.
Tomlin, R. S.  O., and Hassall, M. W.  C. (2005). ‘Roman Britain in 2004:  III. Inscriptions’,
Britannia, 36: 489.
Toynbee, J. M.  C. (1951). ‘The Lullingstone Busts’, in G. W. Meates, E. Greenfield, and E.
Birchenough, ‘The Lullingstone Roman Villa’, Archaeologia Cantiana, 63: 35–​43.
Toynbee, J. M. C. (1962). Art in Roman Britain. London: Phaidon.
Toynbee, J. M. C. (1964). Art in Britain under the Romans. Oxford: Clarendon Press.
Toynbee, J. M. C. (1965). The Art of the Romans. London: Thames and Hudson.
Toynbee, J. M. C. (1986). The Roman Art Treasures from the Temple of Mithras. London and
Middlesex Archaeological Society, Special Paper 7. London: London and Middlesex
Archaeological Society.
Webster, J. (2003). ‘Art as Resistance and Negotiation’, in S. Scott and J. Webster (eds), Roman
Imperialism and Provincial Art. Cambridge: Cambridge University Press, 24–​51.
Wheeler, R. E. M. (1943). Maiden Castle, Dorset. London: Society of Antiquaries.
Woodward, A., and Leach, P. (1993). The Uley Shrines, Excavation of a Ritual Complex on West
Hill, Uley, Gloucestershire: 1977–​9. English Heritage Archaeological Report 17. London:
English Heritage.
 Chapter 30

Names of G od s

Amy Zoll

To invoke the gods was a solemn and powerful act that bridged the mortal and immortal
realms. In Roman religion, appeals for divine assistance were not undertaken lightly, as
the gods expected to be compensated for their favours in the form of sacrifices and gifts
bestowed in elaborate rituals. Priests and officials interceded on behalf of the Roman
state, its emperor and people, but ordinary petitioners also sought to engage the gods
directly in times of personal crisis and uncertainty. Much of the evidence for the array
of deities worshipped in Roman Britain comes from the inscribed offerings of private
individuals and groups, civilians and soldiers, men and women, citizens and freedmen,
immigrants and indigenes.
The eclectic mix of gods attested in Britannia belies the island’s reputation as a remote
outpost. Roman London boasted devotees of the empire-​spanning cults of Isis, Mithras,
and Cybele. Along Hadrian’s Wall, recruits from Germania, Gaul, and as far as Syria
paid homage to the divine patrons of Rome, the emperor, and the army. Soldiers also
carried with them the gods of their homelands and honoured the deities who held sway
in the territories they occupied. The responses of the native populace to the adoption of
local gods by incomers and the introduction of a host of new divinities were undoubt-
edly varied and complex.
Theories of religious transmission and adaptation in the western provinces have gen-
erally echoed broader themes of culture change. The resulting pantheons and practices
have been cast as the inevitable products of benign accommodation (Henig 1984), as
mechanisms for the promotion of imperial ideology (Whittaker 1997) or for local elite
competition and continuity (Derks 1998). Provincial religion has been recognized as an
actively negotiated medium for symbolic domination and resistance (J. Webster 1995),
social mobility (Woolf 1998), and local identity formation (Woolf 2001; Revell 2009).
Few models, however, offer more than a passing critique of the material evidence on
which they are predicated.
Within the diverse communities of Roman Britain, peoples and beliefs mingled and
were transformed. This dynamic interplay of ideas is evident in the depictions of gods
in Romano-​British art, but the imagery can be ambiguous and the actual identity of the
620

620   Amy Zoll

deity portrayed lost to modern interpretation. Seemingly more definitive is the infor-
mation provided by religious inscriptions, which record the names of over 200 differ-
ent gods and goddesses. Many of these entities are well known from elsewhere in the
empire, while others are unique to Britain and appear closely tied to a region or site. But
names, like looks, can be deceiving, masking layers of meaning if taken solely at face
value.
Inscriptions used to express and reproduce religious knowledge did not occur in
isolation; context is essential to understanding the epigraphic evidence. The choice
to inscribe an object, its form, location, setting, and authors, all contributed to a mes-
sage beyond the textual, yet have been largely regarded as normative and unproblem-
atic. The appearance of Greco-​Roman theonyms on Latin-​inscribed, classically styled
monuments goes unquestioned, while localized deities presented in the same fashion
are presumed to be of indigenous derivation, with origins pre-​dating the Roman inva-
sion, even though their names are recorded only on stones erected a century or more
after conquest.
According to Roman belief, a deity had to be addressed properly and precisely and
the rites and prayers in its honour conducted flawlessly to be effective (Scheid 2003: 18).
Determining which god or gods were inclined to heed a request and how they wished to
be called was imperative. Even in Rome identifications were not always certain (Ando
2008: 56), but became more fluid as the empire expanded to encompass a raft of com-
peting cosmologies. New syntheses were born of the socio-​cultural tensions, inequities,
and opportunities of the colonial experience. In Britain, the varying epithets, aspects,
and associations of deities—​both local and imported—​reflect the ongoing efforts of pro-
vincial inhabitants to reconcile disparate influences and aspirations as they constructed
identities for themselves and their gods in a new Roman world.

Sources

The marquee deities of Rome were themselves multifarious, the products of centuries
of cross-​cultural interaction and shifting relationships between the civilizations of the
Mediterranean littoral and beyond. Their precursors and evolution are extensively doc-
umented in art, archaeology, and ancient literature. The geneses of Romano-​British gods
are far murkier, the surviving evidence limited in scope and temporal depth. Nothing
specific is known about Britain’s pre-​Roman pantheon that has not been extrapolated
from later material or projected from continental parallels. Such indirect inferences
assume a widely held and deeply conservative belief system that was able to withstand
a period defined by dramatic changes in nearly every aspect of society, including the
language, rituals, and objects used by many to conceptualize and communicate with the
divine.
Classical authors offer few insights into the beliefs of ancient Britons and are simi-
larly unforthcoming about the post-​conquest conduct of religion in the province. The
 Names of Gods   621

founding and plans of several Roman era towns appear to respect the locations of earlier
sacred, perhaps ancestral, sites on the landscape (Creighton 2006); however, few of the
post-conquest shrines and temples where the presiding deity is known show signs of
unbroken activity stretching back into the Iron Age. Nor does continued use necessar-
ily imply continuity of belief (Woodward 1992: 14–​16). Artistic renderings and inscrip-
tions remain the primary sources of information about the gods of Roman Britain, their
names and purviews.
Few anthropomorphic works and fewer inscriptions date prior to the Claudian inva-
sion, and none can be confidently linked to later divine personae. The scant Iron Age
precedent for Romano-​British sculpture and iconography led Henig (1985) to propose
an entirely Roman impetus for the craft and its symbolic repertoire, although arguments
have been made for a pre-​existing figural tradition, mainly in perishable materials (Green
1998; J. Webster 2003: 42–​46). Representations of gods and goddesses from Roman Britain
owe much to a classical visual vocabulary, but their appearance and attributes were often
reworked to meet local aesthetics and demands (see also Croxford, this volume).
Some scholars seize upon stylistic and iconographic deviations from (presumed)
norms as a means of ascertaining indigenous intent, but Johns (2003) and Stewart (2010)
caution that extant imagery was subject to compounding influences over time, contin-
gent upon not only the changing expectations of artist and patron, but also competency,
materials, location, genre, and modes of transmission. Rather than the embodiment of
a single coherent syncretic process, Romano-​British art was the mature product of con-
tinuous cultural reconciliation that cannot be so easily distilled into its constituent ele-
ments. These caveats can be extended to texts as well.
Inscriptions on stone comprise some of the most recognizable and best-​preserved
evidence for the gods of Roman Britain. The non-​classical theonyms recorded on these
monuments have received particular attention as potential pre-​Roman survivals, with
etymologic and textual associations scrutinized for clues to their origins and charac-
ter (Ross 1967; G. Webster 1986; Aldhouse Green and Raybould 1999). The inher-
ent biases of the epigraphic corpus have been acknowledged for some time. Webster
(1995) contends that religious inscriptions were created by, and for, a literate minority.
Immigrants—​such as soldiers, administrators, and foreign traders—​had a vested inter-
est in the perpetuation of the Roman state, as did members of the local elite, who strove
to align themselves with imperial authority through an adopted language and customs.
Religious inscriptions were products of their place and time, she concludes, repre-
sentative of the views and unequal power relations of certain privileged and materially
conspicuous groups, offering little insight into the Iron Age deities that came before (J.
Webster 1995: 61).
Inscriptions may not be the key to reconstituting earlier belief systems, but neither
were they exclusively or uniformly employed by incomers and the upper echelons of
society. Quantitative and geospatial analyses of Roman Britain’s epigraphic assem-
blage reveal considerable differences in the types and frequencies of votive texts (Cepas
1989; Zoll 1995a, b). Britannia’s inhabitants addressed the gods through the written
word in a variety of media, reflective of their status, aims, and audience. Besides stone
622

622   Amy Zoll

monuments, messages were engraved on vessels, rings, votive plaques, and leaves/​feath-
ers (thin triangular sheets of embossed metal), as well as curse tablets (leaden strips
bearing secret petitions for divine retribution, also known as defixiones). From these
diverse collections of inscriptions, it is possible to discern distinct provincial, regional,
and social patterns of worship.

Distribution and Dating

Biró (1975) notes that the most common type of inscribed monument from Roman
Britain was not the tombstone, but the votive altar; this distinguishes it from most other
western provinces, where funerary inscriptions predominate. Mann (1985) further
observes that the bulk of all inscriptions on stone from the island concentrate in the
north and west, among the settlements and installations of the so-​called military zone,
borderlands in which the Roman army maintained a long-​term presence. Britannia’s
total epigraphic output is shown by Harris (1989: 268–​269) to be remarkably anaemic
compared to the rest of the Roman west. Rather than being due simply to a dearth of
suitable stone in the south-​east, these distributional disparities are attributed to the
civilian populace’s late and lacklustre embrace of a Latin ‘epigraphic habit’.
As described by MacMullen (1982), the Roman epigraphic habit was marked by
an empire-​wide proliferation of both civic and personal inscriptions that was itself
largely a phenomenon of the Principate, developing in concert with—​and in service
to—​imperialist aims and the ambitions of its populace during the first through third
centuries ad (Woolf 1996). The bulk of Britain’s votive altars adhere to this general pat-
tern of production, falling between the mid-​second and early third centuries ad, with
none securely dated beyond the third century ad. Other forms of religious epigraphy,
however, continued unabated after the decline of monumental inscription. Curse tab-
lets found at Bath and Uley date mainly between the third and fourth centuries ad,
whilst the late-​fourth-​century treasure from Thetford, in Norfolk, included thirty-​one
inscribed spoons—​at least twenty of which were dedicated to the god Faunus and/​or
one of seven obscure non-​classical deities (Johns 1986).
Concentrations of monumental inscriptions have also been found among the major
administrative and commercial centres of the south, including London, Colchester,
Silchester, and Richborough (Jones and Mattingly 1990: 153). Towns and rural sanctu-
aries in the vicinity of the Severn Estuary and the Cotswolds—​within the civitas of the
Dobunni—​also exhibit a denser clustering of inscribed works in the locally available
limestone than the surrounding territories. Stewart (2010) detects a similar pattern
in the incidence of stone sculpture in northern military and southern urban contexts,
which he likewise ascribes to differences in geography, geology, and ‘habit’.
The distribution of votive altars, however, does not entirely conform to that of sculp-
ture or monumental epigraphy in general (Figure 30.1). Although abundant on and
around Hadrian’s Wall, and even well represented along the short-​lived Antonine Wall,
 Votive Altars

2–5

>5

>10

>20

0 100 200 km

Figure 30.1  Distribution of inscribed votive altars from Roman Britain.

Source: after Millett (1995a: 110). © Amy Zoll.
624

624   Amy Zoll

inscribed altars are poorly attested at most southern urban sites. Access to resources
may be partly to blame, although adequate stone was available to these same civilian
centres for large-​scale building programmes, statuary, and other types of inscription.
Nor are lower rates of literacy or a weakly internalized epigraphic habit sufficient expla-
nation, as some inhabitants of southern Britain could, and did, employ the written word
in their interactions with the gods in other materials. Differences in choice of offering
may instead relate to dedicator identity and the roles religious institutions played in the
communities of the north and south.
Votive altars are not the only remnants of religious practice irregularly distributed
across Roman Britain. Mattingly (2006:  520–​522) compares the relative frequen-
cies of several categories of material culture related to ritual and finds some artefacts
and features to be more common among urban and rural settlements of the civilian-​
administered south, while being scarce in military areas. Among text-​bearing objects,
votive leaves and curse tablets have been unearthed in greater numbers in civilian
contexts. Votive altars also show little geographic overlap with known temple sites
(Figure 30.2), particularly those of the Romano-​Celtic type, a form thought to pre-​
date the Claudian conquest (King 2007). Some notable exceptions occur in the south-​
west, where the large temple complex at Bath has produced altars, curse tablets, and
inscribed offerings dedicated to the resident goddess, Sulis Minerva, and other deities
(Cunliffe 1988).
Improved recording of stray finds has shed new light on the frequency and distribu-
tion of non-​monumental religious objects, including the regional dispersal of numer-
ous rings bearing the cryptic inscription ‘tot’. The discovery of a specimen with an
expanded legend (deo tota) confirms these items of personal adornment refer to a
deity, possibly the god Toutatis (Daubney 2010). Dated stylistically to the second and
third centuries ad, the rings were produced mainly in silver, with a few in gold or cop-
per alloy. Their find-​spots conform to the civitas of the Corieltauvi in the east Midlands,
centring on the territorium of the Roman colony at Lincoln, although most come from
rural contexts. This region is poorly served by evidence for either temples or altars, but
the scattering of tot rings illustrates an alternate means of expressing religious knowl-
edge and a shared sense of cult embodied by the wearer, rather than rooted in a location
(although these items may themselves signify pilgrimage to a sacred site).
Inscribed objects pertaining to religion and ritual are notoriously difficult to date,
most having been recovered from secondary contexts. Like the tot rings, the major-
ity of British defixiones are either unstratified or redeposited finds, while votive altars—​
composed of solid and handily dressed blocks of stone—​were frequently incorporated
into later structures. Despite their sacred function, these monuments were not immune
to reuse even in antiquity, as evidenced by the series of altars to Jupiter Optimus
Maximus (Best and Greatest), chief god of Rome, found buried in pits near the fort at
Maryport, now interpreted as the repurposed footings for a massive timber building,
possibly dating to the late fourth century ad (Haynes and Wilmott 2012).
The coarse chronologies for religious inscriptions hinder the ability to trace the devel-
opment of the practices they represent. The one major text known to come from the
 Altars

Temples

tot Rings

Hadrian’s Wall

N
0 100 200 km

Figure 30.2  Main concentrations of votive altars.

Source: after Millett (1995a: 110); temples after Millett (1995a: 112); tot rings after Daubney (2010: 112). © Amy Zoll.
626

626   Amy Zoll

first century ad is the famed Purbeck marble plaque from Chichester (RIB 91), which
records the dedication of a temple to Neptune and Minerva, authorized by a British cli-
ent king for the welfare of the emperor’s domus divina (divine house). This example is
extraordinary not only for its early date, but also for its depiction of a leading member of
the native aristocracy involved in the promotion of the gods of Rome and the Imperial
Cult. Local elites in the provinces, as at Rome, are thought to have patronized civic and
private cults and vied for priestly offices as a means of affirming and enhancing their
prestige within the imperial power structure (Gordon 1990; Woolf 1997; Derks 1998). In
Britain, however, epigraphic evidence for the participation of high-​ranking Britons and
civilian authorities in religious institutions is sparse, but this may be less an indication of
an ambivalent attitude towards socio-​political competitiveness than a reflection of the
types of inscriptions being produced, the rituals they commemorate, and the individu-
als and groups responsible for their creation.

Vows and Curses

The majority of inscribed offerings from Britain were votive in the true sense: presented
to the gods not to placate or elicit future favours, but as payment for services requested
and rendered. The vow was a central tenet of Roman religion, articulating the recipro-
cal relationship between mortals and immortals through ritualized exchange (Scheid
2003: 101–​105). At Rome, public vows were formally pronounced by officials and priests
to ensure the well-​being of the emperor and the Roman people at regular intervals and
at times of national crisis or insecurity. Sacrifices, games, temples, statues, and valuable
gifts were pledged in return for divine protection. Private individuals undertook similar
vows on a much reduced scale when confronted with an uncertain fate owing to a grave
illness, perilous journey, risky venture, or other anxiety.
While vows and curses differed in intent—​one to solicit aid, the other to inflict harm—​
both represent specific and finite transactions between gods and mortals that followed
similar ritual structures whereby the obligations and expectations of both parties were
made explicit. In the case of personal vows, little survives of the initial declaration of
terms (nuncupatio), which may have been recorded on sealed wooden tablets stored in
the cult temple (Derks 1998: 224–​231). The sacrificial rites and offerings promised were
deliverable only upon a satisfactory result, to be presented at the ceremonial closure of
the vow (solutio). The legends on votive altars and other inscribed gifts clearly state to
which god they were owed and by whom, often concluding with the phrase ‘votum solvit
libens merito’ (willingly and deservedly fulfilled a vow), indicating the gratitude of the
dedicator at the completion of the compact. These transitory performances were repro-
duced through the monuments and inscriptions left to adorn the shrines, advising visi-
tors of the correct names and rites by which to approach the deity and serving as lasting
testaments to the power of the god and success of the vow.
 Names of Gods   627

Defixiones represent a different kind of pact, in which the conditions have been pre-
served, but the outcome remains a mystery. Continental examples of engraved curses
sought to exercise supernatural influence over a variety of situations. By contrast, British
curse tablets are preoccupied with theft (Tomlin 1988); their authors’ appeals likened to a
form of ‘judicial prayer’ (Versnel 1991; Adams 2006). The creation of a tablet took on the
role of nuncupatio, dedicating the stolen property (or the thief him or herself) in absen-
tia to the god, who was exhorted to torment the perpetrator—​known or unknown—​
until the missing items were returned to the shrine. For this service, the deity would
receive a portion of the recovered goods in reward (solutio). The details of these agree-
ments were concealed in tightly rolled strips of lead alloy, affixed to the sanctuary walls
or base of the cult statue, or cast into sacred waters. No doubt, the prospect of divine
justice held a certain psychological appeal, particularly among those who may have felt
disenfranchised from the terrestrial legal system (Kiernan 2004). The longevity of the
practice attests to its perceived efficacy.
Although the physical and textual remains of vows and curses are readily identifi-
able, these rites were responses to specific circumstances, rather than prescribed parts
of routine religious observance. They were by no means the only ways to communi-
cate with the gods, or indeed the most common. The majority of vows were fulfilled
with consumable sacrifices and perishable gifts, rather than permanent monuments
such as altars. Many more unengraved offerings from countless anonymous donors are
found at sacred sites (see also Smith, this volume), but these tend to be overshadowed
by text-​bearing objects in scholarly discussion. At least half of the over two dozen stone
altars from Coventina’s Well, at Carrawburgh, were uninscribed. Whether unfinished or
originally painted, these anepigraphic monuments were less carefully documented than
their counterparts with legible dedications to the goddess (Allason-​Jones and McKay
1985: 18). From this same deposit, large quantities of pottery, bone, and more than 13,000
coins demonstrate that Coventina’s worshippers expressed their devotion materially in a
variety of ways, not just through votives fashioned and inscribed for the purpose.

Votaries

Whittaker (1997) regards religion and politics as inseparable forces for the imposition
and encouragement of the Roman world view. Public priests and civilian magistrates
were charged with the maintenance of social and spiritual order through the conduct of
sanctioned rites on behalf of the citizenry, in recognition of a communal obligation to
the divine (religio) (Beard 1991; Scheid 2003: 102). This was Roman religion in its strict-
est sense: a unifying set of precepts and practices, integral to law and government, set
out in municipal charters and official calendars (Ando 2007). All else tends to be rel-
egated to the catchall of ‘private’ cult, and it is from this category that the bulk of epi-
graphic evidence for religion in Roman Britain derives.
628

628   Amy Zoll

Identification of both supplicant and recipient deity seems to have been a requisite
component for most votive and defixio texts. The inscribed invocation was a symbolic
act, rooting the god’s presence in space and time. The inclusion of the worshipper’s name
served a corresponding function:  situating the individual within the wider religious
tradition of the community and reinforcing personal and collective relations with the
gods (Beard 1991: 46–​48). Unlike epitaphs, however, religious inscriptions were meant
to commemorate an event rather than celebrate a life and tend to be short on biographi-
cal detail. Even so, some commonalities of participant identity emerge.
The names of petitioners on curse tablets differ from those of altar dedicators, includ-
ing at Bath, where both forms of object are present, possibly denoting the activities
of distinct social groups within a shared sacred space or a shift in clientele over time
(Tomlin 2002: 171–​172). Authors of defixiones exhibit a higher percentage of Celtic per-
sonal names, with few employing the nomenclature of the Roman citizen more com-
monly found on altars. The small sums and basic possessions listed among the items
stolen—​such as articles of clothing, livestock, and agricultural implements—​suggest
rural practitioners of modest means. At Uley, the values of missing goods are somewhat
greater, indicating that even affluent farmers at times felt the need for divine redress
(Adams 2006: 10). Variations in the material and quality of tot rings may also imply
bearers of varying status who expressed a common membership through their visible
commitment to the same deity. Other small objects, such as inscribed plaques (tabu-
lae ansatae) and votive leaves, found mainly in the south, further demonstrate how the
civilian populace adopted and adapted writing in their dealings with the gods.
But it was the Roman army that made the most substantial contribution to Britain’s
corpus of religious epigraphy (Cepas 1989; Raybould 1999). Soldiers routinely included
their unit and rank in inscriptions, attesting to the importance of military affiliation in
self-​representation within this group. The army is the most epigraphically recognizable
of any profession or other subset of the population in Roman Britain, but even without
the abundance of textual clues, the marked distribution of stone inscriptions shows that
soldiers and their associates made disproportionate use of this mode of expression.
Mattingly (2006: 166) estimates Britain’s ‘community of soldiers’, including auxiliaries
and dependent families, merchants, slaves, and freedmen, constituted less than 3 per
cent of the total population. Yet, members of the Roman army, acting individually or as
units, were responsible for well over half of all extant votive inscriptions on stone from
the island. The inordinate number of military dedicators in British religious inscriptions
has been noted elsewhere (Zoll 1995a: 136–​137). Research in progress by this author sug-
gests this was not an isolated phenomenon, but part of a wider pattern in which votive
inscriptions were one of the most prevalent forms of monumental epigraphy along the
empire’s northern limes, in provinces with permanent garrisons and areas of intensive
long-​term military activity (Zoll, forthcoming).
Soldiers observed their own festival calendars and military rites, but also displayed
considerable diversity in belief and practice throughout the empire (see also Haynes,
this volume). The profusion of altars set up by units and their commanders may rep-
resent an alternative to elite-​administered and urban-​focused civic worship, with its
 Names of Gods   629

official priesthoods, euergetism, and monumental temple architecture. Rather than

symbolizing the export of Roman religion, unchanged, to the edges of the empire, the
presence of inscribed altars in large quantities may instead signify its adaptation in
response to regional differences in social structure and administration.

Gods of Empire, Elsewhere,

People, and Place

Convenient labels—​military or civilian, official or unofficial, public or private—​applied

to the evidence for religion have long exerted an undue influence on its interpretation
(Millett 1995b; Goldberg 2009a). Simple classificatory schemes fail to capture the wide
spectrum of ritual and belief, but are especially pervasive in the characterization of pro-
vincial pantheons, where gods are categorized as ‘Roman’ or ‘native’, or some measura-
ble combination of the two. Eastern cults are typically segregated from discussion owing
to the perceived ‘otherness’ of the rites, philosophies, and participant experience. The
majority of practitioners were unlikely to have made such clear distinctions, however.
Treatises on religio by ancient authors and repeated attempts by lawmakers to codify its
correct observance show these definitions were not fixed but required periodic action
and adjustment when confronted with different gods and practices within the empire’s
borders (Ando 2007; 2008: 1–​18).
Even within the army, ‘official’ customs varied. Given the high percentage of mili-
tary dedicators represented in Britannia’s epigraphic assemblage, it is unsurprising that
Jupiter Optimus Maximus, supreme protector of Rome and the empire, is by far the
most frequently named deity on votive altars, with well over 100 surviving examples.
According to Whittaker (1997: 151–​152), Jupiter had become synonymous with the city,
imperial law, and the political order by the second century ad. His monuments occur in
multiples at forts along Hadrian’s Wall and related northern defences, set up primarily
by auxiliary cohorts and their commanding officers. In most instances, Jupiter Optimus
Maximus (often abbreviated IOM) is addressed alone or with the divine authority of
the emperor (numen Augusti). These altars are usually ascribed to a military version
of the annual public vota ceremony performed at Rome for the welfare of the emperor
and Roman people, which also served as an occasion for the renewal of service oaths by
the army (Stoll 2007: 462). Few votive altars have been found associated with Britain’s
three permanent legionary bases at York, Chester, and Caerleon, but comparable series
of IOM monuments from other provinces were erected on different dates to mark the
anniversary of the formation of units under their standards (e.g. CIL II. 2552–​2553)
and the tenures of governors’ adjutants (beneficiarii consulares) at military outposts
(Mirković 1989).
Despite Jupiter Optimus Maximus’ close association with the Roman state, his
identity is not always so clear cut. His sole appearance on a British defixio from
630

630   Amy Zoll

Ratcliffe-​on-​Soar, Nottinghamshire (Turner 1963), seems to confirm a hypothesis

by Versnel (1991: 90) that authors of curses tended to avoid the ‘great gods’, opting
instead to make requests of deities with strong local ties who might be more sympa-
thetic to their plight. However, Jupiter Optimus Maximus is also the most common
recipient of votive altars in parts of Gaul and north-​west Hispania, where he is seen
by some to disguise an indigenous sky or mountain god (Green 1989: 74–​130; Richert
2005: 3–​6). Jupiter’s conflation with a local sky god from Doliche, in southern Turkey,
gained a following across the Roman world, especially among soldiers. Inscriptions
and cult images of Jupiter Optimus Maximus Dolichenus have been found at sev-
eral sites at Hadrian’s Wall, including the unexpected discovery of altars and a shrine
to this ‘non-​traditional’ deity within the walls of Vindolanda fort (Birley and Birley
2010: 25)—​an honour once thought reserved for sanctioned divinities of the empire
and the army.
Just as the definition of ‘Roman’ blurs upon closer examination, so too does that
of ‘native’. The Matres were frequent subjects of votive texts and depictions in stone.
Similar trios of mother goddesses are well known from the continent, but there were
emphatically linked to a particular place or group by use of localized epithets (Derks
1998: 119–​130; Woolf 2001). When the type of Matres is specified in Britain, however,
they are usually of a more inclusive variety and often allude to imported origins for
both dedicators and deities, such as the Matres Domesticae (of the household or home-
land (RIB 652, 2025, 3210)), Tra(ns)marinae (from overseas (RIB 919, 1224, 1318)), or
Ollototae (a Celtic descriptor translated as ‘from other peoples’ or ‘foreign-​born’ (RIB
574, 1030; AE 2009, 738)). Two small votive monuments dedicated to the assorted
Matres of Italy, Gaul, Germania, Africa, and Britannia by military personnel also reflect
the mixed make-​up of the communities in which they were erected (RIB 88, 653).
Most gods were approached in locations—​both built and natural—​where their power
was seen to be most manifest. The placement of altars, niches, and small shrines (aedicu-
lae) in secular and domestic settings shows that temples were not the only venues for
encounters with the divine. Soldiers frequently called upon the protection of the god-
dess Fortuna in the baths (RIB 730, 1212, 2146), while her more judgemental counter-
part, Nemesis, was sought in amphitheatres (RIB 3149). Sacred springs were home to
the Nymphs (RIB 460, 3316) and sometimes presided over by named goddesses, such as
Sulis Minerva at Bath or Coventina at Carrawburgh, whose influence does not appear
to have extended much beyond their watery sanctuaries. When the name of the resident
divinity was not forthcoming, the spirit of the place (genius loci) might be invoked (e.g.
RIB 450, 1984, 3195).
A few gods seem to have held a wider authority, with distributions tying them to the
social and physical geography of a region (Figure 30.3). Belatucadrus is attested from
some twenty-​nine altars confined to the western end of Hadrian’s Wall in northern
Cumbria, an area associated with the civitas of the Carvetii. The god Cocidius is known
from twenty-​three altars, mainly from the mid-​section of the wall and a few forward
installations. He is also schematically depicted in panoply on two silver votive plaques
(RIB 986, 987) from Bewcastle, believed to be the site of a shrine to the god, from which
 Names of Gods   631

Cocidius
1
" 2
5
" Mars Cocidius
"
" " Silvanus Cocidius
# # Vernostonus
Cocidius
Hadrian’s Wall

Belatucadrus
1
2–3
7
Mars Belatucadrus
Hadrian’s Wall

Vitiris/Veteres
1
2–3
4–6
>10
Moguns Vitiris
Hadrian’s Wall

N
0 25 50 km

Figure 30.3  Distribution of votive inscriptions dedicated to the gods Cocidius, Belatucadrus,

and Vitiris/​Veteres, including double-​named variants.
Source: after Zoll (1995a). © Amy Zoll.

it took the name Fanum Cocidi. To the east, more than sixty altars were set up to the
enigmatic deity Vitiris, who is portrayed in a third of examples as the plural Veteres, and
in two instances from Chester-​le-​Street, as female (RIB 1047, 1048).
632

632   Amy Zoll

Vitiris is second to Jupiter in number of extant dedications in Britain, with Mars

his only close rival. The distribution of Vitiris relates to the main eastern watershed
of northern England, and he is most likely a local divinity, despite some seemingly
Germanic renditions of his name (Goldberg 2009b). Falling to opposite sides of the
Pennines, altars to Vitiris and Belatucadrus are consistently among the smallest in
Britain and most are crudely carved. Both suffer from rampant spelling inconsistencies,
suggesting the oral transmission of names and a lack of central authority over their wor-
ship. Devotees of Belatucadrus and Vitiris generally go by only one name, and few men-
tion a military association, but many stones are either too small or incomplete to include
much dedicator information. In contrast to these followers of low or no rank, Cocidius
attracted a higher-​status clientele—​mostly officers and legionary units—​and his monu-
ments tend to be larger and more skilfully crafted.
It is difficult to ascertain just how the attentions of soldiers contributed to the recog-
nition and perpetuation of non-​classical deities. Many local theonyms survive in stone
as a result of epigraphic customs among the military, but this was a selective process,
subject to the concerns of the inscribing community (Zoll 1995a). Among the classi-
cal gods recorded in votive texts from Britain, Juno appears at most twice (RIB 3460,
3505), despite her elevated position as a member of Rome’s Capitoline Triad alongside
Jupiter Optimus Maximus. Venus was a popular subject in Romano-​British decorative
arts, but no written dedications to the goddess are known from the province. The many
mass-​produced pipe-​clay votive figurines modelled in her likeness are seen to represent
an alternate, popular, and indigenous set of beliefs that operated outside classical con-
vention (J. Webster 1997: 332–​334), but there is no assurance that practitioners saw any
inherent contradiction in their actions.
Uninscribed offerings reflect different trajectories of worship by segments of
Romano-​British society for whom epigraphic commemoration may have been inappro-
priate or unavailable. Women are poorly represented in Britannia’s votive epigraphy on
stone, comprising just 3 per cent of known altar dedicators. In other provinces, such
as Hispania Citerior, they are better attested, particularly in areas without a significant
military presence (Zoll, forthcoming). Women participated in the enacting of curses at
Bath and Uley, although not to the same frequency as men. Three women are, however,
among the eight devotees identified on gold and silver votive leaves and other objects
donated to Senuna at her shrine near Ashwell, Hertfordshire (Jackson and Burleigh
2007). While monumental epigraphy does not appear to have been the custom at this
site, worshippers still engaged the local goddess through vows and discharged their
obligations in the accepted fashion of that community.

By Any Other Name …

Much scholarly attention has been lavished upon gods with non-​classical names,
whether in an attempt to decode earlier beliefs or as an indicator of regional ethnicity.
 Names of Gods   633

Deities with standard Roman theonyms often recede into the background, assumed to
be wholesale introductions, uniformly transmitted and received. That Roman and non-​
Roman divinities were invoked side by side on the same monument and even as the
same entity reveals the innovative ways in which local pantheons developed in response
to practitioner need and identity. Appeals were made to those higher powers deemed
most likely to respond, based on prior experience and evidenced by the votive gifts
amassed at sacred sites. How these divinities came to be worshipped in this manner, and
by whom, have been central to the debate over religious change in the Roman west.
Many arguments hinge upon the epigraphic phenomenon of double naming, in
which a Roman theonym was linked with that of a local god and addressed as a single
being. Such dual deities are not unique to Britannia and comprise only a small percent-
age of god names from the island (Zoll 1995b). They have, nonetheless, been imbued
with special significance as intermediaries on the path to assimilation: Tacitus’ inter-
pretatio Romana personified (Germania 43). The equation of autochthonal deities with
Roman counterparts is seen by Henig (1984: 55–​67) as the predictable result of incom-
ers wishing to venerate local divinities in familiar forms, terms, and rites, setting the
example for leading members of native society to follow. For Webster (1995), these
pairings were products of an asymmetric colonial discourse: deliberate superimposi-
tions, privileging one set of beliefs over another and inciting subject peoples to react to
the forcible conversion of their gods with overt rejection or subtler forms of resistance.
Derks (1998: 91–​118), however, disputes the characterization of such hybridized entities
as instruments of cultural imperialism, instead viewing the selectivity of deity pairings
in Gaul as a reflection of a native understanding and application of the Roman pantheon
to further their own aims.
Rather than being transitional and temporary, many twinned deities show remark-
able resilience and portability. At Bath, engraved entreaties to Sulis Minerva, arguably
the most famous doubly named god, occur on stone from the mid-​second century ad
and continue in curse texts into the fourth century, appearing alongside similar exam-
ples addressed solely to Sulis, but never to Minerva alone. Spoons from the late fourth
century ad Thetford hoard bear the names of a series of little-​known divinities; in some
instances coupled with the god Faunus and in others singly, but Faunus himself does not
appear unaccompanied (Johns 1986).
Non-​classical bynames were more than mere descriptive epithets; they were impor-
tant for proper identification and observance. Many deities so named retained their
distinct appellations and character far from their home territories. Well-​travelled gods
such as Hercules Magusanus (RIB 2140), Mars Camulus (RIB 2166, 3014), Loucetius
Mars (RIB 140), and Mars Lenus (RIB 309) were probably already paired when they were
brought to Britain from the European mainland, where they had established followings.
Even Mars Toutatis may have been imported in this form; the two plaques bearing his
name (RIB 219 and AE 2001, 1298) were found in very disparate locations, seemingly
unrelated to the well-​defined distribution of tot rings.
For theonyms exclusive to Britain, Mars occurs in the majority of pairings, followed
(to a much lesser extent) by Apollo and Silvanus. Other classical deities commonly
634

634   Amy Zoll

involved in similar relationships on the continent, such as Mercury and Hercules, are
barely represented. Mars’ appeal among the military communities of northern Britain is
understandable given the god’s martial qualities, but Henig (1984: 50–​55) attributes his
popularity in civilian contexts to the polyvalent and composite nature of most classical
divinities. Derks (1998: 94–​107) makes a cogent argument for why Roman deities such
as Mars and Hercules were affiliated with local gods along the Lower Rhine, there cast as
guardians of field or flock based on regional ecology and subsistence practices.
Webster (1995: 159–​160) posits that some segments of the population, such as native
Britons and low-​ranking soldiers, did not engage in name-​twinning to the degree
of higher-​status devotees. This seems to be borne out in the case of Apollo Maponus
(RIB 1120–​1122, 2063), but in other examples where dedications to the same paired
and unpaired deity survive, as with the numerous inscriptions to Sulis, Cocidius, and
Belatucadrus, little distinguishes votaries employing one or the other naming variant.
Nor do the votive gifts differ markedly in terms of object type, size, or distribution.
Webster (2003: 49) also contends that some deities were more resistant than others to
epigraphic twinning. None of the many votive dedications to Vitiris/​Veteres conflate the
god with a Roman deity, but he is once paired with Mogons, an obscure non-​Roman fig-
ure, some distance from Vitiris’ main territory (RIB 971). Coventina, too, seems to have
possessed no named counterpart, but is twice referred to as ‘nymph’ (RIB 1526, 1527) and
depicted as one in relief (RIB 1534), allying her with those widely worshipped water god-
desses. Visual cues can present a very different understanding of a deity from the textual.
Minerva might have been mistaken for the beneficiary of the votive leaves from Ashwell
on which her image appears, if not for the embossed legends identifying the hitherto
unattested goddess Senuna as the intended recipient (Jackson and Burleigh 2007).
Certain classical divinities also seem less receptive to twinning than others. Although
often invoked in his Dolichene aspect (RIB 1022, 1131, 1725), Jupiter Optimus Maximus
is only once paired with an apparent western god in Britain. Called ‘Tanarus’ on an altar
erected by a legionary officer at Chester (RIB 452), this incarnation is sometimes con-
strued as a variant of the Gallic Taranis mentioned by Lucan (Pharsalia i. 446) and pop-
ularly, if disputably, associated with continental figures of a bearded deity with wheel
and thunderbolt (Green 1986). These same symbols appear on the sides of altars set up
to IOM by the Second Cohort of Tungrians at Castlesteads (RIB 1981–​1983), but lack any
textual indication of a hybrid nature and otherwise resemble the many other altars along
Hadrian’s Wall honouring IOM as per military custom. Jupiter’s own resistance to twin-
ning and his rare occurrence on defixiones suggest it may have been inappropriate to
conceive of this god in localized form in Britain.
Double-​named deities illustrate how votaries grappled with the problem of identi-
fication when a single theonym failed to convey the totality of a divinity or situate it
adequately within the larger cosmological scheme. Variations in name pairings suggest
these associations were contextually contingent rather than centrally dictated. Cocidius
is equated with Mars at the west end of his domain, once with Silvanus in the east (RIB
1578), and again with an otherwise unknown deity, Vernostonus, on an altar set up by a
German at Ebchester, south of Hadrian’s Wall and outside Cocidius’ primary range (RIB
 Names of Gods   635

1102). The deity presiding over the shrine at Uley is twice addressed in curse tablets as
Mars rather than the usual Mercury (Tomlin 1993). Another invokes ‘Mars Mercury,’
while a defixio to ‘Mars Silvanus’ was subsequently corrected to Mercury. Tomlin
(1993: 115) notes that a toponym in several of the Uley inscriptions is applied to Mars as
well as Mercury, implying that all of these names may refer to the same entity.
The persistence of double-​named deities alongside other variants hints at greater com-
plexity and nuance in the development of regional pantheons, reflecting diverse con-
stituencies and concerns, rather than the supplanting of one set of divinities by another.
Deities who escaped twinning cannot be assumed to have been any less changed by the
processes that led their names to be set down in stone and metal. The preservation of
British theonyms is inextricably linked with specific rituals by which religious knowl-
edge and notions of deity were affirmed and propagated across the western Roman
provinces. Local gods must have undergone some degree of conceptual translation in
order to be epigraphically observable (if indeed they possessed pre-​Roman anteced-
ents). Although Derks (1998: 234–​239) argues for a prior votive tradition in northern
Gaul, the rites, language, and covenants recorded in inscriptions are all hallmarks of
Roman practice. Devotees engaged these entities through Latin text and the Roman-​
styled rituals of vows, curses, and formal sacrifice, seeking divine intervention close to
home while referencing a greater shared system of practice and belief.

Conclusions

Roman religious knowledge is often described as grounded in orthopraxy: the proper

conduct of prescribed rites (Scheid 2003; Ando 2008). The enacting of vows and curses
helped create and validate this knowledge. Dedicated objects served as media for com-
munication with the divine and as proof of the efficacy of these modes of address. But
just what constituted ‘correct action’ was subject to setting and circumstance. Woolf
(2003: 149–​150) notes that, for most provincial inhabitants, conceptions of deity were
formed and experienced primarily at local shrines. Variations in the allocation of sacred
space within temple complexes in Britannia and Hispania suggest to Revell (2009) that
different aspects of religious performance were emphasized by their attendant popula-
tions. Selectivity in ritual and its physical expression is also evident in the epigraphy
from Britain. The perception and reception of the new, Roman-​inspired, symbolic and
ceremonial repertoire depended upon the identity of the practitioner, the customs of the
community, and their individual and collective relationships within the larger imperial
system.
The evidence for religious activity in Roman Britain is a study in contrasts. Efforts to
construct a coherent narrative of belief from disparate remains tend to obscure social
and regional diversity while favouring certain subgroups and sources of informa-
tion. Continuity of belief from the pre-​Roman Iron Age is difficult to trace and liter-
ary accounts of provincial practice are lacking, leading to a reliance on epigraphy and
636

636   Amy Zoll

introduced rituals and votive forms. The territory is, however, comparatively inscrip-
tion-​poor, except in areas in the north and west long occupied by the Roman military,
where votive altars prevail. Non-​monumental inscriptions—​such as curse tablets depos-
ited in western sanctuaries and engraved rings from the east Midlands—​offer another
perspective on the gods of Britannia and their followers.
The written sentiments on votive monuments and personal offerings have clear ori-
gins in Roman practice, although differences have been detected in their authors, appli-
cations, and distributions. Vows and curses define the expectations and obligations of
gods and mortals through formalized exchange, but neither was observed by all inhabit-
ants equally. In Britain, inscriptions preserved on stone overwhelmingly represent the
religious activities of soldiers and their dependants, while other subsets of the populace
etched their pleas for divine justice on lead tablets or declared their devotion through
personal adornment. These objects offer an intimate view of how individuals and groups
articulated their social and spiritual identity within their respective communities. The
participants and trappings of a top-​down, centrally imposed, and officially encouraged
state religion may have influenced the form and execution of these modest monuments,
engraved exhortations, and other mementoes of votive practice, but did not dictate them.
The gods themselves were not static or absolute, but took on names and aspects that
accorded with the views and aspirations of their constituents. As such, these deities have
proved difficult to pigeonhole, with past structural approaches resulting in false dichoto-
mies that mask their complexity. Nowhere is this more evident than in the study of the
handful of instances in which divine beings were addressed by two names. These were often
seen to represent a snapshot of syncretism in action rather than one of many possible out-
comes of the multifaceted and ongoing processes of cultural negotiation. All gods of the
provinces, regardless of their origins, were subject to these forces and are best understood
within their social, temporal, and geographic context. So, too, the items that bear their
appellations must be regarded as contextualized objects, with forms, functions, authors,
and audiences, rather than simply disembodied texts. In this way, the kaleidoscopic variety
of divinities that comprise the known pantheon of Roman Britain can be seen to reflect the
needs of the faithful and their efforts to be heard by the gods they called by name.

Abbreviations
AE  L’Année Epigraphique. Paris: Ernest Leroux/​Presses Universitaires
de France, 1888–​to date.
CIL  Corpus Inscriptorum Latinorum.   Consilio et Ductoritate Academie
Litterarum Regiae Borussical Edition. Berlin:
Academieder Wissenschaften, 1862–​to date.
RIB   R. G. Collingwood, and R. P. Wright, The
Roman Inscriptions of Britain. Stroud: Alan
Sutton, 1992–​date.
 Names of Gods   637

References
Primary Sources
Lucanus [M. Annaeus], Pharsalie Libri X, ed. C. H Weise. Leipzig: G. Bassus, 1835.
Tacitus, De Germania, ed. H. Furneaux. Oxford: Clarendon Press, 1894.

Secondary Sources
Adams, G. (2006). ‘The Social and Cultural Implications of Curse Tablets [defixiones] in Britain
and on the Continent’, Studia Humaniora Tartuensia, 7: 1–​15.
Aldhouse Green, M., and Raybould, M. E. (1999). ‘Deities with Gallo-​British Names Recorded
in Inscriptions from Roman Britain’, Studia Celtica, 33: 91–​135.
Allason-​Jones, L., and McKay, B. (1985). Coventina’s Well:  A  Shrine on Hadrian’s Wall.
Hexham: Trustees of the Clayton Collection, Chesters Museum.
Ando, C. (2007). ‘Exporting Roman Religion’, in J. Rüpke (ed.), A Companion to Roman
Religion. Oxford: Blackwell, 429–​445.
Ando, C. (2008). The Matter of the Gods. Berkeley and Los Angeles:  University of
California Press.
Beard, M. (1991). ‘Writing and Religion:  Ancient Literacy and the Function of the Written
Word in Roman Religion’, in J. H. Humphrey (ed.), Literacy in the Roman World. Ann Arbor,
MI: Journal of Roman Archaeology Supplementary Series 3, 35–​58.
Birley, A. R., and Birley, A. (2010). ‘A Dolichenum at Vindolanda’, Archaeologia Aeliana,
39: 25–​51.
Biró, M. (1975). ‘The Inscriptions of Roman Britain’, Acta Archaeologica Academiae Scientiarum
Hungaricae, 27: 13–​58.
Cepas, A. (1989). The North of Britannia and the North-​West of Hispania:  An Epigraphic
Comparison. BAR International Series 470. Oxford: Archaeopress.
Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge.
Cunliffe, B. (1988) (ed.). The Temple of Sulis Minerva at Bath Volume II:  The Finds from
the Sacred Spring. Oxford University Committee for Archaeology Monograph 16.
Oxford: Oxford University Committee for Archaeology.
Daubney, A. (2010). ‘The Cult of Totatis:  Evidence for Tribal Identity in Mid Roman
Britain’, in S. Worrell, G. Egan, J. Naylor, K. Leahy, and M. Lewis (eds). A Decade of
Discovery: Proceedings of the Portable Antiquities Scheme Conference 2007. BAR British Series
520. Oxford: Archaeopress, 109–​120.
Derks, T. (1998). Gods, Temples and Ritual Practices: The Transformation of Religious Ideas and
Values in Roman Gaul. Amsterdam: Amsterdam University Press.
Goldberg, D. M. (2009a). ‘The Dichotomy in Romano-​Celtic Syncretism: Some Preliminary
Thoughts on Vernacular Religion’, in M. Driessen, S. Heeren, J. Hendriks, F. Kemmers,
and R. Visser (eds), TRAC 2008: Proceedings of the Eighteenth Annual Theoretical Roman
Archaeology Conference, Amsterdam 2008. Oxford: Oxbow Books, 187–​202.
Goldberg, D. M. (2009b). ‘Divinities and Ritual Sites of Rivers in Northern England and
Southern Scotland’. Unpublished Ph.D. thesis, University of Glasgow.
Gordon, R. (1990). ‘Religion in the Roman Empire: The Civic Compromise and its Limits’,
in M. Beard and J. A. North (eds), Pagan Priests: Religion and Power in the Ancient World.
Ithaca, NY: Cornell University Press, 235–​255.
638

638   Amy Zoll

Green, M. J. (1986). ‘Jupiter, Taranis and the Solar Wheel’, in M. Henig and A. King (eds),
Pagan Gods and Shrines of the Roman Empire. Oxford: Oxford University Committee for
Archaeology, 65–​75.
Green, M. J. (1989). Symbol and Image in Celtic Religious Art. London: Routledge.
Green, M. J. (1998). ‘God in Man’s Image: Thoughts on the Genesis and Affiliations of Some
Romano-​British Cult-​Imagery’, Britannia, 29: 17–​30.
Harris, W. V. (1989). Ancient Literacy. Cambridge, MA: Harvard University Press.
Haynes, I. P., and Wilmott, T. (2012). ‘The Maryport Altars: An Archaeological Myth Dispelled’,
Studia Universitatis Babes-​Bolyai, Historia, 51/​1: 25–​37.
Henig, M. (1984). Religion in Roman Britain. New York, NY: St Martin’s Press.
Henig, M. (1985). ‘Graeco-​Roman Art and Romano-​British Imagination’, Journal of the British
Archaeological Association, 138/​1: 1–​22.
Jackson, R., and Burleigh, G. (2007). ‘The Senua Treasure and Shrine at Ashwell (Herts)’, in
R. Haeussler, A. C. King, and P. Andrews (eds), Continuity and Innovation in Religion in
the Roman West. Portsmouth, RI: Journal of Roman Archaeology Supplementary Series 67,
i. 37–​54.
Johns, C. (1986). ‘Faunus at Thetford: An Early Latian Deity in Late Roman Britain’, in M. Henig
and A. King (eds), Pagan Gods and Shrines of the Roman Empire. Oxford: Oxford University
Committee for Archaeology, 93–​103.
Johns, C. (2003). ‘Art, Romanisation and Competence’, in S. Scott and J. Webster (eds), Roman
Imperialism and Provincial Art. Cambridge: Cambridge University Press, 9–​23.
Jones, B., and Mattingly, D. (1990). An Atlas of Roman Britain. Oxford: Blackwell.
Kiernan, P. (2004). ‘Did Curse Tablets Work?’, in B. Croxford, H. Eckhardt, J. Meade, and
J. Weekes (eds), TRAC 2003:  Proceedings of the Thirteenth Annual Theoretical Roman
Archaeology Conference, Leicester 2003. Oxford: Oxbow Books, 123–​134.
King, A. C. (2007). ‘Romano-​Celtic Temples in Britain:  Gallic Influence or Indigenous
Development?’, in R. Haeussler, A. C. King, and P. Andrews (eds), Continuity and
Innovation in Religion in the Roman West. Portsmouth, RI: Journal of Roman Archaeology
Supplementary Series 67, i. 13–​18.
MacMullen, R. (1982). ‘The Epigraphic Habit in the Roman Empire’, American Journal of
Philology, 103/​3: 233–​246.
Mann, J. C. (1985). ‘Epigraphic Consciousness’, Journal of Roman Studies, 75: 204–​206.
Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409.
London: Penguin.
Millett, M. (1995a). Roman Britain. London: Batsford/​English Heritage.
Millett, M. (1995b). ‘Re-​Thinking Religion in Romanization’, in J. Metzler, M. Millett, N.
Roymans, and J. Slofstra (eds), Integration in the Early Roman West: The Role of Culture and
Ideology: Papers arising from the International Conference at the Titelberg (Luxembourg)
12–​13 November 1993. Luxembourg:  Musée National d’Histoire et d’Art, Luxembourg,
93–​100.
Mirković, M. (1989). ‘Beneficiarii consularis and the New Outpost in Sirmium’, in V. A. Maxfield
and M. J. Dobson (eds), Roman Frontier Studies (1989): Proceedings of the XVth International
Congress of Roman Frontier Studies. Exeter: University of Exeter Press, 252–​256.
Raybould, M. E. (1999). A Study of Inscribed Material from Roman Britain: An Inquiry into Some
Aspects of Literacy in Romano-​British Society. BAR British Series 281. Oxford: Archaeopress.
Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge
University Press.
 Names of Gods   639

Richert, E. A. (2005). Native Religion under Roman Domination: Deities, Springs and Mountains
in the North-​West of the Iberian Peninsula. Oxford: Archaeopress.
Ross, A. (1967). Pagan Celtic Britain: Studies in Iconography and Tradition. London: Routledge
and Kegan Paul.
Scheid, J. (2003). An Introduction to Roman Religion. Bloomington, IN:  Indiana
University Press.
Stewart, P. (2010). ‘Geographies of Provincialism in Roman Sculpture’, Journal of the
International Association of Research Institutes in the History of Art, 5: 1–​59.
Stoll, O. (2007). ‘The Religions of the Armies’, in P. Erdkamp (ed.), A Companion to the Roman
Army. Oxford: Blackwell, 451–​476.
Tomlin, R. S. O. (1988). ‘The Curse Tablets’, in R. S. O. Tomlin (ed.), Tabellae Sulis: Roman
Inscribed Tablets of Tin and Lead from the Sacred Spring at Bath. Oxford: Oxford University
Committee for Archaeology, 59–​105.
Tomlin, R. S.  O. (1993). ‘The Inscribed Lead Tablets:  An Interim Report’, in A. Woodward
and P. Leach (eds). The Uley Shrines: Excavation of a Ritual Complex on West Hill, Uley,
Gloucestershire:  1977–​9. English Heritage Archaeological Report 17. London:  English
Heritage, 113–​130.
Tomlin, R. S. O. (2002). ‘Writing to the Gods in Britain’, in A. Cooley (ed.), Becoming Roman,
Writing Latin? Literacy and Epigraphy in the Roman West. Portsmouth, RI: Journal of Roman
Archaeology Supplementary Series 48, 165–​179.
Turner, E. G. (1963). ‘A Curse Tablet from Nottinghamshire’, Journal of Roman Studies,
53: 122–​124.
Versnel, H. S. (1991). ‘Beyond Cursing: The Appeal to Justice in Judicial Prayers’, in C. A. Faraone
and D. Obbink (eds), Magika Hiera:  Ancient Greek Magic and Religion. Oxford:  Oxford
University Press, 60–​106.
Webster, G. (1986). The British Celts and their Gods under Rome. London: Batsford.
Webster, J. (1995). ‘“Interpretatio”:  Roman Word Power and the Celtic Gods’, Britannia,
26: 153–​161.
Webster, J. (1997). ‘Necessary Comparisons: A Post-​Colonial Approach to Religious Syncretism
in the Roman Provinces’, World Archaeology, 28/​3: 324–​338.
Webster, J. (2003). ‘Art as Resistance and Negotiation’, in S. Scott and J. Webster (eds), Roman
Imperialism and Provincial Art. Cambridge: Cambridge University Press, 24–​51.
Whittaker, C. R. (1997). ‘Imperialism and Culture: The Roman Initiative’, in D. J. Mattingly
and S. E. Alcock (eds), Dialogues in Roman Imperialism: Power, Discourse, and Discrepant
Experience in the Roman Empire. Portsmouth, RI:  Journal of Roman Archaeology
Supplementary Series 23, 143–​163.
Woodward, A. (1992). Shrines and Sacrifice. London: English Heritage/​Batsford.
Woolf, G. (1996). ‘Monumental Writing and the Expansion of Roman Society in the Early
Empire’, Journal of Roman Studies, 86: 22–​39.
Woolf, G. (1997). ‘Polis-​Religion and its Alternatives in the Roman Provinces’, in H. Cancik
and J. Rüpke (eds), Römische Reichsreligion und Provinzialreligion. Tübingen:  Mohr
Siebeck, 71–​84.
Woolf, G. (1998). Becoming Roman:  The Origins of Provincial Civilization in Gaul.
Cambridge: Cambridge University Press.
Woolf, G. (2001). ‘Local Cult in Imperial Context:  The Matronae Revisited’, in P. Noelke
(ed.), Romanisation und Resistenz in Plastik, Architektur und Inschriften der Provinzen des
Imperium Romanum: neue Funde und Forschungen. Mainz: Zabern, 131–​138.
640

640   Amy Zoll

Woolf, G. (2003). ‘Seeing Apollo in Roman Gaul and Germany’, in S. Scott and J. Webster (eds),
Roman Imperialism and Provincial Art. Cambridge: Cambridge University Press, 139–​152.
Zoll, A. (1995a). ‘A View through Inscriptions:  The Epigraphic Evidence for Religion at
Hadrian’s Wall’, in J. Metzler, M. Millett, N. Roymans, and J. Slofstra (eds), Integration in the
Early Roman West: The Role of Culture and Ideology: Papers Arising from the International
Conference at the Titelberg (Luxembourg) 12–​ 13 November 1993. Luxembourg:  Musée
National d’Histoire et d’Art, 129–​137.
Zoll, A. (1995b). ‘Patterns of Worship in Roman Britain: Double-​Named Deities in Context’, in
S. Cottam, D. Dungworth, S. A. Scott, and J. Taylor (eds), TRAC 94: Proceedings of the Fourth
Annual Theoretical Roman Archaeology Conference. Oxford: Oxbow Books, 32–​44.
Zoll, A. (forthcoming). ‘A Vow Fulfilled: Inscribed Votive Altars from Britannia and Hispania
Citerior’, Ph.D. dissertation, University of Pennsylvania, Philadelphia.
 Chapter 31

Ritual Dep o si t i on

Alex Smith

Introduction

It is increasingly apparent that the nature of sacred space and forms of ritual expression
varied tremendously within the overall geographical and chronological parameters that
define Roman Britain. Even in terms of what may be regarded as ‘formal’ cult sites—​
that is, those that are spatially separate from domestic spheres and often associated with
prescribed architectural principles—​there is still a significant array of variants, which
frequently leads to interpretational issues. In fact, it is often only on the basis of the fre-
quency, type, and patterning of artefactual and ecofactual deposits within such contexts
that any consensus about a religious interpretation can be reached.
The concept of votive deposition within sacred sites was well established within
Roman religious traditions across the empire (Rives 2007: 24–​27) and indeed a num-
ber of classical authors (notably Caesar (6.13; 6.17), Strabo (4.1.13; 12.5.2), and Diodorus
Siculus (5.27)) attested to the display and deposition of objects within sacred loci in
pre-​conquest Gaul (see A. T. Smith 2001: 14, table 1.3). These ritual deposits have long
been recognized archaeologically on late Iron Age and Roman temple sites in France
and Britain (Wait 1985; Brunaux 1988; Woodward 1992; A. T. Smith 2001; King 2007b;
Bagnall-​Smith 2008), whether in the form of ‘structured’ sub-​surface deposits, concen-
trations of material associated with boundaries, or just in the general differentiation of
find types across the site. Nevertheless, interpretation of such deposits even within these
undeniably religious contexts is not always straightforward, and the likelihood is that
they represent the vestiges of a variety of methods of divine propitiation, from formal
cult rituals performed by temple priests to personal devotions, perhaps given in antici-
pation or recompense of answered prayers (Ghey 2007: 25).
Many of these offerings and the rituals surrounding their deposition were probably part
of a broader understanding of ritual expression throughout the Roman Empire, which,
although far from homogenous, had certain underlying shared characteristics quite pos-
sibly linked to the creation of wider cultural identities (Revell 2007: 212). In this respect,
642

642   Alex Smith

although the many different types of votive deposit seen in Romano-​British temple sites
could be viewed as an important element in the construction of local community identi-
ties (see Mattingly 2006: 520–52​2), they may also reflect—​in part at least—​something of
the wider understanding of what it meant to ‘act’ as Roman. Furthermore, this may be
expressed not only in the type and nature of votive offering, but also in the architectural
setting of the temple itself. The very concept of specialized constructed sacred space was,
with a small number of notable exceptions, one that only really developed in Britain after
the Roman conquest (A. T. Smith 2001: 67), and so, despite the extensive physical variety
in site types, many may be seen as local interpretations of a Roman ‘identity’.
In terms of temple offerings, it has been argued that many may have been deposited
as an element in the ritual of the vow, an established and formulaic method of com-
munication with the deity within the Roman world, for which evidence is well attested
in certain parts of Gaul and Britain (Derks 1995; Bagnall-​Smith 2008: 153–​154). The two
parts of this votum, a temporary pact made with the god, comprised the nuncupatio,
the promise to pay a gift in return for divine service, and the solutio, whereby the vow
is paid by the votary. Specific evidence for the nuncupatio part of the vow is generally
lacking in Britain, aside from the inscribed lead curse tablets found in quantity at Bath
and Uley (Bagnall-​Smith 2008), while the solutio part is attested by numerous inscrip-
tions of VSLM (votum solvit libens laetus merito), indicating the vow was paid, found in
parts of Britain, though mainly in military zones (Zoll 1995). The extent to which this
ritual of the vow was explicitly practised in other temples in Britain remains uncertain
(though inscriptions suggesting this have been found at five temples (A. T. Smith 2001:
156)), but the often considerable quantity of objects recovered from such sites that may
be regarded as votive offerings would suggest that at least the basic principle of ‘paying
off a debt’ to the god was acknowledged to a degree.
Of course, votive offering and deposition within a formal temple setting is only one
aspect of ritual expression within Roman Britain, albeit one that until comparatively
recently formed a significant focus of academic study in terms of Romano-​British reli-
gion. The other major body of evidence concerns deposition of metalwork and other
objects within natural watery settings, in particular rivers, lakes, and bogs. This has been
demonstrated to have had a long tradition stretching back into early prehistory (Bradley
2000: 47–​63) and certainly continued into the later Iron Age and Roman periods, with
examples of weapons, ceramic and pewter vessels, jewellery, and other items being
recorded as distinct deposits—​for instance, within the rivers Thames (Booth et al. 2007:
217–​220) and Witham (Field and Parker Pearson 2003: 171–​178).
Aside from such ritual activity associated with what may be regarded as sacred sites,
whether constructed shrines or natural watery features, and following on from the
work of Hill (1995) and Cunliffe (1995) on ritual deposition within Iron Age settlements,
Fulford (2001) recognized that ‘structured’ or ‘special’ deposits were also widespread
within Roman domestic contexts, and not just within rural settlements but also within
urban centres alongside all of the overt trappings of Romanitas. Further work by Clarke
(2001; forthcoming) and Hingley (2006) has sought to go further and identify patterns
within such deposition in terms of physical location (for example, ‘boundary deposits’),
 Ritual Deposition   643

in the types and materiality of objects and the ways in which they were deposited, and in
combinations of ‘object and place’, which, it is argued, may have ultimately determined
the power of the votive action (Hingley 2006: 239). The ritual significance within domes-
tic contexts of different categories of finds (for example, ornamentation: Puttock 2002)
and especially animal species (for example, dogs: K. Smith 2006; corvids: Sergeantson
and Morris 2011; see also Black 2008 and Morris 2011) has also been increasingly dis-
cussed in recent years, and this has all led to a much greater recognition of such unusual
deposits within the archaeological record, as will be shown below.
In the main part of this chapter it is the intention to provide an overview of the prac-
tice of ritual deposition within both formal sacred sites (excluding the body of evidence
from natural watery sites) and the ‘structured deposits’ in ostensibly secular settlements
within Roman Britain. In both cases, there are some similarities in the types of objects
and the ways in which they were deposited, whether this is in terms of articulated ani-
mal remains or deposition of whole pots, and yet there are also clearly many differences,
which may be culturally significant. Furthermore, there is not always a clear differentia-
tion between the different types of ritual expression, with, for example, purpose-​built
ritual pits and shafts being located within and around secular settlements and classic
‘structured deposits’ within disused wells and pits associated with temple sites. As ever,
when it comes to personal religious experience in the Roman world, the boundaries
between secular and divine were consistently blurred.

Ritual Deposition Within

Temple Contexts

As stated above, it has long been recognized that, compared to the post-​conquest
period, evidence for constructed cult sites in late pre-​Roman Iron Age Britain was quite
minimal, with a number of notable and in many ways unusual exceptions (for exam-
ple, Hayling Island, Uley, etc.: A. T. Smith 2001: 67). While there continues to be new
discoveries, such as at Snows Farm (Evans and Hodder 2006) and Marcham/​Frilford
(Kamash et al. 2010), that provide further potential evidence for the existence of pre-​
Roman religious foci (see below), the rarity of Iron Age shrines—​or at least their lack of
recognition in the archaeological record—​remains undeniable. Yet within fifty years of
the conquest, constructed temples were a regular part of the landscape, at least in many
parts of southern and eastern Britain, as ubiquitous as roads, towns, and villas. Indeed,
in looking at the finer chronological details, such sacred sites were often among the first
substantial pieces of architecture to be constructed in these regions during the post-​
conquest period (A. T. Smith 2001: 144).
There are two fundamental factors in the interpretation of formal cult sites in Roman
Britain: recognition of standardized forms of religious architecture (for example, the
Romano-​ Celtic temple) and/​ or architectural elements conforming to underlying
644

644   Alex Smith

religious spatial principles (for example, emphasis on enclosure and entranceway, path-
ways, and so on), and perhaps most importantly in the type, volume, and patterns of
finds associated with the site. The earliest systematic studies of Roman temples in Britain
(e.g. Lewis 1966; Wilson 1975) tended to concentrate on matters of architectural detail
and classification, and it was not really until the publication of Temples, Churches and
Religion in Roman Britain (Rodwell 1980) that the nature of finds within these sacred
sites really began to be explored at any level. Since then there have been various wider
studies of votive deposition within Romano-​British temple sites (e.g. Woodward 1992; A.
T. Smith 2001; King 2007b; Bagnall-​Smith 2008), in addition to specific analyses of par-
ticular categories of deposit within such formal religious contexts (for example, coins:
Haselgrove 2005; King 2007a; Wythe 2007; animal remains: King 2005; K. Smith 2006).
Furthermore, there have been a number of important recent publications on excava-
tions at sacred sites, often with significant details of ritual deposition, such as at Snows
Farm (Evans and Hodder 2006), Hallaton (Score 2011), Higham Ferrers (Lawrence and
Smith 2008), Springhead (Andrews et al. 2011), and Great Chesterford (Medlycott 2011).
Recent excavations at Marcham/​Frilford in Oxfordshire have revealed extensive parts
of a huge rural religious complex mostly outside of the temple and temenos (religious
precinct) area, which have demonstrated the presence of a range of different ritual foci;
this is helping us to understand something of the wider functions and intricacies of such
sanctuaries (Kamash et al. 2010). The following section will first look briefly at the gen-
eral character of votive offerings most commonly found at cult sites, before exploring
some of these newly excavated and/​or published sites in terms of their contribution to
our understanding of votive deposition. In addition, the results of some previous analy-
ses of sites such as Uley will be summarized, as they remain fundamental to studies of
ritual practice within Romano-​British temples.

Votive Offerings Within Cult Sites

In a previous study by the current author of seventy-​five temple sites across southern
Britain, the major categories of votive artefacts were quantified and discussed (see
Table 31.1).
The first important observation to make is that there was a significant variety in the
numbers and types of votive objects between temple sites, undoubtedly reflecting the
nature of the individual cult as well as the personal wealth and aspirations of the sup-
plicant. There is some evidence for wider regional patterns (for example, relative lack of
coins within Sussex temples), but even here there are many discrepancies, and overall
it seems clear that there were no widespread regulations concerning the types of object
that could be offered to the deities. Nevertheless, there are certain artefact classes that
would seem to have been regarded as more appropriate votive material, and the most
ubiquitous find type within temple contexts is coinage, often found deposited in large
quantities (for example, Uley, Harlow, Wanborough). There have been various recent
 Ritual Deposition   645

Table 31.1 Occurrence of selective votive objects in southern Romano-​British

cult sites
Occurrence
Votive Item (no. of sites out of 75)

Coins 60 (29)
Personal ornamentation and toiletries 46 (20)
Martial items 25 (4)
Items associated with healing 8 (1)
Miniature items 20 (3)
Votive/​miniature pots 14 (3)
Votive plaques/​leaves 13 (3)
Iconography 44 (12)
Inscriptions (including curse tablets) 23 (5)

Note: Bold figures in parentheses denote sites where items were found in significant comparative
quantity.
Source: After Smith (2001: 155, figure 5.13).

studies on the deposition of both Iron Age (Haselgrove 2005) and Roman (King 2007a;
Wythe 2007) coins in temple contexts, where it was noted that Iron Age coins in par-
ticular were particularly prevalent deposits at certain sacred sites around the time of the
Claudian conquest, and that, spatially, the coins from these sites were concentrated in
front and to the left of the cult focus (Haselgrove 2005: 417). In terms of Roman coins,
their deposition occurred right up until the end of the fourth century ad, and indeed it
was observed that coins continued to be deposited at temples long after they had struc-
turally fallen to decay, a testament to the sacred memory of such sites (King 2007a: 37).
Other important categories of ‘everyday’ objects used as votive deposits at temple sites
include personal objects such as brooches, finger rings, and hairpins. In a study concerned
with the ritual significance of jewellery in Roman Britain, Puttock (2002: 115) suggested
that such items deposited in ritual situations carried the hopes and prayers of the people,
and it was not only the object type itself that was significant, but also the shapes, colours,
motifs, and materials of the ornaments. At its most obvious, this is probably seen with the
prevalence of horse and rider brooches at a number of religious sites such as Lamyatt Beacon
and Woodeaton (Leech 1986; Bagnall-​Smith 1995). It is further argued that the symbolism
behind the jewellery was often connected with fertility, healing, or protection from evil, with
certain categories (for example, beads, hairpins, and bracelets) being particularly associated
with women and linked with aspects such as childbirth and fertility (Puttock 2002: 115).
Of course, the ritual significance of objects such as coins and personal items is only
readily apparent when they are found in such unequivocal religious contexts as temple
646

646   Alex Smith

sites, except perhaps when they form part of recognized structured deposits within set-
tlements. There are, however, many types of object that seem to have been fabricated
purely with religious associations in mind, and these form the very wide category of
‘religious objects’, ranging from inscribed altars and curse tablets to figurines of deities,
leaf/​feather plaques, and objects of priestly regalia (see Bagnall-​Smith 2008 for a use-
ful summary of these types). The type and occurrence of such items varies enormously
within temple sites across Britain, although there are several characteristics that are
more widespread, most notably that of miniaturization. Miniature objects can take a
variety of forms, with tools and weapons the most common, especially axes that are
found on temples and shrines across Britain. Kiernan (2009: 211–​213) has suggested
that, far from having some universal significance implied by their miniaturization, these
models carried a variety of meanings, with some being direct substitutes for real objects,
others being indirect substitutes for ritual activities or human activities such as sacri-
fice (axes) or travel (shoes, anchors), and finally others being purely symbolic and taken
from the iconography of a divinity (for example, wheels).
An act of ritual expression that seems to occur across a range of intrinsically reli-
gious and ‘normal’ votive objects is that of destruction. This is a widespread phenom-
enon both in Iron Age and Roman contexts and is generally believed to be a sacrificial
act, removing the object from earthly existence and transferring it to the divine realm
(Green 1995: 470–​471). Classic examples of this ritual have recently been found at the
shrine at Higham Ferrers in Northamptonshire, including a bone pin whose stem was
broken and partially split and a brooch whose pin had been detached and deliberately
wrapped around the bow (Lawrence and Smith 2008: 331).
Perhaps the most significant ritual act within many sacred sites would appear to be
the sacrifice of living animals. Animal bone usually forms one of the largest single cate-
gories of find within temple sites, with over 50,000 recorded from Uley, Gloucestershire
(Woodward and Woodward 2004) and over 32,000 from the relatively small shrine
at Snows Hill, Cambridgeshire (Evans and Hodder 2006). In a recent analysis of ani-
mal remains from temples in Roman Britain, King (2005) noted a number of different
characteristics, with some temples having large numbers of bones of a distinctive spe-
cies with evidence for regular periods of sacrifice (presumably at festivals), others with
smaller quantities of bone, often in structured deposits (see below), and some with very
few bones, often those temples associated with healing such as at Bath. The distinction
between those articulated skeletons within structured sub-​surface deposits and the
mass of unarticulated bone (presumably the remains of ritual feasting) is of great impor-
tance, and is explored further later in this chapter.

Offering Zones

There are a number of inherent problems in identifying patterns of votive deposi-

tion within cult sites, including a lack of excavation within the temenos, selective (and
 Ritual Deposition   647

non-​spatial) recording of those finds encountered, and more broader issues of redepo-
sition. Nevertheless, there are sufficient excavated sites with enough spatially specific
information to gain some understanding of the general patterns of finds distribution
(A. T. Smith 2001; King 2007b: 191–​193). In the first instance, there is a broad division
between those finds that were intended to be deliberate deposits in a sub-​surface envi-
ronment (for example, pits, ditches, and so on), and those associated with the occupa-
tion surface, which are by their nature often far more disturbed and difficult to interpret
unless there is an excellent state of preservation. Within one of the Springhead temples
in Kent, for example, there were a number of votive finds recorded as still apparently
lying in situ upon the temple cella floor, including a miniature axe and a Venus figu-
rine (Penn 1962: 116). If this was indeed the case, there is good reason to argue that the
finds were originally displayed here and remained in their primary context, whereas it
is likely that in most cases there was significant redeposition of votive objects that may
have been on display before being removed and disposed of elsewhere, either within the
temple precinct or beyond.
When looking at the overall finds distribution patterns within temple complexes, it
is usually possible to obtain only quite a generalized impression of potential ‘offering
zones’ (that is, areas for display and/​or deposition of votive offerings) that undoubtedly
mask a complex array of rituals enacted at the cult site. Nevertheless, it remains that in
the majority of sites studied, it is—​perhaps unsurprisingly—​the temple building itself,
and especially the central cella, that formed the main focus for the display and/​or final
deposition of votive objects (A. T. Smith 2001: 154). Coins were particularly prevalent
finds in this location, perhaps being displayed before the main cult image (see Wythe
2007: 47–​50 for a detailed analysis of coin deposits in temple buildings). This is seen, for
example, within the Romano-​Celtic temple at Bourton Grounds in Buckinghamshire,
where 90 per cent of the 315 coins came from the southern half of the cella, interpreted
by the excavators as the remains of offerings given to the deity through openings in the
cella walls (Green 1966: 359). Even within less elaborate sacred sites, such as the simple
circular shrine at Claydon Pike, Gloucestershire, which lay close to a modest late Roman
villa, there was a distinct concentration of objects within the interior of the shrine build-
ing, including the largest single collection of coins from the site (Miles et al. 2007: 183).
However, variations to this do of course exist, as King has recently pointed out in his
study of coins from temples in Britain, where he noted concentrations of coin deposi-
tion in zones to the south-​east of the temple in at least three instances, at Hayling Island,
Bath, and Wanborough (King 2007a: 26–​28).
A recently excavated shrine on one side of the Nene valley within a roadside settlement
at Higham Ferrers in Northamptonshire comprised a walled enclosure (though open
towards the river), monumental entrance, and two conjoined precincts, though with-
out any evidence for an actual temple structure (Lawrence and Smith 2008). The inner
precinct in particular contained a substantial volume of votive objects including coins,
personal items, and metal leaves, in addition to animal remains, all lying relatively undis-
turbed and in situ on top of the Roman occupation surface. The finds were concentrated
in the south-​east of the precinct (which is perhaps significant, given the observations of
648

648   Alex Smith

King already noted), particularly around a 3-​metre-square area largely devoid of objects,
which was suggested as space reserved for a cult focus of some kind (ibid., 332). In this
respect the open inner precinct probably acted as a temple cella much the same as in a
traditional Romano-​Celtic temple. The quantity of iron nails within the inner precinct
suggested the presence of wooden structures of sorts, to which offerings would perhaps
have been attached, and the large number of hobnails within the precinct, which show a
similar distribution, may have been used for making such attachments.
In addition to the main cult focus within the heart of the sacred site, other notable
offering zones within sanctuaries, where there were often distinct concentrations of
finds, included entrances to both the temple and the outer precinct, the area in front of
the temple, sometimes on either side of a defined pathway, and in the proximity of the
temenos boundary (A. T. Smith 2001: 154). In many ways this reflects the fundamen-
tally important architectural principles of sacred sites in different cultures, where there
is often a distinct emphasis on the passage from the first transitional point of the outer
boundary through to the prime religious focus of the site (Barrie 1996: 148).
Following metal-​detector discoveries of large numbers of Iron Age coins on a hilltop
location at Hallaton, Leicestershire, excavations revealed what appears to have been an
open-​air shrine dating from the mid to late first century bc to the early post-​conquest
period, with distinct concentrations of finds and animal bone around the east-​facing
entranceway within an outer ditch (Score 2011: 156–1​60). As with Higham Ferrers, there
appears to be no evidence for a temple building, but the excavators have suggested a pro-
cessional way, taking into account prehistoric features in the landscape, leading through
the entrance in the ditch, which has been interpreted as a potential enclosure surround-
ing the hilltop. The finds included 4,943 Iron Age coins (and 149 Roman coins dated
pre-​ad 43), buried in a series of separate deposits on one side of the boundary ditch near
the entranceway, along with a remarkable Roman cavalry parade helmet, a silver bowl,
two ingots, and a substantial animal bone assemblage (6,929 identified fragments), 97
per cent of which was young pig. This was clearly a site of unusual significance, and the
wealth of finds along with the timeframe of deposition may well be associated, in part,
with the stresses of the Roman conquest, but nevertheless the clear concentration of
votive activity along the boundary and entranceway conforms to well-​established pat-
terns in terms of offering zones.
A similar emphasis of ritual deposition has been found during recent excavations at
the vast religious complex at Marcham/​Frilford in Oxfordshire, dating from the late first
to late fourth century ad, but with possible evidence for an earlier ritual focus of Iron
Age date (Kamash et al. 2010: 98–​102). Previous excavations of the temple had shown
that robbing and plough damage had destroyed most of the interior levels, but con-
centrations of coins were noted on the temple pathway (Bradford and Goodchild 1939:
32), and this was again encountered during the recent excavations, in addition to large
quantities of coins being focused on the monumentalized entrance into the temenos
(Kamash et al. 2010: 104–​105). Preliminary analysis of wider finds distribution at this
site has produced interesting results, with concentrations of coins, fineware pottery,
and vessel glass immediately outside the temenos, the latter two categories suggesting
 Ritual Deposition   649

feasting took place in this zone (Kamash et al. 2010: 106). Furthermore, there were also
notable concentrations of personal objects around a small subsidiary shrine and an area
of boggy ground to the south of what is described as a semi-​amphitheatre, lying c. 90
metres east of the temenos entrance, and it is clear that full analysis of the distribution of
objects from this site will be of paramount importance in furthering our understanding
of Romano-​British ritual practices within sacred sites beyond the confines of the temple
and precinct.
A ritual site with similar evidence for longevity, from the late Iron Age to the fifth
century ad, was the exceptionally well-​recorded temple at Uley in Gloucestershire. This
site contained a huge volume of finds that had all been spatially recorded, and in which
a number of offering zones were discerned, despite the inherent problems of disturbed
stratigraphy and significant redeposition (Woodward and Leach 1993; A. T.  Smith
2001: 106). The temple cella itself was clearly a focus for deposition, and the spreads of
votive material (which included miniature pots, lead curse tablets, and votive leaves
along with many coins and personal items) deposited along the demolished buildings
representing the inner boundary to the cult site are argued to have come from a clearing-​
out of the temple when it was substantially modified in its final phase.
Similar acts of redeposition are suggested for other Roman sanctuaries, including at
Marcham/​Frilford, where a large pit and midden in the south-​east corner of the temenos
contained deposits rich in oyster shell, animal bone, and pottery, thought to have orig-
inally been deposited elsewhere within the sacred precinct (Kamash et al. 2010: 103).
At the Henley Wood temple in Somerset, where substantial quantities of votive objects
were also recovered and spatially recorded, several distinct patterns were revealed, with
coins again being concentrated in the temple building, as well as adjacent to a possible
entranceway into the site and on either side of the main axis leading from here towards
the temple (Watts and Leach 1996; A. T. Smith 2001: 92). Items of personal adornment
(brooches, bracelets, and so on) had a slightly different depositional pattern, with most
being found within the outer ditch (taken as the temenos boundary), often in distinct
concentrations, and interpreted by the excavators as representing episodes of rede-
position of objects that had originally been displayed in the temple (Watts and Leach
1996: MF 515). Whether or not this was the case, the spatial distributions of different
object types suggests that these were not random dumps of unwanted material, but care-
fully structured deposits, of a type that are found in sub-​surface features (for example,
ditches, pits, and shafts) within many temple sites.

Structured Deposits Within

Temple Contexts

Determining the location of offering zones within a cult site is reliant upon analysis of
the distribution of votive objects. As already discussed there is a distinction between
650

650   Alex Smith

those objects associated with occupation surface/​spreads, which may be either casual
losses, the remains of items left ‘on display’, or else the product of post-​depositional
processes (alluviation, colluviation, ploughing, and so on), and those objects deliber-
ately deposited in pits, ditches, and so on, probably involving prescribed sets of ritu-
als. In many ways, the structured deposits within sacred sites are easier to interpret,
although may still represent the final stages of a myriad of different rituals associated
with a variety of different beliefs. The items within such deposits, whether the remains
of animals or artefacts, were probably regarded as appropriate messages to the gods, and
placing them within specific sub-​surface contexts may have been part of their transfor-
mation on the journey from the earthly world to that of the divine (Aldhouse-​Green
2001: 24). In the context of the discussion of the animal remains at the shrine at Snows
Farm, Haddenham, Cambridgeshire, the excavators suggested that the different types
of deposit on site might reflect different ritual acts, with those animals deposited in pits
relating to rites of intercession (that is, asking the gods for favour or paying off a divine
debt) and the slaughter of animals alone (without burial) only intended for divination
(Evans and Hodder 2006: 358). The animal remains at this shrine, which dated from the
second to fourth century ad (and with a potential Iron Age precursor) were particularly
prevalent (32,933 fragments), with an emphasis on sheep and evidence for the selection
of particular body parts in places (Evans and Hodder 2006). Two zones were selected
for structured sub-​surface deposits, with the floor area of the shrine containing a single
articulated sheep (interpreted as a foundation deposit) and a number of ‘head and hoof ’
burials with coins placed in the jaws, while the north-​west corner of the precinct con-
tained pits with articulated sheep skeletons associated with complete pots (Evans and
Hodder 2006: 336). These deposits represented different types of rituals undoubtedly
with different purposes, and all again were different from the mass spread of bones and
other objects in the zone around the front of the shrine and perimeter of the enclosure,
which are suggested as the remains of animals consumed in ritual feasting, which had
been carefully redeposited from a previous midden at a time when the shrine under-
went significant structural changes (Evans and Hodder 2006: 413).
Large numbers of animal remains were also recovered from the temple precinct at
Great Chesterford, Essex, which lay around 1 kilometre east of the walled Roman town
(Medlycott 2011). The overwhelming majority of bones were those of lambs, which dis-
played a kill-​off pattern indicative of regular periods of sacrifice, probably at festivals,
similar to the situation at other temples such as Harlow (Medlycott 2011: 84). The great-
est concentration of bones came from a group of pits in the south-​west corner of the
temenos, where it was estimated that there were the remains of over 1,000 sheep. How
far these may be viewed as structured deposits is uncertain, as there was no evidence of
articulation or patterning, and, as they seemed to have been dumped with oyster shell
and pottery (and few other finds), they are more likely to be the remains of ritual feast-
ing, as was suggested for the dump of material in pits in the south-​east corner of the
temenos at Marcham/​Frilford, as already mentioned. They were certainly very differ-
ent in nature to the groups of finds lying apparently just beneath the temple pathway,
which comprised many personal objects and specifically votive items such as bronze
 Ritual Deposition   651

feathers and a silver mask plaque, which had been deliberately folded before deposition
(Medlycott 2011: 85).
A recently excavated and published temple complex with extensive evidence for
structured deposition is that at Springhead in Kent, on the line of Roman Watling Street
(Andrews et al. 2011). The principal walled temple complex, comprising two Romano-​
Celtic temples and at least six other structures, was excavated during the 1950s and
1960s, while the recent excavations revealed substantial additional areas of the site,
which may best be described as a major religious complex incorporating domestic,
commercial, and possibly administrative functions (Andrews et al. 2011: 212). A late
Iron Age religious focus is hinted by the deposition of Iron Age coins in the spring and
structured deposits of pottery vessels and a horse burial in pits below a ‘processional’
way on the hill slope above the spring. During the second century ad, the area around
the spring was transformed into what has been termed the sanctuary complex, with a
shrine, a portico structure, free standing posts, and a number of pits or shafts containing
structured deposits. These included a 4.5-​metre-​deep shaft by the entrance to the sanc-
tuary, which contained at least twenty dogs, several buried with their chains, a number
of near-​complete pots, a human skull, a group of animal skulls, and a cow placed in the
bottom (Andrews et al. 2011: 80–​82, figure 2.55).
The ritual deposition of dogs in pits and shafts is now a well-​known phenomenon in
Iron Age and Roman Britain, occurring in both what may be considered secular and
religious contexts (K. Smith 2006). Perhaps the most striking examples from what may
be regarded as a sacred site is at Cambridge, where a series of at least thirteen ritual
shafts (2–​3 metres deep) dating to the third–​fourth centuries were located in the cen-
tre of the small town, adjacent to an unusual second-​century shrine, which had burnt
down (Alexander and Pullinger 2000). Each shaft contained the articulated remains of
a mature small dog in the southern corner and a rush matt and wicker basket containing
an infant burial laid on large pot sherds. In most shafts two burials occurred one above
the other and in five shafts were the remains of small shoes that would have fitted an
older child. Although these shafts were filled in immediately after the infant burials, it
is likely that they were not elaborate funerary shafts, and yet their deposition along with
the dogs would probably have still been regarded as very powerful ‘magic’ or messages
to the gods.

Ritual Deposition Within

Settlement Contexts

It has already been mentioned that the boundaries between secular and divine were
constantly blurred within Romano-​British society, and that structured deposits of an
assumed ritual nature were not just confined to designated sacred sites. Since Fulford’s
paper in 2001 in particular, the recognition of ritual deposition (often termed ‘special
652

652   Alex Smith

deposits’) in Romano-​British settlement contexts has increased dramatically, to the

point where care must be taken not to over-​interpret every single articulated animal or
complete pot as ritual in nature, to the detriment of potentially more prosaic explana-
tions (see Morris 2011: 182–18​5). Fulford (2001: 201) defined the evidence for ritual as
having both a repetitive nature and ‘irrational’ characteristics, and in this sense there
certainly are considerable numbers of archaeological examples throughout Britain,
though, as Clarke (forthcoming) has recently asserted, the overall banner of ‘ritual’
undoubtedly masks a diverse range of actors and motivations. Broad patterns are cer-
tainly noted in many aspects of these deposits, but, as Hingley (2006: 239) states, ‘each
discovery represents a unique act of deposition and should be studied and interpreted
accordingly’.
Examples of what would appear to have been ritual deposition have been demon-
strated extensively in urban, rural, and military contexts, with Fulford (2001) looking
primarily at such deposits with Romano-British towns, drawing on data from Silchester,
London, Neatham, Baldock, and Verulamium. In the latest Silchester publications on
the mid and late Roman occupation of the town, it has been demonstrated that ‘spe-
cial deposits’ in pits and wells occurred throughout the excavated insula, though there
were chronological differences in spatial patterning and assemblage composition, and
the difficulties in defining them as ritual in nature were duly acknowledged (Fulford et
al. 2006; Fulford and Clarke 2011). One third-​century pit was particularly noteworthy,
containing one complete and five partial dog skeletons, in addition to partial raven skel-
etons and a folding knife or razor with an ivory handle in the shape of two coupling dogs
(Fulford and Clarke 2011: 311). Other possible ritual deposits from Silchester seem asso-
ciated with the construction of buildings, including complete pots (one of considerable
antiquity) in the make-​ups of two adjacent corners of a structure and an iron mason’s
trowel incorporated into one wall. Significantly, some of these pots appear to have been
deliberately pierced, thus perhaps undergoing the rite of sacrifice, as noted above for
many votive deposits in shrines. The motivations behind such deposits would surely
have been different from those associated with the pits, perhaps as suggested by the
excavators ‘to ensure the life, safety and security of the buildings’ (Fulford and Clarke
2011: 333).
In rural contexts, ‘special deposits’ have been interpreted at a growing number of set-
tlements, as seen, for example, in Upper Thames Valley sites such as Horcott Quarry
and Cotswold Community, which have examples of articulated animal remains in pits
and ditches that could be regarded as being deposited for spiritual reasons (Powell et al.
2010: 135; Hayden et al., forthcoming;). Elsewhere, Rudling (2008: 121–​129) has reviewed
the evidence for religious practice across Sussex, including a number of ‘special depos-
its’ within rural settlements, such as a large number of dog remains within a well at
Chilgrove villa and burials of dogs and complete pots in pits at the recently excavated
villa at Barcombe.
One factor to consider throughout all of these examples of ‘special deposits’ within
non-​religious contexts is the extent to which the pits or shafts were dug for the specific
purpose of ritual activity. In the case of the Silchester examples, it would appear that
 Ritual Deposition   653

most were dug originally as wells or latrine pits, and subsequently used for ritual pur-
poses, while, at the other extreme, it is likely the shafts from the Roman small town at
Cambridge noted above were dug specifically to house their unusual deposits before
being rapidly filled in. This would seem in many ways to cross the blurry divide between
sacred site and secular settlement. Ritual shafts, especially in southern Britain, have
long been studied (notably Ross 1968; M. J. Green 1986), and in a sense have come to be
regarded as a sub-​class of sacred site, especially when apparently isolated, such as the 80-​
metre-​deep shaft at Findon in the Sussex Downs, which contained the remains of many
dogs (Rudling 2008: 118). The extent to which other specifically dug ritual pits/​shafts
can be regarded as shrines in themselves is uncertain, but perhaps we should not be
so worried about contemporary interpretational classifications, when it is unlikely that
such divisions were always made in the past. A recently excavated site at Bretton Way,
Peterborough, is illustrative of such blurred divisions, as a substantial late-​third-​century
aisled building lay close by to a contemporary pit lined with reused monumental stone
blocks, which contained a large assemblage of pottery (including complete vessels) and
animal remains (including a dog skull and antler), along with coins, personal items, and
leather shoes (Pickstone 2011). The pit seemed to serve no other obvious purpose than
to receive these deposits, and the use of the huge stone blocks in its construction implies
a considerable undertaking, as would perhaps befit a special offering place to the gods.

The Character of ‘Special Deposits’

Within Settlements

The character of votive deposits within sacred sites has already been discussed, with a
noted concentration of coins and personal items, along with objects of an intrinsic reli-
gious nature. Although the first two categories at least do certainly appear as part of
‘special deposits’ within domestic contexts, they are generally much less common, and
instead there is a far greater concentration of items such as pottery and in particular
animal remains, often articulated. Of course, animal remains are also present in con-
siderable quantities within many sacred sites, but their prevalence in domestic ‘special
deposits’, and more importantly the (apparently) comparative lack of other typical tem-
ple offerings may suggest a different emphasis of ritual expression, perhaps one more
rooted in indigenous practices.
The widespread presence of dog skeletons in ‘special deposits’ is particularly note-
worthy. In addition to the dog remains at Silchester, others in unequivocally ‘non-​
religious’ urban contexts include a fourth-​century pit within a suburb of Roman
Winchester with a minimum of eight dogs along with a raven skeleton and ten com-
plete or near complete colour-​coated beakers (Maltby 2010a). Also, a remarkable col-
lection of over eighty mostly adult dogs were found within pits in ‘backyard’ enclosures
in the Roman town at Dorchester (Dorset), interpreted as urban foundation deposits
654

654   Alex Smith

(Woodward and Woodward 2004; though see Maltby 2010b, for a critique of this inter-
pretation). In a rural context, the villa at Keston in Kent was associated with eight pits
or shafts, many of which contained complete or partial dog skeletons, along with other
animal parts and often complete pots (Philp 1999). Finally, in the context of military
sites, the deposits within 107 deep pits and wells at the late-​first–​second century fort
at Newstead in the Scottish borders have been studied extensively, and include quite
a number of dog skulls, in addition to a huge range of other material, likely to have
been deposited as part of a range of different ritual acts, no doubt for different purposes
(Clarke, forthcoming).
The interpretations of articulated animal remains within ‘secular’ contexts has
recently been discussed by Morris (2011: 159–​162), who noted that the Romano-​British
period had by far the largest number of dog ABGs (Associated Bone Groups) com-
pared to other periods. Although functional interpretations for such deposits such
as culling or simply dying of old age are used occasionally (see Maltby 2010b), Morris
(2011: 162) asserts that most are explained in terms of ritual activity, although whether
or not this implies deliberate sacrifice is often unstated. K. Smith (2006: 24) pointed
out, however, that there does not necessarily have to be strict division between prac-
tical solutions such as disposing of culled or naturally dead animals and religiously
motivated acts.
If we assume that most of the dog ABGs found in pits at places such at Cambridge,
Silchester, Winchester, Dorchester, Keston, and Newstead have been deposited as
the result of ritual activity of some kind (notwithstanding the reservations of Morris
(2011) and Maltby (2010b)), then the question arises as to why such a species was
specifically selected, and why they were particularly associated with pits, wells, and
shafts. In the Greco-​Roman world, dogs have known associations with both the
underworld and healing, and a number of probable healing temples in Britain con-
tained some association (for example, bones or imagery) with these animals (for
example, Lydney Park, Pagans Hill, and Springhead). Black (2008: 2) has recently
suggested that many of the Roman ritual shafts containing dog skeletons were dedi-
cated to the Gallic deity Sucellos (or a local version thereof), patron of crafts and the
underworld, though this is more difficult to substantiate across a wider spectrum of
sites across Britain.
There are, of course, many other animal species noted in ‘special deposits’, including
all the main domesticates. However, other more unusual animals, not often recorded in
temple assemblages, comprise wild birds and in particular corvids (ravens, crows, and
so on). In fact, the combination of articulated canine and corvid remains noted above at
Silchester and Winchester is one that is seen increasingly in other structured deposits at
Iron Age and Roman sites across Britain (Sergeantson and Morris 2011). It is suggested
that the unique character of these birds, including their tolerance of humans, scavenging
habits, and their distinct voice, is what made them particularly suitable as messengers to
the gods (Sergeantson and Morris 2011: 99). A similar link with humans and scavenging
is, of course, also found with dogs, which may go some way to explain their association
with corvids in structured deposits.
 Ritual Deposition   655

Conclusions

It has been the intention of this chapter to provide an overview of the practice of ritual
deposition in Roman Britain, and it is immediately apparent that our understanding of
some aspects has increased significantly over the past 15–​20 years, both because there have
been a number of new excavations and/​or publications and also because a more contex-
tual view has been taken of the evidence. Religious sites of all types, from extensive temple
complexes such as Marcham/​Frilford to simple rural shrines like Snows Farm, and sites
like Hallaton and Higham Ferrers where no actual temple buildings existed, have demon-
strated the huge architectural variety in cult expression, albeit with a few more consistent
underlying conventions. But, perhaps more importantly, detailed analyses of their finds
composition and deposition is enabling us to start asking new questions of the types of
rituals practised at these sites, and even something of the possible beliefs behind them.
Studies on the ritual significance of specific object types and animal remains have demon-
strated their potential to go beyond the label of mere ‘votive deposit’ to gain some insight
into the possible motivations of the supplicant and perhaps even their use in the creation
of cultural identities. In particular, it has been recognized that any understanding of the
motives behind such rituals relies upon detailed analysis of both object and place.
The most significant advance in our understanding of Romano-​British ritual expres-
sion since 2000 has been concerned with looking outside of the sacred site, with the rec-
ognition of ‘special deposits’ within settlement contexts. In some ways, given that it has
often previously been acknowledged that ‘religion was everywhere’ in both the Iron Age
and Roman worlds, it is surprising that the ubiquity of these deposits had not been widely
commented upon before, yet such deposits are now interpreted, or at least speculated
upon, in a great many recent excavation reports on Roman settlements. As with the ritual
deposition observed within sacred sites, it is only through detailed contextual analysis
of object and place, as, for example, has been attempted at Silchester and Newstead, that
further understanding may be gained of the motivations behind such acts.
With the increasing amount of data concerning such ‘special deposits’ within
Romano-​British settlements, it is surely the time to conduct a more in-​depth regional
or even national review of the evidence. Furthermore, comparative analysis in order to
understand the relationship between such ‘special deposits’ within settlement contexts
and those within formal temple sites has not yet been undertaken on a widespread sys-
tematic basis, but would surely be of great benefit in our comprehension of local and
regional religious and cultural expression.

Acknowledgements

I am very grateful to Professor Michael Fulford and Dr Hella Eckardt for reading and
commenting upon earlier drafts of this chapter.
656

656   Alex Smith

References
Primary Sources
Diodorus Siculus, The Library of History, trans. H. Oldfather. Loeb Classical Library.
Cambridge, MA: Harvard University Press, 1947.
Gaius Julius Caesar, The Conquest of Gaul, trans. S. A. Handford. London:  Penguin
Classics, 1951.
Strabo, Geography, trans. H. L. Jones. Loeb Classical Library. Cambridge, MA:  Harvard
University Press, 1917.

Secondary Sources
Aldhouse-​Green, M. (2001). Dying for the Gods:  Human Sacrifice in Iron Age and Roman
Europe. Stroud: Tempus.
Alexander, J., and Pullinger, J. (2000). ‘Roman Cambridge: Excavations on Castle Hill 1956–​
1988’, Proceedings of the Cambridge Antiquarian Society for 1999, Volume 88.
Andrews, P., Biddulph, E., Hardy, A., and Brown, R. (2011). Settling the Ebbsfleet Valley: High
Speed 1 Excavations at Springhead and Northfleet, Kent, Volume 1. Oxford and
Salisbury: Oxford/​Wessex Archaeology.
Bagnall-​Smith, J. (1995). ‘Interim Report on the Votive Material from the Romano-​Celtic
Temple Sites in Oxfordshire’, Oxonensia, 60: 172–​204.
Bagnall-​Smith, J. (2008). ‘Aspects of Votive Offerings in South-​East Britain’, in D Rudling (ed.),
Ritual Landscapes of South-​East Britain. Oxford: Heritage Marketing and Publications Ltd,
153–​170.
Barrie, T. (1996). Spiritual Path, Sacred Place:  Myth, Ritual and Meaning in Architecture.
London: Shambhala.
Black, E. (2008). ‘Pagan Religion in South-​East Britain: Contexts, Deities and Beliefs’, in D.
Rudling (ed.), Ritual Landscapes of South-​East Britain. Oxford: Heritage Marketing and
Publications Ltd, 1–​26.
Booth, P., Dodd, A., Robinson, M., and Smith, A. (2007). Thames through Time: The
Archaeology of the Gravel Terraces of the Upper and Middle Thames. The Early Historical
Period ad 1–​1000. Oxford: Oxford Archaeology Thames Valley Landscapes.
Bradford, J. P.  S., and Goodchild, R. G. (1939). ‘Excavations at Frilford, Berks 1937–​8’,
Oxoniensia, 4: 1–​70.
Bradley, R. (2000). An Archaeology of Natural Places. London: Routledge.
Brunaux, J.-​L. (1988). The Celtic Gauls: Gods, Rites and Sanctuaries. London: Seaby.
Clarke, S. (2001). ‘Probably Ritual:  Assemblage Interpretation at the Roman Military
Complex—​Towards a More Holistic Approach’, in A. T. Smith and A. Brooks (eds), Holy
Ground: Theoretical Issues Relating to the Landscape and Material Culture of Ritual Space.
British Archaeological Report International Series 956. Oxford: Archaeopress, 73–​83.
Clarke, S. (forthcoming). ‘Identifying the Actors and Motivations behind Pit Deposition at
Trimontium Military Complex (Newstead)’, in J Morris (ed.), Ritual in Context: Explaining
Ritual Complexity in Archaeology. Oxford <https://​communities.uhi.ac.uk/​sh01sc/​files/​
363>. [Accessed 2012]
Cunliffe, B. (1995). Danebury:  An Iron Age Hillfort in Hampshire. Volume 6:  A  Hillfort
Community in Perspective. CBA Research Report 102. York: Council for British Archaeology.
 Ritual Deposition   657

Derks, T. (1995). ‘The Ritual of the Vow in Gallo-​Roman Religion’, in J. Metzler et al. (eds),
Integration in the Early Roman West: The Role of Culture and Ideology. Luxembourg: Dossier
d’archéologie du Musée National d’Histoire et d’Art 4, 111–​127.
Evans, C,. and Hodder, I. (2006). Marshland Communities and Cultural Landscape:  The
Haddenham Project. Cambridge: McDonald Institute for Archaeological Research.
Field, N., and Parker Pearson, M. (2003). Fiskerton: An Iron Age Timber Causeway with Iron
Age and Roman Votive Offerings: The 1981 Excavations. Oxford: Oxbow Books.
Fulford, M. (2001). ‘Pervasive “Ritual” Behaviour in Roman Britain’, Britannia, 32: 199–​218.
Fulford, M., and Clarke, A. (2011). Silchester: City in Transition. The Mid-​Roman Occupation of
Insula IX c. ad 125–​250/​300. A Report on the Excavations since 1997. Britannia Monograph
Series 25. London: Society for the Promotion of Roman Studies.
Fulford, M., Clarke, A., and Eckardt, H. (2006). Life and Labour in Late Roman Silchester:
Excavations in Insula IX from 1997. Britannia Monograph Series 22. London: Society for the
Promotion of Roman Studies.
Ghey, E. (2007). ‘Empty Spaces or Meaningful Places? A Broader Perspective on Continuity’,
in R. Haeussler and A. King (eds), Continuity and Innovation in Religion in the Roman West:
Volume 1. JRA Supplementary Series 67. Portsmouth, RI: Journal of Roman Archaeology,
19–​30.
Green, C. W. (1966). ‘A Romano-​Celtic Temple at Bourton Grounds, Buckingham’, Record of
Bucks, 17/​5: 356–​366.
Green, M. J. (1986). The Gods of the Celts. Gloucester: Sutton Publishing.
Green, M. J. (1995). ‘The Gods and the Supernatural’, in M. J. Green (ed.), The Celtic World.
London: Routledge, 465–​489.
Haselgrove, C. (2005). ‘A Trio of Temples: A Reassessment of Iron Age Coin Deposition at
Hayling Island, Harlow and Wanborough’, in C. Haselgrove and D. Wigg-​Wolf (eds), Iron
Age Coinage and Ritual Practices. Mainz am Rhein: Akademie der Wissenschaften und der
Literatur, 381–​418.
Hayden, C., Booth, P., Dodd, A., Smith, A., Laws, G., and Welsh, K. (forthcoming). Horcott
Quarry: Prehistoric, Roman and Anglo-​Saxon Settlement and Burial. Oxford: Oxford Archaeology.
Hill, J. D. (1995). Ritual and Rubbish in the Iron Age of Wessex. BAR British Series 242.
Oxford: Archaeopress.
Hingley, R. (2006). ‘The Deposition of Iron Objects in Britain during the Later Prehistoric and
Roman Periods: Contextual Analysis and the Significance of Iron’, Britannia, 37: 213–​257.
Kamash, Z., Gosden, C., and Lock, G. (2010). ‘Continuity and Religious Practices in Roman
Britain:  The Case of the Rural Religious Complex at Marcham/​Frilford, Oxfordshire’,
Britannia, 41: 92–​126.
Kiernan, P. (2009). Miniature Votive Offerings in the Roman North-​West. Mainz: Harrassowitz.
King, A. (2005). ‘Animal Remains from Temples in Roman Britain’, Britannia, 36: 329–​369.
King, A. (2007a). ‘Coins and Coin Hoards from Romano-​Celtic Temples in Britain’, in R. Haeussler
and A. King (eds), Continuity and Innovation in Religion in the Roman West: Volume 2. JRA
Supplementary Series 67. Portsmouth, RI: Journal of Roman Archaeology, 25–​42.
King, A. (2007b). ‘Characterising Assemblages of Votive Offerings at Romano-​Celtic Temples
in Britain’, in M. Hainzmann (ed.), Auf den Spuren keltisher Gotterverehrung, Vienna: Verlag
der Osterreich. Akademie der Wissenschaften, 183–​196.
Lawrence, S., and Smith, A. (2008). Between Villa and Town:  Excavations of a Roman
Roadside Settlement and Shrine at Higham Ferrers, Northamptonshire. Oxford Archaeology
Monograph 7. Oxford: Oxford Archaeology.
658

658   Alex Smith

Leech, R. (1986). ‘The Excavation of a Romano-​Celtic Temple and a Later Cemetery on Lamyatt
Beacon, Somerset’, Britannia, 17: 259–​328.
Lewis, M. J. T. (1966). Temples in Roman Britain. Cambridge: Cambridge Classical Studies.
Maltby, M. (2010a). Feeding a Roman Town:  Environmental Evidence from Excavations in
Winchester, 1972–​1985. Winchester: Winchester Museums.
Maltby, M. (2010b). Zooarchaeology and the Interpretation of Depositions in Shafts’, in J. Morris
and M. Maltby (eds), Integrating Social and Environmental Archaeologies:  Reconsidering
Deposition. Oxford: BAR British Series 2077. Oxford: Archaeopress, 24–​32.
Mattingly, D. (2006). An Imperial Possession:  Britain in the Roman Empire, 54 bc–​ad 409.
London: Penguin.
Medlycott, M. (2011). The Roman Town of Great Chesterford. Chelmsford:  East Anglian
Archaeology 137.
Miles, D., Palmer, S., Smith, A., and Perpetua Jones, G. (2007). Iron Age and Roman Settlement
in the Upper Thames Valley: Excavations at Claydon Pike and other Sites in the Cotswold
Water Park. Oxford Archaeology Thames Valley Landscapes Monograph 26. Oxford: Oxford
Archaeology.
Morris, J. (2011). Investigating Animal Burials: Ritual, Mundane and Beyond. BAR British Series
535. Oxford: Archaeopress.
Penn, W. S. (1962). ‘Springhead: Temples II and V’, Archaeolgia Cantiana, 77: 110–​132.
Philp, B. (1999). The Roman Villa Site at Keston, Kent:  Second Report. Canterbury:  Kent
Archaeological Rescue Unit.
Pickstone, A. (2011). Iron Age and Roman Remains at Bretton Way, Peterborough:  Post-​
Excavation Assessment and Updated Project Design. Oxford Archaeology East Unpublished
Report.
Powell, K., Smith, A., and Laws, G. (2010). Evolution of a Farming Community in the Upper
Thames Valley. Excavation of a Prehistoric, Roman and Post-​Roman Landscape at Cotswold
Community, Gloucestershire and Wiltshire. Oxford Archaeology Thames Valley Landscapes
Monograph 31. Oxford: Oxford Archaeology.
Puttock, S. (2002). Ritual Significance of Personal Ornament in Roman Britain. BAR British
Series 327. Oxford: Archaeopress.
Revell, L. (2007). ‘Religion and Ritual in the Western Provinces’, Greece and Rome, 54: 210–​228.
Rodwell, W. (1980) (ed.). Temples, Churches and Religion in Roman Britain. BAR British Series
77. Oxford: Archaeopress.
Rives, J. B. (2007). Religion in the Roman Empire. Oxford: Blackwells.
Ross, A. (1968). ‘Shafts, Pits, Wells—​Sanctuaries of the Belgic Britons?’, in J. M. Coles and D. D.
A. Simpson (eds), Studies in Ancient Europe. Leicester: Leicester University Press, 255–​285.
Rudling, D. (2008). ‘Roman-​Period Temples, Shrines and Religion in Sussex’, in D. Rudling
(ed.), Ritual Landscapes of South-​East Britain. Oxford: Heritage Marketing and Publications
Ltd, 95–​138.
Score, V. (2011). Hoards, Hounds and Helmets:  A  Conquest-​Period Ritual Site at Hallaton,
Leicestershire. Leicester Archaeology Monograph 21. Leceister: Leicester Archaeology.
Sergeantson, D., and Morris, J. (2011). ‘Ravens and Crows in Iron Age and Roman Britain’,
Oxford Journal of Archaeology, 30/​1: 85–​107.
Smith, A. T. (2001). The Differential Use of Constructed Sacred Space in Southern Britain from
the Late Iron Age to the 4th Century ad. BAR British Series 318. Oxford: Archaeopress.
Smith, K. (2006). Guides, Guards and Gifts to the Gods:  Domesticated Dogs in the Art and
Archaeology of Iron Age and Roman Britain. BAR British Series 422. Oxford: Archaeopress.
 Ritual Deposition   659

Wait, G. A. (1985). Ritual and Religion in Iron Age Britain. BAR British Series 149 (i and ii).
Oxford: Archaeopress.
Watts, L., and Leach, P. (1996). Henley Wood: Temples and Cemetery Excavations 1962–​69. CBA
Report 99. York: Council for British Archaeology.
Wilson, D. R. (1975). ‘Romano-​Celtic Temple Architecture’, Journal of Roman Archaeological
Association, 38 [3rd Series], 3–​27.
Woodward, A. (1992). Shrines and Sacrifice. London: Batsford.
Woodward, A., and Leach, P. (1993). The Uley Shrines:  Excavation of a Ritual Complex on
West Hill, Uley, Gloucestershire:  1977–​ 9. English Heritage Archaeological Report 17.
London: English Heritage.
Woodward, P., and Woodward, A. (2004). ‘Dedicating the Town: Urban Foundation Deposits
in Roman Britain’, World Archaeology, 36: 68–​86.
Wythe, D. (2007). ‘An Analysis of Coin Finds from 75 Roman Temple Sites in Britain’, in R.
Haeussler and A. King (eds), Continuity and Innovation in Religion in the Roman West:
Volume 2. JRA Supplementary Series 67. Portsmouth, RI: Journal of Roman Archaeology,
43–​66.
Zoll, A. (1995). ‘A View through Inscriptions: The Epigraphic Evidence for Religion at Hadrian’s
Wall’, in J. Metzler et al. (eds), Integration in the Early Roman West: The Role of Culture and
Ideology. Luxembourg:  Dossier d’archéologie du Musée National d’Histoire et d’Art 4,
129–​150.
660

Chapter 32

Christia ni t y i n
Rom an B ri ta i n

David Petts

Introduction

The study of early Christianity has tended to assume a rather peripheral position with
regard to the wider study of ritual, religion, and cult behaviour in Roman Britain. It is
often dismissed as a relative latecomer to the religious landscape of the province, with
its true importance developing only in the early Middle Ages. The wider debate about its
importance is often couched in the terms of ‘success’ or ‘failure’, judgemental terms that
are rarely used in discussion of other Roman religious traditions, such as Mithraism. It
also often implied that the spread of the church in Britain was far less extensive than in
other part of the empire, although often these comparisons are with very different parts
of the empire, and often with different periods (for example, comparing fourth-​century
Britain with fifth-​or sixth-​century southern Gaul). While a proper understanding of
Roman Christianity certainly presents some specific interpretational challenges, the
often oversimplistic comparisons between Christianity and other religious traditions
in Roman Britain often ignore some of the wider conceptual issues presented by the
archaeological study of religious conversion. This chapter attempts to provide a brief
overview of the extant evidence for the early church, but also hopes better to position
Roman Christianity in its temporal and social landscape in fourth-​and fifth-​century
Britain.

Early Christianity

In ad 313, with the Edict of Milan, Christianity finally achieved full recognition as a licit
religion within the Roman Empire. Although it was not until ad 380 that Theodosius
 Christianity in Roman Britain    661

issued the Edict of Thessalonica making Christianity the state religion of the empire, its
adoption by the House of Constantine meant that from that point onwards it became the
de facto official cult of the Roman emperors (excluding a brief return to paganism under
Julian).
During the intermittent persecutions suffered by the church in previous centuries
there was a pressure on the church to maintain a relatively low profile within Roman
society. It is likely that only in Rome and some of the larger towns of the eastern
Mediterranean were there congregations of significant size. This had the consequence
that the church was slow to develop a distinctive set of architectural forms, and there
was relatively little Christian public monumentality (Snyder 2003). This is not to sug-
gest that there was no investment in Christian material culture, but it tended to be artis-
tic endeavours limited to spatially controlled spheres, such as the burial catacombs in
Rome or the third-​century house church at Dura Europos (Baur and Hopkins 1934;
Bowes 2008). In both cases, the Christian symbolism is largely confined to wall-​paint-
ings, a cautionary reminder of the extent to which evidence for pre-​fourth-​century
Christian worship may have been lost.
The practical impact of the Edict of Milan was the emergence of the church from the
shadows, as the newly confident institution had both the legal right and, increasingly,
the economic resources to develop a high profile within Roman society. For the first
time it could invest extensively in large-​scale public expressions of faith, while, at an
individual level, it was only then that it became safe to identify oneself as a Christian
through personal items of material culture. This new freedom is reflected in both the
increase in textual sources for Christianity in the fourth century, and also a massive rise
in the quantity and distinctiveness of Christian art and architecture.
It is not surprising, given this major change of circumstances, that the church groped
for new ways of expressing its identity, and that, owing to the relative lack of existing
Christian modes of representation, it drew more widely on forms and symbols used by
other religious and social groups. For example, although the basilica became one of the
dominant forms of church plan, it was an architectural blueprint that was widely used
by existing secular and religious groups. This relationship between Christian modes of
worship and pagan religious life is one that will be returned to at the end of this chapter.

Documentary Evidence

Contemporary documentary evidence for pre-​Constantinian Christianity in Roman

Britain is limited. Although Christian communities in Britain were mentioned by
early church writers, such as Tertullian and Origen (Thomas 1981:  43–​44), these
occur within very general passages expressing the geographical extent of the church.
As such, these mentions need to be accepted with caution as prima facie evidence for
Christianity in early Roman Britain. There are, though, hints from later textual sources
that there may have been some Christians in second-​and third-​century Britain. The
662

662   David Petts

early sixth-​century British writer Gildas refers to the martyrdom of Aaron and Julius
at Caerleon, the site of a Roman legionary fortress (De Excidio 10.2), although it is not
clear from where he was drawing his information (Knight 2001). However, it is clear that
he derived his information on the martyrdom of St Alban from an earlier Passio Albani,
possibly dating to between ad 430 and 480 (Sharpe 2001). There has been considerable
debate about the dates of the martyrdoms of Aaron, Julius, and Alban; given the paucity
of evidence, it is not possible to reach firm conclusions beyond a broad third-​or early
fourth-​century date.
The records of the Council of Arles, called by Constantine in ad 314 to settle the
Donatist dispute, is the earliest clear evidence for a formally constituted Christian
church in Britain (Rivet and Smith 1979: 49–​50). The Acta Concilii Arelatensis lists the
British delegates:

Eborius episcopus de civitate Eboracensi provincia Brittania

Restitutus episcopus de civitate Londiensi provincia suprascripta
Adelphius episcopus de civitate Colonia Londiniensium
Exinde Sacerdos presbyter Arminius diaconus.
[Eborius, Bishop of the city of York of the province of Britain.
Resitutus, Bishop of the the city of London of the above named province.
Adelphius, bishop of the city of Colonia Londinensium,
Furthermore, Sacerdos the priest and Arminius the deacon.]
(Munier 1963: 15, ll. 54–​58)

The first two names record the presence of bishops in York and London. The iden-
tification of Colonia Londiniensium is less certain; Lincoln and Colchester have
been suggested, with the former the more likely. If the bishops listed represent the
church leadership in three of the four provinces of Roman Britain (York—​Flavia
Caesariensis; London—​Maxima Caesariensis; Lincoln—​Britannia Secunda), it is
possible that Sacerdos and Arminius may have been delegates from the fourth prov-
ince, Britannia Prima, which perhaps lacked a bishop at this point. British bishops
were clearly well integrated into the wider Christian church in the fourth century,
and they are recorded at church councils at Serdica (ad 343) and Ariminum (ad 359)
(Thomas 1981: 121). There were also external interventions into the running of the
British church. Victricius, bishop of Rouen, crossed the channel at some point in the
ad 390s to deal with an unspecified dispute (De Laude Sanctorum 443–​444). Beyond
this, there is very little explicit textual evidence for fourth-​century Christianity in
Britain, although the careers of some administrators clearly indicate that they were
Christian—​the late fourth-​or early fifth-​century vicar of the diocese, Chrysanthus,
became a Bishop of Constantinople, a post that had earlier been held by his father
(Salway 1981: 407–​408).
This gives us a picture of a Romano-​British church with a developed ecclesiasti-
cal leadership who were active in church affairs beyond the confines of its own dio-
cese. However, we know little about the extent to which this leadership changed and
 Christianity in Roman Britain    663

developed over the fourth century. While the evidence from the Council of Arles indi-
cates that there was a hierarchy at provincial level, it is not clear whether the named
individuals were metropolitans with subsidiary bishops, possibly at civitas level. The
names of bishops have been tentatively identified as inscribed on a number of arte-
facts, including a silver dish (the so-​called Risley Park lanx) that mentions a bishop
named Exuperius; a lead salt-​pan from Shavington, which records a possible bishop by
the name of Viventius, and, less certainly, an unnamed bishop on a pewter bowl from
the Isle of Ely. In none of these cases is it possible to localize their spheres of influence
(Clarke 1931; Johns 1981; Johns and Painter 1991; Penney and Shotter 1996).

Christianity and Material Culture

There is evidence for the use of Christian imagery on a wide range of objects and arte-
facts from Roman Britain. However, it is not always easy to move from the simple iden-
tification of a symbol with Christian associations to an identification of it as an object
connected to Christian belief or practice. This knotty problem can be seen most clearly
in the use of the chi-​rho symbol. This basic symbol, bringing together the first two letters
of the Greek word Christos, had clear Christian associations from at least the early fourth
century, and was regularly associated with Constantine’s vision on the eve of the Battle
of the Milvian Bridge. The image appears in two forms in Britain, the ‘Constantinian
chi-​rho’ and the simpler rho-​cross (lacking the chi, but with a cross against the vertical
stroke of the rho). The motif is found alone, but also regularly flanked by the Greek alpha
and omega (an allusion to Revelations 1: 8). The ‘Constantinian’ chi-​rho is more common
in a Romano-​British context, with the rho-​cross beginning to appear only in the later
fourth century (Thomas 1981: 86–​91; Pearce 2008: 197–​201). The chi-​rho can regularly
be found on objects of indisputable religious or ritual function, such as the votive leaves
from the Water Newton treasure (Painter 1999).
However, following the Edict of Milan, the chi-​rho quickly became associated with
emperors in general and the House of Constantine in particular. It is regularly found
being used in seemingly official contexts during the fourth century. The most wide-
spread example of this was on coinage, particularly bronze nummi minted in Amiens
by Magnentius in the mid-​fourth century (Moorhead 2000). It also appears stamped
on pewter ingots (examples have been found in London) and a silver ingot found in
Balline, County Limerick, Ireland, but of Roman origin (Ó Ríordáin 1947: 43–​53; Mawer
1995:  96–​98). In these cases we seem to find the chi-​rho being used as an indicator
that the ingots came from mines under some form of imperial control. A small lead
seal with a chi-​rho from the forum Silchester may have been used in a similar manner
(Joyce 1881: 363). The symbol is also found on a number of personal items, including
a group of finger rings (Fifehead Neville, Dorset; Thruxton, Hampshire; Richborough,
Kent; Bagshot, Surrey) (Middleton 1881–​3: 68; Engleheart 1922: 215; Cunliffe 1968: 98–​
99; Graham 2002). It is not possible to identify their owners, although the items could
664

664   David Petts

clearly have been used by private individuals, state functionaries, and members of the
ecclesiastical establishment.
Christian imagery may also be found on other items, such as a group of buckles, belt
fittings, and strap-​ends of a type likely to have been used by state officials, including
members of both the army and the imperial civil service. The chi-​rho does not appear
on these items—​except for a strap-​end from Sandy, Bedfordshire (Mawer 1995: 65);
instead they carry images of peacocks, which had associations with Christianity in late
antiquity: examples include the belt buckles from Pen-​y-​Corddyn, Clwyd; Tripontium,
Warwickshire; Stanwick, North Yorkshire; and East Challow, Oxfordshire (Sanderson
1993: 2; Hawkes 1973: 145–​159, ­figure 3.1; Collingwood Bruce 1880: 90, fi ­ gure 8; Henig
and Brown 2003). While this general form of belt buckles is widely found in the west-
ern empire, the use of the opposed peacock motif does appear to be a feature found
almost exclusively in Britain, with only one other example known, from Westerwanna,
Germany.
These items were all made from copper alloy; there are few examples of the use of
more precious metals in items with a possible Christian identity. The exceptions are the
large silver buckle from the Traprain Law hoard and an unusual bow brooch of doubtful
authenticity from Shepton Mallet (Johns 2001). However, a series of far more extrava-
gant groups of objects with a probable Christian association has been recovered from
fourth-​century contexts as elements in high-​value silver hoards. The assemblage with
the most extensive range of Christian symbolism is the Water Newton hoard (Painter
1977). This consisted of twenty-​seven silver objects and a gold disc, including sixteen
votive leafs and a range of cups, jugs, and other vessels. One cup carried the inscrip-
tion innocentia et viventia … runt (Innocential and Viventia dedicated/​offered?)
and a chi-​rho flanked by an alpha and omega, while another bore a short text accompa-
nied by a chi-​rho that has resonances with elements of early liturgy (Painter 1999). The
hoard probably dates from the second half of the fourth century, and it has been argued
that this may have been a collection of ecclesiastical plate and ritual items, although the
reason for its deposition is unclear. The hoard from Traprain Law also had a signifi-
cant component of artefacts with Christian imagery, including the buckle already men-
tioned, a strainer with a chi-​rho design, a silver flask with biblical imagery, and a silver
flask with a chi-​rho (Curle 1923; Painter 1999; 2010). Although the circumstances for its
assemblage and deposition were doubtless very different from the Water Newton hoard,
both groups attest to the potential wealth that could be amassed by Christian communi-
ties in Britain. Christian imagery was also found, on a smaller scale, in other major plate
hoards from Britain, including ten spoons adorned with the chi-​rho from the early fifth
century Hoxne hoard and two spoons decorated with a chi-​rho from a hoard found in
Canterbury (Johns and Potter 1985; Johns 2010). These large silver hoards with Christian
components are part of a wider fourth-​century tradition in Britain of silver hoards, with
others, such as the Thetford treasure, containing only pagan imagery (Johns and Potter
1983). However, they are reminders that both pagan and Christian communities in late
Roman Britain were able to amass substantial wealth in the form of silver plate on a scale
largely unparalleled elsewhere in the empire at this period.
 Christianity in Roman Britain    665

Figure 32.1  Silver ring with early Christian symbol (anchor and fish) from the Roman fort at
Binchester, County Durham.
Source: © Durham University.

In addition to the use of the chi-​rho, a range of other early Christian images are known
from Romano-​British objects, such as the combination of fish and anchor found on the
intaglio of a silver ring discovered in 2014 excavated at the Roman fort at Binchester
(County Durham) (Figure 32.1).

Churches

Despite clear documentary evidence of an established Christian church in fourth-​cen-

tury Britain, the archaeological evidence for this is variable. The most problematic area
666

666   David Petts

is in the identification of church structures. There is not a single building of Roman date
that can unequivocally be identified as a Christian church, although there are around a
dozen structures that have a greater or lesser degree of circumstantial evidence. One of
the major challenges is that, as noted above, the most common ground plan for churches
in the western empire in the fourth century was the basilica, a form that was also widely
used in other religious and secular contexts. It is not easy, in the absence of additional
evidence, to identify a structure as a church on morphological grounds alone. In the
absence of an extensive late Roman tradition of epigraphy, and the generally poor pres-
ervation of most of these structures, good supporting evidence is largely lacking.
One of the strongest candidates for a Roman church is the earliest of a sequence of
structures on the site of the medieval parish church of St Paul-​in-​the-​Bail, Lincoln
(Steane 2006:  129–​211). This was located above the central area of the forum of the
Roman town. Beneath the medieval church were three earlier structures. The earliest
was a rectangular building aligned west–​east with a potential rectangular apse at the
eastern end. Above this was a larger timber structure with a substantial eastern apse.
Although the western end of neither of these buildings was found, it is possible that they
abutted the western portico of the forum, which may have acted as a simple narthex.
Above this was a smaller rectangular building with a central burial of probable middle
Saxon date. The dating evidence for this sequence is not precise, with the first two pos-
sible churches belonging anywhere between the very late Roman and the middle Saxon
period on stratigraphic grounds (Steane 2006: 192–​194). The second building was also
surrounded and cut by a series of burials that have a broad seventh/​eighth-​century ad
date, which matches well with the date of the burial within the third structure.
One of the most frequently adduced cases is a small (13 m × 9 m) apsidally-​ended
basilica in Insula IV at Silchester, Hampshire (Frere 1976; King 1983; Ford 1994; Cosh
2004) (Figure 32.2). This building lies to the south-​east of the town’s main forum and
basilica complex. Internally, there are two aisles and an eastern narthex with the apse
lying at the western end of the structure. Most aspects of this building find parallels with
churches elsewhere in the western empire; for example, the western apse echoes the
plan of the early fourth-​century church of St Severin in Cologne (Krämer 1958; Frere
1976: 292). However, the major stumbling block for a certain Christian identity for this
building is its date. It almost certainly post-​dates a nearby wooden structure that prob-
ably fell out of use in the late third century. If the probable pre-​Constantinian date for
this structure militates against identification of the building as a church, then possible
alternatives include use as a schola.
In an urban context, another reasonable candidate for a church is the large-​aisled
building of mid-​fourth-​century date found at Colchester House near Tower Hill,
London (Sankey 1998). Although preservation was poor, excavation revealed the north-​
eastern corner of a basilica structure with double aisles and a short wall marking off the
eastern end of the building. This was a significant structure; substantial in size (c. 50 m ×
80–​100 m) with a stone and tile floor. Fragments of marble and window glass were also
found, although in secondary contexts. The reconstructed plan of this building shows
some parallels with the fourth-​century cathedral of St Tecla in Milan and the church
 Christianity in Roman Britain    667

Forum 0 10m

street

timber building

Well

Church?

Figure  32.2 Potential Roman church from Silchester, Hampshire. Based on Frere

(1975: ­figure 1).
Source: © David Petts.

of St Giovanni Evangelista in Ravenna (Krautheimer 1986: 84, 184–​185). Another pos-

sible function for this building is as a granary for the collection of the annona, although
the evidence for the use of marble in the interior may mitigate against this suggestion.
Intriguingly, a recent analysis of the coin assemblage from the structure suggests that
the coin loss profile has its best parallels with late Romano-​British religious and ritual
sites (Gerrard 2011). Within the context of Romano-​British towns, these are the strong-
est candidates for churches, although other suggestions have been made, including
the apsidal building at Flaxengate, Lincoln; the Roman building beneath the medieval
church of St Mary-​de-​Lode, Gloucester, and the apsidal structure in the late Roman
cemetery at Butt Road, Colchester. All, however, have only limited and circumstantial
evidence for their religious use (Colyer and Jones 1979; Bryant 1980; Crummy et al. 1993:
164–​190; Heighway 2010: 42–​43).
There are also a small number of possible Roman churches from rural contexts. The
most likely example is the small rectangular structure associated with a group of over
forty burials and a possible font from Icklingham, Suffolk (West and Plouviez 1976).
The association of the structure with the burials is suggestive, since Roman pagan tem-
ples were rarely located adjacent to graves. More important, however, is the fact that
three lead tanks carrying overt Christian motifs have been found in the immediate
area. While a strong case can be made for the building being a small church or oratory,
668

668   David Petts

this identification is again made only on the basis of its contextual associations rather
than anything inherent in the form or plan of the building itself. A number of other
small rectangular structures found adjacent to Roman rural temples have also been put
forward as potential churches, possibly replacing an earlier, pagan, focus of worship,
including examples from Lamyatt Beacon, Somerset, Brean Down, Somerset, and Uley,
Gloucestershire (Ap Simon 1964–​5; Leech 1986; Woodward and Leach 1993). While they
all have some features in common with early churches, the strongest parallels are with
early medieval rather than Roman churches, and, in the absence of any internal dating
evidence for most of these structures, an identification as a Roman church is hard to
endorse.
Finally, there are a small group of structures on Roman military sites that may be
churches, although, like other examples, they suffer from problems in interpretation.
A good case in point is the rectangular structure at the major fort at Richborough, Kent
(Brown 1971). Located in the north-​east corner of the fort, this part of the site was exca-
vated using relatively unsophisticated techniques in the 1920s. This work revealed two
lines of possible stone post-​pads, which appeared to form the west and north walls of a
structure. The structure is poorly dated, so there would be no suggestion that this was a
church if it was not for its proximity to a probable masonry stone font that lay between
the structure and the northern wall of the fort.
On the northern frontier, there are a number of potential churches. A simple apsidally
ended building was constructed within the courtyard of the praetorium at Vindolanda.
Dating evidence is sparse, although it was clearly constructed over a build-​up of activity
that overlay the flagstones of the courtyard; a date of early fifth century or later is most
likely (Birley 2009). There is also another small apsidal building at nearby Housesteads
(Crow 1995: 95–​97). As at Richborough, it lay within the north-​eastern corner of the
fort, although the apse was at the west rather than the east end. It was discovered during
Bosanquet’s excavations in the late nineteenth century, and its precise date is unclear,
although it appears to be broadly late Roman.
Slightly different in form is the putative church at the major fort and supply base to the
south of the River Tyne at South Shields, Tyne and Wear (Bidwell and Speak 1994: 103–​
104). This structure lay within the principia of the fort and consists of a possible altar
surrounded on three sides by a stone structure, which might best be understood as ele-
ments of a rectangular apse. This building was quite substantial in size, measuring over
9 metres in width, and there is the potential that the main body of this structure could
have been linked to part of the colonnade surrounding the courtyard.
A final form of church that needs to be considered is the ‘house church’. The dedica-
tion of elements of private houses on a temporary or permanent basis to the celebration
of the Eucharist has its origin in the earliest period of Christian practice (Bowes 2008).
However, it is apparent that this practice continued into the post-​Constantinian era and
that some villas may have had what were, in practice, private estate chapels. The best
example from this in Britain is Lullingstone, Kent (Meates 1979, 1987). Here, exceptional
preservation allowed a series of plaster wall-​paintings from an upper room in building
to be reconstructed. The anti-​room or narthex was decorated with chi-​rhos flanked by
 Christianity in Roman Britain    669

Figure 32.3  Painted wall plaster from Lullingstone Roman villa (Kent) showing chi-​rho sym-
bol flanked by an alpha and omega.
Source: © The Trustees of the British Museum.

an alpha and omega (Figure 32.3). Similar symbols appeared within the main room, as
did a sequence of figures in the orans position (hands raised up), the position associ-
ated with prayer in the early church. At the same time as the paintings were created, it
appears that the access routes to the rooms were changed and access, which had previ-
ously been via the main house, became possible only via an exterior door (Meates 1987).
It is salutary to note that, without the survival of the wall-​paintings, there is nothing that
would have suggested a Christian use or identity to these rooms.
Lullingstone is not the only Roman villa that has produced decorative schemes
with a clearly Christian identity. A number of villas from Britannia Prima (south-​west
England) have produced mosaics that incorporate possible Christian symbolism. The
best known of these is the central roundel from the large pavement from Hinton St
Mary, Dorset, which shows a male head with a chi-​rho symbol behind it (Pearce 2008).
A contiguous mosaic showed hunting scenes and Bellerophon fighting the Chimaera.
The male bust has often been identified as an early (if not the earliest) figurative depic-
tion of Christ (Toynbee 1964; Painter 1967, 1972; Thomas 1981). However, there have also
been strong arguments made that the image is in fact a piece of imperial portraiture,
intended to represent a member of the House of Constantine (Pearce 2008). The chi-​rho
670

670   David Petts

also appears on a mosaic from Frampton, Dorset. As at Hinton St Mary, this was associ-
ated with a depiction of Bellerophon battling the Chimaera with the chi-​rho decorating
the adjacent apse (Cosh and Neal 2006: 130–​140).
From the evidence for urban, rural, and military churches, it is possible to make a
number of general observations. There was the apparent change in the way in which late
Roman formal space was being utilized by the end of the fourth century ad. Whether or
not one accepts the Christian attribution to the buildings at St Paul-​in-​the-​Bail, Lincoln,
Vindolanda, and South Shields, they were all located within the centre of what had
formally been open courtyards. There was also a clear move away from architectural
expressions of power drawn from Mediterranean architecture, which had an emphasis
on courtyards and open spaces. It is also notable how little good dating evidence there
is for these structures, which were often excavated with unsophisticated techniques, so
that detailed stratigraphic insight is frequently lacking. Even when excavations were to
modern standards, the lack of good diagnostic material culture of a fifth-​century date
makes it extremely difficult to distinguish between buildings that were built broadly
in the fourth century and those that were first constructed in the fifth century ad. It is
also important to consider the British structures in their wider context. In the western
Roman Empire very few churches can be securely dated to the fourth rather than the
fifth century, and thus the evidence from Britain is, in fact, broadly comparable with
developments in neighbouring provinces.

Baptism

Churches are not the only built structures associated with early Christianity in Britain.
A number of masonry structures that may be fonts or font bases are known. While the
evidence for the church at Richborough may not be strong, the hexagonal masonry struc-
ture to its north is almost certainly a font (Brown 1971). This stone-​built basin has strong
parallels with other certain examples from Boppard and Cologne, Germany (Brown 1971:
227–​228, plate XXXI). An octagonal example, built of tile rather than stone with a gravel
drain or soakaway, has been excavated at Chimneys, Witham, Essex; this was found on
the site of a Romano-​Celtic style temple built in the third century ad (Turner 1999: 51–​
55). However, it appears to post-​date the phase of activity at the temple and may represent
the Christianization of a pagan focus of worship. Unlike the Richborough example, there
are hints of an external superstructure surrounding the font, with a series of post-​holes
placed at each corner of the tank; this was succeeded by a later reworking to convert it
into a four-​post structure. Another masonry cistern-​type font comes from Icklingham,
Suffolk, where a small tile-​built tank was found close to the possible small church already
discussed (West and Plouviez 1976: 71). A final possible example is the small stone-​lined
tank that stood to the west of the possible church at Housesteads (Crow 1995: 96–​97).
Although it is rectangular in form, similar simple forms are known from the continent,
such as at Zurzach, Switzerland (Brown 1971: 228, plate XXXII).
 Christianity in Roman Britain    671

In addition to complete font cisterns, there are also a number of potential font bases,
the footings or platforms on which smaller tanks could have stood. Tile platforms, on an
axial alignment with putative church structures, stood to the east of the small basilica at
Silchester already mentioned and to the west of the apsidal structure in the cemetery at
Butt Road, Colchester (Frere 1976: 290–​291; Crummy et al. 1993: 176–​177). Obviously,
the case for accepting these as font bases is closely connected to the acceptance of the
related buildings as churches.
It is possible that some ornamental pools or cisterns from Roman villas may also
have had a baptismal or at least quasi-​ritual function. At Chedworth, Gloucestershire,
a small apsidal pool to the north-​west of the main building was fed by natural springs
(Goodburn 2000: 24). Although constructed in the pre-​Constantinian period, it may
have had a later Christian use, as a number of chi-​rho symbols were found carved onto
stone slabs that probably originally lined the pool. It has been suggested that a group of
octagonal pools from villas in south-​west England (Dewlish, Dorset; Lufton, Somerset;
Holcombe, Devon) may also have had a similar ritual function (Perring 2003; Todd
2005). They were all located within existing bath-​suites but also had separate external
access. However, although there is evidence for Christianity on mosaics and other mate-
rial culture from villas in the general area, this functional identification of the pools
must remain speculative (Henig 2006).
An important class of objects is a group of large circular lead tanks some of which
carry overt Christian imagery, with over twenty examples known. Circular in shape and
constructed from sheets of cast lead, they vary in size from 0.46 metres to 0.97 metres
in diameter. A number of these have been decorated with chi-​rho symbols (for exam-
ple, Ashton, Northants; Pulborough, West Sussex; Icklingham, Suffolk), in some cases
associated with the Greek letters alpha and omega. Two also carry figural images. A tank
from Flawborough, Nottinghamshire, has four figures in the orans posture, while the
most complex is a figurative frieze on a tank from Walesby, Lincolnshire. Within an
architectural frame it shows three groups of figures; the central group consists of two
clothed women flanking a naked woman, while the other two groups depict clothed
males in cloaks and tunics. This has often been interpreted as the depiction of a bap-
tism, with the naked catechumen being assisted to the font by two supporters (Thomas
1981: 221–​225; though see now Crerar 2012). These are usually associated with baptism,
despite the more widespread tradition of baptism by immersion, and seen either as fonts
or vessels associated with other elements of the baptismal liturgy, such as ritual foot-
washing (pedilavium) (Watts 1991: 171–​173). However, solid evidence for their precise
function is far from clear. Although a small number of broadly similar lead tanks have
been found outside Britain, these have not been found carrying Christian imagery, and
they do seem to be a genuinely insular phenomenon.
An important dimension to the distribution of the lead tanks is their specific dep-
ositional context. Many appear to have been placed in ‘watery’ contexts, including
ditches (Flawborough), wells (Caversham; Ashton), rivers (Pulborough; Oxborough;
Huntingdon), and water holes (Heathrow). This appears to parallel a long-​term north-
ern European tradition of utilizing such contexts for acts of ritualized deposition. In
672

672   David Petts

Roman Britain, particularly, there is a clear tradition of placing hoards of pewter vessels
and tableware in similar locations, a rite that becomes increasingly popular from the
later third century ad (Petts 2002). There is also evidence for regional focus to this rite,
with the deposition of both lead tanks and other suites of items being most common
in the east Midlands and East Anglia. This can even be seen scaled down to the local
level, such as in the area around Icklingham, which has produced evidence for four lead
tanks, as well as a cluster of other hoards, including coins and deposits of ironwork, with
evidence that this local tradition had its origins in the early Roman period (Petts 2002).

Death and Burial

As with other aspects of fourth-​century Christian belief, identifying a diagnostic

Christian burial rite in the late Roman period is challenging. There are undoubtedly
major changes in the late Romano-​British burial rite from the third century onwards
(Philpott 1991; Quensel-​von Kalbern 1999; Petts 2004). Broadly speaking, there is
a transition from cremation to inhumation as the predominant rite, although within
this general pattern there is certainly considerable regional variation. The inhumation
rite is itself far from homogenous, but it is possible to recognize two general traditions.
One rite continued to see the deposition of a range of grave goods, including vessels
and items of personal ornamentation. These burials were often aligned north–​south,
and there was considerable variation in the positioning of the body, with some unu-
sual treatments of the corpse, including prone burial and post-​mortem decapitation.
The spatial arrangement of these burials within a cemetery often showed evidence for
clustering in possible family groups or around focal graves. This rite contrasted with a
tradition in which grave goods were largely absent. These graves were usually aligned
approximately west–​east, and the graves within cemeteries were often laid out in regular
rows. There is also a tendency for a greater proportion of bodies in these graves to be
protected by a coffin, cist, or stone-​lining. These two traditions could appear contempo-
raneously within a single settlement but were usually spatially distinct.
This distinction can be seen within an individual cemetery in the case of Poundbury,
Dorset, where the bulk of the cemetery population was aligned west–​east and had few
grave goods, with a significant number being placed in wooden, stone, or lead coffins
(Farwell and Molleson 1993). Within this group were a number placed in stone-​lined
mausolea with internal plaster wall-​painting, some of which may have had Christian
symbolism (Sparey-​Green 1993). However, at the periphery of the cemetery were
smaller plots where the graves were aligned mainly north–​south and contained grave
goods. At a different spatial resolution, this distinction between the two rites is apparent
at Ilchester, Somerset, where the main cemetery to the north-​east of the town was domi-
nated by west–​east-​aligned graves with few grave goods (P. Leach 1994: 91–​103). To the
south-​east of the town, situated in the back lots of roadside buildings, burials character-
istic of the other tradition were situated (P. Leach 1982: 82–​88).
 Christianity in Roman Britain    673

Though the two rites were clearly broadly contemporary through the fourth century,
it is not clear precisely when the west–​east aligned inhumation tradition developed.
The lack of grave goods means that they are often inherently hard to date. However, at
Poundbury the cemetery appears to have been founded in the early to mid-​fourth cen-
tury, while at Butt Road, Colchester, there is clearly a major cemetery reorganization of
the burial ground, replacing a more heterogenous tradition that featured a wide range of
grave goods and a north–​south alignment, at some point in the late third or early fourth
century ad (Crummy et al. 1993: 4–​163; Millett 1995). These changes in burial rite appear
generally to have occurred around the turn of the fourth century or perhaps a little later,
but it is not easy to show for certain whether these developments are pre-​Constantinian
or not.
It would be tempting to relate the advent of this burial rite with the spread of
Christianity, but the situation is doubtlessly more complicated than this. A major obsta-
cle for such a simple interpretation is the complete lack of any textual evidence for any
extensive church interest in defining the nature of the Christian burial rite in this period,
with no recorded prescriptions about orientation, spatial organization, or treatment of
the body beyond a broad stricture against cremation.
However, in the fourth century it is likely that, with the increasing Christianization
of the imperial civil service, the church stepped into the gap left by the secular elite’s
withdrawal from involvement in this sphere and became increasingly involved in civil
administration. This would have allowed the church increasingly to control a range
of activities, including burial in major cemeteries. This need not mean that all those
buried within the formally arranged cemeteries were Christians; many may simply
have been conforming to a burial rite advocated by a socially and politically powerful
element within Roman society. Those who were not able to come to an accommoda-
tion with this may instead have opted to bury their dead in more peripheral locations.
Indeed, the increase of unusual rites, such as post-​mortem decapitation, may even
have been an attempt to develop more overtly ‘pagan’ rites in the face of an increasing
Christian orthodoxy (for a more developed version of this argument, see Petts 2003:
146–​149).
Attempts to distinguish diagnostic evidence for Christian burial are hampered by the
lack of overtly Christian epigraphy from fourth-​century Britain. This reflects the wider
decline in the epigraphic habit in the third and fourth centuries throughout the west-
ern Empire (Handley 2001). A number of fourth-​century funerary inscriptions have
been identified as potentially Christian on the basis of parallels between particular epi-
graphic formulae used in similar Christian contexts elsewhere in the empire (for exam-
ple, vixit plus minus found on inscriptions from Carlisle and Brougham; titulum posuit
found on inscriptions from Hadrian’s Wall, Old Penrith, Templeborough, and York)
(Handley 2001: 181–​183; RIB 620, 689, 786-​7, 934, 955, 1667). The extent to which these
phrases are peculiarly Christian are open to debate, and Knight (2010) has argued for
the need to look for broad ‘Christian symptoms’ rather than direct Christian identity
in the epigraphic record. Across the empire there is a slight increase in the use of epig-
raphy towards the end of the fourth century and into the early fifth century (Handley
674

674   David Petts

2001: 181–​183). There is a small group of late gravestones from the western end of
Hadrian’s Wall that appear to date from this period; however, it is still unclear whether
they belong in a fourth-​or fifth-​century context, and they have no overt Christian sym-
bolism (Dark and Dark 1996; Todd 1999). In the sub-​Roman/​early medieval period
there is a clearer revival of funerary epigraphy in the western fringes of Roman Britain
and, to a lesser extent, in lowland Scotland (Thomas 1993; Edwards 2001). While these
later stones are clearly being erected in a Christian context, they too carry remarkably
limited evidence for overt religious belief. There is still some debate over the precise
relationship between these later insular inscriptions and the latest Roman examples
(Handley 2001; Knight 2010). These insular stones differ from contemporary continen-
tal stones in their notable lack of Christian symbols or imagery (for example, chi-​rho
symbols, peacock, or doves).

Conclusions

The intractable nature of the evidence for Roman Christianity in Britain means that
there is very little consensus on its extent during the fourth century. The success or oth-
erwise of the Romano-​British church has important implications for understanding the
growth of Christianity in western and northern Britain in the fifth and sixth centuries.
Some scholarship has regarded the fourth century as a period of consolidation, with
a firmly established church within the Roman province forming a springboard for its
rapid extension beyond the frontiers of the empire in the fifth century (e.g. Thomas 1981;
Petts 2003). An alternative perspective sees the church in Roman Britain as remain-
ing institutionally very weak and reliant on direct imperial support. As a consequence,
the abdication of the political control of Britannia by the Roman state in the early fifth
century resulted in the final failure of an attenuated church (e.g. Frend 1955, 1968, 1979;
Watts 1991). This view then requires that early medieval Britain undergoes a period of
reconversion carried out by missionaries from mainland Europe. While the limited
nature of the evidence is capable of multiple readings, it is clear from the early documen-
tary evidence that, while the Gallic church clearly had an involvement with the British
church, this appears to be in the context of a relationship between two established bod-
ies, and there is no indication of any missionary activity. It is also notable that the earliest
significant historic sources written within early medieval Britain situated the island’s
conversion during Roman rule rather than the post-​Roman period. Nonetheless, it is
clear that, despite the likelihood of continuity, there was a major realignment of the geo-
graphic focus of the British church, with a probable decline or collapse of the church in
the east of the province in the face of Anglo-​Saxon takeover. The best evidence for the
church can be found in western Britain and southern Scotland, regions that had previ-
ously been peripheral to the focus of Roman control. However, such shifts in emphasis
must be understood within the context of wider changes in the geography of power in
sub-​Roman Britain. It is important, though, to emphasize that later fourth-​century and
 Christianity in Roman Britain    675

fifth-​century sub-​Roman paganism has been under-​researched, and the low material
visibility of the sub-​Roman church is paralleled in the lack of a diagnostic archaeology.
As well as debates about the extent and nature of continuity, another key area of debate
about Romano-​British Christianity is centred on its relationship with contemporary
paganism. This relationship can be recognized in a number of separate cultural spheres.
It is clear from the juxtaposition of Christian and Gnostic imagery on some mosaics that
there may have been a positive interaction between the church and the more literate
and theologically sophisticated aspects of classical paganism (Perring 2003). However,
it is not always easy to distinguish the deliberate association of compatible aspects of
Gnostic and Christian imagery from a more general expression of classical learning
(paideia) that even Christian elites were expected to maintain in this period (Petts 2003:
116–​118). Pagan influence can also be seen in other aspects of Christian practice. This
can be seen clearly in the silver repoussée votive leaves that formed an element of the
Water Newton hoard. These have strong parallels with votive leaves with overt pagan
imagery found elsewhere in Britain (Henig 1993; Painter 1999; Crerar 2006). The prac-
tice of depositing such votive offerings found a similar material expression in pagan and
Christian practice in Roman Britain. An argument has also been made for wider simi-
larities between pagan and Christian depositional practices in the fourth century (Petts
2002; cf. Fulford 2001). Relationships between Christian and pagan may, of course, also
have been tense. Evidence for the destruction of pagan shrines, temples, and statuary
in the fourth century has often been related to Christian iconoclasm (for example, the
Wallbrook Mithraeum: Green 1976: 46; Merrifield 1977: 375). More recently, though, it
has been recognized that the evidence for seemingly arbitrary destruction may actu-
ally have been more carefully structured and that the fragmentation and preservation
of elements of statues may fit into wider attitudes to the preservation and destruction
of holy objects (Croxford 2003). While some destruction of pagan sites and objects may
have occurred in the context of Christian-​on-​pagan violence, the evidence for complex
artefact biographies for much of the cult material is a reminder that there may be other
motivations for damaging and dismembering such items.
A final dimension to the wider analysis of Christianity in Roman Britain is the need
for a better understanding of the extent of regional variation. This needs to encompass
a better appreciation of variation in the evidence within Britain, but also a clearer appre-
ciation of how Romano-​British Christianity compares with the evidence for the church
elsewhere in the empire in the fourth century. It is now clear that simple cumulative dis-
tribution maps, showing the presence of diagnostic Christian material, are not a satis-
factory way of mapping the extent of Christianity (Petts 2003: 26–​27). A range of social
factors can influence such distributions and maps can mask spatial variability. For exam-
ple, it is apparent that the majority of hoards with Christian items are found in the east
Midlands and East Anglia, whereas the evidence for high-​status Christianity, as found
on mosaics, is largely found in the south-​west. However, in both cases there are com-
plex social processes at play. The south-​west of England also has the largest number of
late Roman temples, and we are most likely simply seeing an area where religious belief
was expressed through the construction of buildings, whereas in East Anglia and the
676

676   David Petts

Midlands there is clearly a long-​term tradition of depositing substantial votive hoards

(Petts 2003). However, this does not necessarily mean that Christians in the south-​west
of England were not using silver plate for religious purposes; it may simply never have
entered the archaeological record, instead perhaps being recycled. Such variation in reli-
gious practice at an intra-​provincial level should come as no surprise; nor should the fact
that such variation can be recognized between provinces (Petts 2014). For example, the
distribution of lead tanks with Christian imagery is almost entirely restricted to Britain,
as is the use of peacock imagery on belt-​buckles. However, the use of Christian imagery
on gravestones is largely a continental phenomenon, and there is a lack of stone sarcoph-
agi bearing distinct Christian symbolism in Britain. There are also similarities between
provinces—​for example, both Britain and northern Gaul are largely lacking good archae-
ological evidence for substantial fourth-​century church structures, with little to compare
with the more spectacular evidence from further south or east (Petts 2014).
In conclusion, the evidence for Christianity from Roman Britain is not extensive or easy
to interpret, although it is useful to remember that it is still the most widely represented
individual cult in fourth-​century Britain. While it may never have been the dominant reli-
gion in terms of numbers, it was clearly adopted by wealthy individuals, many of whom
were likely to have been in positions of political and social power. There is still debate over
the extent to which this Roman church was the foundation for the rapid blossoming of
Christianity in the insular world in the fifth and sixth centuries. However, simply focusing
on the ‘success’ or otherwise of the Roman church should not prevent us from developing
a wider understanding of the institution both within and beyond Britannia.

Abbreviation
RIB R. G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume
I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Primary Sources
De Laude Sanctorum, ed. J.-​P. Migne 1845, Patrologia Latina 20 ( Ridgewood, NJ: Gregg, 1965).
Gildas, De Excidio (The Ruins of Britain), ed. and trans. M. Winterbottom (Chichester:
Phillimore, 1978).

Secondary Sources
Ap Simon, A. (1964–​5). ‘The Roman Temple on Brean Down, Somerset’, Proceedings of the
Speleological Society, 10: 195–​258.
Baur, P. V. C., and Hopkins, C. (1934). The Christian Church at Dura Europos. New Haven,
CT: Yale University Press.
 Christianity in Roman Britain    677

Bidwell, P., and Speak, S. (1994). Excavations at South Shields Roman Fort, Volume 1.
Newcastle: Society of Antiquaries of Newcastle upon Tyne.
Birley, R. (2009). Vindolanda: A Roman Frontier Fort on Hadrian’s Wall. Stroud: Amberley.
Bowes, K. (2008). Private Worship, Public Values and Religious Change in Late Antiquity.
Cambridge: Cambridge University Press.
Brown, P. D. C. (1971). ‘The Church at Richborough, Kent’, Britannia, 2: 225–​231.
Bryant, R. (1980). ‘Excavations at St Mary De Lode, Gloucester 1978–​1979’, Glevenis, 14: 4–​12.
Clarke, L. G. C. (1931). ‘Roman pewter Bowl from the Isle of Ely’, Proceedings of the Cambridge
Antiquarian Society, 31: 66–​75.
Collingwood Bruce, J. (1880). A Descriptive Catalogue of Antiquities Chiefly British at Alnwick
Castle. Newcastle: Andrew Reid.
Colyer, C., and Jones, M. (1979). ‘Excavations at Lincoln, Second Interim Report. Excavations
in the Lower Town’, Antiquaries Journal, 59: 50–​91.
Cosh, S. (2004). ‘A possible date for the Silchester “Church”’, Britannia, 35: 229–​233.
Cosh, S., and Neal, D. (2006). Roman Mosaics of Britain. Volume II. South-​West Britain.
London: Society of Antiquaries of London.
Crerar, B. (2006). ‘Votive Leaves from Roman Britain’, in M. Henig (ed.), Roman Art, Religion
and Society:  New Studies from the Roman Art Seminar, Oxford 2005. BAR International
Series, 1577. Oxford: Archaeopress, 71.
Crerar, B. (2012). ‘Contextualising Romano-​ British Lead Tanks:  A  Study in Design,
Destruction and Deposition’, Britannia, 43: 135–​166.
Crow, J. (1995). The English Heritage Book of Housesteads. London: Batsford.
Croxford, B. (2003). ‘Iconoclasm in Roman Britain?’, Britannia, 34: 81–​95.
Crummy, P., Crummy, N., and Crossan, C. (1993). Colchester Archaeological Report
9: Excavations of Roman and Later Cemeteries, Churches and Monastic Sites in Colchester
1971–​88. Colchester: Colchester Archaeological Society.
Cunliffe, B. W. (1968) (ed.). Fifth Report on the Excavations of the Roman Fort at Richborough,
Kent. London: Society of Antiquaries of London.
Curle, A. O. (1923). The Treasure of Traprain Law. Glasgow: Maclehose, Jackson.
Dark, K. R., and Dark, P. (1996). ‘New Archaeological and Palynological Evidence for a Sub-​
Roman Reoccupation of Hadrian’s Wall’, Archaeologia Aeliana, 24: 57–​72.
Edwards, N. (2001). ‘Early Medieval Inscribed Stones and Stone Sculpture in Wales: Context
and Function’, Medieval Archaeology, 45: 15–​39.
Engleheart, G. H. (1922). ‘On Some Roman Buildings and Other Antiquities in a District of North-​
West Hants’, Proceedings of the Hampshire Field Club and Archaeological Society, 9: 214–​218.
Farwell, D. E., and Molleson, T. I. (1993). Poundbury Volume 2:  The Cemeteries.
Dorchester: Dorchester Natural History and Archaeological Society.
Ford, S. D. (1994). ‘The Silchester Church: A Dimensional Analysis and a New Reconstruction’,
Britannia, 25: 119–​126.
Frend, W. H. C. (1955). ‘Religion in Roman Britain in the 4th Century’, Journal of the British
Archaeological Society. 3rd ser. 18: 1–​18.
Frend, W. H.  C. (1968). ‘The Christianisation of Roman Britain’, in M. W. Barley and R.
P.  C. Hanson (eds), Christianity in Britain 300–​700:  Papers Presented to the Conference
on Christianity in Roman and Sub-​Roman Britain, 17–​20 April 1967. Leicester:  Leicester
University Press, 37–​50.
Frend, W. H. C. (1979). ‘Ecclesia Britannica: Prelude or Dead End?’, Journal of Ecclesiastical
History, 30: 129–​144.
678

678   David Petts

Frere, S. S. (1976). ‘The Silchester Church:  The Excavations by Sir Ian Richmond in 1961’,
Archaeologia, 105: 277–​302.
Fulford, M. J. (2001). ‘Links with the Past: Pervasive “Ritual” Behaviour in Roman Britain’,
Britannia, 32: 199–​218.
Gerrard J. (2011). ‘Cathedral or Granary? The Roman Coins from Colchester House (PEP89)’,
Transactions of the London and Middlesex Archaeological Society, 62: 81–​88.
Goodburn, R. (2000). Chedworth Roman Villa. London: National Trust.
Graham, T. (2002). ‘A Rho-​Cross Engraved on a Jet Finger-​Ring from Bagshot, Surrey’, Oxford
Journal of Archaeology, 21: 211–​216.
Green, M. (1976). A Corpus of Religions Material from the Civilian Areas of Roman Britain. BAR
British Series, 24. Oxford: Archaeopress.
Handley, M. (2001). ‘The Origins of Christian Commemoration in Late Antique Britain’, Early
Medieval Europe, 10/​2: 177–​199.
Hawkes, S. C. (1973). ‘A Late Roman Buckle from Tripontium’, Transactions of the Birmingham
and Warwickshire Archaeological Society, 85: 145–​159.
Heighway, C. (2010). ‘Christian Origins at Gloucester: A Topographical Inquiry’, in M. Henig
and N. Ramsay (eds), Intersections: The Archaeology and History of Christianity in England,
400–​1200. BAR British Series, 505. Oxford: Archaeopress, 39–​48.
Henig, M. (1993). ‘Votive Objects: Weapons, Miniatures, Tokens and Fired Clay Accessories’, in
A. Woodward and P. Leach, The Uley Shrines: Excavation of a Ritual Complex on West Hill,
Uley, Gloucestershire, 1977–​9. London: English Heritage and British Museum Publications,
130–​147.
Henig, M. (2006). ‘Neither Baths nor Baptisteries’, Oxford Journal of Archaeology, 25: 105–​107.
Henig, M., and Brown, C. (2003). ‘A Romano-​British Buckle Plate from East Challow, near
Wantage’, Oxoniensis, 67: 363–​365.
Johns, C. (1981). ‘The Risley Park Lanx: A Lost Antiquity from Roman Britain’, Antiquaries
Journal, 61: 53–​72.
Johns, C. (2001). ‘Silver Pendant with Monogram Cross’, in P. Leach, Fosse Lane, Shepton Mallet
1990. Britannia Monograph Series 18. London: Britannia, 257–​260.
Johns, C. (2010). The Hoxne Late Roman Treasure:  Gold Jewellery and Silver Plate.
London: British Museum Publications.
Johns, C., and Painter, K. (1991). ‘The Risley Park Lanx Rediscovered’, Minerva, 2/​62: 6–​13.
Johns, C., and Potter, T. (1983). The Thetford Treasure:  Roman Jewellery and Silver.
London: British Museum Publications.
Johns, C. M., and Potter, T. W. (1985). ‘The Canterbury Late Roman Treasure’, Antiquaries
Journal, 65: 312–​352.
Joyce, G. (1881). ‘Third Account of Excavations at Silchester’, Archaeologia, 46: 344–​365.
King, A. (1983). ‘The Roman Church at Silchester Reconsidered’, Oxford Journal of Archaeology,
2: 225–​237.
Knight, J. (2001). ‘Britain’s Other Martyrs: Julius, Aaron and Alban in Caerleon’, in M. Henig
(ed.), Alban and Saint Albans:  Roman and Medieval Art, Architecture and Archaeology.
Leeds: British Archaeological Association, 30–​35.
Knight, J. K. (2010). ‘An Inscription from Bavai and the Fifth Century Christian Epigraphy of
Britain’, Britannia, 41: 283–​292.
Krämer, W. (1958). Neue Ausgrabungen in Deutschland. Berlin: Verlag Gebr. Mann.
Krautheimer, R. (1986). Early Christian and Byzantine Architecture. 4th edn. New Haven,
CT: Yale University Press.
 Christianity in Roman Britain    679

Leach, P. (1982). Ilchester: Excavations 1974–​5. Bristol: Western Archaeological Trust.

Leach, P. (1994). Ilchester Volume 2: Archaeology, Excavations and Fieldwork to 1984. Sheffield
Excavation Reports, 2. Sheffield: Collis.
Leech, R. (1986). ‘The Excavation of a Romano-​Celtic Temple and a Later Cemetery on Lamyatt
Beacon, Somerset’, Britannia, 17: 259–​328.
Mawer, F. (1995). Evidence for Christianity in Roman Britain: The Small Finds. Oxford: BAR
British Series, 243. Oxford: Tempus Reparatum.
Meates, G. W. (1979). The Roman Villa at Lullingstone. Volume 1: The Site. Maidstone: Kent
Archaeological Society.
Meates, G. W. (1987). The Roman Villa at Lullingstone. Volume 2: The Wall Paintings and Finds.
Maidstone: Kent Archaeological Society.
Merrifield, R. (1977). ‘Art and Religion in Roman London: An Inquest on the Sculptures of
Londinium’, in J. Munby and M. Henig (eds), Roman Life and Art in Britain II. BAR British
Series, 41. Oxford: Archaeopress, 375–​406.
Middleton, J. H. (1881–​3). [Proceedings:  2 February  1881]. Proceedings of the Society of
Antiquaries of London. 2nd ser. 9: 66–​70.
Millett, M. (1995). ‘An Early Christian Community at Colchester?’, Archaeological Journal,
152: 451–​454.
Moorhead, S. T. (2000). ‘An Inspiration for the Hinton St Mary Head of Christ’, British Museum
Magazine, 36 (Spring): 22.
Munier, C. (1963). ‘Concilia Galliae A.  314–​A. 506’, Corpus Christianorum CXLVIII Series
Latina. Turnholti, Belgium: Brepols.
Ó Ríordáin, S. P. (1947). ‘Roman Material in Ireland’, Proceedings of the Royal Irish Academy,
47: 77–​150.
Painter, K. S. (1967). ‘The Roman Site at Hinton St Mary, Dorset’, British Museum Quarterly, 32/​
1–​2: 15–​31.
Painter, K. S. (1972). ‘Villas and Christianity in Roman Britain’, in Actas del VIII Congresso
Internacional de Arqueologia Cristiana 1969. Barcelona: Consejo Superior de Investigaciones
Cientificas Egipciaces, 149–​166.
Painter, K. S. (1977). The Water Newton Early Christian Silver. London:  British Museum
Publications.
Painter, K. S. (1999). ‘The Water Newton Silver: Votive or Liturgical?’ Journal of the British
Archaeological Association, 152: 1–​23.
Painter, K. (2010). ‘A Roman Silver Jug with Biblical Scenes from the Treasure Found
at Traprain Law’, in M. Henig and N. Ramsay (eds), Intersections:  The Archaeology
and History of Christianity in England, 400–​ 1200. BAR Reports British Series, 505.
Oxford: Archaeopress, 1–​24.
Pearce, S. (2008). ‘The Hinton St Mary Mosaic Pavement:  Christ or Emperor?’, Britannia,
29: 193–​218.
Penney, S., and Shotter, D. C. A. (1996). ‘An Inscribed Salt-​Pan from Shavington, Cheshire’,
Britannia, 27: 360–​365.
Perring, D. (2003). ‘Gnosticism in Fourth-​ Century Britain:  The Frampton Mosaics
Reconsidered’, Britannia, 34: 97–​128.
Petts, D. (2002). ‘Votive Deposition in Late Roman Britain: Pagan or Christian?’, in M. Carver
(ed.), The Cross Goes North: Processes of Conversion in Northern Europe. Woodbridge: Boydell
Press, 109–​118.
Petts, D. (2003). Christianity in Roman Britain. Stroud: Tempus.
680

680   David Petts

Petts, D. (2004). ‘Early Medieval or Late Antique? Burial in Western Britain ad 410 to 600’, in
R. Collins and J. Gerrard (eds), Debating Late Antiquity ad 300–​700: Proceedings of the York
Conference. BAR British Series, 365. Oxford: Archaeopress, 77–​87.
Petts, D. (2014). ‘Christianity and Cross-​Channel Connectivity in Late and Sub-​Roman Britain’,
in R. Collins, F. Haarer, K. Matthews, S. Moorhead, and D. Petts (eds), Late and Post-​Roman
Britain. London: Society for the Promotion of Roman Studies, 73–​88.
Philpott, R. (1991). Burial Practices in Roman Britain. BAR British Series, 219. Oxford: Tempus
Reperatum.
Quensel-​von Kalbern, L. (1999). ‘The British Church and the Emergence of the Anglo-​Saxon
Kingdoms’, Anglo-​Saxon Studies in Archaeology and History, 10: 89–​98.
Rivet, A. L. F., and Smith, C. (1979). Place Names of Roman Britain. London: Batsford.
Salway, P. (1981). Roman Britain. Oxford: Clarendon Press.
Sanderson, I. (1993). ‘Exciting Find Helps Throw Light on “Dark Ages”’, Clwyd Archaeology
News, 92–​3 (Winter): 2.
Sankey, D. (1998). ‘Cathedrals, Granaries and Urban Vitality in Late Roman London’, in B.
Watson (ed.), Roman London: Recent Archaeological Work. JRA Supplementary Series 24.
Portsmouth, RI: Journal of Roman Archaeology, 78–​82.
Sharpe, R. (2001). ‘The Late Antique “passio” of St Alban’, in M. Henig (ed.), Alban and Saint
Alban’s: Rome and Medieval Art, Architecture and Archaeology. Leeds: British Archaeology
Association, 13–​34.
Snyder, G. F. (2003). Ante Pacem: Archaeological Evidence of Church Life before Constantine.
Macon, GA: Mercer University Press.
Sparey-​Green, C. (1993). ‘The Mausolea Painted Plaster’, in D. E. Farwell and T. I. Molleson
(eds), Poundbury Volume 2:  The Cemeteries. Dorchester:  Dorset Natural History and
Archaeological Society, 135–​150.
Steane, K. (2006). The Archaeology of the Upper City and Adjacent Suburb.
Oxford: Oxbow Books.
Thomas, C. (1981). Christianity in Roman Britain to ad 500. London: Batsford.
Thomas, C. (1993). ‘The Early Christian Inscriptions of Southern Scotland’, Scottish
Archaeological Journal, 17: 1–​10.
Todd, M. (1999). ‘The Latest Inscriptions of Roman Britain’, Durham Archaeological Journal,
14–​15: 53–​59.
Todd, M. (2005). ‘Baths or Baptisteries? Holcombe, Lufton and their Analogues’, Oxford
Journal of Archaeology, 24: 307–​311.
Toynbee, J. (1964). ‘The Christian Roman Mosaic, Hinton St Mary, Dorset’, Proceedings of the
Dorset Archaeological Society, 85: 116–​121.
Turner, R. (1999). Excavations of an Iron Age Settlement and a Roman Religious Complex at
Ivy Chimneys, Witham, 1978–​83. East Anglian Archaeology, 88. Chelmsford: East Anglian
Archaeology.
Watts, D. (1991). Christians and Pagans in Roman Britain. London: Routledge.
West, S. E., and Plouviez, J. (1976). ‘The Roman Site at Icklingham’, East Anglian Archaeology,
3: 63–​125.
Woodward, A., and Leach, P. (1993). The Uley Shrines:  Excavation of a Ritual Complex on
West Hill, Uley, Gloucestershire, 1977–​9. London:  English Heritage and British Museum
Publications.
 Chapter 33

Memories of t h e Past
in Roman Bri ta i n

Zena Kamash

Introduction

In recent years there has been a boom in memory studies, with a particular focus on
collective memory. In what follows I will first present a brief overview of the history of
memory studies and then discuss the role memory has played in archaeological thought
in the 1990s and 2000s. I will then move onto look at the place of memory in the Roman
world, before considering the role of memory in Roman Britain.
Several threads will run through this discussion, including the temporal direction of
memory, the manipulation of memory, and competing memories. These themes revolve
around the idea that people tend to look back in order to address concerns for their pre-
sent and future and, in so doing, are apt to manipulate memories to suit these present
concerns. This can lead to competing memories, whereby there is not a single, uncon-
tested version of the past. This has numerous ramifications for our understandings of
identity in Roman Britain. Working across all these themes will be an exploration of
how memories were enacted in Roman Britain and how these memory practices may be
similar to or different from the wider Roman world.

The History of Memory

The traditional focus of the western episteme, including Roman writers on memory, was
on the individual, private nature of memory. The key Roman texts on memory—​mostly
from works on rhetoric—​can be found in the writings of Cicero (de Oratore 2.74.299–​
300; 2.86.351–​4), Quintilian (Institutio Oratoria 11.2.17–​26), the author of the Rhetorica
ad Herrenium (3.16–​24), and Augustine (e.g. Confessions 10.8–​12). The focus of these
682

682   Zena Kamash

authors is on how we remember and the methods by which we can enhance our ‘natural’
memory through training and mnemonic techniques. The most famous of the exam-
ples used by these authors is that of Simonides of Ceos, who is said to be the inven-
tor of mnemonics. According to the story, Simonides was able to recall who was sitting
where at dinner at the House of Scopas at Crannon (Thessaly), and so able to identify
the remains of those killed after the roof fell in (for a detailed discussion of this story,
see Farrell 1997). In addition, Cicero (de Or. 2.74.299–​300) also foreshadows Borges’s
‘Funes the Memorious’ (2000: 87–​95) and modern neuroscience in his discussion about
how it is kinder to be taught how to forget than how to remember (see, e.g., Parker et al.
2006). The main thread running though all these accounts is that memory was seen as
a storehouse of faithful representations of past sensory experiences that could be drawn
upon as and when required. This was the accepted and unchallenged view of memory
throughout western thought until the twentieth century.
In the twentieth century, there was a major rethinking of the nature of memory. In the
1950s Maurice Halbwachs (1992), a student of Emile Durkheim, proposed the radical
theory of collective memory (Gedächtnisgeschichte). In this new account, memory was
not restricted to individual minds, but could be shared and distributed across groups
and communities. Collective memory could be seen, then, as ‘what remains of the past
in the lived reality of groups, or what these groups make of the past’ (Le Goff 1992: 95; see
also Nora 1978).
This work has been highly influential in memory studies, in particular in the work of
scholars such as Pierre Nora (1984–​92), Aleida Assman (1999) and Jan Assmann (1999).
Much of this work shifted the focus away from ‘memory as storehouse’ and instead
moved towards the idea that memory is malleable and open to change, and so, by impli-
cation, not necessarily a faithful representation of one particular true past. This new
malleable, collective memory opened up the possibility of numerous competing nar-
ratives of past events. (For a full history of collective memory, see Olick et al. 2011. For
critiques of the ‘memory industry’, see Todorov 1995; Confino 1997; Klein 2000.)

The Material Turn: Archaeology

and Memory

Although the important role the physical world has to play in the creation and summon-
ing of memories was implicit in, for example, Pierre Nora’s concept of ‘lieux de mémoire’,
its full potential was largely sidelined in favour of written and oral histories (see, e.g., Le
Goff 1992: 52). Towards the end of the twentieth century, however, collective memory
studies saw a turn towards materials, considering the ways in which material culture is
implicated in both remembering and forgetting. The arrival of a material element to col-
lective memory brought with it a space in which archaeologists could make a contribu-
tion (e.g. Radley 1990; Rowlands 1993; Tatum 1995; Williams and Bradley 1998; Alcock
 Memories of the Past in Roman Britain    683

2002; Van Dyke and Alcock 2003; Jones 2007; Borić 2010; Lillios and Tsamis 2010). In
particular, archaeologists have noted the ability of objects to traverse time and, through
their physical and sensual qualities, to make the past present (Jones 2007: 61).
In addition, this archaeological work—​particularly that of Williams and Bradley
(1998)—​has brought with it the concept of ‘the past in the past’ that is now key in archae-
ological thought. This concept introduced the idea that past peoples were also conscious
of the past and made reference to it in numerous, material ways that we as archaeolo-
gists might detect and interpret. This is a theme to which I will return later in this chap-
ter. Implicit in this idea is the relationship between memory and time. Here, I follow
Schwartz (1982: 395) in thinking that references to the past, material or otherwise, are
often made because of concerns in the present. Furthermore, I would argue that memo-
ries are used not just to create present identities, but also, potentially, to project visions
and hopes for the future. This facet of memory is made clear by Holtorf and Williams
(2006), who introduce the idea of ‘retrospective’ and ‘prospective’ memories in relation
to accumulative and created landscapes, respectively. An example of the latter would be
a war memorial, where the memories being created are intended for the future. In the
first type, retrospective memories, people are looking back at, or to, older elements of
the landscape; in cases where the actual ‘meaning’ of that place or ‘content’ of that mem-
ory is lost, we may find the creation of landscapes of the imagination and concomitant
legendary topographies:

This kind of ‘looking back’ is not necessarily about accurately recalling past events as
truthfully as possible: it is rather about making meaningful statements about the past
in the given cultural context of a present as well as evoking aspirations for the future.
(Holtorf and Williams 2006: 238)

While we can point to distinct examples of these kinds of memorial processes in land-
scapes, they are often found acting together in a single landscape, as will be demonstrated
later in this chapter in my discussion of examples of prospective and retrospective mem-
ories from Roman Britain. They also point to the fact that the deliberate selection of
points referenced in a landscape can be highly manipulative, hiding some aspects of the
past and encouraging others to be forgotten. This links with some caveats made by Hella
Eckardt (2004), who reminds us that, when we, as archaeologists, look at landscapes,
we should not assume that all past objects and monuments held equal significance (an
easy trap into which to fall when considering the past in the past). She encourages us to
search not ‘simply for re-​use, but for more complex interactions with the past, which
include nostalgia, exclusion, destruction and forgetting’ (Eckardt 2004: 46).
On the topic of destruction and forgetting, Susanne Küchler’s studies (1987, 1988) of
malangan sculpture in Papua New Guinea has demonstrated that the creation and delib-
erate destruction of sculpture were powerful methods not only of remembering, but
also of forgetting. Küchler neatly reminds us that remembering and forgetting can often
not be separated. Also central to Küchler’s work is the concept that the action of remov-
ing an object from memory can be as important as the object itself (see also Connerton
684

684   Zena Kamash

1998). This highlights the fact that memory cannot be of objects alone, but must be
thought about in conjunction with human experience and action—​that is, that memory
is distributed between people and things (Renfrew et al. 2004: 2; Jones 2007: 26). Some
of the most powerful examples of these actions of remembering and forgetting through
objects are found in Holocaust memorials. The Stolpersteine, for example, are small,
cobble-​sized bronze plaques that are set into pavements across many cities in northern
Europe. Each gives the name of a victim of the Holocaust. One of the most striking com-
mentaries on these Stolpersteine—​and of direct relevance to this discussion of objects,
action, and forgetting—​comes from Joseph Pearson (2010):

It is not what is written which intrigues me, because the inscription is insufficient to
conjure a person. It is the emptiness, void, lack of information, the maw of the forgot-
ten, which gives the monuments their power and lifts them from the banality of a sta-
tistic. (emphasis added)

Memory in the Roman World

Most recently Karl Galinsky’s Memoria Romana project (2009-​2013) has explicitly
sought to explore the role concepts such as Gedächtnisgeschichte might play in our
understandings of memory in the Roman world (Galinsky 2009). The project funded
numerous Ph.D. scholarships and postdoctoral research projects, whose fruits are
beginning to make themselves felt, including my own work on memory and religion in
Roman Britain (see Fabrizi 2012; Li Causi 2012; Farrell and Nelis 2013; Raymond 2013;
Rebeggiani 2013; Seider 2013; Kamash 2016; Barchiesi, forthcoming; Kamash, forth-
coming). The project also organized major conferences in Rome (Memory in Rome
and Rome in Memory, which included a contribution from Daniel Libeskind (Galinsky
2014)) and at the Getty Villa (Cultural Memories in the Roman Empire (Galinsky 2016)).
This project has been able to use a wealth of previous work on Roman memory as its
springboard. In many ways the important role of memory in Roman society has long
been recognized in studies of the Roman world and pre-​dates the attention paid to mem-
ory in other areas of archaeology. This is due, partly, to the writings of Roman authors
on memory and, partly, to the much-​discussed idea of damnatio memoriae—​the act of
deliberately removing from public view all images of, and inscribed references to, dis-
graced emperors; for example, Domitian, or Caracalla’s unfortunate brother, Geta (on
this in Britain, see Barrett 1993). Of course, it is also now widely recognized that such
actions serve less to make people forget than to act as powerful reminders of that person
and his or her actions (Flower 1998, 2006). The act of damnatio memoriae does, however,
bring to the fore two main features of memory in the Roman world: first, that memory
was central to the lives—​and perhaps most importantly the afterlives—​of Roman people,
and, secondly, that the Romans were well practised in the art of manipulating memories.
 Memories of the Past in Roman Britain    685

A few, highly selective, examples from in and around the city of Rome will serve to
demonstrate these features (a useful and authoritative set of bibliographies on the sub-
ject can be found on the Memoria Romana website:  Galinsky 2014). The Via Appia,
which connects Bovillae and Rome, is renowned for the elaborate tombs that lined its
course. Here the memory of those who had died was kept alive in tombs that bore their
names, faces, and sometimes inscriptions detailing their careers and genealogies. These
tombs and their prominent position highlight the centrality of vision and visibility in
the maintenance of memory in Rome (see Koortbojian 1996).
Of all monuments, the Arch of Constantine in Rome perhaps best illustrates the role
that monuments could play in memory in the Roman world. This monument was dedi-
cated to the emperor Constantine by the senate for his decennial celebrations. The arch
is well known for its use of spolia—​reused architectural components—​including exam-
ples of earlier imperial portraiture being recut into the face of Constantine. In his analy-
sis of the monument, and particularly the use of spolia, Gutteridge (2010) suggests that
this is a ‘trans-​temporal’ monument that defies the law of time. The memories of van-
ished buildings and dead emperors cannot be completely erased in this monument but,
rather, are recombined into a new set of remembrances that look forwards and back-
wards simultaneously.
Perhaps the best example of Roman manipulation of memory can be seen in
Augustus’ building programme in Rome (see Gowing 2005: 132–​159). Augustus deliber-
ately chose to maintain what could be considered as antiques in the forum, such as the
umbilicus urbis (the central pit dug by Romulus when he laid out the sacred boundary of
the city) and the lapis niger (the spot in front of the senate house where Romulus disap-
peared—​either murdered or taken to the heavens), reminding the people of his com-
mitment to protecting the old traditions of Rome. He also constructed new buildings,
such as the Temple to Divus Julius, linking himself both to the recent past and to a divine
lineage. In these ways, Augustus made use of spaces and buildings in Rome to create
both a museum of the past and a dynastic monument to his family, and so to manipulate
strongly the memories of Rome’s inhabitants.

Remembering in Roman Britain

Looking at memory in Roman Britain presents somewhat different challenges and

opportunities from studies of the remainder of the empire. As well as the distance of
Britain from Rome, major differences lie in the nature of the archaeological evidence and
in the history of Romano-​British studies. Although Britain does have some examples
of still extant monumental architecture (for example, the amphitheatres at Silchester,
Cirencester, and Chester), these are in no way on the same scale as the Mediterranean.
This has clear ramifications: not for us the detailed analyses of sculpture or streets lined
with funerary monuments. This does not mean, however, that we cannot engage with
the debates over memory in the Roman world. On the contrary, I am of the opinion
686

686   Zena Kamash

that, because of these differences, scholars of Britain under the Roman Empire have, for
a long time, allied themselves closely with prehistoric archaeology and so have taken a
more theoretically driven approach, putting us ahead of the memory game in compari-
son to other parts of the Roman world. Foremost, here, is the work of Howard Williams
(1998) on the past in the past in relation to Roman Britain. In addition, we also have the
major advantage of access to an ever-​increasing amount of high-​quality excavation data
(Fulford and Holbrook 2011), from which to make detailed analyses of the processes of
memory in Roman Britain.
In what follows I hope to capitalize on these substantial opportunities. I will look at
three realms of memory in Roman Britain: the burial of memories; the reorganization
of landscapes and memories; and the building of religious memory. First, I will explore
how burying people and objects, and in so doing making them invisible, can be part of
the memory process. The focus here will be on embodied actions and the use of legend-
ary topographies in the landscape. I will then move on to examine how different mem-
ory communities responded to major periods of landscape reorganization with the aim
of demonstrating how these analyses can contribute to wider discussions about the crea-
tion and maintenance of identity in Roman Britain. Finally, I will explore the Romano-​
Celtic temple phenomenon. I will argue that their construction—​and the memories
generated and evoked by them—​was linked to the kinds of memory-​making that are
particularly prevalent in times of social instability and that this phenomenon should be
seen as part of a broader set of processes that began in the late Iron Age.
My aim here is not to be exhaustive, but rather to demonstrate the potential of taking
a memory approach in understanding Roman Britain. In particular, I wish to demon-
strate that focusing on memory can contribute to, and enhance, wider topical discus-
sions about continuity and change and identity. In so doing, I will be drawing on many
of the insights about memory outlined above, with an emphasis on the temporal direc-
tions of memory practices in Roman Britain and what these may suggest about the peo-
ple enacting these practices. In particular, I will make reference to the active interplay
between people and things in processes of remembering and forgetting, and to the ret-
rospective and prospective directions of memory.
As noted above, one of the biggest differences between Roman Britain and the
Mediterranean is the scale of monumentality. This challenges us to consider the ways in
which invisibility and absence might be implicated in remembering in Roman Britain.
In relation to this, Howard Williams (2004) considers the ephemeral nature of crema-
tion practices in Roman Britain. In this analysis, he rightly points out that the practices
associated with cremating the dead may have been more significant as a ‘technology of
remembrance’ than the physical presence of a permanent, enduring monument. As with
the example of malangan sculpture, embodied action is key here for such ephemera to
be endowed with the power of prolonged memory creation.
Also important in thinking about the absent and what is not seen is the hold
that these may take on people’s imaginations, as noted, for example, in the crea-
tion of legendary topographies (Holtorf and Williams 2006) and in the power of the
Stolpersteine. Evidence for such can be seen across Britain at the sites of Neolithic
 Memories of the Past in Roman Britain    687

and Bronze Age monuments where we find the burial of people and objects in the
Roman period (see Aitchison 1988; Dark 1993; Williams 1998; Darvill 2004; Hutton
2011; Vejby 2012). One example of this phenomenon is at Silbury Hill, a Neolithic
artificial chalk mound near Avebury, Wiltshire. Field (2003) has noted that each time
a part of Silbury Hill has been investigated more Roman than Neolithic finds have
been discovered. While this may, in some ways, not be surprising, given that Roman
sites are almost always more productive than those dating to the Neolithic, there is,
nevertheless, a considerable concentration of Roman activity at this hill. This activ-
ity includes the deposition of material in at least three shafts. The finds from these
shafts included pottery, coins, bronze objects, glass beads and vessels, iron objects and
sculpted masonry, as well as deer, sheep, cattle, pig, fox, and dog bones (Brooke and
Cunnington 1896; Brooke 1910; Hutton 2011). This assemblage bears a striking resem-
blance to the deposits found in the pits and wells of Silchester (Fulford 2001), sug-
gesting that those at Silbury Hill too may have been of ritual significance. In addition
to these shafts, ninety-​four coins, a bronze bracelet, and Roman pottery were found
in the ditch of the mound itself (Moorhead 2001: 100; Gillings and Pollard 2004: 97–​
100). These finds are significant because they suggest that, while those engaging in
activity at Silbury Hill may not have known its detailed history, and indeed may well
have thought it a natural landscape feature, they may have been aware of previous
human activity at the mound in the shape of the ditch. Such associations may have
leant this curious mound mythological and legendary power in the Roman period,
making it suited to the activity seen in the shaft deposits.
Chronology seems to be key to the activities at these legendary topographies. Hutton
(2011: 15–​16) notes three elements of recurrent chronological patterning in these acts
of long-​term memory creation at Neolithic chambered tombs, the Wessex monuments
and caves in the Bristol Channel region: (1) deposition of objects seems to have been
more widespread in the Romano-​British period than in other periods; (2) there was an
apparent reduction of activity in later prehistory, followed by a ‘revival’ under Roman
rule; and (3)  the datable finds are overwhelmingly from the mid-​third century ad
onwards. This patterning seems to be highly significant. The apparent break in activity
in later prehistory adds weight to the idea that we are looking at a form of retrospective
memory-​making focused around landscapes of the imagination and legendary topog-
raphies. Secondly, the concentration of these activities in the third century ad demon-
strates that memory-​making can be more prevalent at certain times than at others. It is
difficult to be sure what the concentration in the third century indicates. One possibility
is that, as the third century ad was a period of prosperity in Roman Britain, we may be
seeing here activities linked to self-​confidence and optimism about the future.
In contrast, reactions to Roman domination from ad 43 onwards are widely accepted
to be multiple and varied (see Mattingly 2007 on discrepant identities). Jeremy Taylor’s
Atlas of Roman Rural Settlement (2007) ably demonstrates, for example, the variety of
reactions we find in rural settlements and landscape organization across England. In
Norfolk, for example, dramatic changes in the organization of the landscape occurred
in the years immediately after the conquest (Taylor 2007: 66, 97–​101, 110). The story is
688

688   Zena Kamash

very different, however, in Cornwall and the Mersey Basin, where there seems to be
prolonged continuity from the Iron Age into the Roman period, with the major period
of reorganization not occurring until the third century ad (Taylor 2007: 57–​59, 69–​7 1,
110–​111). In the Upper and Middle Thames Valley we see a similar pattern of continuity
from the Iron Age into the Roman period, followed by a period of major reorganization
and change—​this time in the second century ad. This reorganization of the landscape
resulted in the construction of large settlement enclosures, trackways, and timber-​
aisled buildings—​for example, at Claydon Pike and Neigh Bridge, Somerford Keynes
(Gloucestershire) (Miles et al. 2007: 377). One exception to this general pattern for the
Upper and Middle Thames Valley is a small group of sites near Bicester that seem to have
been abandoned in the conquest period (Henig and Booth 2000: 106). One possibility
is that these sites lay within the tribal area of the Catuvellauni, rather than the Dobunni
like the rest of the Upper Thames Valley (Miles et al. 2007: 376), and so the variation that
we see here may be the tangible consequence of different reactions to the conquest from
one tribe to another.
Clearly, whenever they occurred, these episodes of major landscape reorganiza-
tion would have been unsettling periods of upheaval and disruption that would have
required significant social renegotiation of the landscape. Before I look at this in more
detail, a caveat must be sounded here, related to the limits of the explanatory power of
a memory approach. Memory is a powerful tool for suggesting potential impacts and
effects on people and how they have created and re-​created their identities, but it cannot
so easily provide an explanation of the causal factors in the past. In other words, while
I may be able to suggest some of the effects of the second century ad reorganization
on memory communities, I will not be explaining why or how this occurred—​this is a
question that is worthy of more attention elsewhere.
In addition to ‘angst depots’, another major method of social renegotiation in times
of stress is through the medium of religion. At Claydon Pike, the reorganization of the
landscape was accompanied by the delineation of a probable sacred precinct in trench
19 (Miles et al. 2007: 164–​165). Some of the activities that seem to have been performed
here included the burning of incense—​the strong sensory impacts of which may have
been central to memory creation here (on smell, see Drobnik 2006). As the area lacked
any clear religious or sacred precursor and seems to have been created anew in the sec-
ond century ad, it would seem reasonable to suggest that the memory practices asso-
ciated with this site would have been prospective in nature. So, we may suggest that
the response of the inhabitants of Claydon Pike to this upheaval was to create some-
thing new around which they could focus memories for the future. At Neigh Bridge,
Somerford Keynes, however, there is evidence for a shrine at the river crossing, marked
by extraordinarily high numbers of first–​second century ad brooches (Miles et al. 2007:
388). In contrast to Claydon Pike, this religious focus seems to have traversed the period
of continuation and reorganization, suggesting that both retrospective and prospective
memories may have been at play. The retrospective nature of the memories is further
strengthened by the focus on water, which is widely recognized as a long-​term fascina-
tion in Britain (see Kamash 2008). What this suggests is that similar-​seeming sets of
 Memories of the Past in Roman Britain    689

processes, when probed, may expose divergent memory practices, similar in nature and
probably contributing to, Mattingly’s discrepant identities (2007).
The question of continuity from the Iron Age into the Roman period touched on here
is one of the main debates in studies of religion in Roman Britain. This debate centres
mostly on whether there were precursors to Romano-​Celtic temples on particular sites
(see, e.g., Hingley 1982, 1985; Henig 1984; Watt 1998; Smith 2001: 13–​16, 33–​75; Kamash
et al. 2010). In addition, Williams (1998) has carried out a survey on the reuse of ancient
monuments by Romano-​Celtic temples, pointing to numerous instances of Roman
period ritual activity in and around Bronze Age barrows and Iron Age hillforts. In what
follows, instead of thinking about reuse or continuity on particular sites, I will consider
the phenomenon of building Romano-​Celtic temples more generally and explore fur-
ther how memories of the past could be used to create new social constellations.
The earliest known temples of this type in Britain, such as those at Harlow (Essex),
Heathrow (Middlesex), and Hayling Island (Hampshire), date to the latest Iron Age
immediately preceding the Roman conquest (for Harlow: Wheeler 1928; France and
Gobel 1985; Bartlett 1988; for Heathrow: Grimes and Close-​Brooks 1993; for Hayling
Island: Downey et al. 1980; King and Soffe 1991, 1994, 1998; Smith 2001: 40–​44). There is
no obvious architectural precedent in Britain for this building form; the nearest parallels
seem to come from the continent (see, e.g., King 1990). Interestingly, the later Iron Age
is a period that sees numerous shifts in the uses of, and practices surrounding, material
culture—​particularly that imported from the continent.
In the 1980s and early 1990s, these shifts and imports were explained and discussed
through the framework of world-​systems theory and core-​periphery models, leading to
a focus on elite desire for goods from the Mediterranean and the all-​pervasive allure of
Rome (e.g. Cunliffe 1988). Such perspectives have been increasingly challenged of late,
with more agency being given to the native Britons in these exchanges. Most recently, J.
D. Hill (2007) and Niall Sharples (2010) have championed the idea that the shifts we see
in the archaeological record were in response to significant changes in the social con-
text in Britain, and that the various permutations of these adoptions and changes were
used to reinforce particular identities and bind people together into distinctive groups.
The usual list of ‘new things’ in the late Iron Age includes distinctive burial practices,
such as the Aylesford–​Swarling cremations in south-​eastern England, the use of a coin-
age in some areas, and the use of material culture imported from the continent. At the
same time as this introduction of new practices and forms of material culture, many of
the hillforts were being abandoned, leaving a large social vacuum. This social vacuum
may have been partially filled by oppida settlements, but may also have been filled by
Romano-​Celtic temples. Furthermore, by placing the initial impetus for Romano-​Celtic
temples within these wider late Iron Age processes of change and flux, we can also give
context to the otherwise unanchored appearance of Romano-​Celtic temples in Britain.
The shrines in the late Iron Age religious complex at Westhampnett Area 2 are interest-
ing in this regard (Fitzpatrick 1997: 229; 2008: 180, 254, figures 83–8​4). The site com-
prises 2–​4 shrines, numerous pyre sites, and c. 160 cremations. The closest parallels for
the rectangular and sub-​rectangular shrines are found on hillforts, such as Danebury
690

690   Zena Kamash

(Structures 2 and 3) and Cadbury Castle (Cunliffe 1984), suggesting a possible link
between this style of shrine and hillforts.
Sharples (2010: 116–​124) has argued, with specific reference to Wessex, that hillforts
were powerful mechanisms for creating communities in the early Iron Age through a
combination of employing labour as potlatch and the concomitant feasting that would
accompany these building, and rebuilding, episodes (on memory and food, see Sutton
2001). The construction, and any subsequent remodelling, of Romano-​Celtic temples
and their complexes would have required a similar mobilization of human effort (which
may have been coerced or voluntary) and of feasting. At Marcham/​Frilford, for exam-
ple, we find large deposits of animal bone and pottery in the top fills of the Iron Age dou-
ble-​ditched enclosure that lay to the south of the Roman temenos. The pottery in these
ditches includes some very unusual imported forms, including raspberry rouletted
wares, which are dated to the period immediately post-​conquest. It seems reasonable
to suggest that these deposits were the remains of feasting episodes associated with the
builders of the temenos and temple.
At Hayling Island (Hampshire) the timber shrine was replaced in c. ad 50–​60 by a cir-
cular temple (Downey et al. 1980; King and Soffe 1991, 1994, 1998; Smith 2001: 40–​44). The
finds from this phase were dominated by animal bones and pottery, which also suggests
that the people involved in this rebuilding indulged in grand feasts, similar to those argued
for by Sharples (2010) at hillfort sites. Likewise, at Harlow (Essex) a masonry Romano-​
Celtic temple was constructed around 7 metres north of a pre-​Flavian circular feature
(Wheeler 1928; France and Gobel 1985; Bartlett 1988). As well as coins being deposited
repeatedly and continuously in both phases, there was also repeated deposition of animal
bones (Haselgrove 1989: 76; Smith 2001: 85–​86). The animal bone evidence demonstrated
continued preference for sheep across both phases and selectivity in side of animal (right-​
over left-​handed elements) (Smith 2001: 86). The sheep also seem to have been killed at
particular times of year. This evidence suggests that the feasting may originally have been
associated with episodes of building and rebuilding. The apparently annual kill cycle sug-
gests that feasting may then have been used as a device to evoke the memory of labouring
episodes and, in turn, to create new memory ties for younger generations.
Anthony King (2005: 357) has also noted that the high percentages of sheep/​goat
remains at his Group A sites (e.g. Hayling Island, Uley, and Lowbury) show a tempt-
ingly similar pattern to Iron Age sites. Although he ultimately goes on to dismiss this
pattern as being evidence for cultural similarity—​citing further similarities with sites in
Gaul and Germany—​this actually seems to reinforce the argument for linking these tem-
ple sites not only to earlier Iron Age practices, but also to the late Iron Age processes of
continental borrowing—​that is, this borrowing may have been facilitated by fitting into
established patterns of practice that included feasting habits (see also T. Moore, this vol-
ume, for discussion of continental influences on Britain). In these ways, then, Romano-​
Celtic temples seem to have been created as a new space for gathering together people
and memories.
This was, however, no simple renegotiation and replacement. First, while patterns
of practice (such as feasting) may be repeated over time, this does not mean that they
 Memories of the Past in Roman Britain    691

are immutable and monolithic; rather, they can be simultaneously both old and new,
both retrospective and prospective. Secondly, although there are a handful of verified
instances of Romano-​Celtic temples that were situated within Iron Age hillfort enclo-
sures (Williams 1998; Smith 2001:  151), the vast majority of Romano-​Celtic temples
(over eighty out of ninety known sites) positioned themselves in relation to deeper
Bronze Age and Neolithic pasts, avoiding the more recent Iron Age past (Kamash,
forthcoming a). One implication behind this spatial configuration and patterning is
that the collective memories generated by these sites were being directed physically
away from the more controversial recent past and towards an earlier, possibly more
idealized, past. Another possibility is that by not building directly over the direct
past this may have been a powerful form of protest in itself: leaving behind a physi-
cal reminder of what had to be left behind in the face of major social upheavals. These
spaces are both abandoned and empty and yet still living in tension with the newly cre-
ated spaces. We see, then, selectivity and manipulation of memory in referencing the
past at, and through, Romano-​Celtic temples—​in terms of both material practices and
landscape placement.

Conclusions

Overall, using memory as an approach to understanding the patterns of practice that

we see archaeologically in Roman Britain opens up a space for exploring some of the
concerns (and the effects of those concerns) of the people living in Roman Britain. To
some extent we may also be able to use memory to probe the motivations that lay behind
those practices, though, as I noted earlier, the power of a memory approach probably
does not lie in its ability to provide direct explanations or causal factors—​memory is
too slippery for such specificity. Memory is very useful, however, in demonstrating the
variety of ways in which people have engaged with the past and the variety of pasts that
may have been created through those enacted memories. What people chose to remem-
ber or selected as the focus of their memory practices could vary widely even between
neighbours, such as the inhabitants of Claydon Pike and Neighford Bridge. What seems
to link these choices seems invariably to be related to present concerns and so we often
find evidence for a rich interplay between retrospective and prospective memories. In
this way, the enactment of memories in Britain seems similar to that in the wider Roman
world. The major difference that we see between Britain and the rest of the empire
revolves around the links between visibility and memory. Visibility of memories is, per-
haps implicitly, central to most discussions of memory in the Roman world. In Britain,
however, we find numerous instances of invisible, buried memories whose power lies
precisely in their lack of visibility and their absence. Whether this represents an actual
difference between Britain and elsewhere in the Roman Empire, or whether this belies
a difference in evidence and approach, remains to be seen. This suggests, then, that we
may be only at the beginning of memory in the Roman world.
692

692   Zena Kamash

Acknowledgements

I owe a great debt of thanks to Professor Karl Galinsky and the Max Planck Society for
providing generous funding and support for this work through the Memoria Romana
project. My thanks also go to my graduate students for helping me develop various ideas
expressed here in our numerous discussions. In addition, I am very grateful to the edi-
tors of this volume for their insightful and helpful comments on an earlier version of this
chapter.

References
Primary Sources
Augustine, Confessions, trans. E. B. Pusey. New York: Pocket Books, 1907.
Cicero, De Oratore, trans. E. W. Sutton. Loeb Classical Library. Cambridge, MA:  Harvard
University Press, 1967.
Quintilian, Institutio Oratoria, trans. H. E. Butler. Loeb Classical Library. Cambridge,
MA: Harvard University Press, 1920–​2.
Rhetorica ad Herrenium, trans. H. Caplan. Loeb Classical Library. Cambridge, MA: Harvard
University Press, 1954.

Secondary Sources
Aitchison, N. B. (1988). ‘Roman Wealth, Native Ritual: Coin Hoards within and beyond Roman
Britain’, World Archaeology, 20: 270–2​83.
Alcock, S. E. (2002). Archaeologies of the Greek Past: Landscape, Monuments and Memories.
Cambridge: Cambridge University Press.
Assmann, A. (1999). Erinnerungsräume: Formen und Wandlungen des kulturellen
Gedächtnisses. Munich: C. H. Beck.
Assmann, J. (1999). Moses the Egyptian:  The Memory of Egypt in Western Monotheism.
Cambridge, MA: Harvard University Press.
Barchiesi, A. (forthcoming). The War for Italia:  Conflict and Collective Memory in Vergil’s
Aeneid. Berkeley and Los Angeles: University of California Press.
Barrett, J. (1993). ‘Chronologies of Remembrance: The Interpretation of Some Roman
Inscriptions’, World Archaeology, 25/​2: 236–​247.
Bartlett, R. (1988). ‘The Harlow Celtic Temple’, Current Archaeology, 112: 163–​166.
Borges, J. L. (2000). Labyrinths. London: Penguin Modern Classics.
Borić, D. (2010) (ed.). Archaeology and Memory. Oxford: Oxbow Books.
Brooke, J. W. (1910). ‘The Excavation of a Roman Well near Silbury Hill’. Wiltshire Archaeological
and Natural History Magazine, 36: 373–​375.
Brooke, J. W., and Cunnington, B. H. (1896). ‘Excavation of a Roman Well near Silbury Hill’,
Wiltshire Archaeological and Natural History Magazine, 29: 166–​171.
Confino, A. (1997). ‘Collective Memory and Cultural History: Problems of Method’, American
Historical Review, 102: 1386–​1404.
 Memories of the Past in Roman Britain    693

Connerton, P. (1998). How Societies Remember. Cambridge: Cambridge University Press.

Cunliffe, B. (1984). Danebury:  An Iron Age Hillfort in Hampshire. Volumes 1 and 2.  CBA
Research Report 52. London: Council for British Archaeology.
Cunliffe, B. (1988). Greeks, Romans and Barbarians: Spheres of Interaction. London: Batsford.
Dark, K. R. (1993). ‘Roman-​Period Activity at Prehistoric Ritual Monuments in Britain and
the Armorican Peninsula’, in E. Scott (ed.), Theoretical Roman Archaeology First Conference
Proceedings. Aldershot: Avebury, 133–​146.
Darvill, T. (2004). Long Barrows of the Cotswolds and Surrounding Areas. Stroud: Tempus.
Downey, R, King, A., and Soffe, G. (1980). ‘The Hayling Island Temple and Religious
Connections across the Channel’, in W. Rodwell (ed.), Temples, Churches and Religion in
Roman Britain. BAR British Series 77(i). Oxford: Archaeopress, 289–​302.
Drobnik, J. (2006). ‘Preface: Part VII: Sublime Essences’, in J. Drobnik (ed.), The Smell: Culture
Reader. Oxford: Berg, 371–​374.
Eckardt, H. (2004). ‘Remembering and Forgetting in the Roman Provinces’, in B. Croxford,
H. Eckardt, J. Meade, and J. Weekes (eds), TRAC 2003: Proceedings of the Thirteenth Annual
Theoretical Roman Archaeology Conference, Leicester 2003. Oxford: Oxbow Books, 36–​50.
Fabrizi, V. (2012). Mores Veteresque Novosque: La Rappresentazione del Passato e del Presente di
Roma negli Annales di Ennio. Pisa: ETS.
Farrell, J. (1997). ‘The Phenomenology of Memory in Roman Culture’, Classical Journal,
92: 373–​383.
Farrell, J. and Nelis, D. (2013) (eds). Augustan Poetry and the Roman Republic. Oxford: Oxford
University Press.
Field, D. (2003). ‘Silbury Hill’, British Archaeology, 70: 15–​17.
Fitzpatrick, A. (1997/2008). Archaeological Excavations on the Route of the A27 Westhampnett
Bypass, West Sussex, 1992. Volumes 1 and 2. Salisbury: Trust for Wessex Archaeology.
Flower, H. (1998). ‘Rethinking damnatio memoriae: The Case of Cn. Calpurnius Piso in ad 20’,
Classical Antiquity, 17: 155–​186.
Flower, H. (2006). The Art of Forgetting: Disgrace and Oblivion in Roman Political Culture.
Chapel Hill, NC: University of North Carolina Press.
France, N. E., and Gobel, B. M. (1985). The Romano-​British Temple at Harlow. Gloucester: Alan
Sutton.
Fulford, M. (2001). ‘Links with the Past:  Pervasive “Ritual” Behaviour in Roman Britain’,
Britannia, 32: 199–​218.
Fulford, M., and Holbrook, N. (2011). ‘Assessing the Contribution of Commercial
Archaeology to the Study of the Roman Period in England, 1990–​2004’, Antiquaries
Journal, 91: 323–​345.
Galinsky, K. (2009) Memoria Romana website. <http://​www.utexas.edu/​research/​memoria/​)>
(accessed May 2013).
Galinsky, K. (2014) (ed.). Memoria Romana:  Memory in Rome and Rome in Memory.
Rome: Memoirs of American Academy in Rome.
Galinsky, K. (2016) (ed.). Cultural Memories in the Roman Empire. Los Angeles: Getty
Publications.
Gillings, M., and Pollard, J. (2004). Avebury. London: Duckworth.
Gowing, A. M. (2005). Empire and Memory:  The Representation of the Roman Republic in
Imperial Culture. Cambridge: Cambridge University Press.
Grimes, W. F., and Close-​ Brooks, J. (1993). ‘Excavation of Caesar’s Camp, Heathrow,
Harmondsworth, Middlesex, 1944’, Proceedings of the Prehistoric Society, 59: 303–​360.
694

694   Zena Kamash

Gutteridge, A. (2010). ‘The Depiction of Time on the Arch of Constantine’, in D. Borić (ed.),
Archaeology and Memory. Oxford: Oxbow Books, 158–​170.
Halbwachs, M. (1992). On Collective Memory, ed. and trans. L. A. Coser. Chicago: University of
Chicago Press.
Haselgrove, C. (1989). ‘Iron Age Coin Deposition at Harlow Temple, Essex’, Oxford Journal of
Archaeology, 8/​1: 73–​88.
Henig, M. (1984). Religion in Roman Britain. London: Batsford.
Henig, M., and Booth, P. (2000). Roman Oxfordshire. Stroud: Sutton Publishing.
Hill, J. D. (2007). ‘The Dynamics of Social Change in Later Iron Age Eastern and South-​Eastern
England c.300 bc–​ad 43’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain
and Beyond. Oxford: Oxbow Books, 16–​40.
Hingley, R. (1982). ‘Recent Discoveries of the Roman Period at the Noah’s Ark Inn, Frilford,
South Oxfordshire’, Britannia, 13: 305–​309.
Hingley, R. (1985). ‘Location, Function and Status: A Romano-​British “Religious Complex” at
the Noah’s Ark Inn, Frilford (Oxfordshire)’, Oxford Journal of Archaeology, 4/​2: 201–​214.
Holtorf, C., and Williams, H. (2006). ‘Landscapes and Memories’, in D. Hicks and M. C.
Beaudry (eds), The Cambridge Companion to Historical Archaeology. Cambridge: Cambridge
University Press, 235–​254.
Hutton, R. (2011). ‘Romano-​British Reuse of Prehistoric Ritual Sites’, Britannia, 4: 1–​22.
Jones, A. (2007). Memory and Material Culture. Cambridge: Cambridge University Press.
Kamash, Z. (2008). ‘What Lies Beneath? Perceptions of the Ontological Paradox of Water’,
World Archaeology, 40/​2: 224–​235.
Kamash, Z (2016) ‘From the Individual to the Collective: Memory Practices on Religious Sites
in Roman Britain’, in K. Galinsky (ed.) Cultural Memories in the Roman Empire. Los Angeles:
Getty Museum Press, 153–169.
Kamash, Z. (forthcoming). ‘Memory, Materiality and Religion in Roman Britain’, in T. Clack
(ed.), Archaeology, Syncretism and Creolisation. Oxford: Oxford University Press.
Kamash, Z., Gosden, C., and Lock, G. (2010). ‘Continuity and Religious Practices in Roman
Britain:  The Case of the Rural Religious Complex at Marcham/​Frilford, Oxfordshire’,
Britannia, 41: 95–​125.
King, A. (1990). Roman Gaul and Germany. Berkeley and Los Angeles:  University of
California Press.
King, A. (2005). ‘Animal Remains from Temples in Roman Britain’, Britannia, 36: 329–​369.
King, A., and Soffe, G. (1991). ‘Hayling Island’, in R. F. L. Jones (ed.), Britain in the Roman
Period: Recent Trends. Sheffield: J. R. Collis Publications, 111–​113.
King, A., and Soffe, G. (1994). ‘The Iron Age and Roman Temple on Hayling Island’, in A.
Fitzpatrick and E. L. Morris (eds), The Iron Age in Wessex: Recent Work. Salisbury: Trust for
Wessex Archaeology, 114–​116.
King, A., and Soffe, G. (1998). ‘Internal Organisation and Deposition at the Iron Age Temple
on Hayling Island’, Proceedings of the Hampshire Field Club Archaeological Society,
53: 35–​47.
Klein, K. (2000). ‘On the Emergence of Memory in Historical Discourse’, Representations,
69: 127–​150.
Koortbojian, M. (1996). ‘In commemorationem mortuorum: Text and Image along the “Streets
of Tombs”’, in J. Elsner (ed.), Art and Text in Roman Culture. Cambridge:  Cambridge
University Press, 210–​233.
Küchler, S. (1987). ‘Malangan: Art and Memory in a Melanesian Society’, Man, 22: 238–​255.
 Memories of the Past in Roman Britain    695

Küchler, S. (1988). ‘Malangan: Objects, Sacrifice and the Production of Memory’, American

Ethnologist, 15: 625–​637.
Le Goff, J. (1992). History and Memory, trans. S. Randall and E. Claman. New York: Columbia
University Press.
Li Causi, P. (2012). Il Riconoscimento e il Ricordo: Fama e Memoria nel De Beneficiis di Seneca.
Palermo: Palumbo Editore.
Lillios, K. T., and Tsamis, V. (2010) (eds). Material Mnemonics: Everyday Memory in Prehistoric
Europe. Oxford: Oxbow Books.
Mattingly, D. (2007). An Imperial Possession: Britain in the Roman Empire. London: Penguin.
Miles, D., Palmer, S., Smith, A., and Perpetua Jones, G. (2007). Iron Age and Roman
Settlement in the Upper Thames Valley: Excavations at Claydon Pike and Other Sites within
the Cotswold Water Park. Thames Valley Landscapes Monographs 26. Oxford: Oxford
Archaeology.
Moorhead, T. S. N. (2001). ‘Roman Coin Finds from Wiltshire’, in P. Ellis (ed.), Roman Wiltshire
and After. Devizes: Wiltshire Archaeological and Natural History Society, 85–​107.
Nora, P. (1978). ‘Mémoire collective’, in J. Le Goff, R. Chartier, and J. Revel (eds), La Nouvelle
Histoire. Paris: Retz.
Nora, P. (1984–​92). Lieux de Mémoire. Paris: Gallimard.
Olick, J., Vinitzky-​Seroussi, V., and Levy, D. (2011) (eds). The Collective Memory Reader.
Oxford: Oxford University Press.
Parker, E. S., Cahill, L., and McGaugh, J. L. (2006). ‘A Case of Unusual Autobiographical
Remembering’, Neurocase, 12: 35–​49.
Pearson, J. (2010). ‘Nazi Victims and Stumbling Blocks to Memory’ <http://​needleberlin.com/​
2010/​08/​23/​nazi-​victims-​and-​stumbling-​blocks-​to-​memory/​> (accessed May 2013).
Radley, A. (1990). ‘Artefacts, Memory and a Sense of the Past’, in D. Middleton and D. Edwards
(eds), Collective Remembering. London: Sage Publications, 46–​59.
Raymond, E. (2013) (ed.). Vox Poetae: Manifestations Auctoriales dans l’Epopée Gréco-​Latine.
Lyon: CECOR.
Rebeggiani, S. (2013). ‘Reading the Republican Forum: Virgil’s Aeneid, the Dioscuri and the
Battle of Lake Regillus’, Classical Philology, 108/1: 53-​69.
Renfrew, C., Gosden, C., and DeMarrais, E. (2004). ‘Introduction: Art as Archaeology and
Archaeology as Art’, in C. Renfrew, C. Gosden, and E. DeMarrais (eds), Substance, Memory,
Display: Archaeology and Art. Cambridge: McDonald Institute Monographs, 1–​6.
Rowlands, M. (1993). ‘The Role of Memory in the Transmission of Culture’, World Archaeology,
25: 141–​151.
Schwartz, B. (1982). ‘The Social Context of Commemoration: A Study in Collective Memory’,
Social Forces, 61/​2: 374–​402.
Seider, A. (forthcoming). Memory in Vergil’s Aeneid: Creating the Past. Cambridge: Cambridge
University Press.
Sharples, N. (2010). Social Relations in Later Prehistory: Wessex in the First Millennium bc.
Oxford: Oxford University Press.
Smith, A. (2001). The Differential Use of Constructed Sacred Space in Southern Britain from the
late Iron Age to the 4th Century ad. BAR British Series 318. Oxford: Archaeopress.
Sutton, D. (2001). Remembrance of Repasts:  An Anthropology of Food and Memory.
Oxford: Berg.
Tatum, J. (1995). ‘Aunt Elvie’s Quilt on the Bed of Odysseus: The Role of Artifacts in Natural
Memory’, Helios, 22/​2: 167–​174.
696

696   Zena Kamash

Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151.
York: Council for British Archaeology.
Todorov, T. (1995). Les Abus de la Mémoire. Paris: Arlea.
Van Dyke, R., and Alcock, S. E. (2003) (eds). Archaeologies of Memory. Oxford: Blackwell.
Vejby, M. (2012). ‘Reinterpreting the Iron Age and Roman Reuse of Megalithic Tombs in
Atlantic Europe’. Unpublished Ph.D. thesis, University of Reading.
Watt, D. (1998). Religion in Late Roman Britain: Forces of Change. London: Routledge.
Wheeler, R. E.  M. (1928). ‘A Romano-​Celtic Temple near Harlow and a Note on the Type’,
Antiquaries Journal, 8: 300–​326.
Williams, H. M. R. (1998). ‘The Ancient Monument in Romano-​British Ritual Practices’, in C.
Forcey, J. Hawthorne, and R. Witcher (eds), TRAC 97: Proceedings of the Seventh Annual
Theoretical Roman Archaeology Conference. Oxford: Oxbow Books, 71–​86.
Williams, H. M.  R. (2004). ‘Ephemeral Monuments and Social Memory in Early Roman
Britain’, in B. Croxford, H. Eckardt, J. Meade, and J. Weekes (eds), TRAC 2003: Proceedings
of the Thirteenth Annual Theoretical Roman Archaeology Conference, Leicester 2003.
Oxford: Oxbow Books, 51–​61.
Williams, H. M. R., and Bradley, R. (1998) (eds). The Past in the Past: The Re-​use of Ancient
Monuments (World Archaeology 30(1)). London: Routledge.
 PA RT  I V

LANDSCAPE AND
E C ON OM Y
698
 Chapter 34

‘By Sm all T h i ng s
Reveal e d ’
Rural Settlement and Society

Martin Millett

Introduction

It is a paradox that, while the vast majority of the population of Roman Britain were
farmers, and certainly the bulk of its landscape was given over to some form of agricul-
ture, rural settlement has attracted comparatively little attention, at least until recently.
Perhaps even more surprising is that most past scholarship is heavily biased towards one
particular social group—​those who dwelt in villas. This is like looking at the modern
countryside through the lens of those who live in the Cotswolds and drive Range Rovers!
To gain a more balanced perspective, we are faced with a serious problem of a lack of
good evidence from the full range of ordinary rural sites. This imbalance has been to
some extent rectified in recent years by archaeological work that has been undertaken in
advance of building development (see Wilson, this volume), since the locations of such
projects have been determined by external factors and hence they provide something
that may approximate to a random sample of the landscape. The results of such work
have already contributed to one major review of the rural landscape (Taylor 2007) and
they are currently being analysed in considerable detail in a project run by the University
of Reading and Cotswold Archaeology, which has yet to produce its definitive results
(http://​www.cotswoldarchaeology.co.uk/​developer-​funded-​roman-​archaeology-​in-​
england/​).
Pending the completion of this work, the present chapter provides an overview of
some key issues together with a discussion of research from one small area that perhaps
provides some complementary insights. However, this should be read in an awareness
of the issue of regional variation, for it is now very clear that there were tremendous
differences in the character of rural settlement across Roman Britain, so much so that
700

700   Martin Millett

there is really no such thing as the Romano-​British countryside. It is also important to

understand that such geographical variation is not simply a matter of environmental
constraints but reveals fundamental differences in the social organization and lifestyles
of the populations.
At the outset we should acknowledge that Roman Britain was primarily a rural soci-
ety. Assessing the total size of the population is very difficult, but at a rough estimate
it amounted to perhaps 3.6 million people of whom about 3.4 per cent were associ-
ated with the military, 6.5 per cent were urban dwellers (using the broadest definition
of towns), with the remaining 90 per cent living in the countryside (Millett 1990: 185;
Mattingly 2006: 356 suggests a total population of about two million). These numbers
are based on the evidence from a series of rural surveys undertaken in the latter part
of the twentieth century and they probably need to be updated. Taylor, for instance,
recorded about 28,000 known rural settlements in England alone (Taylor 2007: 12,
table 3.2). Whatever the exact figure, the point of importance is that much of the coun-
tryside was teaming with people and very different from some past images of a largely
wooded place with clearings for the occasional settlement. The whole landscape had
been exploited in a range of different modes by a large population since c. 1000 bc,
and, given the reliance on manual labour for farming and other rural activities, any
traveller passing along a Roman road through Britain will have been aware of living and
working people almost everywhere. Furthermore, as Taylor (2103) has argued, these
people must be recognized as active agents in the construction of the cultures of Roman
Britain.

The Character of the Countryside

Rural areas were a product of a long and varied history, with the types of buildings
and the character of the settlement patterns and fields being the result of this develop-
ment and the inhabitants’ intricate relationship with the natural environment. Thus the
countryside was a patchwork onto which we have only comparatively small windows
of knowledge. The best are generally provided by landscape-​scale evidence, mostly the
product of aerial photography or large-​scale excavation (for instance, in gravel quar-
ries), and now occasional large-​scale geophysical surveys. All these methods are lim-
ited by a focus on the features—​such as settlements, fields, and trackways—​which are
a product of human exploitation. Although, as we shall see in a moment, these provide
key insights into the character of the countryside, it is important to appreciate that there
was more to the rural world than those things associated with farming and habitation.
On the one hand, there were focal structures that had other functions: forts and fron-
tiers were located in the countryside and there were also a significant number of temples
and sanctuaries that will have acted as key points for the gathering of people. On the
other, paths, roads, trackways, and navigable rivers created linkages along which people,
animals, and things constantly moved back and forth.
 ‘By Small Things Revealed’: Rural Settlement and Society    701

There were also substantial tracts that were not used for arable farming. Aside from
areas for pasture—​largely unenclosed—​there were also moorlands, marshlands, and
woodlands, all of which were exploited in different ways. There is, for instance, strong
evidence for the production of salt, the digging of peat, and the management of wood-
land through coppicing and pollarding. Wood and charcoal were, after all, key prod-
ucts, although they are very difficult to trace archaeologically. Equally, the ground itself
was mined and quarried for its resources. Metal-​ore extraction, although geologi-
cally circumscribed, was of key importance (see also Dungworth, this volume). More
widespread was the quarrying of stone for building and other purposes, and the dig-
ging for clay for the production of tiles and pots. These examples demonstrate two key
points. First, much of what we think of today as industry was rurally located, and, sec-
ond, the boundary between farming and other modes of exploiting the land was much
more blurred than we might now think. It is especially important to appreciate that,
in a pre-​industrial society, a large supply of labour was vital to many rural activities.
Furthermore, labour requirements were seasonally variable, with many more people
needed for harvest than, for instance, for ploughing. Hence, it is very likely that there
were complex systems that allowed these demands to be met. Aside from social struc-
tures, which may have involved the payment of some dues in labour, we can envisage
estates or other communities where activities like quarrying or potting or woodmanship
were seasonal and complementary with the demands of arable farming or stock-​raising.
This further highlights an aspect of Roman Britain that remains obscure—​the nature
of land tenure. The existence of occasional land-​sale documents (Turner 1956, RIB
2443.13; Tomlin 1996) and the odd textual reference (Theodosian Code XI. 7, 2) con-
firm that at least some property was privately owned, although the extent to which this
was the case remains unclear. Some have argued for the existence of forms of multi-
ple ownership or ‘Celtic’ systems of partible inheritance (Stevens 1966)—​although this
is somewhat undermined by the discovery of a will formulated using Roman law in
Trawsfynydd, Merionethshire (Tomlin and Hassall 2004: 347–​349), while many assume
the growth of large private estates in the hands of a later Roman aristocracy (Scott 2000:
106–​120). All or any of these may have applied; we simply do not know, and we certainly
cannot assume that there was any single application of supposed Roman legal norms.
Indeed, wide variations in the patterns of rural settlement witnessed in the archaeo-
logical evidence suggest that there was a range of practice. It is important to acknowl-
edge this level of ignorance, since many past discussions of villas, in particular, have
been founded on very simplistic notions of landownership that are almost certainly
anachronistic.
It is worth pausing for a moment to look at some examples of well-​explored landscapes
that illustrate the variation across the country. One of the most remarkable pieces of
archaeological work in recent years has been Dominic Powleslend’s geophysical survey
along the southern side of the Vale of Pickering in Yorkshire (http://​thelrc.wordpress.
com/​lrc-​digital-​atlas/​). This project involved mapping 1,250 hectares of a multi-​period
landscape using magnetometry (Figure 34.1). From the point of view of understand-
ing Roman rural settlement, the results are stunning. Although many had previously
702

Figure 34.1  Magnetometer survey of part of the Roman landscape in the Vale of Pickering,
North Yorkshire. Across the top of the image a trackway runs east–​west, flanked on either side by
settlement enclosures and small fields. Beyond these to the south are the fainter traces of a series
of fields bounded by ditches. Superimposed on these, in the northern zone, the back dots repre-
sent grubenhaüser (sunken-​featured buildings) of Early Medieval date. To the south the land rises
towards the Yorkshire Wolds and the ancient landscape is obscured by deposits of wind-​blown
sand.
Source: courtesy of Dominic Powlesland, Landscape Research Centre.
 ‘By Small Things Revealed’: Rural Settlement and Society    703

concluded that the landscape was densely occupied and wholly exploited during this
period, there had previously been no such large-​scale survey to allow a full apprecia-
tion of this. A snapshot of part of the landscape is shown in Figure 34.1, where we see a
trackway with adjacent enclosures running from east to west across the image—​this is
part of a system that has been traced for many kilometres, and many of the enclosures
were used for human habitation, as is illustrated by the density of features revealed. To
the north lie the wetlands at the centre of the Vale, which were seasonally flooded yet
provided important pasture. To the south the land rises towards the Yorkshire Wolds,
and is partly obscured (and protected) by a blanket of wind-​blown sand. Visible are field
enclosures and other features, as well as spots representing Anglo-​Saxon grubenhaüser.
The remarkable thing is not the density of landscape use shown, but the fact that it is
continuous across the whole area surveyed. This shows how we should now understand
the whole of Roman Britain.
A contrasting example, based on aerial photographic evidence, is provided by the
regional study of the Magnesian Limestone ridge in South and West Yorkshire (Roberts
2010). Here there is also evidence for the use of the whole landscape, but with an exten-
sive system of small fields and enclosures that originated in the later Iron Age, domi-
nating the landscape and indicating a system of dispersed settlements. A few of these
developed into villas in the third and fourth centuries, but for the most part this land-
scape too continued to be dominated by indigenous-​style farmsteads.

The Significance of the Villa

It is in the context of this patchwork of intensive and varied rural activities that we should
understand villas. The term is derived from Latin but has gained an archaeological mean-
ing of its own, so almost any Roman-​style rural house is labelled with the term—​although
very few indeed of those found in Roman Britain would have been seen as villas by a visi-
tor from Italy. In archaeological terms, their physical remains are fairly clearly defined,
and many of the houses share common architectural characteristics ultimately derived
from continental (although rarely Mediterranean) models. In broad terms, there are per-
haps 2,000 such sites, representing perhaps 1 per cent of known rural sites, largely distrib-
uted over the southern and eastern counties of England, to one side of a line between the
mouths of the rivers Tees and Exe. In this sense they clearly represent both an aspiration
towards a particular set of cultural values—​arguably associated with Roman power—​
and also a decision by their owners to expend significant capital resources on this sort of
permanent building. We can rarely tell whether their owners were indigenous people, or
incomers, or those returning after a period elsewhere: what is clear is that the vast major-
ity show continuity of site from previously occupied settlements. It is equally clear that
the buildings themselves provide little or no information about how the wealth deployed
to construct them was created, only how it was spent. The resources may have been gen-
erated from economic activity (for example, farming, mining, trade) whether locally
704

704   Martin Millett

or elsewhere, or they may have been inherited, or extorted: we simply cannot tell. Thus
the common jump from villa house to the assumed associated villa estate, and thence
to a particular reconstruction of the Romano-​British economy (usually in terms of the
exploitation of marketing opportunities in towns generated by the Pax Romana), is self-​
evidently unjustifiable. It is also difficult to reconcile this common explanation with evi-
dence that shows that, in Britain, villas start to be built early in the Roman period (often
at a large scale) but spread relatively slowly, becoming on average smaller through time.
This suggests the spread of a fashion for a particular type of elite display, not the develop-
ment of a particular economic mode (Taylor 2011). Equally, the distribution of villas is
uneven, with some regional concentrations (for instance, in the Cotswolds) and a general
spatial association with towns that we understand to be important local (not provincial)
political centres. As such they may perhaps be associated with the houses of magnates
who fulfilled roles in civic administration. This is not to deny that most villas had farms
and estates attached to them: it is merely to assert that this association is not necessarily
causative. Members of the council of a civitas will often have owned farms, for this was
the Roman mode of demonstrating wealth: that does not mean that they derived their
principal wealth from farming.
This suggests that we should be looking at villas, not as a crude economic indicator,
but rather for the insights they provide about the social organization of Roman Britain.
Attempts to approach this on the basis of a study of villa plans has generated some
insights (Smith 1997; cf. Rippengal 1993), in particular through encouraging us to look
at the building plans critically, rather than on the basis of loose analogies with aristo-
cratic country houses of more modern periods. In particular, interesting studies have
suggested how some major villas of the later Roman period may have worked within
the framework of elite Mediterranean classical culture, with an especial focus on dining
rituals and a familiarity with classical mythology and literature (Scott 2000; Gerrard
2013: 141–​142). Other studies (Smith 1997), which have emphasized possible patterns of
multiple occupancy at certain villas, have met with less acclaim, although they do high-
light the difficulty in reading plans in terms of a simple nucleated family.
Two contrasting examples illustrate some of the key issues. At one extreme, we have
Fishbourne, near Chichester in West Sussex—​often misleadingly referred to as a ‘palace’
(Figure 34.2). Although there is some debate about the precise date of its construction
(Cunliffe 1971: 77; Black 1987: 84–​86), it is clear that an enormous villa was constructed
over a brief period in the period around ad 70. It was within a complex best interpreted
as a late Iron Age oppidum, which seems to have been used for an invasion period mili-
tary base. The exact relationship between an indigenous ruler, the Roman invaders, and
the establishment of civitas government here is contentious, but it is very clear that the
villa represents a symbol of acceptance of Roman ways at a very early date. Unlike any
other known villa in Britain, its opulent architecture displays close links with the con-
temporary Mediterranean, suggesting both the expenditure of immense wealth within
a very short time and close personal ties between its owner and the imperial household.
The principal excavated courtyard is only part of a much larger complex, which includes
ancillary structures that are very difficult to understand, but indicate a large-​scale
 Str
eam
Ditch

Aqueduct

Road/path

Masonry building

Road/path

0 10 50m

Figure 34.2  Plan of the Roman villa at Fishbourne, Sussex, in the later first century ad showing the approach from the east and associated structures.
Source: drawn by Lacey Wallace, after Cunliffe (1971); Cunliffe et al. (1996); Manley and Rudkin (2003).
706

706   Martin Millett

monumentalization of the landscape (Cunliffe et al. 1996; Manley and Rudkin 2003).
Interestingly, although the architectural detail and decoration signify Italian influence,
the plan is not easy to parallel in Italy, and includes features such as an audience hall that
surely betray use by a broader community and not just a pro-​Roman potentate.
By contrast, the villa at Frocester is much more typical of other villas in Britain,
except that it has been subject to a programme of very thorough long-​term research,
which has included a close study of the surrounding landscape (Gracie 1970; Gracie
and Price 1979; Price 2000–​10) (Figure 34.3). The site, which lies on the edge of the
Severn valley in Gloucestershire just below the Cotswold escarpment, is one of a large
group in this part of England. It shows a more gradual evolution from an enclosed Iron
Age farm, which was remodelled in the second century, but only became a recognizable
villa towards the end of the third century ad. The initial villa house comprised a simple
range, with a large reception room near its centre. This was added to in the fourth cen-
tury with the creation of a façade comprising portico and pavilions, and the addition
of a second range of rooms along the back of the house, as well as perhaps a hall at one
end. There is very good evidence to suggest that the house was of more than one storey
and thus built to create a strong impression to the visitor. This gentrification is also
seen within, as several rooms were decorated with mosaic pavements (Neal and Cosh
2010: 136–​139, nos 430.1–​4). The frontal aspect of the architecture was further enhanced
around ad 340 with the laying-​out of a formal garden. Thus we see a relatively ordinary
farm monumentalized and made into a show piece within a few generations late in the
Roman period. The building plan, although showing Roman influence, was charac-
teristic of Britain and the north-​western provinces, and the form again suggests that
receiving, entertaining, and impressing visitors were crucial to its owners.

The Diversity of Farmsteads

In contrast to villas, ordinary farmstead sites are much more difficult to evaluate. This is
partly simply because they were mostly built in timber with the result that the houses,
especially the less monumental types, are much less legible as structures. It is clear that,
while pre-​Roman Iron Age housing was dominated by the round-​house, a broader
range of mostly rectilinear structures occurred in the Roman period (e.g. Roberts 2010:
73–​76). There were pockets of continued round-​house use, mostly in association with
other structural forms (Taylor 2007: 31, table 4.4). Equally, there is much variety in other
types of architecture, with few really distinctive and widespread types, like the aisled
hall or the stone-​built hall-​house. Furthermore, farms vary from small places with a few
buildings to more extensive settlements with multiple structures. A common theme is
the use of enclosure to define space within settlements, suggesting that the definition of
space was important to the inhabitants.
The most sustained regional study of rural settlement in Roman Britain is probably
the Danebury Environs project (Cunliffe 2008). Here, the main focus was on examining
 Late Iron Age c. ad 50–100

c. ad 100–150 c. ad 150–200

c. ad 200–275 c. ad 275–360
N

Building
Gravel pit
Ditch
Ditch postulated or disused
Road
0 100m
c. ad 360–400 0 300ft

Figure 34.3  Plans showing the development of the Roman villa and associated structures at
Frocester Court, Gloucestershire.
Source: drawn by Lacey Wallace, based on Price (2000–​10).
708

708   Martin Millett

a series of villa sites, but the work included the excavation of one ordinary rural site at
Rowbury Farm, which was located within a field system of Bronze Age origin; the set-
tlement itself did not have any intelligible structures surviving from the Roman period,
and it appears that it was used for livestock management rather than occupation, going
out of use early in the Roman period. The limited evidence from the Roman period led
to the suggestion that the site was either a subsidiary of a nearby villa or a peasant settle-
ment (Cunliffe and Poole 2008: 132–​133). What the whole project provided was an inten-
sive review of the development of rural society in one particular area, revealing a range
of different histories and structural types, even within the category of the smaller villas,
but with apparent evidence for a broad continuity in agrarian traditions from the Iron
Age with little evident Roman impact.

The Issue of Nucleated Sites

While accounts have traditionally given emphasis to single farms or smaller clusters
of building, recent work has shed more light on the existence of larger nucleated set-
tlements. Taylor’s Atlas has demonstrated that this is partly a regional matter, with a
pronounced preference for settlement nucleation is some areas like the East Midlands
(Taylor 2007: 76, figure 34.6.5). This is reminiscent of the patterns of later periods, when
there is a marked contrast between landscapes dominated by villages, on the one hand,
and dispersed farmsteads, on the other, associated with different social formations.
While this distinction may hold some value for Roman Britain, it would be a mistake
to impose such a simple division or to transpose models, derived from the study of the
medieval peasant, onto the interpretation of Roman Britain (cf. McCarthy 2013). The
social structure of those living in Roman Britain was the product of particular historical
circumstances that were very different from those of later periods, so we should be cau-
tious even of referring to nucleated sites as villages.
The recently excavated site at Higham Ferrers, beside the river Nene in
Northamptonshire, provides an interesting example of an east Midlands nucleated site
(Lawrence and Smith 2009). The Roman settlement, which overlies an Iron Age farm, was
laid out in the second century but reached its full size in the later third and fourth centuries
(Figure 34.4). It comprises a series of rectangular stone-​founded houses, set within enclo-
sures with paddocks and horticultural plots, which lined a minor road that followed the
river valley. There was a small shrine at one side of the excavated area, although it is seems
most likely that this served the community rather than being a more significant rural sanc-
tuary. The buildings (Figure 34.5) are otherwise relatively undifferentiated, and the care-
fully analysed evidence from the excavations suggests that this was a primarily agrarian
settlement. There is no suggestion of its dependence on any villa or other major house
that might have been occupied by the landowner or their agent. This perhaps implies a
settlement of the ‘middling sort’, although it is interesting to note the find of slave shackles,
which might indicate that we are dealing with a different type of estate centre.
 ‘By Small Things Revealed’: Rural Settlement and Society    709

Figure 34.4  Plan of the nucleated rural settlement at Higham Ferrers, Northamptonshire, in

the late third–​fourth centuries ad. A series of buildings set within enclosures flank the Roman
road, which runs along the valley of the river Nene just above the flood plain.
Source: illustration from Lawrence and Smith (2009), reproduced courtesy of Oxford Archaeology.

There are many general similarities with a larger example, at Kingscote in

Gloucestershire (Timby 1998), which is also dominated by simple rectangular houses,
this time organized in a couple of broadly rectilinear groupings. This site is larger, with
as many as seventy-​five houses indicated by survey work. In contrast to Higham Ferrers,
there is some evidence for a single larger house, which was located to one side of the main
group of buildings and seems to date to the fourth century. This had a main hall, not unlike
710

710   Martin Millett

Figure 34.5  Photograph of a domestic building (Figure 34.4, No. 10810) within the nucleated
rural settlement at Higham Ferrers, Northamptonshire. The Roman road and the river Nene are
visible in the background.
Source: illustration from Lawrence and Smith (2009), reproduced courtesy of Oxford Archaeology.

those found in the other houses, but to one side there were a series of private apartments,
including one at the focal point of the building, which is perhaps a reception chamber.
This had a mosaic floor with central medallion showing Venus (Neal and Cosh 2010: 168–​
171, no. 438.1). Although differentiated from the other houses and with an apparent func-
tion for gatherings, this building does not look like the main residence of a major landowner,
although it could have been the base for his or her agent. A possible villa close by might per-
haps represent a resident landowner. As with so many such sites, this leaves questions about
its social and economic role completely open. The original suggestion that it might have been
the centre for an imperial estate seems most improbable, especially now that such nucleated
centres are evidently so very widespread in Roman Britain.

Towards a More Dynamic

Understanding of the Countryside

Traditional approaches to the countryside have very often been based on the study of
particular sites, generally starting from a study of their structures, but now generally
also involving the integration of evidence from artefacts, seeds, and animal bones (see
 ‘By Small Things Revealed’: Rural Settlement and Society    711

also Maltby, this volume, and Van der Veen, this volume). The best such studies provide
excellent insights into these individual settlements, but they too often fail to place them
into a broader provincial context. The real challenge for us today is to obtain a fuller
understanding of the human societies who lived in Roman Britain. This requires the
peopling of the whole landscape rather than just looking at particular structures, and
seeing these people’s lives as dynamic, in terms of both day-​to-​day existence and devel-
opment through time. Such an understanding can only ever derive from a combination
of the analysis of detailed evidence from excavations with a broader reading of features
within the landscape. An example of this type of approach is provided by our work in
the lowlands on the edge of the Vale of York, in northern England (Halkon and Millett
1999; Millett 2006; Halkon et al. 2014). Here long-​term research has combined land-
scape study with selective excavation on a series of sites with the aim of exploring more
fully the transition from Iron Age to Roman society.
The study area is itself geographically varied, with the chalk uplands of the Yorkshire
Wolds to the north-​east, its steep scarp slope dominating the lower ground to the south
and west, cut into by a series of small valleys of streams that flow into the Humber. The
foot of the Wolds is skirted by a bench of good arable land, which merges into a zone of
wetland moors beside the Humber estuary. During the Roman period these lowlands
were interspersed by the fingers of a now-​silted tidal inlet, the margins of which were
dominated by sand ridges, a relic of the area’s glacial history. This topography has gen-
erated a highly varied environment that was differentially exploited by its population.
Two key themes emerge from its study; first, the strength of the connections between
the exploitation of uplands and lowlands, which were not separate, but formed an inte-
gral part of the same landscape, and, second, the varied use of the landscape for differ-
ent types of agriculture and industrial production. Integration into the Roman world
affected aspects of the lives of the inhabitants in complex ways.
Iron Age settlement in the area had been dominated by a pattern of dispersed rural
settlement, with most farms apparently occupied by single nuclear families. At Hayton,
on the arable lands at the Wold foot, enclosed farmsteads with round-​houses were
mostly focused on the stream valley, along which ran a major droveway that connected
to the Wolds. The droveway was defined on either side by substantial earthen dykes
against which many of the farmsteads were built. These had small fields or paddocks
attached, on the side away from the dyke, and the excavated evidence shows the adop-
tion of mixed arable farming with the exploitation of the usual range of domestic ani-
mals. Beyond the stream valley there were areas of larger fields to the south-​east and
apparently a zone of open pasture to the north-​west. The droveway that followed the
valley up onto the Wolds seems to have been used for the seasonal movement of flocks
of sheep, which were kept in the valley over winter and for lambing in spring, but were
moved to the uplands for summer pasture. The intensive exploitation of these flocks is
witnessed by the widespread consumption of sheep-​meat in feasting, as shown by bone
deposits within the excavated settlement sites.
In the wetlands towards the Humber, there were similar farm sites, some of which
were aggregated into slightly larger hamlets, but significantly these are generally without
712

712   Martin Millett

evidence for associated field systems. There was large-​scale smelting of iron ore, which
was found in the boggy ground locally. This industrial production required substantial
supplies of wood as fuel, and it would appear that much of the wetland landscape was
given over to managed woodland to provide this. The woods also provided grazing ani-
mals, probably cattle and pigs, creating a range of resources that complemented those of
the mainly arable lands and could be exchanged with these nearby communities.
Roman intervention in the landscape is marked by the arrival of the military some-
time after ad 70, seen locally with the construction of an auxiliary fort at Hayton. The
military brought new influences with them, as well as effecting a wholesale restructur-
ing of power relationships. The latter can, to a limited extent, be seen in their preferential
consumption of the prime lamb, produced by the local farms. New power relations are
also suggested by the way that an Iron Age-​style settlement with an exceptionally large
round-​house was soon constructed close up against the front of the fort, suggesting that
a local was cosying-​up to the newly dominant authority. Elsewhere, the impact of Rome
is less clearly visible, with only very gradual changes in lifestyles on local farms. We see
the adoption of some aspects of a new material culture. The previously modest material
culture repertoire, with just a limited range of ceramic cooking pots and beehive querns
used within timber buildings, was gradually augmented, starting during the phase of
military contact. Dress accessories, especially brooches, began to be used. There were
also indications of changes in cooking and eating with the introduction of new types of
quern, which imply a new way of making the daily bread, and a broadening of the vari-
ety of pot shapes and changes in the patterns of use-​deposits on jars, which imply altera-
tions in ways of cooking.
After the period of direct military intervention, which ended towards the end of
the first century ad, more major changes were brought about by integration into the
broader Roman world, which is represented by the construction through the area of a
strategically important Roman road which linked Lincoln and York. The route crossed
the Humber via a ferry at Brough, and followed the foot of the Wolds through our area,
before continuing towards the frontier by way of the legionary base at York. This new
road, which in part seems to have followed an existing network of trackways, completely
changed the world of the local inhabitants, bringing a vast array of connections to an
outside world.
This road rapidly became the focus for new ways of life for some who migrated to live
beside it, forming new roadside settlements at Hayton, and, within a generation or so,
also at Shiptonthorpe nearby. Excavations at Shiptonthorpe give a very clear impres-
sion of the occupants’ transition. The earliest phases of the excavated household closely
resemble an Iron Age-​style enclosed farmstead, with a variant on the round-​house built
right next to the new road within a square enclosure: in other words, a traditional indig-
enous settlement relocated into an imperial context. This also applies to the modest
character of the occupants’ uses of material culture, which remained in an Iron Age-​
style for at least a generation or two. It seems likely that most of the people moving to
live by the road at both Shiptonthorpe and Hayton were coming from nearby, and there
is a hint in the survey evidence of the abandonment of certain farmsteads, which may
 ‘By Small Things Revealed’: Rural Settlement and Society    713

reflect this move. However, there are also indications in the ceramics that some might
have moved from further afield, from what is now Lincolnshire, perhaps taking advan-
tage of economic opportunities created by the new road system.
There is a suggestion of a slightly more rapid change in lifestyles at Hayton, in sur-
vey evidence both from the roadside settlement and on the excavated farmstead. Here
the round-​houses and enclosures were transformed during the second century with the
construction of two new-​style rectangular timber buildings, a small courtyard house,
and an aisled hall within a new enclosure. This transformation, which respected the
topography of the earlier houses, was accompanied by a broader change in the use of
material culture. This included a diversification both in the types of object represented,
to include a range of tools, architectural fittings and furniture, as well as in the range of
pots used. The latter indicates a wider change in domestic habits, of storage and cook-
ing as well as eating, which shows that the inhabitants had by now adopted a lifestyle
typical of many in Roman Britain. The one category of material that remains conspicu-
ously absent is that associated with personal grooming. A comparable overall pattern is
repeated at Shiptonthorpe, but a little later in the middle of the third century, illustrat-
ing how different families adopted new life-​ways on their own terms and for their own
means. As at Hayton, this change in material use was accompanied by a new domestic
context, here with the construction of a huge aisled hall, the size of which is surely a
strong statement of the social status of its owners (Figure 34.6).

Figure  34.6  Cut-​away reconstruction drawing of the third-​century aisled hall excavated at
Shiptonthorpe, East Yorkshire.
Source: drawing by Mark Faulkner, from Millett (2006).
714

714   Martin Millett

This highlights one of the more striking features revealed by work in this landscape,
for, while there is very strong evidence for the integration into the new life-​ways of the
peoples who lived close to the Roman road, the same is not true for some of those liv-
ing further from it. While sites along the road have a profusion of ceramics, coinage,
and other metal artefacts (as revealed in particular by finds reported to the Portable
Antiquities Scheme), most who lived away from the road seem not to have used such
artefacts. Although there are exceptions, the overwhelming pattern is of a continued
modest use of material culture. This is particularly surprising in the settlements in the
wetlands in the southern part of our study area, as here the Iron Age pattern of large-​
scale iron production seems to have been superseded by pottery manufacture as iron
ore resources became depleted. There is thus only the slightest of evidence for Roman-​
period iron-​working, but, by the third century, a number of sites were producing house-
hold pottery in kilns that lay on the margins of their settlements. The wide distribution
of this pottery illustrates how successful these people were in negotiating new economic
opportunities as it spread widely across what is now East Yorkshire and continued to
be produced until the later fourth century. It is thus unclear why these producers did
not also take up the use of the new material culture. It may be the result of rejection
and the maintenance of other value systems, or it might be because the producers were
in some way subservient, so that the profits from their pot sales did not come into the
production zone but instead stayed elsewhere. There are also hints of a more nuanced
pattern of rejection of Romano-​British material culture elsewhere in the area, with some
sites quite close to the Hayton road apparently using imported pottery, but not receiving
coins or other metal artefacts in any quantity. This reinforces the idea that the people of
the region were selective in their adoption of material culture and there were significant
variations in their attitudes to different elements of what is too often seen as a single cul-
tural package.
On the excavated farm at Hayton personal grooming seems eventually to have come
into fashion in the later part of the third century when the enclosure system was remod-
elled to allow a small bathhouse to be added at one corner of the courtyard building.
This was a tiny structure, with rooms only 2 metres × 2 metres, but includes the full
basic suite for Roman-​style bathing and used the best of materials, including window
glass and painted wall-​plaster. The adoption of both a new architecture and the habit of
bathing suggests a very significant change in the daily lives of at least some of the inhab-
itants. Finally, it was during the middle of the fourth century that the well-​established
timber buildings were replaced by a rather rambling range of stone structures, which
brought the new architecture fully into the domestic domain. Too little survives of these
buildings to understand details of their form, but the scraps of wall-​plaster and decora-
tion indicate that some of the elements traditionally associated with villas were being
adopted here. Most surprisingly, a fragment of wood recovered from a well backfilled
when the site went out of occupation shows that the building’s furnishing included a
cupboard inlaid with bone of a type known from the Mediterranean.
So there is evidence for the ways in which the people in this area become more
entwined with material and activities associated with the broader empire. But they also
 ‘By Small Things Revealed’: Rural Settlement and Society    715

continued with a series of their own traditional practices, which were evidently key to
their cultural identity. Most obvious of these are certain patterns of burial practice that
can be shown to have originated in the Iron Age. In particular, the inhabitants had par-
ticular rituals associated with the death of infants at around the time of birth—​a com-
mon occurrence in pre-​industrial societies. Those infants who died at this time were
given careful burial within a domestic setting (a context from which older infants,
children, and adults seem to have been systematically excluded). Some of them were
buried around the walls of the houses—​and at Shiptonthorpe this seems to have been
particularly associated with the eastern end of the aisled hall. The others were interred
at boundaries in the general area of the house. There is also an apparent association
between these neonatal burials and a series of feasting deposits, which were found
across the site, possibly representing funerary ceremonies.
The animals represented in the feasting deposits found at the farm at Hayton were
mostly sheep, which is rather unusual in comparison with the evidence for other sites,
including Shiptonthorpe. This brings us back to a consideration of how the people liv-
ing here interacted with society more broadly. In terms of the farming economy, there is
very little evidence for substantial change from the Iron Age into the Roman period, so
the role of sheep in feasting may be a function of the close connection between the com-
munity in this valley and their flocks, which were regularly moving seasonally between
lowland streamside pasture and the uplands (Figure 34.7). Other aspects of movement
between the different local environments are illustrated by the exploitation of heather
healthlands—​presumably nearer the Humber—​perhaps for roofing or for animal bed-
ding, as reflected in the environmental samples from excavated sites. We may also spec-
ulate that this link also brought pots manufactured in that area together with woodland
products like charcoal to Hayton. Such seasonally influenced movements need to be
set in the context of the connectivity created by the Roman road, for both Hayton and
Shiptonthorpe were situated at junctions between this and ancient routes from the low-
lands up to Wolds. Given that there was demand for animals to supply meat to both
the military and those living in the towns, we can envisage that stock would have been
gathered at these points before being driven along the road to centres of consumption.
Moving animals via the roads would have required frequent watering places as well as
regular stops for feeding. Clear evidence of this is provided by a waterhole right beside
the road in the excavated area at Shiptonthorpe, and we may envisage such features
placed regularly along the route. Less direct evidence is provided by the environmental
evidence for the production of hay noted at the Hayton farmstead, which derived from
the exploitation of the meadows beside the stream for growing animal fodder, perhaps
to provide feed for animals passing along the road. It certainly implies a more intensive
stock regime.
Linked with the idea of the roadside settlements becoming funnels for supplies
that were henceforth moved along the road, we have two different sets of evidence. At
Shiptonthorpe there is an unusual focus of corn-​grinding querns, which implies that
flour was being milled on a scale beyond the usual domestic level. As this area is not
one with much evidence for large-​scale intensive cereal farming, we might tentatively
716

716   Martin Millett

Figure 34.7  Aerial photograph of part of the Roman landscape at Burnby Lane, Hayton, East
Yorkshire, looking north-​east. The valley of the Burnby Beck runs down the centre of the pho-
tograph, with its relict course showing as a light crop-​mark against the darker green of the flood
plain. On either side of this damp ground can be seen a series of settlement enclosures. On the left
a length of boundary ditch is also visible. This runs parallel with the stream, and formed one side
of a droveway that linked the lowlands to the south-​west with the Wolds to the north-​east, allow-
ing animals to be moved for seasonal grazing.
Source: photograph by Peter Halkon, courtesy of the Hayton Project.

suggest that it acted as a gathering point between smaller-​scale producers and con-
sumers without access to their own grain supplies (perhaps from the woodland to the
south), so that the small-​scale consumer could trade and take away milled flour. At
Hayton, beside the Roman road in an area of probable pastureland that is devoid of set-
tlement features, there is a substantial concentration of metal-​detected finds, which has
been interpreted as evidence for the site of a season market or fair, where stock brought
down the valley might have been sold. The composition of the assemblage of metal finds
includes many objects associated with horse harnesses and carts, as well as a number
associated with the military, among which is a lead seal of Legio VI Victrix, which was
based at York from the time of Hadrian. These finds indicate that there was probably a
lot more to the fair than just the selling of sheep, and we may well envisage a more sub-
stantial and varied event, which perhaps involved military buyers.
Such a context will have widened the network of contacts of those living in the valley
and provided them with the opportunities to be active participants in a broader cultural
 ‘By Small Things Revealed’: Rural Settlement and Society    717

milieu. As such it provides an unusual insight into the sorts of social contact that must
have been at the centre of the lives of most rural inhabitants of Roman Britain.

Conclusions

It is clear that the study of rural settlement in Roman Britain is at a key turning point.
The proliferation of new evidence provided both by developer-​funded work and long-​
term field research has provided a catalyst for new thinking about the subject which has
at long last moved the focus away from villas. It is equally important that emphasis is
now being placed not on individual sites but on the study of the overall landscape, not
simply as a background canvas, but as a highly variegated framework within which dif-
ferent people lived their own lives. Finally, as Jeremy Taylor (2013) has shown, the rural
population must be recognized as the active creators of their own culture, not simply as
passive recipients of that imported from elsewhere in the empire or, more locally, from
the towns of the province.

Abbreviation
RIB  R
 . G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume I:
Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Primary Sources
Theodosian Code. The Theodosian Code and Novels and the Sirmondian Constitutions.
A Translation with Commentary, ed. C. Pharr. Princeton: Princeton University Press, 1952.

Secondary Sources
Black, E. W. (1987). The Roman Villas of South-​ East England. BAR British Series 171.
Oxford: Archaeopress.
Cunliffe, B. W. (1971). Excavations at Fishbourne. Leeds: Research Reports of the Society of
Antiquaries of London.
Cunliffe, B. W. (2008). The Danebury Environs Roman Programme: A Wessex Landscape during
the Roman Era. Volume 1: Overview. Oxford University School of Archaeology Monograph
70. Oxford: Oxford University School of Archaeology.
Cunliffe, B., Down, A., and Rudkin, D. (1996). Chichester Excavations IX: Excavations at
Fishbourne 1969-​88. Chichester: Chichester District Council.
Cunliffe, B. W., and Poole, C. (2008). The Danebury Environs Roman Programme: A Wessex
Landscape during the Roman Era. Volume 2 Part 5: Rowbury Farm, Wherwell, Hants. 2003.
718

718   Martin Millett

Oxford University School of Archaeology Monograph 71. Oxford: Oxford University School

of Archaeology.
Cotswold Archaeology 2014 ‘Realising the research potential of developer-​funded Roman
Archaeology in England’ http://​www.cotswoldarchaeology.co.uk/​developer-​funded-​
roman-​archaeology-​in-​england/​. Accessed 27/​7/​14.
Gerrard, J. (2013). The Ruin of Roman Britain:  An Archaeological Perspective.
Cambridge: Cambridge University Press.
Gracie, H. S. (1970). ‘Frocester Court Roman Villa, Gloucestershire. First Report’, Transactions
of the Bristol and Gloucestershire Archaeological Society, 89: 15–​86.
Gracie, H. S., and Price, E. (1979). ‘Frocester Court Roman Villa, Gloucestershire. Second
Report’, Transactions of the Bristol and Gloucestershire Archaeological Society, 97: 9–​64.
Halkon, P., and Millett, M. (1999) (eds). Rural Settlement and Industry:  Studies in the Iron
Age and Roman Archaeology of Lowland East Yorkshire. Yorkshire Archaeological Report
4. Leeds: Yorkshire Archaeology.
Halkon, P., Millett, M., and Woodhouse, H. (2014) (eds). Hayton, East Yorkshire: Archaeological
Studies of the Iron Age and Roman Landscapes. Leeds: Yorkshire Archaeology.
Landscape Research Centre: Digital Atlas. http://​thelrc.wordpress.com/​lrc-​digital-​atlas/
(accessed 27 July 2014).
Lawrence, S., and Smith, A. (2009). Between Villa and Town: Excavations of a Roman Roadside
Settlement and Shrine at Higham Ferrers, Northants. Oxford Archaeology Monograph
7. Oxford: Oxford Archaeology.
McCarthy, M. (2013). The Roman Peasant: Towards a Study of People, Landscape and Work dur-
ing the Roman Occupation of Britain. Macclesfield: Windgather Press.
Manley, J., and Rudkin, D. (2003). Facing the Palace: Excavations in Front of the Roman Palace
at Fishbourne, Sussex, UK 1995–​99. Lewes: Sussex Archaeological Society.
Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Allen Lane.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Millett, M. (2006) (ed.). Shiptonthorpe, East Yorkshire: Archaeological Studies of a Romano-​
British Roadside Settlement. Yorkshire Archaeological Report 5.  Leeds:  Yorkshire
Archaeology.
Neal, D. S., and Cosh, S. R. (2010). Roman Mosaics of Britain. Volume 4: Western Britain.
London: Society of Antiquaries.
Price, E. (2000–​10). Frocester: A Romano-​British Settlement, its Antecedents and Successors.
Volumes 1–​4. Stonehouse: Gloucestershire and District Archaeological Research Group.
Rippengal, R. (1993). ‘Villas as a Key to Social Structure? Some Comments on Recent
Approaches to the Romano-​British Villa and Some Suggestions towards an Alternative’,
in E. Scott (ed.), Theoretical Roman Archaeology:  First Conference Proceedings.
Aldershot: Avebury, 79–​102.
Roberts, I. (2010). Understanding the Cropmark Landscapes of the Magnesian Limestone.
Leeds: West Yorkshire Archaeological Services.
Scott, S. (2000). Art and Society in Fourth-​Century Britain: Villas in Context. Oxford University
School of Archaeology Monograph 53. Oxford: Oxford University.
Smith, J. T. (1997). Roman Villas: A Study in Social Structure. London: Routledge.
Stevens, C. E. (1966). ‘Social and Economic Aspects of Rural Settlement’, in A. C. Thomas (ed.),
Rural Settlement in Roman Britain. London: Council for British Archaeology, 108–​128.
 ‘By Small Things Revealed’: Rural Settlement and Society    719

Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151.
York: Council for British Archaeology.
Taylor, J. (2011). ‘The Idea of the Villa:  Reassessing Villa Development in South-​ East
Britain’, in N. Roymans and T. Derks (eds), Villa Landscapes in the Roman North.
Amsterdam: Amsterdam University Press, 179–​194.
Taylor, J. (2013). ‘Encountering Romanitas:  Characterising the Role of Agricultural
Communities in Roman Britain’, Britannia, 44: 171–​190.
Timby, J. R. (1998). Excavations at Kingscote and Wycomb, Gloucestershire.
Cirencester: Cotswold Archaeological Trust.
Tomlin, R. S. O. (1996). ‘A Five Acre Wood in Roman Kent’, in J. Bird, M. Hassall, and H. Sheldon
(eds), Interpreting Roman London: Papers in Memory of Hugh Chapman. Oxford: Oxbow
Books, 209–​216.
Tomlin, R. S.  O., and Hassall, M. W.  C. (2004). ‘Roman Britain in 2003 III. Inscriptions’,
Britannia, 35: 335–​349.
Turner, E. G. (1956). ‘A Writing Tablet from Somerset’, Journal of Roman Studies, 46: 115–​118.
720

Chapter 35

Rur al Transf ormat i on i n

the Urbanized L a nd s c a pe

Martin Pitts

Introduction: Questions and Models

Considerable attention has been devoted to questions of rural change in Roman

Britain. To date, this has tended to crystallize around two interrelated issues: the eco-
nomic relationship between town and country, and the extent to which ‘Roman culture’
was adopted in rural areas. Particularly influential on both counts is the following pas-
sage by R. G. Collingwood: 

Economically, the towns were parasitic on the country-​side. They had to be fed by it,
and the goods they produced, together with the services they rendered as markets
and trading-​centres, were no adequate return for the food they consumed and the
expenditure which they demanded for the upkeep of their public services. They had
their industries; but these consisted only to a small extent in the production of goods
needed in the country; most of them were luxury-​trades whose produce was mostly
used in the towns themselves. They did a large business in retail trade, selling pottery
made in Gaul and other imports, but here again, the total quantity of these goods
which found its way into the country districts was the barest fraction of what the
towns consumed. From a strictly economic point of view the towns were a luxury.
Their function was cultural and political. They stood for the decencies and elegances
of civilized life, and they provided a link between the Roman government and the
mass of the people, to whom those decencies and elegances were things out of reach.
Their populations, rich and poor alike, thus formed a privileged section of the peo-
ple, privileged to enjoy the blessings of romanization at the expense of the country-​
folk. (Collingwood and Myres 1937: 198–​199)

Seventy-​five years on, Collingwood’s model has endured remarkably well. One
reason for this is the lack of data available to previous generations properly to test
 Rural Transformation in the Urbanized Landscape    721

Collingwood’s statement (Millett 1982). Nonetheless, as discussed below, the most

recent data fit Collingwood’s model rather well. While there have been critics over the
years, notably Wacher’s restatement of the positive contribution of Romano-​British
towns in the development of rural settlement and agricultural improvements (Wacher
1995: 70), the core idea of cities surviving as rural parasites continues to strike a chord
with the extant evidence. The basic logic of the parasitical model resonates with attempts
to explain Romano-​British rural change in response to integration into imperial sys-
tems and the transformative pressures of taxation (Hingley 1982; Millett 1990a), as well
as post-​colonial emphasis on the exploitative dimensions of Roman imperialism, culmi-
nating in sharp divisions in culture and identity experienced between urban and rural
communities (Mattingly 2006).
Despite an emerging degree of consensus on Collingwood’s vision, the general issue
of the character and extent of rural transformation in the urbanized landscapes of cen-
tral, southern, and south-​east Britain is far from settled. A shifting research agenda in
recent decades has led to a change from studies prioritizing the economy (for example,
villa development and pottery marketing) to those favouring cultural change (for exam-
ple, the social use of space and consumption), resulting in necessarily partial explana-
tions of Rome’s impact. Given this state of play, the most compelling general models
of rural change that integrate both economic and cultural processes remain those that
were first proposed in the 1980s and early 1990s. Notable examples include Haselgrove’s
account (1982) of political centralization in late Iron Age south-​east Britain, Hingley’s
structural analysis (1982) of the impacts of Roman imperialism, and Millett’s interpreta-
tive essay (1990a) on the Romanization of Britain. These are typical examples of proces-
sual grand narrative, drawing upon Immanuel Wallerstein’s ‘World-​Systems Analysis’
(1974), which invokes the concepts of ‘centre’ and ‘periphery’ as analytical categories.
While the most explicit centre–​periphery models have fallen out of regular usage—​
owing to a combination of significant elements being refuted by further archaeologi-
cal investigation and a shift towards post-​processual studies eschewing macroeconomic
models in favour of local and symbolic perspectives—​such models deserve fresh exami-
nation in the light of new data being unearthed.
The most relevant processual models for understanding rural change in Roman
Britain are those of Hingley (1982) and Millett (1990a). Both are effective in linking
the global pressure of Rome’s imperial regime through taxation to local developments
at the tribal or civitas level that dictated the subsequent character of urban–​rural rela-
tionships. For Hingley, Rome’s mission as a colonizing power was to exploit its subject
territory, to facilitate a flow of surplus goods and produce to be directed to the ‘core’
of the empire. To achieve this, particular forms of social organization and economic
relationships were fostered by Rome, involving recurring relations of dominance and
dependency at multiple scales—​between the provinces and the ‘core’ at the inter-​
provincial scale and between town and country at the tribal level. Hingley’s model
envisaged the empire in terms of a pyramid structure, with relations between the
imperial core at the pyramid’s apex and rural locations at the base, being articulated
through tribal centres (for example, Verulamium), which were, in turn, subordinate
722

722   Martin Pitts

to provincial centres (for example, London). Flows of basic agricultural produce, raw
materials, and slaves moved up the pyramid in response to the stimulus of taxation,
with quantities being retained or converted to cash at the intermediate levels for local
consumption, and a reciprocal downward flow of luxury and manufactured goods
from the apex to the base. Crucially, the reciprocal exchange of manufactured goods
for agricultural raw materials was unequal, with the latter produced in the periph-
ery requiring greater expenditure of energy (largely in the form of labour) than the
exchange commodities produced in the core. In short, this is a more formalized and
systemic realization of Collingwood’s statement of major Romano-​British towns (that
is, the civitas/​tribal capitals) functioning as parasites. Major towns were required ade-
quately to facilitate the exploitation of the rural populace, serving as nodal points for
the conversion of agricultural surplus to tax and the wider dissemination of imported
luxury and mass-​produced items.
In contrast to Hingley, Millett (1990a, b) highlighted Rome’s laissez-​faire attitude to
provincial administration, emphasizing the role of pre-​conquest elites as the primary
agents of change. Providing that sufficient centralized social organization existed prior
to Roman annexation, friendly members of the pre-​existing tribal aristocracy could
retain their local power and govern on Rome’s behalf in return for the adoption of a
Roman constitution and the guarantee of tax revenues, which were payable by civitas
(Millett 1990a: 125). As the power of the local elite became increasingly dependent on
Roman patronage, the desire to use and appropriate Roman symbols became more
prevalent throughout society, eventually becoming status indicators that were to be
progressively emulated by the lower social orders (Millett 1990b). Millett (1990a: 117–​
123) argued that the ‘trickle-​down’ of such Roman aspiration was evident in rural areas
through the flourishing of villas in the second and third centuries, in addition to signifi-
cant distributions of imported samian pottery outside the urban sphere (see also Evans,
this volume). Despite the reduced emphasis placed on imperial exploitation in Millett’s
model, it does not contradict Hingley (1982). Both accounts recognize the role of taxa-
tion in structuring relations between town and country and both suggest that the adop-
tion of Roman goods and symbols in the countryside was an inevitable and desirable
outcome. Whereas the early formulations of Hingley and Millett are broadly consist-
ent with one another, they appear to diverge from Collingwood on the rural penetra-
tion of Roman material culture, who was at pains to stress the failure of Romanization
in the countryside (either material or ideological), a view that has been more recently
expressed by David Mattingly (2006) and others.
Clearly, questions of rural change in Roman Britain cut across several issues of wider
importance and controversy—​the extent to which the colonial regime was exploitative
or facilitating; whether change in material culture was a by-​product of economic pro-
cess or cultural participation; the degree to which non-​elite communities had agency
and choice; and the precise mechanisms by which rural communities were socially and
economically connected to the larger whole of the Roman Empire. Addressing these
concerns requires not only detailed empirical exploration of the available data (of which
a burgeoning corpus has been amassed since the debates of the 1980s and early 1990s),
 Rural Transformation in the Urbanized Landscape    723

but also an integrated approach that considers such evidence in the light of both eco-
nomic and cultural terms (cf. Fulford 1982; Millett 1982).

Models and Data

There currently exists a large and diverse array of archaeological data for the evaluation
of theories of rural transformation in Roman Britain. However, this is a comparatively
recent development in the history of the subject. In the mid–​late twentieth century, for
example, attempts to understand rural change were predominantly focused on chart-
ing the origins and development of ‘villa’ buildings in the countryside, which came to
be seen as evidence for the Romanization of the indigenous population (e.g. Rivet 1964:
99–​130; Richmond 1969). Rivet’s statement that ‘we cannot have a villa without a town’
(Rivet 1964: 105) was, in part, reified by the emergent pattern of the earliest villas devel-
oping in close proximity to major towns in the south and south-​east. The excavations
at early villas such as Park Street (near Verulamium, modern St Albans) and Lockleys
(Welwyn, both in Hertfordshire and in the civitas of the Catuvellauni) became type-​
sites to illustrate the Romanization of the ‘native’ elite. This process was characterized
at both sites by the direct replacement of Iron Age round-house dwellings by rectilinear
‘Romanized’ house forms within a generation of the Claudian conquest. Such explana-
tions remain broadly current, with Millett (1990a: 94) regarding the phenomenon as an
outgrowth of the desire by their local owners to appear Romanized rather than an indi-
cator of increased prosperity and living standards under the new regime (contra Frere
1987: 258).
The privileged position of villas in accounts of rural change was to change irrevocably
in the 1970s and 1980s. In the first place, these decades witnessed an upsurge in the num-
bers of non-​villa rural settlements excavated, reflecting the historical reality that villas
would have been vastly outnumbered by their non-​villa counterparts (Hingley 1989: 4–​
5). The persistence of rural settlements lacking overtly ‘Romanized’ architectural fea-
tures highlighted a gap in traditional narratives as well as in explanations of cultural
change amongst non-​elite communities. In addition to a more balanced coverage of set-
tlement types, advances in the recording and quantification of small-​finds assemblages
(artefacts and ecofacts) provided glimpses of a more holistic picture of urban–​rural
relationships in Roman Britain, in terms of patterns of diet and ‘marketing’ distribu-
tions. For example, King (1978, 1984) showed entrenched differences in the proportions
of major domesticates in the animal bone assemblages of so-​called Romanized settle-
ments (towns and villas, favouring cattle) and ‘un-​Romanized’ settlements (low-​status
farmsteads, favouring sheep and goat) (for further discussion of animal husbandry, see
Maltby, this volume). King attributed these changes to Romanization as a form of cul-
tural preference, with the influence of garrison units previously stationed in Gaul and
Germany providing the basis for local emulation. In terms of the pottery, there were still
insufficient data properly to evaluate potential differences in urban and rural supply at
724

724   Martin Pitts

this time (Millett 1982). Nevertheless, Swan’s corpus (1984) of Romano-​British kiln sites,
in addition to distribution studies of later Roman products such as Oxfordshire (Fulford
and Hodder 1975) and Farnham wares (Millett 1979), suggested consistent patterning,
notably a shift away from the urban marketing of ceramics in the early Roman period
to the development of late Roman rural-​based industries that took advantage of river
transport to tap into markets further afield (for further discussion of the pottery evi-
dence, see Evans, this volume).
Building on the developments of the previous two decades, scholarship from the 1990s
onwards shows a greater level of sophistication and specialization. On a theoretical level,
the influence of postcolonial literature (e.g. Said 1993) and its applications in Roman
archaeology (e.g. Webster and Cooper 1996; Mattingly 1997) prompted a rethinking of
historical models of Romanization as a top-​down civilizing process, with debates over
the contextual meaning of material culture taking increased prominence (e.g. Webster
1997). Whereas the issue of rural change remains firmly on the agenda, with increased
emphasis on ‘bottom-​up’ perspectives and non-​elite communities (e.g. James 2001),
questions of urban–​rural interdependence, economic growth, and marketing have been
arguably sidelined in favour of the dominant agenda of cultural concerns—​for exam-
ple, Romanization, creolization (Webster 2001) and identity (Mattingly 2004). At the
same time, the ever-​increasing corpus of assemblage data from individual excavations
demanded quantitative study and the application of new methodologies for appropri-
ate analysis and synthesis (e.g. Evans 1995, 2001; Cool and Baxter 1999, 2002; Lockyear
2000). As a consequence, while archaeological data from Roman Britain are routinely
examined in order to shed light on cultural difference, the economic underpinning of
such patterns is seldom prioritized. The most sophisticated models that address eco-
nomic concerns remain, for the most part, those formulated in the early 1990s (e.g.
Millett 1990a)—​ironically an era in which barely sufficient data were available to test the
models being proposed.
Despite the relative lack of attention to the integration of findings and model test-
ing, the patterns emerging from more recent synthetic studies of Romano-​British finds
assemblages are remarkably consistent. While the scope for significant variation within
settlements may be accepted (e.g. Eckardt 2007; Willis 2007), most striking is the dif-
ferentiation of major urban settlements (and military sites) from lesser nucleated (for
example, small towns and villages) and rural settlement classes—​the latter tending to
be less distinctive as a discrete category and subject to greater variation depending on
regionality and/​or the particular classes of evidence under scrutiny. A brief summary
of the basic patterns is as follows. In terms of coin loss, the major towns show a consist-
ently higher rate than all other classes of settlement, apart from sites associated with
the army (Reece 1993). Faunal assemblages from major urban centres and military sites
tend to have higher proportions of pig and cattle remains than other settlement types,
with progressively lower percentages in assemblages in the order of secondary urban
sites, villas, and rural sites (King 1999, 2001). Analysis of butchery practice shows a high
frequency of cut-​marks indicative of specialist animal carcass processing in towns, with
a comparative paucity of equivalent marks in rural assemblages (Maltby 1989, 1994,
 Rural Transformation in the Urbanized Landscape    725

2007). The synthesis of available data on plant remains shows military communities and
major urban centres (especially London) to have been the biggest consumers of exotic
and imported plants, with access to such foodstuffs being more limited on rural sites in
the south-​east, and virtually non-​existent for rural sites in the rest of the province (van
der Veen et al. 2008). A similar pattern emerges for the spread of lighting equipment,
with little uptake of lamps outside military communities and the major civilian centres
of London and Colchester (Eckardt 2002). Finally, the increasingly wide-​ranging and
statistically robust analyses of pottery consumption in Roman Britain highlight the sup-
ply of both imported wares (such as samian and amphorae) and specialist consump-
tion technology (drinking and dining forms as opposed to basic cooking wares) to be
most prevalent in assemblages from major towns (and military sites), with a progressive
fall-​off in frequency further down the settlement hierarchy (e.g. Willis 1996, 2011; Evans
2001, 2005; Pitts 2008).

Urban Hinterlands—​Beyond
Resistance and ‘Romanization’

The overarching picture that emerges from recent surveys of consumption in Roman
Britain (as briefly summarized above) is, arguably, one involving the maintenance of
a wide gap between urban and non-​urban populations. It should not be underesti-
mated that the principal evidence for such a gap consists of differences in the supply of
foodstuffs and everyday manufactured objects—​such patterns are most likely to have
been experienced in the guise of entrenched differences in social practice and culture
between urban and rural communities. This observation is acutely apparent in recent
attempts to investigate urban hinterlands in Roman Britain. Gaffney and White’s sur-
vey-​driven study of Wroxeter’s hinterland revealed negligible material change in the
countryside in terms of the spread of urban forms of consumption: ‘the town appears
rich, dynamic, and “Romanized”; the countryside appears impoverished, static and
conservative’ (Gaffney and White 2007: 282–​283). While the scale and character of such
results are likely to be, in part, attributable to the social organization (or lack thereof) of
the Cornovii prior to conquest, described as ‘decentralized and egalitarian’ by Millett
(1990a: 100) (a classification disputed by Gaffney and White), it is, nonetheless, arguable
that comparable patterns emerge in areas with centralized and hierarchical pre-​con-
quest societies, once such pre-​existing differences are accounted for.
In contrast to the Wroxeter hinterlands project, the study of the East Anglian hin-
terlands of Roman London and Colchester (Perring and Pitts 2013) privileged the
analysis of consumption patterns from excavated finds assemblages, taking particular
advantage of the region’s numerous large and well-​dated published pottery groups. The
results of this project revealed that, although there was a strong uptake of new forms of
material culture in the Essex countryside (relative to regions outside the south-​east),
726

726   Martin Pitts

in quantitative terms the urban–​rural difference remained striking. Correspondence

analysis of nearly 100 stratified pottery assemblages (quantified by estimated vessel
equivalent) from London, Essex, and Cambridgeshire shows very little change in the
pronounced separation of patterns of urban and rural pottery use over c. 200 years,
despite a moderate uptake of supposedly ‘Romanized’ vessel forms throughout the
region and period (Figure 35.1 and Table 35.1). The earliest assemblages from settlements
with pre-​conquest origins are distinguished by the presence of imported Gallo-​Belgic
fine wares and older Gallic-​influenced pedestalled vessels (for example, Heybridge and
Kelvedon, upper-​left quadrant of Figure 35.1), with the tendency for large drinking ves-
sels in such ceramic repertoires fitting with the regional predilection for feasting activi-
ties (Pitts 2005, 2010). However, such colourful practices appeared to die out within a
generation of the Claudian conquest. From the Flavian period onward, communities
living outside the major urban centres appear to have been content to use a seemingly
watered-​down, locally produced version of what was being used by city-​dwellers—​far
fewer imports, fewer vessels associated with the presentation and serving of food and
drink, and, compared to the earlier first century, a relative shift in emphasis from drink-
ing vessels to dining vessels within the subset of pottery used for consumption. In terms
of pottery usage, Figure 35.1 shows no significant or consistent differences between
assemblages from minor urban settlements, villas, and low-​status rural sites. Such rela-
tive homogeneity outside the ‘true’ urban sphere echoes findings in other classes of evi-
dence, such as plant remains, whereby lesser urban centres showed greater affinity with
rural sites and little differentiation between villas and lower-​status sites (van der Veen et
al. 2008).
Further indication of the character of impact of urbanism on the wider settlement
landscape can be obtained by considering particular types of material culture that can
act as proxies for other more illuminative materials (for example, foodstuffs). One
particularly striking pattern from the Essex urban hinterlands project (Perring and
Pitts 2013) was the relatively high levels of lids in pottery assemblages from London
and Colchester (plotted in Figure 35.1 alongside other diagnostic urban staples
such as flagons and mortaria), a trend that is apparent throughout the early Roman
period. Further investigation showed lids made up around 4–​10 per cent of the typi-
cal ‘true’ urban assemblage, with less than 4 per cent on most non-​urban sites (Figure
35.2). What makes this a seemingly odd finding is that the vast majority of lids were
made of local sandy grey coarse wares, probably at multiple kiln sites throughout the
region (for example, production of this fabric is attested at Ardleigh, near Colchester
and Heybridge, somewhat further afield) (Going and Belton 1999; Biddulph et al.
2015). Although such grey wares invariably form a substantial component of most
coarse ware assemblages regardless of settlement-​type in the region during the early
Roman period, the most common products in the fabric (various cooking jars) are
more synonymous with rural and minor urban settlements than major centres such
as London and Colchester. Such patterns, at the least, suggest a particular prefer-
ence for lids among urban communities, perhaps for cultural reasons linked to
the preparation of food. However, if this was the case, the scarcity of imported lids
 1.0
WDW2-4 WIT3-4 KEL1-3 STN5 Major urban
WAM3-5 BOR7-8 STN3-5 STN3 RHM4-5
HEY2-3
KEL5-6
STN6-7 Minor urban
CHM-TP4-6 STN7-8 High-status rural
CAN3-6
HEY3-6 CHM5 BRT5-6 CHM-TP8-9 Low-status rural
FDT3-5 BRT5-9 WAM10-11
KEL3-4 ARD3-5 RET6-7 NSH8-10
HEY3 WDW3-4 BRT8-9 ARD8
HEY3-4 CHM-MA3-6 HEY8-10
KEL3-5 BRT10-11 STH8-9 BOR10
0.5 STN3-7
CHM-MA8 HEY7-8
CAN3-7 CHM-TP9-10
CHG3-6 RET8-9
CHM-MA6
CHM6-7 WIT8-10 CHM-MA10-11 CHM-MA9-10
HEY4-6 ARD6-7
HEY10-11 BOR9-10 LOK11
ARD4-6
HEY9-10
KEL4-6 CHM8-10
RET10-11
0.0
COL8
COL8-10
COL4 LON-E6-8
LON-E8-9
COL4-6 COL9
COL6-8 COL6-7 COL6-9 COL9-10
COL4-5
COL5-6 COL7
LON-W5-8 LON-W5-6
COL6 LON-E5-7
LON-W5 LON-E5-6 LON-W6-8
–0.5 LON-W5-7 LON-E9
LON-W7-9
LON-E7-8 LON-W6-7
LON-W3-4 LON-W9-10
LON-W7-10

–0.5 0.0 0.5 1.0 1.5

1.5 Pedestal jars

GB beakers
Pedestal tazze
GB jars
1.0
GB platters
Platters Non-urban and
minor urban
0.5 Jars
Miniatures
Storage jars
GB flagons
0.0
Cups
Beakers Dishes

Major urban
IMP platters
–0.5 Bowls SMN dishes
IMP mortaria
SMN platters Lids IMP beakers
IMP flagonsSMN cups
IMP cups SMN bowls
Amphorae Flagons
IMP lids Mortaria
–1.0 IMP bowls IMP dishes

IMP jars
–0.5 0.0 0.5 1.0 1.5

Figure 35.1  Correspondence analysis highlighting differences in the composition of pottery

assemblages (top) by vessel form (bottom) in the hinterland of Roman London and Colchester, c.
50 bc–​ad 250.
Note: Details of abbreviations are outlined in Table 35.1. © Martin Pitts.
728

728   Martin Pitts

Table 35.1 Codes used in Figures 35.1 and 35.2

Code Site Ceramic phase Date range

ARD Ardleigh 1 50–​15 bc

BOR Great Holts Farm, Boreham 2 15 bc–​ad 20
BRT Braintree 3 ad 20–​55

CAN Cambridge, New Addenbrookes 4 ad 55–​60

CHG Chignall 5 ad 60–​75/​80

CHM Chelmsford, street frontage 6 ad 75/​80–​100

CHM-​MA Chelmsford, Mansio 7 ad 100–​125

CHM-​TP Chelmsford, temple 8 ad 125–​140

COL Colchester 9 ad 140–​160/​70

FDT Fen Ditton 10 ad 160/​70–​210

HEY Elms Farm, Heybridge 11 ad 210–​250

KEL Kelvedon Fabric code Fabric

LOK Little Oakley SMN Samian ware
LON-​E London, East IMP Misc. imports
LON-​W London, West GB Gallo-​Belgic
ware
NSH North Shoebury
RET Curry Hill, Rettendon
RHM Moor Hall Farm, Rainham
STH Strood Hall
STN Stansted
WAM Wendens Ambo
WDW Woodham Walter
WIT Ivy Chimneys, Witham

from the continent in early urban assemblages is unusual. An alternative explanation

is that lids were required specifically to aid the transportation of foodstuffs or other
products from the country to the city, facilitating urban consumption. This hypoth-
esis accords with the suggestion that major towns would have formed the primary
market for new varieties of fresh fruit, vegetables, and herbs that became widespread
in the south-​east (van der Veen et al. 2008) (see also van der Veen, this volume).
 Rural Transformation in the Urbanized Landscape    729

18 LON-W9-10
COL9 Major urban
Minor urban
16 LON-E6-8 High-status rural
COL6-9 Low-status rural
14 LON-W7-9

12 LON-W6-7
LON-E9
10
% lids

LON-W6-8 COL7
LON-W5-8 COL4
HEY4-6
8 LON-E5-6 COL6-7
LON-E5-7 COL9-10 LON-E8-9 COL8 ARD6-7
LON-W5-7 LON-W5-6 COL4-6
LON-E7-8
6
COL6 ARD8
COL6-8 CHG3-6
ARD4-6
4 LON-W7-10 COL5-6
COL4-5 HEY7-8
HEY3 FDT3-5 ARD3-5
CHM6-7 WDW3-4 CHM-TP4-6 RHM4-5
LOK11 HEY3-4
2 KEL3-5 STN5
HEY2-4
LON-W5 HEY10-11 HEY8-10 BRT5-9
CHM8-10 CHM-TP9-10 BRT8-9 KELI-3
CHM-MA10-11 CAN3-6
20 30 40 50 60 70 80 90
% jars

Figure 35.2  Pottery assemblages from Roman London, Essex, and Cambridgeshire plotted
according to the percentage prevalence of lids versus jars, c. 50 bc–​ad 250.
Source: after Perring and Pitts (2013). © Martin Pitts.

Returning to the processual models of Hingley (1982) and Millett (1990a), the situ-
ation described for the south-​east fits well with the idea of rural produce (contain-
ers and/​or their contents) being sent to major towns in response to the stimulus of
taxation—​either to be converted to cash in urban markets or to be taken directly in kind.
Furthermore, although imports and more specialist mass-​produced goods available in
urban markets clearly did make it into the countryside, the relative weakness of rural
consumption in emulating urban consumption highlights the essential asymmetry and
inequality of urban–​rural exchange as postulated by Hingley. Similarly, in terms of the
consumption of regional produce, Colchester appeared much more dependent on the
output of its immediate hinterland, whereas London received flows of goods from a
wider variety of sources within the province, seemingly confirming the latter’s elevated
position in Hingley’s pyramidal structure of imperial relationships. Although the gap
between urban and non-​urban consumption was largely maintained into the third cen-
tury, relations of urban–​rural interdependence were not static. Indeed, the available
evidence suggests a tendency for increased integration in pottery supply across all settle-
ment types by the late second century (Perring and Pitts 2013). While this period coin-
cides with a wave of villa construction in the region, the villa assemblages themselves are
not greatly differentiated from those of non-​villa rural settlements. It thus remains to be
seen whether or not the appearance of villas represented the ‘trickle-​down’ of Roman
culture into the countryside or simply the investment in rural property by absentee
landlords who relied on local tenants to manage their holdings.
730

730   Martin Pitts

Assessing the cultural implications of the various changes in rural production and
consumption in East Anglia is not straightforward. The relatively weak rural consump-
tion of imports must be set in the context of sweeping changes to coarse ware produc-
tion, which, although heavily influenced by continental forms (via Gaul) prior to the
conquest, took on a decidedly new ‘provincial’ character by the late first century ad. In
the light of the infamous Boudiccan uprising of ad 60/61, it is unlikely that such changes
in production reflected a widespread local desire to adopt the ways of the colonial set-
tlers at Colchester. Instead, as the case of the coarse ware lids suggests, such changes
probably reflected the realignment of production to suit the needs of the imposed urban
communities—​implemented either via the remaining indigenous elites or by the next
generation of veteran colonists. The impact on the countryside of fulfilling the basic cal-
orific demands of the new urban communities should not be underestimated—​as well
as being provided with the basic foodstuffs, the urban populations also required appro-
priate vessels in which to cook and serve food and drink in concordance with their own
cultural needs and practices.
Against the backdrop of a new impetus for rural production comes the question of
the agency of communities living outside the major urban centres. Were they content
to adopt the new types of pottery used by urban dwellers and carry on with their lives
as before, or did they do so consciously to emulate the tribal aristocracy and to iden-
tify themselves with the new provincial order (as per Millett 1990b)? On this issue the
models of Hingley (1982) and Millett (1990a, b) are less in tune with the available evi-
dence. Here it is crucial to evaluate both the extent and meaning of the so-​called trickle-​
down of imported ceramics that enjoyed conspicuous display in the graves of local
elites in the initial post-​conquest period, such as Folly Lane, St Albans (Niblett 1999),
and Stanway, Colchester (Crummy et al. 2007). While it is true that imports such as
samian ware clearly did exhibit widespread penetration outside the major towns, the
quantities involved rarely rivalled those of urban assemblages. On this point, Millett
(1990a: 123) expressed the view that:

Larger-​scale distributions like that of samian ware were less heavily centred on the
towns than might be expected. This indicates an ample sufficiency of supplies and
so the demands of any consumer could be filled. Where these objects were not being
received the deficiency was of demand, not supply.

In the light of data now available, it appears that the gap between urban and non-​
urban consumers was greater than Millett originally envisaged. Either the supply of
imports was not sufficient to meet rural demand, the pots themselves were too expen-
sive for a large proportion of rural consumers, or the demand for such goods in the
countryside was simply not as strong as that of the major towns. These are not necessar-
ily mutually exclusive possibilities. While samian was evidently imported in quantity to
Britain, supplies were not limitless. As the premium ceramic table ware of the period,
such imports would inevitably have carried higher prices than basic coarse wares. A full
suite of samian vessels probably involved a substantial outlay of income for the average
 Rural Transformation in the Urbanized Landscape    731

Romano-​British peasant and could hardly have been a priority for people living at, or
close to, the subsistence level.
Addressing the uses to which imported ceramics were put in the Romano-​British
countryside is especially pertinent to understanding the desire for such commodities. In
cultural terms, the growth in the consumption of Gallo-​Belgic wares (a repertoire featur-
ing many forms that imitated samian originals) at a wide range of locations in Essex up
to the time of the conquest demonstrates a pre-​existing taste for sophisticated ceramic
forms beyond the circle of elite usage (e.g. Cool 2006: 152–​171). Although it might be
expected that this demand for imported fine wares would be maintained in rural areas
after the conquest, the shift in elite dining and display to the increasingly built-​up and
rarefied urban environment effectively raised the bar of exclusivity that only a minority
could hope to emulate. Whereas the use of imports in late Iron Age feasting practices
seemingly placed few restrictions on appropriate architectural settings, an elegant ban-
quet held in a British round-house would have done little to impress Roman officials
accustomed to the levels of luxury experienced in the Mediterranean. Despite evidence
for the more widespread uptake of imports, the changed political landscape neverthe-
less afforded limited opportunities for true emulation and empowerment on the part
of provincials living outside the major towns. This did not preclude the possibility of
imports being used to engender a sense of social distinction, luxury, and sophistication
within local rural communities. Indeed, where continuity in the uptake of Gallo-​Belgic
to samian fine wares is apparent at indigenous settlements post-​conquest (for exam-
ple, Baldock, Hertfordshire, and Heybridge, Essex), evidence for the deposition of such
vessels suggests continuity with pre-​Roman and early Roman food ways into the late
second century ad (Pitts 2007, 2008). A wholesale shift towards a greater emphasis on
dining forms (for example, dishes and bowls) across all assemblages is more likely to
reflect the changed priorities of the surviving local elite and the needs of the urban mar-
ket, rather than the widespread adoption of Roman dining practices and mores by Essex
peasants (Perring and Pitts 2013). As Cooper (1996: 95) succinctly stated, ‘many of the
people who used Roman-​style pottery had no particular desire to emulate Roman ways
but were simply taking what was available to them and busily adapting it as part of the
material expression of their own culture’.

Health, Well-​being, and

Social Inequality

It is undeniable that Roman imperialism had a profound impact on the lives of people
who resided in the British countryside. Although the arrival of Rome presented new
opportunities for plurality and cultural expression—​for example, through new forms of
dress and items of personal adornment (Crummy and Eckardt 2003; Eckardt 2005)—​it
is assumed that the new demands of taxation and the provisioning of imposed urban
732

732   Martin Pitts

and military populations meant that for the majority this impact would have been more
exploitative than facilitating (e.g. Mattingly 2006). This idea forms another component
of R. G. Collingwood’s model of the parasitical city, but is again one that remains largely
unproven. After all, the entrenched differences in urban and rural patterns of consump-
tion do not provide direct evidence of the degree of human exploitation nor of the extent
to which the consequences of Roman imperialism may have affected the quality of rural
life. An important question in this respect is whether or not the observed discrepan-
cies in consumption patterns between town and country are indicative of differences in
social inequality and the distribution of wealth between urban and rural communities.
As Jongman (2007: 596) puts it:

social inequality matters for an understanding of the lives of many ordinary Romans.
It matters for our understanding of the [economic] growth that did occur, because
the wealth of the elite (often paraded by ancient historians as a sign of prosperity)
may not have been a sign of a prospering economy after all, but instead of effective
exploitation of the poor.

Moreover, recent epidemiological research highlights the startling reality that mod-
ern societies with high levels of inequality in the distribution of income also show higher
rates of poor health and lower life expectancies, even in wealthy industrialized nations
with advanced healthcare provision such as the United States and the United Kingdom
(Wilkinson 2005; Wilkinson and Pickett 2010). The implications of these findings with
respect to the ancient world are clear: if significant social inequality existed, it should be
manifest in patterns of poor health, even if evidence of income is not available to meas-
ure inequality directly.
Unfortunately, effectively gauging the level of poor health in ancient popula-
tions is fraught with challenges. In addition to the familiar archaeological problems
of ensuring that the various biological traits visible through the scrutiny of human
remains are consistently recorded and quantified between different excavated cem-
eteries, it is also necessary to mitigate for the effects of other factors that could have
a major impact on interpretation. For example, many skeletal conditions that might
be associated with hard labour during life are also more likely to occur in particularly
long-​lived individuals as a general sign of ageing. Conversely, rather than indicating
an individual who had lived a healthy life, a skeleton showing no apparent evidence
of disease or trauma could simply have resulted in the case of an unhealthy individ-
ual rapidly succumbing to infectious disease, leaving no time for lesions to form on
the skeleton—​the so-​called osteological paradox (Wood et al. 1992). In addition, the
prevailing ritual of cremation in the early Roman period severely limits the potential
insights into changing patterns of health from human remains in Roman Britain,
owing to the destructive nature of the rite. Nevertheless, if such problems are taken
into account, an initial study of the potential relationship between social inequality
and poor health in late Romano-​British cemeteries has yielded some striking results
(Pitts and Griffin 2012).
 Rural Transformation in the Urbanized Landscape    733

Ancaster

0.5
Boscombe
Kempston
Baldock 4
Baldock 3
Cassington
Baldock 1
Stanton Harcourt Rural

Cirencester, North
Chignall
Colchester
Lankhills St Albans
0.0 Derby
Dimension 2

Minor urban Radley, Barrow Hills

Alchester
Queensford Mill, Dorchester-on-Thames
Cirencester, South

Icklingham
Gloucester, Kingsholm
Dorchester, Tolpuddle Hall
Bletsoe
Dunstable
–0.5

Gloucester, Gambier-Parry Lodge

Major urban
London, West Tenter St Radley
Urban
Minor Urban
Leicester, Newarke St Rural
–0.5 0.0 0.5 1.0
Dimension 1

Figure  35.3  Multidimensional scaling analysis of the prevalence of non-​work-​related health

conditions in late Roman Britain.
Source: after Pitts and Griffin (2012). © Martin Pitts.

Figure 35.3 presents a picture of the overall similarities and differences in health in
late Roman Britain by measuring the statistical similarity of 30 late Roman cemetery
populations from central and southern England in terms of multiple health indicators,
and plotting the results using multidimensional scaling. Each cemetery selected was
composed of at least 20 individuals, with data available for at least three of the following
eight palaeopathological conditions chosen for their relationship with diet, infectious
disease, and/​or as general indicators of poor health: rib periostitis, caries, dental cal-
culus, ante-​mortem tooth loss (AMTL), dental abscesses, periodontal disease, enamel
hypoplasia, and cribra orbitalia (for discussion of medicine, see Baker, this volume).
Although the results exhibit a degree of overlap when considered by settlement type, a
general pattern is visible whereby urban cemeteries are plotted towards the lower left,
minor urban cemeteries in a line following the central vertical axis, and rural cemeteries
734

734   Martin Pitts

to the upper right. Regression analysis indicates that dimension 1, accounting for the
greatest variability in the sample, has a statistically significant relationship with settle-
ment type as well as the prevalences of all the dental health indicators considered apart
from caries, including enamel hypoplasia, an indicator of general health. In short, the
results show that, as rates of poor health increase, so too does the likelihood that the
cemetery in question is associated with a rural settlement.
If we take stock of the results outlined in Figure 35.3, the emergent pattern of (rela-
tively) healthy urban populations in contrast to their rural counterparts is somewhat
at odds with the squalid picture of Roman urban life painted by ancient historians (e.g.
Scobie 1986; Garnsey 1999) and palaeopathologists (Redfern and Roberts 2005). At this
point it is important to consider that late Roman towns in Britain were probably not
especially densely occupied, which thus reduced the scope for infectious disease and
other problems related to inadequate sanitation. In contrast to their predecessors and
continental counterparts, late Romano-​British towns were probably relatively sparsely
occupied centres of government with large private residences (Millett 1990a: 221–22​3;
Mattingly 2006: 325–3​50). In this context, it is not unlikely that the provision of urban
amenities outweighed the negative health implications of urban living—​the benefits of
Roman civilization highlighted by Collingwood that only the privileged urban dwell-
ers were able to access. Given that dental differences linked to diet form an important
component of the patterns displayed in Figure 35.3, it is likely that a particular benefit of
urban living was access to better food and dental treatment. The dental differences imply
a greater dependency on carbohydrates (that is, cereals) in the countryside and suggest
a higher meat component in the urban diet, which is consistent with the higher propor-
tions of animals bred solely for meat (that is, pigs) and evidence for butchery practices
indicating the production of filleted beef in urban contexts (Maltby 2007). Access to
better food is difficult to measure other than in terms of general health, although this
again is consistent with what might be expected in a system involving the transfer of
agricultural surplus to urban centres. Indeed, the patterns fit with Galen’s (De alimento-
rum facultatibus 1.13) second-​century observations on Roman Asia Minor, where rural
populations were forced to subsist on poorer-​quality cereals, having sent their best pro-
duce to the towns.
In the context of recurrent differences between urban and rural settlements in virtu-
ally every class of material evidence available, it should come as no surprise that pat-
terns of health in Roman Britain similarly indicate that rural populations were worse
off. This did not mean that the country was not a healthy place to be for the wealthy few,
but, rather, that the numbers of people enjoying such a lifestyle would have been negli-
gible in the wider context. The wretched patterns of health observed among the small
cemetery population associated with the extensive villa at Chignall (Essex) are entirely
in keeping with the picture of an overworked workforce of slaves or bonded tenants, as
opposed to a wealthy Romano-​British family (Clarke 1998).
So far the available data suggest that the major fault-​lines of inequality in the urban-
ized landscape of Roman Britain existed between urban and rural populations, with
minor towns forming something of a compromise between the two extremes. However,
 Rural Transformation in the Urbanized Landscape    735

if inequality itself was a cause of such health differences, according to the epidemiologi-
cal literature (e.g. Wilkinson 2005) we would expect such differences to be underpinned
by higher levels of inequality within the most badly affected (rural) communities. This is
especially relevant to the strong relationships between settlement type and conditions
such as neoplastic disease (cancer) in late Roman Britain observed by Pitts and Griffin
(2012), which were likely to have been influenced by differences in living conditions con-
sidering the absence of modern carcinogens in the environment. Unfortunately, there is
no straightforward means of measuring the level of income inequality between different
communities in Roman Britain. To explore this problem, Pitts and Griffin (2012) calcu-
lated the level of inequality in the provision of grave furnishings in a sample of late Roman
cemeteries, with the basic assumption that a more unequal society would treat its mem-
bers unequally in death, thus providing a theoretical proxy for inequality in life. Although
somewhat speculative, the results outlined in Figure 35.4 show that, on the whole, late
Roman urban cemeteries were characterized by lower inequality than their non-​urban
counterparts, measured according to the Gini coefficient. The Gini coefficient is a com-
monly used indicator of inequality in modern societies, yielding a number between 1 and
0, where 0 denotes perfect equality and 1 a scenario of perfect inequality, with one per-
son having everything and everyone else nothing. Significantly, the most notable result to
buck this trend was from the urban cemetery of Cirencester, which behaved more like a
rural settlement according to indicators such as coin loss (Reece 1993).
In the wider context, the picture of inequality in late Roman Britain outlined by
Pitts and Griffin (2012) is certainly consistent with Willem Jongman’s argument that
the material benefits of the Roman system were increasingly unequally distributed
under the late Empire (Jongman 2007: 597). Also included in Figure 35.4 are three early
Roman cremation cemeteries from Hertfordshire, which typically show higher levels
of furnishing as well as lower levels of inequality in their distribution compared to most
late Roman cemeteries. While at face value these patterns appear to validate Jongman’s
vision, caution is warranted in the absence of corroborative health data. Indeed,
although the early Roman cemeteries in question are not small (with at least 100 indi-
viduals in each), the rite of cremation appears less widespread than inhumation in the
late Roman period, and it is likely that, in some cases, cremation may have been the
exclusive preserve of particular subsets of society, such as collegia (Biddulph 2005) (see
also Weekes, this volume, for discussion of burial practice). As we do not know exactly
how representative the cemetery populations were of living communities, it is difficult
to assess the true significance of the results in Figure 35.4. In material terms, the earlier
rite of cremation was intrinsically more egalitarian, with incremental differences in the
numbers of pots and brooches per cremation (perhaps indicative of group member-
ship), whereas later Roman inhumation often involved a much wider gulf between the
large numbers of simple coffin burials and those interred in more expensive stone cof-
fins or mausolea.
Despite the necessary cautions and qualifications, the broad overlap with patterns of
health and other evidence for differences between urban and rural communities in late
Roman Britain is unlikely to be simple coincidence. One of the main elements contributing
736

736   Martin Pitts

3.5 Braughing
C1-C2 cremation
Urban
Minor urban
3.0 Rural

Welwyn
2.5
Lankhills
Mean no. of furnishings per grave

King Harry Lane

2.0

1.5
Ilchester
London, East
Colchester London, West Tenter St
Poundbury
Baldock
1.0 Bradley Hill
Stanton Harcourt
Chignall
Alchester
Kempston Dunstable
0.5 Cirencester
Radley, Barrow Hills
Queenford Farm, Dorchester-on-Thames

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Gini (all furnishings)

Figure  35.4  Comparison of the Gini coefficient of inequality with the mean number of fur-
nishings per grave for selected Romano-​British cemeteries
Source: after Pitts and Griffin (2012), with additions. © Martin Pitts.

to the differences in inequality was whether or not a coffin was provided for the deceased.
The inability for rural communities to provide what appears to have been regarded as a
standard component of the prevailing burial rite reinforces the picture of exploitation
beyond the urban sphere. Against this backdrop, it is tempting to view the longer-​term
effects of Roman imperialism in Britain in terms of increased oppression and inequal-
ity for the rural majority and greater wealth and luxury for the globalized and mobile
minority—​most notably evidenced in the flourishing of increasingly palatial villa com-
plexes in the fourth century (Millett 1990a: 227; Perring 2002; contra Hingley 1982: 34–​41).
This, of course, remains an impressionistic view that requires further testing. Nevertheless,
as the determinants of economic and societal well-​being in the modern world are ever more
based on the distribution of wealth, so should the success or failure of former historical eras.
Indeed, the degree of social inequality in the Romano-​British countryside should be a firm
priority for future research, especially given the tendency for postcolonial narratives to
assume high levels of exploitation.
 Rural Transformation in the Urbanized Landscape    737

Conclusions

R. G. Collingwood’s statement on the parasitical nature of Roman urbanism in

Britain was for the most part a deduction made on the basis of limited evidence.
Despite its somewhat prophetic nature in the light of the wealth of information
now available, it was, above all, a product of its time. In the same book Collingwood
explained the identification of contemporary early twentieth-​century Britons with
Rome as a civilizing power (and not the ‘native’ Britons), owing to the nineteenth-​
century British experience as a colonial power in central Africa (Collingwood and
Myres 1937: 178–​179). In a similar vein, the present account of rural transformation
in Roman Britain has deployed two yardsticks of early twenty-​first-​century glo-
balization—​consumption and social inequality—​to make sense of the Roman past.
Indeed, it is inevitable that archaeologists and historians will continue to interpret
the past through the lens of their contemporary social contexts, and this should not
be regarded as an obstacle to understanding, so long as it is done thoughtfully and
critically. Most importantly, Collingwood acknowledged the inherent inequalities in
the manifestation of Roman imperialism in newly conquered provincial territories—​
not unlike the effects of modern globalization in the developing world (e.g. Bauman
1998), the different forms and underlying mechanisms notwithstanding. That such
inequalities would have been felt most acutely in the countryside is unsurprising
given the pre-​eminent role of agriculture in the Roman economy and is in broad
concordance with the available archaeological evidence—​if not yet proven outright.

References
Primary Sources
Galen, De alimentorum facultatibus (On the Properties of Foodstuffs. Introduction, Translation
and Commentary), trans. O. Powell. Cambridge: Cambridge University Press, 2003.

Secondary Sources
Bauman, Z. (1998). Globalization: The Human Consequences. Cambridge: Polity Press.
Biddulph, E. (2005). ‘Last Orders:  Choosing Pottery for Funerals in Roman Essex’, Oxford
Journal of Archaeology, 24: 23–​45.
Biddulph, E., Compton, J., and. Martin, T. S. (2015). ‘The Late Iron Age and Roman Pottery’,
in M. Atkinson and S. J. Preston (eds), Heybridge: A Late Iron Age and Roman Settlement.
Excavations at Elms Farm 1993–5. Volume 2. Internet Archaeology 40. http://dx.doi.
org/10.11141/ia.40.1 (accessed 30 May 2016).
Clarke, C. P. (1998). Excavations South of Chignall Roman Villa, Essex, 1977–​81. East Anglian
Archaeology Report 83. Chelmsford: East Anglian Archaeology.
Collingwood, R. G., and Myres, J. N. L. (1937). Roman Britain and the English Settlements.
Oxford: Oxford University Press.
738

738   Martin Pitts

Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge
University Press.
Cool, H. E. M., and Baxter, M. J. (1999). ‘Peeling the Onion: An Approach To Comparing Vessel
Glass Assemblages’, Journal of Roman Archaeology, 12: 72–​100.
Cool, H. E.  M., and Baxter, M. J. (2002). ‘Exploring Romano-​British Finds Assemblages’,
Oxford Journal of Archaeology, 21: 365–​380.
Cooper, N. (1996). ‘Searching for the Blank Generation:  Consumer Choice in Roman and
Post-​Roman Britain’, in J. Webster and N. Cooper (eds), Roman Imperialism: Post-​Colonial
Perspectives. Leicester Archaeology Monographs 3. Leicester: Leicester Archaeology, 85–​98.
Crummy, N., and Eckardt, H. (2003). ‘Regional Identities and Technologies of the Self: Nail
Cleaners in Roman Britain’, Archaeological Journal, 160: 44–​69.
Crummy, P., Benfield, S., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An Élite
Burial Site at Camulodunum. Britannia Monograph Series 24. London: Britannia.
Eckardt, H. (2002). Illuminating Roman Britain. Montagnac: Editions Monique Mergoil.
Eckardt, H. (2005). ‘The Social Distribution of Roman Artefacts: The Case of Nail-​Cleaners
and Brooches in Britain’, Journal of Roman Archaeology, 18: 139–​160.
Eckardt, H. (2007). ‘Contexts in Colchester’, in R. Hingley and S. Willis (eds), Roman
Finds: Context and Theory. Oxford: Oxbow Books, 142–​149.
Evans, J. (1995). ‘Roman Finds Assemblages, towards an Integrated Approach?’, in P. Rush (ed.),
Theoretical Roman Archaeology: Second Conference Proceedings. Aldershot: Avebury, 33–​58.
Evans, J. (2001). ‘Material Approaches to the Identification of Different Romano-​British Site
Types’, in S. James and M. Millett (eds), Britons and Romans: Advancing an Archaeological
Agenda. CBA Research Report 125. York: Council for British Archaeology, 26–​35.
Evans, J. (2005). Pottery in Urban Romano-​British Life’, in A. MacMahon and J. Price (eds),
Roman Working Lives and Urban Living. Oxford: Oxbow Books, 145–​166.
Frere, S. S. (1987). Britannia: A History of Roman Britain. London: Pimlico.
Fulford, M. G. (1982). ‘Town and Country in Roman Britain–​a Parasitical Relationship?’, in D.
Miles (ed.), The Romano-​British Countryside: Studies in Rural Settlement and Economy. BAR
British Series 103. Oxford: Archaeopress, 403–​419.
Fulford, M. G., and Hodder, I. (1975). ‘A Regression Analysis of Some Late Romano-​British
Pottery: A Case-​Study’, Oxoniensia, 39: 26–​33.
Gaffney, V., and White, R. [with Goodchild, H.] (2007). Wroxeter, the Cornovii, and the Urban
Process. Final Report on the Wroxeter Hinterland Project 1994–​1997. Volume 1: Researching the
Hinterland. JRA Supplementary Series 68. Portsmouth, RI: Journal of Roman Archaeology.
Garnsey, P. (1999). Food and Society in Classical Antiquity. Cambridge:  Cambridge
University Press.
Going, C., and Belton, J. (1999). ‘Roman Pottery’, in N. R. Brown (eds), The Archaeology of
Ardleigh, Essex: Excavations 1955–​1980. East Anglian Archaeology Report 90. Chelmsford:
East Anglian Archaeology, 125–​157.
Haselgrove, C. C. (1982). ‘Weath, Prestige and Power: The Dynamics of Late Iron Age Political
Centralisation in South-​East England’, in C. Renfrew and S. Shennan (eds), Ranking,
Resource and Exchange. Cambridge: Cambridge University Press, 79–​88.
Hingley, R. (1982). ‘Roman Britain: The Structure of Roman Imperialism and the Consequences
of Imperialism on the Development of a Peripheral Province’, in D. Miles (ed.), The Romano-​
British Countryside:  Studies in Rural Settlement and Economy. BAR British Series 103.
Oxford: Archaeopress, 17–​52.
Hingley, R. (1989). Rural Settlement in Roman Britain. London: Seaby.
 Rural Transformation in the Urbanized Landscape    739

James, S. (2001). ‘“Romanization” and the Peoples of Britain’, in S. Keay and N. Terrenato (eds),
Italy and the West: Comparative Issues in Romanization. Oxford: Oxbow Books, 187–​209.
Jongman, W. (2007). ‘The Early Roman Empire:  Consumption’, in W. Scheidel, I. Morris,
and R. Saller (eds), The Cambridge Economic History of the Greco-​ Roman World.
Cambridge: Cambridge University Press, 592–​618.
King, A. C. (1978). ‘A Comparative Survey of Bone Assemblages from Roman Sites in Britain’,
Institute of Archaeology Bulletin, 15: 207–​232.
King, A. C. (1984). ‘Animal Bones and the Dietary Identity of Military and Civilian Groups
in Roman Britain, Gaul and Germany’, in T. F. C. Blagg and A. C. King (eds), Military and
Civilian in Roman Britain. BAR International Series 136. Oxford: Archaeopress, 187–​217.
King, A. C. (1999). ‘Diet in the Roman World:  A  Regional Inter-​Site Comparison of the
Mammal Bones’, Journal of Roman Archaeology, 12: 168–​202.
King, A. C. (2001). ‘The Romanization of Diet in the Western Empire’, in S. Keay and N.
Terrenato (eds), Italy and the West: Comparative Issues in Romanization. Oxford: Oxbow
Books, 210–​223.
Lockyear, K. (2000). ‘Site Finds in Roman Britain:  A  Comparison of Techniques’, Oxford
Journal of Archaeology, 19: 397–​423.
Maltby, M. (1989). ‘Urban–​Rural Variations in the Butchering of Cattle in Romano-​British
Hampshire’, in D. Serjeantson and T. Waldron (eds), Diet and Crafts in Towns. BAR British
Series 199. Oxford: Archaeopress, 75–​106.
Maltby, M. (1994). ‘The Meat Supply in Roman Dorchester and Winchester’, in A. Hall and
H. Kenward (eds), Urban–​Rural Connexions: Perspectives from Environmental Archaeology.
Oxford: Oxbow Books, 85–​102.
Maltby, M. (2007). ‘Chop and Change:  Specialist Cattle Carcass Processing in Roman
Britain’, in B. Croxford, N. Ray, R. E. Roth, and N. White (eds), TRAC 2006: Proceedings
of the Sixteenth Annual Theoretical Roman Archaeology Conference. Oxford:  Oxbow
Books, 59–​76.
Mattingly, D. J. (1997) (ed). Dialogues in Roman Imperialism. JRA Supplementary Series 23.
Portsmouth, RI: Journal of Roman Archaeology.
Mattingly, D. J. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of
Roman Archaeology, 17: 5–​25.
Mattingly, D. J. (2006). An Imperial Possession: Britain in the Roman Empire. London: Penguin.
Millett, M. (1979). ‘The Dating of Farnham Pottery’, Britannia, 10: 121–​138.
Millett, M. (1982). ‘Town and Country: A Review of Some Material Evidence’, in D. Miles (ed.)
The Romano-​British Countryside:  Studies in Rural Settlement and Economy. BAR British
Series 103. Oxford: Archaeopress, 421–​431.
Millett, M. (1990a). The Romanization of Britain. Cambridge: Cambridge University Press.
Millett, M. (1990b). ‘Romanization:  Historical Issues and Archaeological Interpretation’, in
T. F. C. Blagg and M. Millett (eds), The Early Roman Empire in the West. Oxford: Oxbow
Books, 35–​41.
Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium.
London: Society for the Promotion of Roman Studies.
Perring, D. (2002). The Roman House in Britain. London: Routledge.
Perring, D., and Pitts, M. (2013). Alien Cities: Consumption and the Origins of Urbanism in
Roman Britain. Spoilheap Publications Monograph 7. London: Spoilheap Publications.
Pitts, M. (2005). ‘Pots and Pits: Drinking and Deposition in Late Iron Age South-​East Britain’,
Oxford Journal of Archaeology, 24: 143–​161.
740

740   Martin Pitts

Pitts, M. (2007). ‘Consumption, Deposition and Social Practice: A Ceramic Approach to Intra-​

Site Analysis in Late Iron Age to Roman Britain’, Internet Archaeology.
Pitts, M. (2008). ‘Globalizing the Local in Roman Britain: An Anthropological Approach to
Social Change’, Journal of Anthropological Archaeology, 27: 493–​506.
Pitts, M. (2010). ‘Artefact Suites and Social Practice:  An Integrated Approach to Roman
Provincial Finds Assemblages’, Facta:  A  Journal of Roman Material Culture Studies,
4: 125–​152.
Pitts, M., and Griffin, R. (2012). ‘Exploring Health and Social Well-​Being in Late Roman
Britain: An Inter-​Cemetery Approach’, American Journal of Archaeology, 116: 253–​276.
Redfern, R., and Roberts, C. (2005). ‘Health in Romano-​ British Urban
Communities: Reflections from the Cemeteries’, in D. Smith, M. Brickley, and W. Smith
(eds), Fertile Ground: Papers in Honour of Susan Limbrey, Symposia of the Association of
Environmental Archaeology. Oxford: Oxbow Books, 115–​129.
Reece, R. (1993). ‘British Sites and their Roman Coins’, Antiquity, 67: 863–​869.
Richmond, I. (1969). ‘The Romano-​British Countryside’, in P. Salway (ed.), Roman Archaeology
and Art: Essays and Studies by Sir Ian Richmond. London: Faber and Faber, 133–​180.
Rivet, A. L. F. (1964). Town and Country in Roman Britain. London: Hutchinson.
Said, E. (1993). Culture and Imperialism. London: Chatto and Windus.
Scobie, A. (1986). ‘Slums, Sanitation, and Mortality in the Roman World’, Klio, 68: 399–​433.
Swan, V. G. (1984). The Pottery Kilns of Roman Britain. London: Royal Commission on
Historical Monuments.
Van der Veen, M., Livarda, A., and Hill, A. (2008). ‘New Plant Foods in Roman
Britain: Dispersal and Social Access’, Environmental Archaeology, 13: 11–​36.
Wacher, J. (1995). The Towns of Roman Britain. London: Routledge.
Wallerstein, I. (1974). The Modern World System I. Capitalist Agriculture and the Origins of the
European World-​Economy in the Sixteenth Century. London: Academic Press.
Webster, J. (1997). ‘A Negotiated Syncretism: Readings on the Development of Romano-​Celtic
Religion’, in D. J. Mattingly (ed.), Dialogues in Roman Imperialism. Portsmouth, RI: JRA
Supplementary Series 23. Portsmouth, RI: Journal of Roman Archaeology, 165–​184.
Webster, J. (2001). ‘Creolizing the Roman Provinces’, American Journal of Archaeology,
105: 209–​225.
Webster, J., and Cooper, N. (1996) (eds). Roman Imperialism:  Post-​Colonial Perspectives.
Leicester Archaeology Monographs 3. Leicester: Leicester Archaeology.
Wilkinson, R. G. (2005). The Impact of Inequality:  How to Make Sick Societies Healthier.
London: Routledge.
Wilkinson, R. G., and Pickett, K. (2010). The Spirit Level: Why Equality is Better for Everyone.
London: Penguin.
Willis, S. (1996). ‘The Romanization of Pottery Assemblages in the East and North-​East of
England during the First Century ad: A Comparative Analysis’, Britannia, 27: 179–​221.
Willis, S. (2007). ‘Roman Towns, Roman Landscapes:  The Cultural Terrain of Town and
Country in the Roman Period’, in A. Fleming and R. Hingley (eds), Prehistoric and Roman
Landscapes: Landscape History after Hoskins Volume 1. Macclesfield: Windgather Press, 143–​164.
Willis, S. (2011). ‘Samian Ware and Society in Roman Britain and Beyond’, Britannia,
42: 167–​242.
Wood, J. W., Milner, G. R., Harpending, H. C., and Weiss, K. M. (1992). ‘The Osteological
Paradox:  Problems of Inferring Prehistoric Health from Skeletal Samples’, Current
Anthropology, 33: 343–​370.
 Chapter 36

The Devel opme nt

of Tow ns

Adam Rogers

Introduction

Towns and urban development have formed a major part of the history of studies of
Roman Britain, and some of the towns are among the most heavily investigated Roman
settlements. For centuries Roman material has been encountered within living cities,
and, in some instances, there are Roman remains still standing. The nineteenth century
saw the commencement of some large-​scale excavation programmes, especially at town
sites unoccupied by later settlement, including Silchester, Caerwent, Verulamium, and
Wroxeter. The investigation of these green-​field sites was instrumental in the forma-
tion of models for urban development, but interpretations were also influenced by the
cultural context and social preoccupations of the day (cf. Hingley 2008). The post-​war
years saw a massive increase in knowledge relating to those towns now lying beneath
modern settlements through urban redevelopments and the formation of urban archae-
ological units. Much of this work was collated for the first time in Wacher’s important
narrative (1975) on the Roman towns of Britain. There are now many important stud-
ies and syntheses of urban development taking different perspectives (e.g. Millett 1990;
Wacher 1995; M. J. Jones 2004; Hingley 2005; Creighton 2006; Mattingly 2006; Rogers
2008), but our knowledge and understanding of town development remains open to
constant revision as new discoveries are made and analyses are undertaken.
Especially important in more recent years has been the development of theoreti-
cal approaches in Roman archaeology and the potential that this has for studying set-
tlement. Towns were new to Britain in the Roman period, although there were also a
range of large settlements and complexes in prehistory. There has, however, tended to
be a divide between the approaches taken by prehistorians and Romanists in address-
ing social themes relating to processes of settlement development, their relationship
to human identities and experiences, and other issues. While hillforts were important
742

742   Adam Rogers

for much of the Iron Age, in the second and first centuries bc another group of sites
appeared—​often referred to collectively as oppida and conventionally divided into
‘enclosed’ and ‘territorial’ oppida based on the nature of the earthworks evident (e.g.
Haselgrove 1999). Some Roman towns were located in or near oppida, but there is still
considerable uncertainty as to their real nature and function and whether we can think
of oppida as a clear category of site or whether there was more variety and variation of
sites and circumstances than has so far been fully appreciated (cf. Haselgrove and Moore
2007). While it would be unwise to attempt to apply Classical or later urban models to
these sites, moreover, including in debating the role of the earthworks, it might be pos-
sible to think of them as taking another form of trajectory in urban development (cf.
Sharples 2010: 173); oppida may also have had more impact on the urban geographies of
Roman Britain than is often acknowledged, and it is important for Roman archaeology
to draw on landscape theory. As such it is important for studies of Roman urban devel-
opment to engage in theoretical debates relating to landscape archaeology as well as the-
ories concerning interconnected issues including identity and experience. The divide
between Roman and Iron Age is as much about archaeological perspectives as historical
events, and theoretical frameworks can help us to adopt more nuanced approaches to
the transition from Iron Age to Roman Britain.

Categorizing Towns: Regionality

Central to archaeology have been attempts to categorize the past:  be it by periods,

objects, or settlements. Romano-​British studies have created lists of characteristics of
what defines certain types of settlements, drawing on knowledge of archaeology, tex-
tual sources, and the Roman Empire as a whole. Inevitably, however, these attempts will
also have imposed modern perspectives and preoccupations onto the past, and there
is always a danger in oversimplifying our understanding of settlements through these
definitions. The towns of Roman Britain are usually defined by their planned and organ-
ized appearance, their size, the presence of a variety of buildings, a concentration of
population, and the administrative and economic roles that they had (e.g. Mann 1996;
Burnham et al. 2001). But it is undoubtedly the legal categorizations known through
written records that have been most instrumental in defining urbanism and urban
development in Roman Britain. It is important to recognize, however, that there may
well also have been other forms of urbanism in Britain at this time, alongside the towns
included in more conventional studies.
It is known that there was a legal hierarchy of urban settlements, with coloniae at the
top, municipia, and then civitas-​capitals or centres. The colonia was a chartered town, by
the first century ad generally founded for the settlement of discharged veterans, and it
had a constitution that was modelled on Rome and adopted Roman law. In Britain, the
three initial colonia were founded at Colchester (Camulodunum), Gloucester (Glevum),
and Lincoln (Lindum), on the sites of legionary fortresses that have traditionally
 The Development of Towns    743

formed important parts of the narrative of the conquest and occupation of the prov-
ince (e.g. Frere 1967; Wacher 1975). Legio XX is conventionally assigned to the fortress
at Colchester, founded around ad 43–​4 and replaced by the colonia from around ad 49.
In the ad 60s Legio IX Hispania was assigned to Lincoln (used later by Legio II Adjutrix)
and another, as yet undetermined, legion to Gloucester, although Legio II Augusta is
thought to have occupied the fortress at some stage (Wacher 1995: 150). While the dates
have been the subject of much debate, the colonia at Lincoln was probably founded in
the ad 80s (M. J. Jones 2002) and at Gloucester in the 90s (Copeland 2011). Even this
most familiar of narratives of development, however, is not safe from reassessment, not
least because so little is actually known about the composition of the army in Britain at
any one time or of its movements (Mattingly 2006: 130). At Lincoln, for example, the
possibility of an earlier Claudian military base has been raised, which is based on some
evidence for an early military cemetery (M. J. Jones 2003a: 38–​39), but until more posi-
tive structural evidence comes to light this must remain a subject for debate.
The coloniae in Britain have generally been treated as Classical-​style settlements with
Roman populations; they have been described as ‘bulwarks of loyalty’ (Collingwood and
Richmond 1969: 95) and ‘instrument(s) of civilisation’ (Richmond 1963: 55). Many of the
inhabitants will have been Roman citizens, but there will also have been other inhabit-
ants, especially local indigenous people, who would not have been citizens, indicating a
mix of people and perspectives with social implications. Below the colonia in the hier-
archy was the municipium, which was also a chartered town, but far fewer inhabitants
automatically received citizenship, with only the ex-​magistrates usually having the right
to acquire citizen status. Verulamium was described by Tacitus (Ann. XIV.33) as a muni-
cipium, but exactly when, and even if, the town actually acquired this status and the rea-
sons behind it are uncertain; it may well have been an upgrade from civitas-​capital made
in the ad 70s (Niblett 2001: 66). York (Eboracum) was described by the fourth-​century
writer Aurelius Victor as a municipium at the time of Septimius Severus’ visit in the early
third century (Aur. Vict. Caes. XX; Ottaway 2004: 83). It is not entirely clear how much
value can be placed on the terms used in these accounts, since they may simply have
been used for convenience or individual preference. However, an inscription on an altar
dating to ad 237 (AE 116 [1922]), set up in Bordeaux by Marcus Aurelius Lunaris for
the goddess Tutela Boudiga, indicates that by this date at least York had become a colo-
nia. He described himself as a sevir Augustalis (priest of the deified emperor) from the
coloniae of both Lincoln and York. It is possible in the case of York that the settlement
was later granted the status of colonia by the emperor, perhaps after petitioning for it, as
occurred elsewhere (cf. Boatwright 2000), and was not founded in the same sense as the
earlier coloniae in Britain.
The largest group in the urban hierarchy was what are usually termed civitas-​capitals
or civitas centres and regarded as the centres of the civitates: the means by which land
and people were divided up for administration. Although lower in status, the consti-
tution, like the other towns, was also ultimately modelled on Rome, with laws, annual
magistracies, and a town council. We still know frustratingly little about the nature of
town councils or other forms of organization and management within the towns of
744

744   Adam Rogers

Roman Britain. It seems likely that the circumstances differed within each town depend-
ing on local peoples, histories, and the nature of the urban development. The province
was large, and, because of this, the regions, local conditions, and peoples are important
factors to consider in the development of towns and the meanings that will have been
attached to them as places. It is the recorded names of the civitas-​capitals that have con-
ventionally been taken as evidence of processes of negotiation with local people as civi-
tates and towns were formed: for example (Figure 36.1), Silchester (Calleva Atrebatum);
Winchester (Venta Belgarum); Caistor-​by-​Norwich (Venta Icenorum); Leicester (Ratae
Corieltauvorum); Cirencester (Corinium Dobunnorum); Dorchester (Durnovaria);
Exeter (Isca Dumnoniorum); Wroxeter (Viroconium Cornoviorum). The existing ‘tribal’

Coria
Luguvalium
Carvetiorum

Isurium
Brigantum

Cataractonium
Eburacum

Petuaria
Parisiorum

Colonia
Domitiana
Lindensium

Ratae Corieltauvorum
Viroconium Venta
Cornoviorum Durobrivae Icenorum
Colonia Claudia
Victricensis
Colonia Nervia Camulodunensium
Moridunum Demetarum Glevensium
Caesaromagus
Corinium
Venta Silurum Dodunnorum Verulamium
Calleva Atrebatum Londinium
Durovernum
Venta Belgarum Cantiacorum
Lindinis Durotrigum Noviomagus
Durnovaria Reg(i)norum
Isca Durotrigum
Dumnoniorum

Figure 36.1  Map of southern Britain with the main settlements mentioned in the chapter.
Source: © A. C. Rogers.
 The Development of Towns    745

organization of Britain prior to the conquest, however, remains a focus of debate,

since the main evidence comes from texts written by Roman authors and inscriptions
that date to after ad 43. It seems likely that Rome had an influence on the creation of
these entities either before or after ad 43, but the nature of this remains uncertain (e.g.
Moore 2006, 2011; Sharples 2010). Group identities and allegiances may well have been
more fluid in the late Iron Age, perhaps based on individuals’ fluctuating capabilities
of encouraging the support of others, and more localized settlement and household
identities will also have been important. Ongoing work at the Bagendon/​Ditches com-
plex north of Cirencester is certainly indicating that the negotiations for power and the
beginnings of urban development at Corinium are likely to have been more complex
than traditional models of tribal oppidum to town (Moore 2011). In the second century
further towns appear to have been founded: Caerwent (Venta Silurum), Carmarthen
(Moridunum Demetarum), and the less-​understood settlements of Brough-​on-​Humber
(Petuaria Parisiorum) and Aldborough (Isurium Brigantum). These remind us of the
complex ongoing developments and changes that took place in Roman Britain, although
the nature of the processes and negotiations that may have taken place with their foun-
dation remain obscure to us.
Settlement size has also been used to categorize towns and also to comment on the
perceived success of Roman urbanism in Britain. The towns of Roman Britain are
generally fairly small in size, with many of the main enclosed areas falling to around
40 hectares, including Canterbury, Leicester, Exeter, Silchester, and Lincoln. Larger
towns included Cirencester at around 97 hectares, Verulamium at 80 hectares, and
London at 133 hectares; the enclosed area at Gloucester, however, was only around
16.6 hectares. Settlement, however, will also have extended beyond the enclosures (cf.
Esmonde Cleary 1987). Another category of settlement in Romano-​British archae-
ology is what is generally referred to as the ‘small town’ (e.g. Burnham and Wacher
1990), although these are not always any smaller than the civitas-​capitals or coloniae.
Although it seems clear that they did not have the same legal status (at least in terms
of that imposed by the empire) within the province, they may have had a more promi-
nent role within the civitates than has generally been acknowledged (cf. Laurence
2001; Hingley 2005). Mattingly (2006: 281) has suggested that the territories assigned
to the civitas-​capitals may have been smaller than is conventionally attributed to them,
although there remains frustratingly scant evidence to support arguments either way.
Since it is likely that all land was attributed to some kind of authority for the purposes
of tax-​gathering, it would mean that other forms of settlements and landownership
with the civitates were also important. These issues remain debatable and highlight
the constant need for discussion without taking the evidence for granted. Hingley
(1997) has also argued that ‘small towns’ may reflect more local indigenous interpre-
tations and expressions of urbanism, perhaps not unlike how we might be able to
interpret some of the oppida. This might help to explain the apparent diverse range
of contexts of ‘small-​town’ development and function and reminds us of the possibil-
ity of different forms of urbanism in the Roman period. The debate is a reminder of
the necessity to avoid the ‘one-​model-​fits-​all’ approach to urban development and
746

746   Adam Rogers

instead examine the complex processes and negotiations that must have occurred in the
local context of each site.
As noted for Verulamium and York, and possibly also London (cf. Tomlin 2006),
urban status could change. It has been argued that some ‘small towns’ possibly sought
and were granted status as civitas-​capitals in the later Roman period, including Water
Newton (Durobrivae) and Ilchester (possibly the Lindinis in the Ravenna Cosmography;
or the Civitas Durotrigum Lindinesis on RIB 1672 and 1673 from Cawfields on Hadrian’s
Wall) (Leach 1982; Fincham 2004; Putnam 2007; cf. Fulford 2006). These arguments
are based principally on settlement size and the buildings they contained, as there is no
textual evidence for these changes. The town at Chelmsford in Essex (Caesaromagus) is
often included in discussions of civitas-​capitals (e.g. Wacher 1995), but in other accounts
it is treated as a ‘small town’ (e.g. Mattingly 2006: 259). Wacher (1995: 207) suggested
that it may have been a largely unsuccessful civitas-​capital, because of his belief that the
colonia at Colchester would not have controlled the civitas. It might be that attempting
to fit the settlement within the category of either civitas-​capital or ‘small town’ is inad-
equate, especially since there is little consensus as to what we mean by ‘small town’ and
how these settlements functioned. In the north the military establishment of Carlisle
(Luguvalium Carvetiorum) may have become a civitas-​capital in the third century
(Burnham and Wacher 1990: 58; M. J. Jones 2004: 174; Mattingly 2006: 261).

Town Development and Texts

One of the earliest forms of text relating to places in Britain are the names incorporated
into Iron Age coin designs, which have long been associated with the oppida known in
the vicinity of some of the Roman towns (cf. Hingley 2008). Coin distributions have been
studied to assess ownership of oppida, and the work of antiquarians has been instrumen-
tal in our recognition of Iron Age and Roman sites and the development of concepts of
‘Romanization’ and ‘civilization’. The name Calleva (Silchester) first appears on the coin-
age of Eppillus dating to the beginning of the first century ad (Creighton 2006: 64), the
self-​proclaimed son of Commius (who is mentioned in Caesar’s Bellum Gallicum IV.21).
The coin distribution then changes to those that refer to Epatticcus, son of Tasciovanus
and brother of Cunobelin (Creighton 2006: 64), which is often taken to imply some kind
of violent conquest of the area or a dynastic union, perhaps through marriage (ibid.).
Verulamium is also first named on coinage, with the issues of Tasciovanus inscribed
with VER, and Camulodunum on the coinage of Cunobelin as CAMV (Potter 2002: 21;
Williams 2005: 33) (see also Walton and Moorhead, this volume, for further discussion
on coinage). The names Verlamion and Camulodunon are sometimes suggested for the
Iron Age period settlements in order to distinguish them from the Roman towns (Potter
2002: 21). There is no definite evidence for the exact form of these names, but it seems
likely that they did differ in the Iron Age. Cassius Dio mentions that Camulodunum was
Cunobelin’s capital in his Histories (LX.21), but caution is needed as to whether he would
 The Development of Towns    747

have understood the social context of this place at the time he was writing about (cf.
Haselgrove 2004; Haselgrove and Moore 2007; Sharples 2010).
There are few surviving texts that mention the towns of Roman Britain, and there is
very little in the way of any descriptions of the towns in the sources that do exist. The
narrative of the Boudiccan Revolt of ad 60/​61 refers to the towns of Camulodunum,
Verulamium, and Londinium (Tac. Ann. XXXI–​XXXIII) and, in so doing, provides a
brief comment on Camulodunum with its statue of Victory, a curia, a theatre, and a tem-
ple where much of the population sought safety during the revolt, but which was ulti-
mately destroyed. Tacitus also recorded that there was no ditch or rampart surrounding
the town at this time, which made it easier to attack. As part of the same narrative, he
mentions that Londinium was at that time full of merchants and trading vessels (Ann.
XXXIII), although he did not see this for himself and was writing at a later date, the
Annales probably being written in the early second century ad.
There are also other forms of texts that refer to the towns of Roman Britain but pro-
vide little or no information as the nature of the settlements. Ptolemy’s Geographia,
probably written in the ad 140s, consists of a list of place names and other geographi-
cal features from across the empire, which had probably been drawn from a number
of sources from different periods. Another collection of place names is the Ravenna
Cosmography compiled around ad 700 and using a number of maps that are now lost.
Despite the sparse details, its late date, and the many mistakes and ambiguities, perhaps
through copying errors, it has been a hugely influential text in Romano-​British stud-
ies, since it lists the colonia and assigns tribal names to the civitas-​capitals. It was used
by Haverfield (1912: 60–​61) and others to create their reconstructions of Roman urban
development in Britain with the civitas system and the processes of conquest, tribal
domination, and Romanization. There is also the Antonine Itinerary, long studied in
relation to Britain (Hingley 2008), probably compiled in the early third century. It con-
sists of a collection of itineraries or routes across the provinces, with fifteen relating to
Britain and mentioning over 100 place names. Many of these names can be recognized,
and have also been identified archaeologically, but there are also a number that remain
enigmatic. While these sources are important, as lists they provide no information on
the individual circumstances surrounding the creation of each settlement or the lives of
the people involved.

Debating Urbanization
and Romanization

Many antiquarians and scholars have been important in developing our understand-
ing of Roman urban development, but it could be argued that Francis Haverfield was
one of the first in the context of modern archaeological scholarship (see also Millett and
Swift, this volume). His The Romanization of Roman Britain (1912) described the towns
748

748   Adam Rogers

of Roman Britain in terms of the ‘five municipalities of the privileged Italian type’ and
their dependent territories (Colchester, Verulamium, Lincoln, Gloucester, and York)
and what he termed the cantonal system, which was one of the important ‘influences
which aided the Romanization of the country’ and ‘permitted the complete remodel-
ling of the native institution by the interpenetration of Italian influences’ (Haverfield
1912: 58). Though he did recognize that urbanization would have been ‘a subtler and
more complex process than mere absorption in Rome’ (Haverfield 1912: 64), with towns
containing native as well as Roman elements, his arguments did focus principally in the
Roman perspective with an emphasis on Romanization: ‘Britain shared in that expan-
sion of town-​life which formed a special achievement of the Roman Empire’ (Haverfield
1912: 64). Frere’s analysis (1967) of towns also drew on the Classical perspective and
argued that urbanism was not particularly successful: Britain ‘never produced towns
of fully Mediterranean character’. Frere’s analysis, however, considered Roman towns
to have been important for the ‘spread of education and Roman civilisation’ (Frere
1987: 192) and the preceding oppida as no more than ‘amorphous collection(s) of round-
houses and unorganised squalor’ (Frere 1987: 192). He also thought of Camulodunum
and Verulamium as acting as models for urban development where buildings, town
life, and organization could be copied by local peoples in other areas of Britain, and this
meant that these two towns received government assistance in their construction.
Wacher’s monumental The Towns of Roman Britain (1975) was keen to place an
emphasis on the observed pattern of urban development as the result of a well-​planned
programme of conquest and expansion. In his book, the towns were divided into a series
of chronological chapters entitled ‘The Coloniae’ (Colchester, Lincoln, Gloucester, and
York); then ‘The First Civitas Capitals’ (Canterbury, Chelmsford, and Verulamium);
‘From Client Kingdom to Civitas’ (Caistor-​by-​Norwich, Chichester, Silchester, and
Winchester); ‘Flavian Expansion’ (Cirencester, Dorchester, Exeter, Leicester, and
Wroxeter); and finally ‘Hadrianic Stimulation’ (Caerwent, Carmarthen, Brough-​on-​
Humber, and Aldborough). This was the predominant model of the time, which placed
an emphasis on the foundation of towns when forts were abandoned and the army
moved on (e.g. G. Webster 1966). The creation of towns was predominantly the result
of military strategy, even though at not every town is there evidence of any preceding
military settlement.
Millett’s The Romanization of Britain (1990) can be seen as an important work of tran-
sition in Roman urban studies that sought to place more significance and distinction
between the towns with definite military origins and those with indigenous involve-
ment, emphasizing, in particular, the role of the oppida in the development of the civitas
system. For Millett, the four coloniae and London had an essentially anomalous position
within the settlement system, with the three early coloniae representing deliberate acts
of policy by the Roman administration. The civitas system, on the other hand, devel-
oped with the aid of the local elite of tribal groups. He constructed a detailed argument
contrasting military and native origins of urbanism, but, as the title of the book indi-
cates, his work was not a critique of Romanization but rather a reanalysis of the way in
which it took place.
 The Development of Towns    749

More recent studies are suggesting that the distinction between military and indig-
enous origins, moreover, was probably in fact more blurred (cf. Hingley 2005:  271).
Mattingly (2006) has argued against there having been an overall plan in urban devel-
opment in Britain; instead he sees the current picture of urban development as one that
resulted from a variety of negotiations between Rome and the peoples of Britain that
contributed to the complex individual histories of each settlement. But he did not deny
Rome’s power to make decisions:  for Mattingly, the pre-​Roman political geography
would not necessarily have been of primary importance; instead what was significant
was the resultant circumstances, as people may have surrendered, made alliances, or
were subdued. The need to create a new communications network may also have been
considered a priority, and settlements were able to develop that were close to this net-
work. Placing more emphasis on pre-​conquest activity, Creighton’s work (2006) has
provided an important reanalysis of the chronological sequence of a number of towns,
arguing for the importance of the relationship between Rome and local elites vying for
power and the development of some of these places before ad 43. The observed devel-
opments at these places, however, were not simply about Romanization, since the elites
were able to draw on Roman elements and adapt them to emphasize their own power
and achievements. Our understanding of urban development, then, is clearly influenced
by what perspective is taken, and so it would, perhaps, be most productive to acknowl-
edge the actions and viewpoints of the full range of people associated with each town.
Any study of Roman urbanism, then, must tie into debates relating to ‘Romanization’
(Haverfield 1912; Millett 1990; Hingley 2000; Mattingly 2006). Increasingly, alternative
terms and ways of understanding the processes of change, continuity, and development
that took place in Britain during the Roman period are being sought (e.g. J. Webster
2001; Mattingly 2004, 2006; Hingley 2005) and the historiographical context of the
term scrutinized (e.g. Hingley 2000, 2008). Individual negotiations with the presence
of Rome indicate that change and continuity were far from uniform processes across
Britain, and these non-​uniformities must take into account not only elite but also non-​
elite perspectives. There were peoples who did not wish to engage in the Roman pres-
ence, and they may have been either encouraged or forced to live in a town or left more
to their own devices in the countryside. There is a tendency in studies of the develop-
ment of Roman urbanism in Britain to focus on the Roman perspective or that of the
indigenous elite, but other perspectives of these monumental processes taking place
within the landscape must also be considered.

Town Location and Place

Most descriptions of Roman town development in Britain include a discussion of the

locations of the settlements, including the strategic, economic, and practical benefits
of these sites. Drawing on the empiricist traditions of landscape archaeology, these
descriptions have tended to emphasize the practical value and rationality of landscapes
750

750   Adam Rogers

relating to urban development. At Colchester, for example, the fortress was in a well-​
chosen location occupying a ridge of land overlooking the Colne Valley and (on a neu-
tral site) within the oppidum, because it was believed to have been placed away from any
existing settlement (e.g. Wacher 1995; Crummy 1997); Roman Gloucester ‘overlooks the
lowest point where the Severn could be bridged until the Industrial Revolution’ and was
the main route into south Wales so was of ‘key strategic importance’ (Hurst 1988: 48).
These practical considerations are important, but they provide an external perspective
and give little or no consideration to the social value of these locations as places and their
longer-​term biographies as foci of human actions, histories, experiences, emotions, and
stories. Landscapes were not simply functional, rational, and mundane: ‘natural’ fea-
tures in the landscape, for example, could have cultural values associated with them and
form just as important elements of the settlement record as earthworks and buildings
(cf. Rogers 2008). More emphasis can be given to the meanings associated with the loca-
tions in which towns were sited through a study of the history of these places and the
cultural values that could be given to features in that landscape. Regardless of the politi-
cal geography created in the Roman period, meanings associated with specific places
will also have been drawing on the existing significance associated with them.
Known place names may be able to provide some initial hints as to the possible exist-
ing meanings attached to some of the places where the towns were founded and the way
in which this had an impact on the urban experience. The name Calleva (Silchester),
for example, has been interpreted as meaning ‘(the settlement in the) woods’ (Rivet and
Smith 1979: 291). It is possible that the oppidum developed here through clearing a space
in the woodland or there was perhaps even a sacred grove here. The name Durovernum
(Canterbury) appears to relate to wet conditions here, with ‘verno-​’ meaning alder
swamp or marsh, suggesting that this location was also culturally meaningful (Rivet and
Smith 1979: 53). The meaning of Corinium (Cirencester), though not entirely clear, may
relate to a botanical feature also in a wetland area, while Isca (Exeter) is derived from the
river (Rivet and Smith 1979: 353, 376). These features were not just mundane points of
identification but will have been imbued with meanings and cultural values.
Studies of the coloniae have tended to focus on the fortress or town itself rather than
their wider setting. The colonia at Colchester was placed within a ‘territorial’ oppidum
consisting of a number of dykes, enclosures, and other foci (Hawkes and Crummy 1995).
Although the town developed on the site of the fortress, there were other foci connected
with the oppidum that continued to be important and formed part of the urban land-
scape, including what became a temple and theatre complex at Gosbecks, where there
was already a number of enclosures, including what is often termed ‘Cunobelin’s farm-
stead’, and a foci of temples at Sheepen (Crummy 1984, 1997; Niblett 1985). The Roman-​
period urban landscape drew upon and became part of the pre-​existing geography.
There has also been one tentative suggestion that the Iron Age coins and pottery found
on excavations within the town may indicate some earlier form of settlement or activity
here, perhaps even a religious sanctuary (Brooks 2006).
The coloniae not located near or within obvious oppida will also have been placed
within landscapes of pre-​existing use and meaning. The Roman fortress at Lincoln
 The Development of Towns    751

overlooked the valley where the river Witham met the river Till, and there was a large
natural body of water known as the Brayford Pool where the two rivers met. There were
a number of causeways across the Witham valley in prehistory, which encouraged ritual
activity, including at Fiskerton (Field and Parker Pearson 2003) and at Stamp End (Jones
and Stocker 2003), where the famous Iron Age Witham shield and sword were found,
having been deposited in the water. There also appears to have been late Iron Age and
early Roman activity on what was once an island within the Brayford Pool (Steane 2001;
Jones and Stocker 2003), as well as some traces of Iron Age activity beneath the for-
tress itself (Jones and Stocker 2003: 28–​30). To neglect the importance of this landscape
before the fortress was constructed will have an impact on the way in which we under-
stand the town and the urban experience here.
The civitas-​capitals are often considered according to whether there was a fort
(Exeter, Wroxeter, and Cirencester) or oppidum (Verulamium, Winchester, Silchester)
preceding the development of the town, but, in some cases, the forts are not proven (for
example, Cirencester), and in no cases is the Iron Age activity yet fully understood. At
Leicester there is also growing evidence being documented of Iron Age activity concen-
trating near the river Soar beneath the later town, and this may have been some kind
of oppidum or other settlement, including round-houses and evidence of coin manu-
facture (Clay and Mellor 1985; Kipling et al. 2007). It is knowledge of activity preceding
towns, and the understanding of these places that it brings, that is most likely to change
through archaeological investigation, as ongoing excavation and survey work at and
around Silchester, Cirencester, Aldborough, London, and other towns are indicating.
Work at Caistor-​by-​Norwich, for example, has revealed considerable traces of activity
beneath the town, including gullies and ring ditches suggestive of round-houses and
other features (Bowden and Bescoby 2008).
Structural or activity remains need not be the only indications of the importance of
places in the past. Rivers, lakes, woods, and hills were not mundane but integral parts
of the cultural landscape, and they will also have had an impact on the nature of urban
development and the urban experience (cf. Rogers 2011a, b, 2013). For example, the loca-
tion of London at the highest point where the river was still tidal allowed the develop-
ment of important port facilities (Milne 1985). Apart from economic or strategic factors
relating to London’s location and development (cf. Perring 1991; Milne 1995; Wacher
1995), however, it is possible to analyse the landscape setting from additional perspec-
tives, which will also have formed an important part of the urban biography. Although
there does not appear to have been much settlement here in the late Iron Age, it does
not mean that there would not have been cultural values attached to this place, espe-
cially in relation to its watery setting and islands situated across the Thames in the vicin-
ity of Southwark (Cowan et al. 2009). There are some traces of late Iron Age activity
on the islands at Southwark, and the rivers and streams here appear to have invoked
religious activity, which continued during the Roman period, especially noticeable in
the Walbrook, which ran through the centre of the later town and the marshy area of
the Upper Walbrook valley (Bradley and Gordon 1988; Merrifield 1995; Butler 2006;
Merrifield and Hall 2008). Towns could be disruptive, even built over pre-​existing
752

752   Adam Rogers

features, but also provide continuities of place by being incorporated into the existing
geographies and meanings in the landscape.

Town Construction Processes

As well as being placed within landscapes of existing meaning, the foundation processes
of the towns themselves will not have been mundane. There are likely to have been foun-
dation rituals drawing on both indigenous and Roman traditions. They may not now be
easily recognized in the archaeological record, but excavations in the centre of Roman
Dorchester have discovered what appear to have been pits and shafts containing unu-
sual collections of objects and animal remains, which have been interpreted as repre-
senting ritualized deposits, perhaps even connected with the foundation of the town
itself (Woodward and Woodward 2004) (Figure 36.2). Certainly within the Roman tra-
dition there were rituals associated with urban foundation including taking the auspices
and the construction of an enclosure or templum (cf. Rykwert 1976); an interpretative
possibility which has also been raised for the evidence of an early enclosure beneath the
forum at Verulamium (Creighton 2000: 210). Towns were ritualized places through their
foundation, creation, definition of space, and the actions of people within and around
them. While the exact nature and interpretation of the pits and shafts at Dorchester
are still open to speculation, they remind us of the need to recognize activities that we
might not necessarily expect or assume in urban development or urban life in Roman
Britain, drawing on both local traditions and those introduced from elsewhere in the
empire. This includes the ritual and religious activities that took place in many differ-
ent contexts outside temples and shrines. It is also likely that there will also have been
origin myths drawing on the histories and stories that were attached to the landscapes
and places where towns developed and perhaps in some cases were created in order to
provide legitimacy for the towns in the landscape.
Where towns emerged on the sites of fortresses at Colchester, Lincoln, and Gloucester,
there is now good archaeological evidence of the physical processes involved in the con-
versions of these settlements, including the demolition of fortress structures, the con-
struction of new buildings, and also the reuse of others. Excavations within Gloucester
in the late 1960s and 1970s, especially at 13–​17 Berkeley Street, have uncovered remains
relating to the reuse of fortress buildings as well as the construction of new half-​
timbered and clay-​walled structures (Hurst 1988, 1999a, b). In some cases it appears
that the barrack blocks were demolished but no new buildings were constructed in their
place for some time. The main area of the town remained small, moreover, since it con-
tinued to lie within the original confines of the fortress, although there are remains of
settlement known beyond it (Heighway 1983; Heighway and Bryant 1999). Excavations
within Colchester have revealed a similar process of conversion, although here the town
did expand beyond the original area of the fortress in a more planned way. Excavations
at Lion Walk demonstrated that some of the barrack blocks were converted to other
 The Development of Towns    753

Durnovaria

River Fr N
ome
Forum

Site Baths

Limit of
excavation

Pits and
shafts

Timber
fencing

Enclosure?

Enclosure?

Stone footings
0 10 20 30
m

Figure 36.2  Plan of pits, shafts, and timber-​fenced enclosures excavated within the centre of
Roman Dorchester (Durnovaria) at the Greyhound Yard site, 1981–​4.
Source: adapted from Woodward and Woodward (2004: figure 1). © A. C. Rogers.

uses, while further buildings or parts of structures were demolished and the land put to
other uses, including cultivation (Crummy 1984, 1988, 1992) (Figure 36.3). These towns
were clearly not grand and monumental in the Classical style from the outset.
The first public buildings at Colchester were constructed in the area of the fortress
annexe outside the main fortress, probably because the fortress was too crowded, and
this may be the reason why the defences were levelled. This conversion and construc-
tion process, however, was not speedy and may have taken a number of years, since
excavations indicated that the fortress ditch had been neglected for some time before it
was levelled (Crummy 1988: 42); the Temple of Claudius, moreover, may well not have
been begun until Claudius’ death in ad 54, although this need not necessarily have been
the case (Drury 1984). Like Gloucester, the defences of the fortress were maintained at
754

754   Adam Rogers

River Col Duncan’s Gate

ne
N Possible site
of gate (St Peter’s
North Gate Street)

THEATRE
TEMPLE

East Gate

Balkerne Gate

Lion Walk
Culver excavations
Street
excavations
Possible site
South Gate of gate (Priory
Street)
Head Gate

Outline of fortress ditch

0 100 200 300 400

m

Figure 36.3  Plan of the relationship between the fortress and later colonia at Colchester; the
walls around the town were constructed in the early second century ad; the theatre and temple
were constructed within the fortress annexe.
Source: adapted from Crummy (1993: figure 2.9). © A. C. Rogers.

Lincoln, and there was probably a fairly similar process of conversion (M. J. Jones 1988,
2002, 2003b; Steane 2006). What differs here, however, is that there is more evidence
known regarding the settlement that developed outside the walls (often referred to as
the ‘Lower City’), and it appears to have been more formalized, eventually itself being
walled, running down the hill from the fortress to the river Witham (M. J. Jones 2002).
The development of the colonia at York was an exception in that the fortress appears
to have remained in use throughout the Roman period, and a civilian settlement, or
canabae, developed beside it. Across the river Ouse a much larger settlement developed,
which was eventually promoted to the status of colonia (Ottaway 1999, 2004). Nothing is
yet known about the forum/​basilica complex within the town; one suggestion is that the
fortress principia may have had this role, but it may still await discovery, since there have
not been many opportunities for excavation in the central area of the town.
Some of the civitas-​capitals appear to have had a fairly similar process of development
to that of the coloniae. At both Exeter and Wroxeter the towns were founded on the sites
of fortresses, and they adapted and expanded the street grids (Bidwell 1979; Henderson
1988; White and Barker 1998). Cirencester differed in there only having been a possi-
ble small fort (Leaholme) and little is known about its interior, if it did exist (Wacher
 The Development of Towns    755

and McWhirr 1982; McWhirr 1988; Holbrook 1998). The processes that sparked the
developments at Canterbury, Leicester, Caerwent, Caistor-​by-​Norwich, Dorchester,
Carmarthen, Aldborough, and Brough-​on-​Humber are less well understood. While
there is likely to have been a military presence at Carmarthen and Brough-​on-​Humber,
more archaeological investigation is needed at these towns (Wacher 1969; James 2003).
The possibility of a military base at the other towns is often raised, although there is,
so far, no conclusive proof; significant pre-​Roman activity, however, is known at
Canterbury, Leicester, and Caistor-​by-​Norwich, which will have formed a significant
element of these urban biographies.
It has often been the conventional argument in discussing urban development that
there was direct involvement from Rome in urban planning and construction, includ-
ing through the utilization of the military. One of the difficulties is that there are very
few inscriptions surviving from towns in Roman Britain to provide information as to
who funded and built the urban infrastructure. Frere (1967) emphasized the role of the
military and drew on his own excavations at Verulamium, where he identified an early
(c. ad 49) rectangular timber-​built row of shops in insula XIV (Frere 1972, 1983). He
argued that they were similar in plan and appearance to military barrack blocks, and
so they must have been designed by military architects, built by craftsmen lent by the
government, and military stockpiles would have been used in their construction. Millett
(1990: 70–​7 1), however, argued that the comparison with barrack blocks is only superfi-
cial and that there was no evidence that the row was planned and roofed as a single unit.
He also emphasized the fact that there will have been sophisticated technology already
available in Britain, along with established sources of timber supplies, which would have
made such construction possible. This evidence, therefore, cannot be used to support
the emphasis on official involvement and assistance in town construction.
Creighton’s work (2006) has refocused attention to the evidence and role of phases of
activity that pre-​dated ad 43. At Silchester, for example, there is evidence of early elite
activity but which was probably influenced by Rome or the near continent. Excavations
on the site of the basilica have revealed a number of phases of activity that pre-​date the
foundation of the town. The earliest evidence related to traces of at least three round-
houses, dating to around 15 bc–​ad 25, which were then replaced by rectangular struc-
tures and metalled streets and apparently surrounded by a timber palisade (Fulford and
Timby 2000: 8, 20–​24). What might be regarded as fairly high-​status material came
from this area, including coins, copper-​alloy brooches, and toilet instruments. Pottery
included terra nigra and terra rubra wares and some early samian ware dating to around
ad 30 onwards. There were also large numbers of pig bones and oyster shells (Fulford
and Timby 2000: 20-2​4), perhaps indicating special feasting activities. Upon reaching
the early levels in the excavations within insula IX of the town, there appears to have
been a fairly dense occupation with Iron Age timber rectangular buildings and streets
(Fulford 2003, 2011). The elites here may well have had contact with Rome, but their use
of the material need not necessarily mean that they considered themselves to have been
Romanized. While there are earthworks at Silchester conventionally associated with the
oppidum, there remains general uncertainty about their nature and purpose and their
756

756   Adam Rogers

relationship with the settlement evidence at the site. It would probably be a mistake to
make too many assumptions about their nature and function without further evidence,
including dateable material; our understanding of these sites and sequences of develop-
ment can still clearly be advanced through further work. It could be said that the inner
earthwork enclosed an area of around 32 hectares and partly underlies the Roman town,
while the outer earthwork is more discontinuous and later in date, perhaps representing
the continuation of Iron Age forms of monumentalization and special organization in
the Roman period; it may also have been part of further dyke systems to the north-​west
of the town (Fulford 1984: 79–​83) (Figure 36.4).
At Camulodunum a small fort has been identified at Gosbecks (Hawkes and Crummy
1995), which has been interpreted as having been part of the ad 43 invasion pre-​dating
the fortress here or, perhaps, built after the fortress had been converted into a colonia
around ad 49 or following the Boudiccan Revolt (Hawkes and Crummy 1995: 101). The
interpretation remains difficult, because the fort has not been excavated, but its early
plan and position led Creighton (2006: 61–​64) to suggest instead that it may even have
been pre-​ad 43 in date and influenced by the Roman army. The local leader may have
been organizing his/​her own forces in the Roman style. Alternatively, the fort may have
been garrisoned with Roman auxiliaries before Roman annexation. Either way, it is clear
that there remain many ambiguities in our understanding of urban origins and con-
struction than conventional narratives have allowed for.
It appears that London’s early and large-​scale development may have been largely an
exception among the towns. Here it appears that there may have been some military
involvement in the development of the site, with the construction of a monumental
street system and timber port facilities (Milne 1985; Millett 1994; Brigham 1998). The
large size of London compared with other towns and the number of early public build-
ings also indicates its importance. In much of the literature on Roman Britain, London
is known as the provincial capital, but, although there is evidence of Roman officials liv-
ing here, such as the procurator, there is no direct evidence that it was also the home of
the provincial governor; a permanent fort incorporated as part of the town (identified at
Cripplegate) dating to the late first or early second century ad, however, might support
this claim (Perring 1991; Milne 1995; Shepherd 2012; cf. Millett 1998).
Development was slow in most towns, and it is likely that there was little monetary
input from Rome; official funding was limited even for the development of the colo-
nia. With even less imperial money going to the civitas-​capitals, it seems most likely
that local elites would have been involved in funding the projects. An inscription (RIB
3123) from Verulamium dating the completion of the forum/​basilica to the end of ad
79 may have been an exception among towns in terms of its early date. Documents and
inscriptions from other parts of the empire indicate that slaves, prisoners, and the mili-
tary could all be used for construction projects. Surviving sections of the statute from
the town of Urso (modern Osuna) in Spain (the Lex Coloniae Genetivae Juliae), more-
over, stipulates that the citizens of the town had to be engaged in five days of construc-
tion work a year (Hardy 1912: 7; González and Crawford 1986; Duncan-​Jones 1990: 174),
but it is unclear to what extent this applied to other areas, though it was apparently also
 The Development of Towns    757

West End Brook

rk’
wo th
Ear

‘In
ne
ter

rE
‘Ou

ar
th
wo
rk
’
Town wall circuit

Brook
Silchester

0 100 200 300 400

m

Figure 36.4  Plan of the relationship between the town wall circuit and enigmatic earthworks at
Silchester (Calleva Atrebatum) possibly associated with the earlier oppidum here.
Source: Adapted from Fulford (1984: figure 85).

the case in Egypt (Duncan-​Jones 1990: 174). If the military was present in a number of
towns for the creation of road networks and other tasks, there is little direct evidence
for their involvement, and it is unlikely that they will have remained in towns for long
(unless retired).
758

758   Adam Rogers

The client ruler Togidubnus has formed a part of the narratives of urban construction
of Chichester and also, to a lesser extent, Silchester and Winchester. He is mentioned
by Tacitus (Agr. XIV): ‘quaedam civitates Cogidumno regi donatae (is ad nostrum usque
memoriam fidissimus mansit) [some of the estates were given to King Cogidubnus, who
lived down to our day a most faithful ally]’. The difficulty here is that the discrepancy in
the name may in fact indicate different people or that an error may have got into the text;
confusingly, also, the inscription of the name is also damaged, with the first letter miss-
ing. He appears to have been awarded land, but nothing else is known about his life or
career (Todd 2004: 47); there were clearly disruptions in landownership here, though.
The dedication slab was found at Chichester in the eighteenth century with what appears
to be his name (RIB 91), but there is no evidence that he was also involved in Winchester
or Silchester, these suggestions being based mainly on the similar dates in which the
towns developed and a number of datable fragments of architectural stonework from
Chichester. As well as RIB 91, which is a pre-​Flavian dedication of a temple to Neptune
and Minerva in honour of the imperial house built on behalf of Togidubnus and the
guild of town craftworkers, there was another inscription (RIB 92) dedicated to Nero,
perhaps from a major building or statue, and a dedication to the Matres Domesticae,
which may have been another early temple (Fulford 2006: 97). Unfortunately, very little
is really known about the early development of the town, and the street grid does not
appear to have been laid down until the ad 70s. There may, instead, have been some
kind of early monumental complex here (cf. Fulford 2006; Mattingly 2006: 269).
Another form of indigenous elite involvement in the physical creation of the town
has been argued by Creighton (2006) in his analysis of the relationship between urban
spaces and monumental burial chambers. At Camulodumum there was the Lexden
tumulus dating to the late first century bc and a larger burial complex at Stanway, just
west of Gosbecks, and dating from the late first century bc to around ad 60 (Crummy
1997: 22–2​4; Crummy et al. 2007). At Verulamium there was a rich burial chamber,
which lay uphill across the river Ver and dated to around the ad 40s–​50s (Niblett 1999),
and there were Iron Age burial mounds outside Cirencester in the Tarbarrow Field, in
an area that also appears to have been used for Roman period burials and religious activ-
ity (O’Neil and Grinsell 1960; Holbrook 2008; Biddulph and Welsh 2011). The recent
survey work at Silchester has also produced traces of what may have been elite burials
associated with the oppidum, although as yet no excavation has taken place here, and
more work is needed to confirm this. The burials that have been excavated demonstrate
knowledge of, and contact with, the Roman Empire, as indicated by the items within the
burials; but it must be recognized that they may have been as much about using Roman
goods and settlement forms for expressing their own identity as of adopting Roman
ways. Creighton (2006) has suggested that perhaps even the urban planning and estab-
lishment of street grids acted as a form of commemoration associated with the elites that
succeeded in achieving power.
Within a number of towns, including the colonia, many of the earliest public build-
ings were in fact constructed in timber rather than stone. At Lincoln, for example,
excavations on the site of the fortress principia have demonstrated that, once that it was
 The Development of Towns    759

demolished, the stone forum/​basilica complex did not appear until the second century,
following the construction of an earlier timber building (Steane 2006). An early timber
forum/​basilica complex has also been identified at Silchester (Fulford and Timby 2000),
as well as a theatre at Canterbury (Bennett 1988) and amphitheatres at Cirencester
(Holbrook 1998) and London (Bateman et al. 2008). There were also many timber
domestic structures in the early phases of towns, often in the form of ‘strip’ buildings, as
at London (e.g. Hill and Rowsome 2011). At London, however, a number of sites in the
western area of the town have also provided evidence of round-houses, perhaps indicat-
ing the continuation of local building traditions as people moved into the town from the
surrounding area (Perring and Roskams 1991; Hill and Rowsome 2011; Perring 2011).
The occupants of these early round-houses may even have come into the town in order
to work as labourers during its construction. It would be overly simplistic to interpret
these structures as inferior, or of lower status, to rectangular buildings and necessarily
representing the houses of poor labourers or slaves. These round-houses formed part
of a long tradition of construction, which was fully integrated into human expressions
of being and identity; the structures and spaces within them formed a microcosm of
the way in which their world was organized, experienced, and understood (cf. Harding
2009; Sharples 2010). It was, of course, not just monumental buildings that formed these
towns, moreover, but the people using and interacting in these spaces.
These early phases of town development formed significant parts of the biography
of these places and the creation of space. It is important not to apply modern expecta-
tions and perceived Roman values to our treatment of these sites, many of which draw
on Iron Age traditions. That it could be argued that towns never reached a high level
of monumentality in Roman Britain may also relate to a continuation of pre-​existing
ways of expressing wealth and using space. Local traditions in building and organizing
labour and the significance behind using and manipulating different materials may well
also have continued to be important, even if they were now building different styles of
structures. Sharples (2010: 116–​124), for example, has argued that the construction of
Iron Age oppida, hillforts, and other monuments would have been important cultural
events, with people encouraged to come together and contribute through motives relat-
ing to social interaction, feastings, and rituals. Construction processes and methods
themselves could also be highly symbolic, often involving the deliberate choice of the
use of specific materials, sometimes brought from long distances, to incorporate into
the structures, as seen at hillforts and other prehistoric monuments (Miles et al. 2003;
Sharples 2010: 116–​124). Building construction had more than simply practical and eco-
nomic meaning, since it involved the act of creation of both physical structures and new
spaces and could reflect the identities and expressions of those carrying out the work (cf.
Gardner 2007; Revell 2009). The construction processes would have formed an impor-
tant part of the biography of the building and just as significant a part of its meaning as
the finished structure itself (cf. McFadyen 2006). Towns can be seen as major construc-
tion events, and they also transformed the land through moving earth, rock, and water
in terracing, land drainage, reclamation, and other activities—​as identified through
excavation at London (e.g. Hill and Rowsome 2011; Leary and Butler 2012), Winchester
760

760   Adam Rogers

(e.g. Ford and Teague 2011), and other towns (Rogers 2012). These actions altered places
and formed an important part of the urban reality with an impact on the way in which
these landscapes were used and experienced.

Conclusion

It is becoming increasingly difficult to write about the development of towns in Roman

Britain, because no single narrative is possible. Though there were clearly processes
involved relating to the expansion of Roman control and military strategy, emphasis also
needs to be placed on the importance of each individual urban biography starting before
the Roman conquest and the different people involved in and experiencing each town
development, from local people to incomers and the military. More nuanced accounts
of each town are required, drawing on changing understandings of late Iron Age and
Roman period activity at both local and regional levels. It is this that makes urban devel-
opment such a challenging and important subject for understanding Roman Britain.

Abbreviations
AE    L’Année Epigraphique. Paris: Ernest Leroux/​Presses Universitaires des France,
1888–​to date).
RIB  R
 . G. Collingwood and R. P. Wright, Roman Inscriptions of Britain. Oxford:
Clarendon Press, 1965–​to date.

References
Primary Sources
Aurelius Victor, Caesares, trans. H. W. Bird. Liverpool: Liverpool University Press, 1994.
Caesar, Bellum Gallicum, trans. H. J. Edwards. London: Heinemann, 1917.
Cassius Dio, Roman History, trans. E. Cary. London: Heinemann, 1961.
Ptolemy, Geographia, trans. E. L. Stevenson. Mineola, NY: Dover, 1991.
Tacitus. Agricola, trans. M. Hutton. London: Heinemann, 1970.
Tacitus, Annales, trans. J. Jackson. London: Heinemann, 1937.

Secondary Sources
Bateman, N., Cowan, C., and Wroe-​Brown, R. (2008). London’s Roman Amphitheatre: Guildhall
Yard, City of London. MoLAS Monograph 35. London: Museum of London Archaeology
Service.
Bennett, P. (1988). ‘No. 25 Watling Street’, Archaeologia Cantiana, 106: 129–​130.
 The Development of Towns    761

Biddulph, E., and Welsh, K. (2011). Cirencester before Corinium: Excavations at Kingshill North,
Cirencester, Gloucestershire. Oxford: Oxford Archaeology Ltd.
Bidwell, P. T. (1979). The Legionary Bath-​House and Basilica and Forum at Exeter. Exeter
Archaeology Reports Volume I. Exeter: Exeter City Council and the University of Exeter.
Boatwright, M. L. (2000). Hadrian and the Cities of the Roman Empire. Princeton: Princeton
University Press.
Bowden, W., and Bescoby, D. (2008). ‘The Plan of Venta Icenorum (Caistor-​by-​
Norwich): Interpreting a New Geophysical Survey’, Journal of Roman Archaeology, 21: 324–​334.
Bradley, R., and Gordon, K. (1988). ‘Human Skulls from the River Thames, their Dating and
Significance’, Antiquity, 62: 503–​509.
Brigham, T. (1998). ‘The Port of Roman London’, in B. Watson (ed.), Roman London: Recent
Archaeological Work. Journal of Roman Archaeology Supplementary Series 24. Portsmouth,
RI: Journal of Roman Archaeology, 23–​34.
Brooks, H. (2006). ‘Colchester before Colchester’, in P. Ottaway (ed.), A Victory
Celebration: Papers on the Archaeology of Colchester and Late Iron Age–​Roman Britain pre-
sented to Philip Crummy. Colchester: Colchester Archaeological Trust, 5–​10.
Burnham, B., and Wacher, J. (1990). The Small Towns of Roman Britain. London: Batsford.
Burnham, B., Collis, J., Dobinson, C., Haselgrove, C., and Jones, M. (2001). ‘Themes
for Urban Research, c.100 bc to ad 200’, in S. James and M. Millett (eds), Britons and
Romans: Advancing an Archaeological Agenda. CBA Research Report 125. York: Council for
British Archaeology, 67–​76.
Butler, J. (2006). Reclaiming the Marsh:  Archaeological Excavations at Moor House,
City of London. Pre-​ Construct Archaeology Monograph 6.  London:  Pre-​ Construct
Archaeology Ltd.
Clay, P., and Mellor, J. E. (1985). Excavations in Bath Lane, Leicester. Leicestershire Museums
Archaeological Report 10. Leicester: Leicester Museums.
Collingwood, R. G., and Richmond, I. (1969). The Archaeology of Roman Britain.
London: Methuen.
Copeland, T. (2011). Roman Gloucestershire. Stroud: The History Press.
Cowan, C., Seeley, F., Wardle, A., Westman, A., and Wheeler, L. (2009). Roman Southwark
Settlement and Economy:  Excavations in Southwark 1973–​ 91. MoLAS Monograph 42.
London: Museum of London Archaeology Service.
Creighton, J. (2000). Coins and Power in Late Iron Age Britain. Cambridge:  Cambridge
University Press.
Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge.
Crummy, P. (1984). Excavations at Lion Walk, Balkerne Lane and Middleborough, Colchester,
Essex. Colchester Archaeology Report 3. Colchester: Colchester Archaeological Trust.
Crummy, P. (1988). ‘Colchester (Camulodunum/​ Colonia Victriciensis)’, in G. Webster
(ed.), Fortress into City:  The Consolidation of Roman Britain First Century ad.
London: Batsford, 24–​47.
Crummy, P. (1992). Excavations at Culver Street, the Gilberd School and Other Sites in Colchester
1971–​85. Colchester Archaeology Report 6. Colchester: Colchester Archaeological Trust.
Crummy, P. (1997). City of Victory:  The Story of Colchester—​Britain’s First Roman Town.
Colchester: Colchester Archaeological Trust.
Crummy, P., Benfield, S., Crummy, N., Rigby, V., and Shimmin, D. (2007). Stanway: An Élite
Burial Site at Camulodunum. Britannia Monograph Series 24. London: Society for the
Promotion of Roman Studies.
762

762   Adam Rogers

Drury, P. J. (1984). ‘The Temple of Claudius at Colchester Reconsidered’, Britannia, 15: 7–​50.

Duncan-​Jones, R. (1990). Structure and Scale in the Roman Economy. Cambridge: Cambridge
University Press.
Esmonde Cleary, A. S. (1987). Extra-​Mural Areas of Romano-​British Towns. BAR British Series
169. Oxford: Archaeopress.
Field, N., and Parker Pearson, M. (2003). Fiskerton: An Iron Age Timber Causeway with Iron
Age and Roman Votive Offerings. Oxford: Oxbow Books.
Fincham, G. (2004). Durobrivae: A Roman Town between Fenland and Upland. Stroud: Tempus.
Ford, B. M., and Teague, S. [with Biddulph, E., Hardy, A., and Brown, L.] (2011). Winchester—​
a City in the Making:  Archaeological Excavations between 2002 and 2007 on the Sites of
Northgate House, Staple Gardens and the former Winchester Library, Jewry St. Oxford: Oxford
Archaeology Ltd.
Frere, S. S. (1967). Britannia: A History of Roman Britain. London: Routledge and Kegan Paul.
Frere, S. S. (1972). Verulamium Excavations, Volume I. Reports of the Research Committee of
the Society of Antiquaries of London 28. London: The Society of Antiquaries of London.
Frere, S. S. (1983). Verulamium Excavations, Volume II. Reports of the Research Committee of
the Society of Antiquaries of London 41. London: The Society of Antiquaries of London.
Frere, S. S. (1987). Britannia: A History of Roman Britain. 3rd edn. London: Routledge and
Kegan Paul.
Fulford, M. (1984). Silchester Defences 1974–​80. Britannia Monograph Series 5. London: Society
for the Promotion of Roman Studies.
Fulford, M. (2003). ‘Julio-​ Claudian and Early Flavian Calleva’, in P. Wilson (ed.), The
Archaeology of Roman Towns: Studies in Honour of John S. Wacher. Oxford: Oxbow Books,
95–​104.
Fulford, M. (2006). ‘Corvées and civitates’, in R. J.  A. Wilson (ed.), Romanitas:  Essays on
Roman Archaeology in Honour of Sheppard Frere on the Occasion of his Ninetieth Birthday.
Oxford: Oxbow Books, 65–​7 1.
Fulford, M. (2011). ‘“Calleva: Silchester in the Iron Age”: From Julius Caesar to Claudius: The
Late Iron Age/​Roman Transition’. Unpublished conference paper, University of Leicester,
November.
Fulford, M., and Timby, J. (2000). Late Iron Age and Roman Silchester: Excavations on the Site
of the Forum-​Basilica 1977, 1980–​86. Britannia Monograph Series 15. London: Society for the
Promotion of Roman Studies.
Gardner, A. (2007). Archaeology of Identity:  Soldiers and Society in Later Roman Britain.
Walnut Creek, CA: Left Coast Press.
González, J., and Crawford, M. H. (1986). ‘The Lex Irnitana:  A  New Copy of the Flavian
Municipal Law’, Journal of Roman Studies, 76: 147–​243.
Harding, D. W. (2009). The Iron Age Round-​House: Later Prehistoric Building in Britain and
Beyond. Oxford: Oxford University Press.
Hardy, E. G. (1912). Roman Laws and Charters. Oxford: Clarendon Press.
Haselgrove, C. (1999). ‘The Iron Age’, in J. Hunter and I. Ralston (eds), The Archaeology of
Britain. London and New York: Routledge, 113–​134.
Haselgrove, C. (2004). ‘Society and Polity in Late Iron Age Britain’, in M. Todd (ed.), A
Companion to Roman Britain. Oxford: Blackwell Publishing, 12–​29.
Haselgrove, C., and Moore, T. (2007). ‘New Narratives of the Later Iron Age’, in C. Haselgrove
and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 1–​15.
 The Development of Towns    763

Haverfield, F. (1912). The Romanization of Roman Britain. Oxford: Clarendon Press.

Hawkes, C. F. C., and Crummy, P. (1995). Camulodunum 2. Colchester Archaeology Report 11.
Colchester: Colchester Archaeological Trust.
Heighway, C. (1983). The East and North Gates of Gloucester. Gloucester:  Western
Archaeological Trust.
Heighway, C., and Bryant, R. (1999). The Golden Minster: The Anglo-​Saxon Minster and Later
Medieval Priory of St Oswald at Gloucester. CBA Research Report 117. York:  Council for
British Archaeology.
Henderson, C. (1988). ‘Exeter (Isca Dumnoniorum)’, in G. Webster (ed.), Fortress into City: The
Consolidation of Roman Britain in the First Century ad. London: Batsford, 91–​119.
Hill, J., and Rowsome, P. (2011). Roman London and the Walbrook Stream Crossing: Excavations
at 1 Poultry and Vicinity, City of London. MoLAS Monograph 37. London:  Museum of
London Archaeology Service.
Hingley, R. (1997). ‘Resistance and Domination:  Social Change in Roman Britain’, in D.
J. Mattingly (ed.), Dialogues in Roman Imperialism:  Power, Discourse and Discrepant
Experiences in the Roman Empire. Journal of Roman Archaeology Supplementary Series 23.
Portsmouth, RI: Journal of Roman Archaeology, 145–​166.
Hingley, R. (2000). Roman Officers and English Gentlemen. London and New York: Routledge.
Hingley, R. (2005). Globalizing Roman Culture:  Unity, Diversity and Empire. London and
New York: Routledge.
Hingley, R. (2008). The Recovery of Roman Britain 1586 to 1906:  A  Colony so Fertile.
Oxford: Oxford University Press.
Holbrook, N. (1998). Cirencester:  The Roman Town Defences, Public-​Buildings and Shops.
Cirencester: Cotswold Archaeological Trust.
Holbrook, N. (2008). ‘Cirencester and the Cotswolds: The Early Roman Evolution of a Town
and Rural Landscape’, Journal of Roman Archaeology, 21: 305–​323.
Hurst, H. (1988). ‘Gloucester (Glevum)’, in G. Webster (ed.), Fortress into City: The
Consolidation of Roman Britain in the First Century ad. London: Batsford, 24–​69.
Hurst, H. (1999a). ‘Topography and Identity in Glevum colonia’, in H. Hurst (ed.), The Coloniae
of Roman Britain: New Studies and a Review. Papers of the Conference held at Gloucester
on 5–​6 July, 1997. Journal of Roman Archaeology Supplementary Series 36. Portsmouth,
RI: Journal of Roman Archaeology, 113–​135.
Hurst, H. (1999b). ‘Civic Space at Glevum, in H. Hurst (ed.), The Coloniae of Roman
Britain: New Studies and a Review. Papers of the Conference held at Gloucester on 5–​6 July,
1997. Journal of Roman Archaeology Supplementary Series 36. Portsmouth, RI: Journal of
Roman Archaeology, 152–​160.
James, H. (2003). Roman Carmarthen. London: Society for the Promotion of Roman Studies.
Jones, M. J. (1988). ‘Lincoln (Lindum)’, in G. Webster (ed.), Fortress into City: The Consolidation
of Roman Britain in the First Century ad. London: Batsford, 145–​166.
Jones, M. J. (2002). Roman Lincoln: Conquest, Colonia and Capital. Stroud: Tempus.
Jones, M. J. (2003a). ‘The Roman Military Era’, in D. Stocker (ed.), The City by the Pool: Assessing
the Archaeology of the City of Lincoln. Oxford: Oxbow Books, 36–​53.
Jones, M. J. (2003b). ‘The colonia Era: Archaeological Account’, in D. Stocker (ed.), The City by
the Pool: Assessing the Archaeology of the City of Lincoln. Oxford: Oxbow Books, 56–​138.
Jones, M. J. (2004). ‘Cities and Urban Life’, in M. Todd (ed.). A Companion to Roman Britain.
Oxford: Blackwell Publishing, 162–​192.
764

764   Adam Rogers

Jones, M. J., and Stocker, D. (2003). ‘Settlement in the Lincoln Area in the Prehistoric
Era:  Archaeological Account’, in D. Stocker (ed.), The City by the Pool:  Assessing the
Archaeology of the City of Lincoln. Oxford: Oxbow Books, 19–​33.
Kipling, R., Parker, D., and Cooper, L. (2007). ‘Leicester: Bath Lane, Merlin Works’, University
of Leicester Archaeological Services News (November), 5–​6.
Laurence, R. (2001). ‘The Creation of Geography: An Interpretation of Roman Britain’, in C.
Adams and R. Laurence (eds), Travel and Geography in the Roman Empire. London and
New York: Routledge, 67–​94.
Leach, P. (1982). Ilchester:  Volume 1 Excavations 1974–​ 1975. Bristol:  Western
Archaeological Trust.
Leary, J., and Butler, J. (2012). Roman Archaeology in the Upper Reaches of the Walbrook
Valley:  Excavations at 6–​8 Tokenhouse Yard, London EC2. Pre-​Construct Archaeology
Monograph 14. London: Pre-​Construct Archaeology Ltd.
McFadyen, L. (2006). ‘Building Technologies, Quick Architecture and Early Neolithic Long
Barrow Sites in Southern Britain’, Archaeological Review from Cambridge, 21/​1: 117–​134.
McWhirr, A. (1988). ‘Cirencester (Corinium Dobunnorum)’, in G. Webster (ed.), Fortress into
City: The Consolidation of Roman Britain in the First Century ad. London: Batsford, 74–​90.
Mann, J. C. (1996). Britain and the Roman Empire. Aldershot: Variorum.
Mattingly, D. (2004). ‘Being Roman: Expressing Identity in a Provincial Setting’, Journal of
Roman Archaeology, 17: 5–​25.
Mattingly, D. (2006). An Imperial Procession: Britain in the Roman Empire. London: Penguin.
Merrifield, R. (1995). ‘Roman Metalwork from the Walbrook: Rubbish, Ritual or Redundancy?’,
Transactions of the London and Middlesex Archaeological Society, 46: 27–​44.
Merrifield, R., and Hall, J. (2008). ‘In its Depths what Treasures: The Nature of the Walbrook
Stream Valley and the Roman Metalwork Found Therein’, in J. Clark, J. Cotton, J. Hall,
R. Sherris, and H. Swain (eds), Londinium and Beyond: Essays on Roman London and its
Hinterland for Harvey Sheldon. York: CBA Research Report 156. York: Council for British
Archaeology, 121–​127.
Miles, D., Palmer, S., Lock, G., Gosden, C., and Cromarty, A. M. (2003). Uffington White Horse
and its Landscape:  Investigations at White Horse Hill Uffington, 1989–​95 and Tower Hill
Ashbury, 1993–​4. Oxford: Oxford Archaeological Unit.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Millett, M. (1994). ‘Evaluating Roman London’, Archaeological Journal, 151: 427–​435.
Millett, M. (1998). ‘Introduction: London as Capital?’, in B. Watson (ed.), Roman London:
Recent Archaeological Work. Journal of Roman Archaeology Supplementary Series 24.
Portsmouth, RI: Journal of Roman Archaeology, 7–​12.
Milne, G. (1985). The Port of Roman London. London: Batsford.
Milne, G. (1995). Roman London. London: Batsford/​English Heritage.
Moore, T. (2006). Iron Age Societies in the Severn-​Cotswolds: Developing Narratives of Social
and Landscape Change. BAR British Series 421. Oxford: Archaeopress.
Moore, T. (2011). ‘Detribalizing the Later Prehistoric Past: Concepts of Tribes in Iron Age and
Roman Studies’, Journal of Social Archaeology, 11: 334–​360.
Niblett, R. (1985). Sheepen: An Early Roman Industrial Site at Camulodunum. CBA Research
Report 57. York: Council for British Archaeology.
Niblett, R. (1999). The Excavation of a Ceremonial Site at Folly Lane, Verulamium. Britannia
Monograph Series 14. London: Society for the Promotion of Roman Studies.
Niblett, R. (2001). Verulamium: The Roman City of St Albans. Stroud: Tempus.
 The Development of Towns    765

O’Neil, H., and Grinsell, L. V. (1960). ‘Gloucestershire Barrows’, Transactions of the Bristol and
Gloucestershire Archaeological Society, 79: 5–​149.
Ottaway, P. (1999). ‘York: The Study of a Late Roman colonia’, in H. Hurst (ed.), The Coloniae of
Roman Britain: New Studies and a Review. Papers of the Conference held at Gloucester on 5–​6
July, 1997. Journal of Roman Archaeology Supplementary Series 36. Portsmouth, RI: Journal
of Roman Archaeology, 136–​150.
Ottaway, P. (2004). Roman York. Stroud: Tempus.
Perring, D. (1991). Roman London. London: Seaby.
Perring, D. (2011). ‘Two Studies on Roman London’, Journal of Roman Archaeology, 24:
249–​282.
Perring, D., and Roskams, S. [with Allen, P.] (1991). Early Development of Roman London West
of the Walbrook: The Archaeology of Roman London, Volume 2. CBA Research Report 70.
London: Council for British Archaeology.
Potter, T. W. (2002). ‘The Transformation of Britain from 55 bc to ad 61’, in P. Salway (ed.), The
Roman Era: The British Isles: 55 bc–​ad 410. Oxford: Oxford University Press, 11–​36.
Putnam, B. (2007). Roman Dorset. Stroud: Tempus.
Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge
University Press.
Richmond, I. A. (1963). Roman Britain. London: Jonathan Cape.
Rivet, A. L. F., and Smith, C. (1979). The Place-​Names of Roman Britain. London: Batsford.
Rogers, A. (2008). ‘Religious Place and its Interaction with Urbanization in the Roman Era’,
Journal of Social Archaeology, 8/​1: 37–​62.
Rogers, A. (2011a). ‘Reimagining Roman Ports and Harbours: The Port of Roman London and
Waterfront Archaeology’, Oxford Journal of Archaeology, 30/​2: 207–​225.
Rogers, A. (2011b). Late Roman Towns in Britain:  Rethinking Change and Decline.
Cambridge: Cambridge University Press.
Rogers, A. (2012). ‘Water and the Urban Fabric: A Study of Towns and Waterscapes in the
Roman Period’, International Journal of Nautical Archaeology, 41/​2: 327–​339.
Rogers, A. (2013). Water and Roman Urbanism: Towns, Waterscapes, Land Transformation and
Experience in Roman Britain. Leiden: Brill.
Rykwert, J. (1976). The Idea of a Town: The Anthropology of Urban Form in Rome, Italy and the
Ancient World. London: Faber and Faber.
Sharples, N. (2010). Social Relations in Later Prehistory: Wessex in the First Millennium bc.
Oxford: Oxford University Press.
Shepherd, J. (2012). Discovery of the Roman Fort at Cripplegate, City of London: Excavations by
W. F. Grimes 1947–​68. London: Museum of London Archaeology.
Steane, K. [with Darling, M. J., Mann, J. Vince, A., and Young, J.] (2001). The Archaeology of
Wigford and the Brayford Pool. Oxford: Oxbow Books.
Steane, K. (2006). The Archaeology of the Upper City and Adjacent Suburbs. Lincoln
Archaeology Studies 3. Oxford: Oxbow Books.
Todd, M. (2004). ‘The Claudian Conquest and its Consequences’, in M. Todd (ed.), A
Companion to Roman Britain. Oxford: Blackwell Publishing, 42–​59.
Tomlin, R. S. O. (2006). ‘Was Roman London ever a colonia? The Written Evidence’, in R. J. A.
Wilson (ed.), Romanitas: Essays on Roman Archaeology in Honour of Sheppard Frere on the
Occasion of his Ninetieth Birthday. Oxford: Oxbow Books, 49–​64.
Wacher, J. S. (1969). Excavations at Brough-​on-​Humber, 1958–​1961. Reports of the Research
Committee of the Society of Antiquaries of London 25. London: The Society of Antiquaries
of London.
766

766   Adam Rogers

Wacher, J. S. (1975). The Towns of Roman Britain. London: Batsford.

Wacher, J. S. (1995). The Towns of Roman Britain. 2nd edn. London: Batsford.
Wacher, J. S., and McWhirr, A. (1982). Early Roman Occupation at Cirencester. Cirencester
Excavations 1. Cirencester: Cirencester Excavations Committee.
Webster, G. (1966). ‘Fort and Town in Early Roman Britain’, in J. S. Wacher (ed.), The Civitas
Capitals of Roman Britain. Leicester: Leicester University Press, 31–​45.
Webster, J. (2001). ‘Creolizing the Roman Provinces’, American Journal of Archaeology,
105: 209–​225.
White, R., and Barker, P. (1998). Wroxeter: Life and Death of a Roman City. Stroud: Tempus.
Williams, J. (2005). ‘“The Newer Rite Is Here”: Vinous Symbolism on British Iron Age Coins’,
in C. Haselgrove and D. Wigg-​Wolf (eds), Iron Age Coinage and Ritual Practices. Mainz am
Rhein: Van Zabern, 25–​42.
Woodward, P., and Woodward, A. (2004). ‘Dedicating the Town: Urban Foundation Deposits
in Roman Britain’, World Archaeology, 36: 68–​86.
 Chapter 37

Urban Monum e nta l i t y

i n Roman Bri ta i n

Louise Revell

Introduction

One of the most conspicuous changes to the landscape of Britain after the Roman con-
quest was the appearance of towns. Constructed from stone, standing two or more
storeys high, the public buildings of these towns bore witness to a new form of mon-
umentality. More than this, they announced a new ideology of power, not only of the
distant authority of Rome, but also of the British elite. The construction of towns began
shortly after the conquest, and, with evidence for rebuilding through to the fourth cen-
tury, these towns proved a long-​term landmark in the lives of the provincial populations.
In recent years, work on Roman urbanism has moved away from an aesthetic approach
to a more spatial enquiry, the so-​called spatial turn. This has not been confined to the
Roman period; drawing on geography, anthropology, and sociology, among others, the
archaeology of monuments and buildings has focused more on the concept of inhabited
spaces. Furthermore, it is this act of dwelling and the social practices which these spaces
allow, which form part of the short-​term and long-​term processes of social reproduc-
tion. Urban buildings are not only symbols of power, but part of the process of the crea-
tion and maintenance of varying relationships of power.
These ideas have forced us to rethink the Roman townscape as a whole, and how peo-
ple inhabited it or moved through it (MacDonald 1986; Zanker 1998; Kaiser 2000; Revell
2009; Laurence and Newsome 2011). Concurrently there has been a recognition of the
variability in the urban experience within the empire, such as Gros’ monumental sur-
vey (1996) of public architecture, or Laurence, Esmonde C ​ leary, and Sears’ examination
(2011) of the development of the urban landscape within Italy and the western provinces.
One strand of these new approaches has been the role towns played in the maintenance
of an idea of Roman-​ness, linking the peoples of the provinces into wider ideologies at
play within the empire, and creating a sense of belonging in the empire. A central idea
768

768   Louise Revell

in the work of Laurence, Esmonde ​Cleary, and Sears is that one role of the Roman town
was the production of citizens. While the idea of Roman citizenship as such is less rel-
evant within the western provinces, and Roman Britain in particular, I have argued else-
where that the daily experience of inhabiting the Roman town created a particular form
of Roman ideology of urbanism among the provincial communities which tied them
into a wider Roman identity (Revell 1999, 2009).
The analysis of the towns of Roman Britain has played a limited role in these new
approaches to the Roman townscape. Often viewed as a failed experiment, the British
towns have been seen as a poor relation of their continental cousins. Debates have
focused on their origins (military or civilian) and how early they can be said to have
declined (if they ever had a high point). The character of the mature Romano-​British
town has been of little interest. In spite of the numbers of books focusing on individual
towns (e.g. Niblett 2001; Jones 2002; White 2002), syntheses are few and far between.
Wacher’s monumental survey (1995) of the towns of Roman Britain has not been repro-
duced, although there have been certain works which have focused on particular ques-
tions such as the origins of towns (Creighton 2006) or the significance of water in the
town (Rogers 2013). This chapter goes some way to addressing the lack of application
of these new strands of investigation to Roman towns. In particular, it concentrates on
the monumental buildings of the town, the kinds of activities they fostered, and how
far they were implicated in the re-​creation of a particular Roman ideology of Roman
urbanism. It focuses on the larger chartered towns rather than including the so-​called
small towns, as the latter generally lack the range of monumental architecture symbolic
of Roman urbanism and Roman power.

Space for Politics and Administration

The first form of activity to consider is political activity and the role of the forum (this
material is discussed in greater detail in Revell 2007a). Within the Roman system of
imperialism from the time of Augustus until late antiquity, the town was the means of
organizing the provinces and the focus for provincial administration through activities
such as the collection of taxes, the administration of justice, and the taking of the census.
Through the institutions of the town council and local magistrates, the town was central
to the more localized administration of the empire. The forum was the symbolic heart of
the Roman town, and the stage for the enactment of the political rituals which held the
town together. This was the case in ancient Rome, and all our evidence suggests that this
was similarly the case for the provinces.
The importance of the forum to the urban experience can be demonstrated through
its location at the heart of the street grid. At the newly founded town of Caerwent, which
arguably most conforms to the stereotype of the Roman street grid, the forum lay at
the point where the two major roads through the town intersected. This centrality is
also evident in towns which arose out of Iron Age settlements, such as Silchester. At
 Urban Monumentality in Roman Britain    769

Verulamium, where there was a more complex street grid, the forum was given added
significance by being constructed on the site of the Central Enclosure, which had been
either a royal enclosure, or some form of ceremonial or religious precinct (Niblett
2001: 42–​43). The size of the fora would have meant they dominated the townscape at
key locations within passages of movement.
The Romano-​British forum generally took the form of a large open courtyard,
enclosed by porticoes on three sides and a basilica running the full length of the fourth.
The layout and decoration of the basilica emphasized its association with political activi-
ties, as opposed to other possible functions. The raised platform of the tribunal was the
site of magisterial authority and the application of the law throughout the empire, and
these were located within the basilica, at one or both ends of the nave. Architectural fea-
tures distinguished them from the rest of the nave, such as pillars to the sides, or differ-
ent styles of decoration. Within the second-​century basilica at London, there seems to
have been a single tribunal located in a semi-​circular apse at the eastern end of the nave,
set apart from the main part by some form of antechamber (Milne 1992). The area of the
antechamber was decorated with at least four phases of multi-​coloured wall-​paintings,
including one with floral and possibly figural motifs. The apse itself may have been lined
with marble wall plaques. At Silchester, there was a tribunal at either end of the basil-
ica, and, following the second-​century rebuilding of the basilica, the tribunals were set
within square apses, the northern one faced with thin slabs of Purbeck marble (Fulford
and Timby 2000). At Cirencester, a tribunal has been uncovered in the west end of the
basilica, set within a large semi-​circular apse, and decorated with imported marbles and
architectural decoration (Martin 1898). At Caerwent, the tribunal was accentuated by
engaged columns and more than one phase of wall-​painting (Brewer 1993).
The second area for political activity was the curia or council chamber; however, these
are difficult to identify from the archaeological evidence alone. Balty (1991) uses as cri-
teria the identification of benches, architectural form, position, and internal decora-
tion. From these criteria, the curia at Caerwent can be plausibly identified (Ashby 1906;
Frere 1989: 274). Here, one room showed evidence of benches along two opposite walls,
a small platform between them, elaborate wall-​painting, and a later mosaic. Others have
proved more controversial (Figure 37.1). At Verulamium, it is likely that one of the three
adjoining buildings was a curia, but various commentators have argued for each one of
the three, all with plausible supporting arguments. At Caistor-​by-​Norwich, Frere (1971)
argued that the curia lay to the south end of the basilica nave, based on its position and
size (9.1m × 9.14m). The other alternative is room 2, which lay behind the basilica, with
an anteroom separating it from the bustle, as at Caerwent, and a hypocaust adding to its
comfort and status. Other political requirements, such as an archive room and offices
for the town clerks, were met through the other rooms within the rear range of the basil-
ica, or within the porticoes surrounding the forum courtyard.
Elsewhere in the empire, the forum was used as a venue for the display of inscrip-
tions and statues of both leading local magistrates and also the emperor and impe-
rial family (Lefebvre 2004; Revell 2013). These are usually argued to have been absent
from the fora of Britain, but a review of the evidence suggests this may not be the case.
770

770   Louise Revell

Curia

0 5 10 m

Figure 37.1  Plan of Caerwent forum, with location of probable curia marked.

Source: adapted from Brewer (2006: 39).

The forum at Silchester was decorated with a series of statues and inscriptions (Figure
37.2). In the northern apse of the basilica, traces of statue bases were found, and nearby
a fragment of a larger-​than-​life statue of a military commander, almost certainly an
emperor (Joyce 1873; Fox and Hope 1893; Boon 1973; Cunliffe and Fulford 1982; Isserlin
1998; Fulford and Timby 2000). A bronze eagle was found in the room adjoining the
southern tribunal, a stone head, argued to represent the guardian deity of the town,
was found in the central apsed room, and there were another four fragments of human
figures without firm location, at least one of which may also be of an emperor. Further
statue bases stood in the northern range and eastern ambulatory of the forum. A num-
ber of inscriptions probably stood in the basilica and the ambulatories. The majority
 Urban Monumentality in Roman Britain    771

Approximate locations of:

Statues
Possible statues
Inscriptions 0 25 m
Possible inscriptions

Figure  37.2  Plan of Silchester forum with known and approximate locations of inscriptions
and statues indicated.
Source: P. Copeland; adapted from Isserlin (1998: figure 9.1) with additions.

are fragmentary, but one contains the phrase ob honorem (RIB 67), suggesting it was
in honour of a magistrate or other leading figure. Two more were monumental in size,
forming part of the fabric of the building. No other forum has quite the wealth of evi-
dence, but most have the remains of the podia for statues, or fragments of bronze stat-
ues likely to be of emperors. At Cirencester, the eye from a bronze statue was found
within the tribunal apse (Martin 1898), and fragments of an inscription found unstrati-
fied on the site refer to the Res Publica (RIB 114). At Gloucester, a statue base approx-
imately 4.1m × 3m was found in front of the basilica with fragments from a bronze
statue around it (Hurst 1972; Henig 1993: no. 177). From Verulamium and Wroxeter, we
have the inscriptions for the fora, dedicating them to the reigning emperor (RIB 3123,
RIB 288, respectively). At the majority of the fora in Britain which have seen systematic
excavation, there is evidence, however fragmentary, for their decoration with statues
and inscriptions.
772

772   Louise Revell

While the forum–​basilica in Britain clearly provided an appropriate space for carry-
ing out specific political activities, the lack of temples within them is a notable feature.
The exclusion of a temple from the forum is not unique to Britain, and is also seen in the
North African provinces (Euzennat and Hallier 1986). One problem with understand-
ing how the British examples fit into the wider history of the development of the forum
is that we lack some of the earliest structures built, in particular that at Cirencester,
while the forum at Colchester is somewhat problematic and will be discussed later in
the chapter. Nevertheless, the earliest fora do show more evidence for incorporating
temples. The forum–​basilica at Verulamium, dedicated in ad 79, in its first phase had a
single building opposite the basilica, possibly a temple (RIB 3123; Niblett and Thompson
2005: 78–​83). In the second phase, probably mid-​second century, this was flanked by
two further buildings, potentially although not definitely temples. Less problematic is
the Flavian forum from London, which was constructed with a small classical temple,
20m × 10m, immediately to the west (Milne 1992). Most tentative is the argument for
a temple as part of the Flavian timber basilica at Silchester (Fulford and Timby 2000;
Millett 2001). However, in the second-​century rebuild, the shrine was moved into the
basilica.
Nevertheless, even without the evidence for free-​standing temples, it would be
erroneous to see the fora of Britain as completely devoid of religious activity. In reality,
most examples seem to have an internal shrine opening onto the nave of the basilica,
in some cases as part of the rear range of rooms. Their identification as shrines gener-
ally rests on the fact that they were open to the rest of the basilica, and lay on the cen-
tral axis of the forum–​basilica. This is the case at the rebuilt forum at Silchester, where
the central room of the rear range was open, and raised up above the level of the rest of
the nave (Fulford and Timby 2000). It was decorated with thin slabs of white marble,
and the original excavator commented that it appeared more richly decorated than
the rest of the building. This was the room in which the tutela head was discovered,
adding weight to the identification of the room as a shrine. A raised, open room on
the dominant axis is also present in the fora at Caistor-​by-​Norwich and Caerwent,
again suggesting that these were shrines. The dedication of such shrines is uncertain,
although they are often assumed to be associated with the imperial cult through com-
parison with others elsewhere in the empire. However, this obscures the fact that very
few shrines or temples to the imperial cult have been unambiguously identified in
the west, and creates an element of circularity to their identification (Alföldy 1996;
Wallace-​Hadrill 2011).
The forum had many purposes: economic, religious, social, but arguably the most
important was the political. It was the product of a new political organization, in which
the pre-​Roman system of social organization and elite status was replaced by one based
upon the system of magistracies, with power distributed among a wider group, and
supreme power held for a restricted period. This was the ideal of Roman politics, and it
was the ideal used to formulate Roman administration systems. How far it is applicable
in reality is unclear, and, instead of competition, it is possible that power was concen-
trated in the hands of a few families (Blagg 1990). Nevertheless, the wider community
 Urban Monumentality in Roman Britain    773

was brought into this through the idea of the citizen community, and practices such
as voting, and the taking of the census, provided a formal, if infrequent, opportunity
to engage with the town as a political centre. While certain activities were applicable
only to Roman citizens, voting was extended to the urban citizenship. Applying the pic-
ture to Britain is not uncontentious. The four colonia of Gloucester, Colchester, Lincoln,
and York and the one municipium of Verulamium may well have had similar charters to
those from elsewhere in the western provinces (González 1986; Crawford 1996), but we
have no evidence for the form of charters given to the civitas capitals, which comprised
the majority of the towns. Therefore we have to assume that the political activities car-
ried out there would be similar, although not necessarily identical, to those we know
about from other provinces. Nevertheless, the evidence for their ongoing alterations to
suit better the needs to the community demonstrates their significance, and that these
were not white elephants in the urban skyline.

Space for Worship

As discussed in the previous section, one notable feature of the Romano-​British forum
was the lack of a freestanding temple, with, instead, a shrine leading from the basilica.
Nevertheless, religion formed an important part of the urban experience within the
province. Most towns provide evidence for one or more temples of varying monumen-
tality and size. There were few classical-​style temples, with Romano-​Celtic temples
being the main type, both in the centre of the towns and on the periphery (Goodman
2013). Nevertheless, this should not detract from the function of temples as the focus
for religious activity, and there is little evidence for them being used for non-​Roman or
resistant religious practices. Instead of focusing on the building type, it is more fruitful
to explore how they were used in the articulation of urban space.
In certain towns, temples formed part of a monumental urban centre, associated
with the forum and/​or a theatre. This is best expressed in Verulamium, with the pair-
ing of the temple and theatre, linked to the forum by a road out of alignment with the
main street grid. Much has been written about the location of the temple in insula 16 in
Verulamium and its relationship with the theatre and the Folly Lane burial (Haselgrove
and Millett 1997; Niblett 2001; Creighton 2006). The temple was constructed in the late
first century ad on what seems to have been a pre-​conquest sacred site (Niblett and
Thompson 2005). It was set within a large porticoed temenos or sacred precinct, 100m
× 50m, and took the form of a square Romano-​Celtic temple. In the later second cen-
tury ad, the shrine was further elaborated with side wings, and a double portico was
added to the wall surrounding the temenos, along with a monumentalized porch. This
porch faced south, towards the theatre, which had been constructed ad 140–​160, and
was also altered in the late second century (Kenyon 1935). There is some indication of
an open area between the theatre and the temple, and this may have had some form of
ritual purpose. This temple–​theatre complex was linked to the forum by a street leading
774

774   Louise Revell

to the side entrance of the forum, creating a focus for public religious and political
activity within the town.
It seems likely that a similar monumental centre can be identified at Canterbury,
although here the picture is not straightforward (Figure 37.3). A series of excavations
have uncovered a large portico enclosing a large central court, constructed c. ad 100–​120
(Bennett and Nebiker 1989). There has been less excavation of the area within this court,
but in one corner stood a timber Romano-​Celtic temple, built when the enclosure itself
was first constructed, and demolished at some point in the third century ad (Bennett
1981). It is usually assumed that this was one of a series of temples within the temenos,
possibly the large classical temple suggested by Blagg (1984) based on the architectural
fragments. A theatre stood to the east of the temenos complex. It was constructed in the
late first century ad, earlier than the temple, and was altered in the early third century
(Frere and Simpson 1970). To the north of the theatre, a substantial baths building was
constructed in the early second century (Blockley et al. 1995). The insula directly to the
north of the temenos has produced evidence for monumental structures, and this has
led to the hypothesis that it was the site of the forum. A large quantity of marble was
excavated in this area, approximately two-​thirds of which was imported, predominantly
from Carrera, but also from other eastern Mediterranean quarries (Blagg 1984: 66–​7 1).
Fragments of Corinthian columns were also among these remains. Thus, by the early
second century, the heart of Roman Canterbury consisted of four insula forming a mon-
umental core of sacred shrine, adjoining theatre, baths, and possibly the forum, linked
by a wide road out of alignment with the main grid. This association has led Millett to
argue that the town could be seen as a religious sanctuary (Millett 2007). However, in
terms of monumentality, there is an interesting contrast between the overall appearance
of the theatre, baths, and the exterior of the temenos, with the Roman-​style architec-
ture and use of marble, and the only known temple inside the temenos, which was built
of timber and not reconstructed in masonry before its demolition in the third century.
Even if there was the large temple suggested by Blagg, this deliberate curation of an
increasingly ‘archaic’ building material can only be seen as creating a particular symbol-
ism for the shrine.
At Colchester, we also have the situation of a monumental temple complex which
has been identified as the Temple to Deified Claudius, mentioned in Tacitus’ account
of the Boudiccan Revolt (Tac. Ann. 14.31). There is evidence of an associated theatre,
but, again, no certain location for the forum. One possibility is the insula opposite the
entrance to the temenos. This irregular shaped insula (insula 30) has produced remains
for a substantial masonry building, possibly relating to some form of basilica. There are
also traces of a foundation wall crossing the street between them. This could have been
a monumental arch, echoing that at the Balkerne Gate, or, alternatively, it could indicate
that the two insulae formed a single complex. It has been argued that insula 30 was the
forum, and, if linked, the two together could have formed a Gallic-​style double forum,
with the main road from the Balkerne Gate leading straight to it. This complex formed
a monumental focus for the city. The remains of the podium survive below the Norman
castle, and it was 23.5m wide and 32m long (Drury et al. 1984). Nothing remains of the
Figure 37.3  Plan of Canterbury.
Source: Millett (2007: figure 5.15). Reproduced by kind permission of the Kent History Project, Kent County Council,
from plans provided by the Canterbury Archaeological Trust.
776

776   Louise Revell

superstructure, so it is impossible to reconstruct its full appearance, but it was presum-

ably a classical-​style temple. Raised up on a 4m-​high podium, it must have been a domi-
nant feature on the skyline. The large podium temple was set within an architecturally
elaborate complex. In front of the temple stood an altar, set within a paved area. This was
approached through a monumental gateway, set within an arched screen wall fronting
onto the main road. The complex seems to have undergone a major refurbishment in
the late second century, and it may be that at this point it was linked with a newly con-
structed theatre in the adjoining insula (Crummy 1982), which theatre has been dated to
the second or third century, with no evidence for earlier structures.
While these three examples provide the clearest evidence of temples as part of a
monumental centre, we see hints of this in other towns. At Caistor-​by-​Norwich, two
Romano-​Celtic temples stood in the insula to the north of the forum (Atkinson 1931),
and at Caerwent, when a temple was constructed within the town walls in the fourth
century, it was in the insula adjoining the forum (Brewer 1993). At Gloucester and
Wroxeter, it has been suggested that large colonnaded insulae adjacent to or opposite
the forum represent the temenos of a temple (Hurst 1999; White et al. 2013: 92–​93). At
Cirencester, the relationship is more problematic. It has been suggested that insula 6 was
a temple from the evidence of a large open courtyard surrounded by porticoes on at least
three sides, constructed in the mid-​second century (Holbrook 1998). However, while it
was in the insula adjoining the forum and so looks related on a plan, in reality there was
no immediate link between them at street level, as this complex was placed behind the
basilica, and so was at the opposite end to the forum entrance.
These ritual centres formed one way temples featured prominently within the urban
experience. A second way was to mark key points and junctions within the town grid.
One was the entrance to the town, moving through the peripheral zone of outside
the boundary, with its cemeteries and industrial working, and moving into the town
itself (Goodman 2007). As the traveller approached the western Balkerne Gate into
Colchester, they would see a square Romano-​Celtic temple separated from the road by
a gravelled court containing the plinth for a statue or an altar (Crummy 2006). This was
constructed in the first quarter of the second century, although there is a suggestion of
an earlier shrine, and it was probably dedicated to Mercury. There may have been a sec-
ond, smaller temple further outside the town, although the evidence for this is problem-
atic. At Silchester, a temenos stood just inside the eastern entrance to the town, which
contained at least two temples and associated structures of unknown date. An alterna-
tive location was on key roads through the towns, as can be seen at Wroxeter (Figure
37.4), where three of the four known temples run along the main road through the town
(White et al. 2013). At Verulamium, the triangular temple marked the splitting of the
main road into the town on the eastern side, and a temple stood slightly back from the
road in insula 21 (Niblett and Thompson 2005). Other complexes stood further away
from the town, potentially connected through processions, but still a distinct extra-​
mural ritual focus. This is most clearly seen at the Gosbecks complex at Colchester, which
formed a sanctuary complex of temple and theatre (Hull 1958: 259–​269). A similar phe-
nomenon can be seen at Caistor-by-Norwich, where a temple stood to the north-​east
 Urban Monumentality in Roman Britain    777

Figure 37.4  Plan of Wroxeter.

Source: White et al. (2013: figure 4.21).

of the town, reached from the town by a street diagonal to the main street grid, linking
this temple with the pair of temples in the town centre (Bowden 2013).
One characteristic of some of these religious areas is the clustering of Romano-​Celtic
temples in groups of two or more. We have seen this already at Silchester, where the
778

778   Louise Revell

insula 30 complex comprised two or possibly three shrines within a single temenos wall,
and possibly at the Balkerne Gate at Colchester. This is repeated at Caistor-​by-​Norwich,
where two Romano-​Celtic temples were found in insula 9, probably dating to the late
second century (Atkinson 1931). Recent excavations at Tabard Square in London, on
the south banks of the Thames, have revealed a pair of temples constructed in the late
first or early second century (Durrani 2004). An open space with a mortar floor stood
between them, with evidence for stone plinths decorating it. An inscription from the site
recorded a dedication to the Numina of the emperors and Mars Camulos (RIB 3014). At
Verulamium, two Romano-​Celtic temples were located in insula 17, although these are
only known from survey, and so it is difficult to understand the sequence of their con-
struction. However, in both cases, architectural differences demonstrate that neither was
envisaged as an identical pair. This is clearest at Caistor, where the two temples were of
different dimensions and different construction methods. Temple B was approximately
13.7m × 12.8m, with walls constructed of irregularly shaped stones, with more regular
pieces used for quoins at the corners. Temple A was smaller, at approximately 6.1m ×
5.5m, and was constructed of rubble walls faced with flints, and bonding courses of tile.
It is possible that temple A was constructed first, although the dating evidence is impre-
cise, and could equally indicate simultaneous construction. This difference in size is also
evident at Silchester; however, the London temples have been interpreted as a deliberate
pair, both seeming to be 11 metres × 11 metres and sharing the same alignment.
The variability in religious experience is reinforced by the presence of temples to the
so-​called oriental deities within the some towns. Perhaps the most well-​known evi-
dence for this comes from London, with the temple of Mithras on the eastern bank of
the Walbrook, although this was not constructed until the mid-​third century. A tem-
ple to Isis is hinted at within the epigraphic evidence, with a graffito on a pot reading
londoni ad fanum isidis (at the temple of Isis, London; RIB 2.8.2503.127; discussed
in Haynes 2000), and an inscription recording the rebuilding of the temple to Isis by
the imperial governor in the third century (RIB 3001). A further urban Mithraeum may
have been built in Leicester during the second century, although this has not been fully
published, and its identification has been questioned (Crompton 1971: 75; Sauer 2004).
Nevertheless, the preliminary report of a basilica-​style structure with the aisles raised
slightly higher than the nave does echo the form of that found in London (Shepherd
1998). There is also a temple to Serapis at York (Ottaway 1993:  88). Also associated
with the eastern deities is the cult of Cybele, which has been argued to be present in
Verulamium. Here, an unusual shaped triangular temple was excavated by Wheeler, and
consisted of an internal portico, with a series of shrines or rooms first constructed in the
early second century.
While temples within towns are usually overlooked as a group (although see Esmonde
Cleary 2005; Creighton 2006), they form an important feature in the monumental-
ity of the townscape. The rituals carried out in the urban temples, such as animal sac-
rifice and the burning of incense, would have added to the sensory experience of the
town: the noises of the animal victims, and the smell of the burning meat, or the wafting
of incense. Processions, and spectacles in the theatres (whatever form they took), would
 Urban Monumentality in Roman Britain    779

have punctuated the daily grind, adding a sense of spectacle to the urban experience.
One element which is clear is the variability in religious experience. Within the town,
there were the monumental centres, single or double shrines, potentially set at strategic
points within the towns, and the shrines of the so-​called oriental cults. Taken together,
they provide a picture of a vibrant and varied religious experience within the towns.

Space for Spectacles and Bathing

The relationship between the temples and both baths and theatres in the previous sec-
tion raises the question of a third form of monumental architecture within the town—​
buildings, which might be loosely grouped together as those for entertainment. These
might be considered somewhat trivial compared to fora and theatres, but they were
an important feature of Roman urbanism. Buildings for spectacles (theatrical games,
gladiatorial games, and chariot-​racing) and the large public baths were spaces where
core tenets of Roman-​ness were enacted. Whether they encouraged the ideals of cleanli-
ness or the competition between gladiators or charioteers, these buildings allowed their
users to participate in an ideal of urbanism which spread throughout the empire. The
feature all such buildings shared was the idea of public living.
Beginning with buildings for spectacles, amphitheatres dominate, with fewer thea-
tres and only one known stadium. However, the unexpected discovery of the Colchester
circus in 2004 allows the possibility that other theatres and amphitheatres may yet be
revealed. This is reinforced by the recent geophysical survey of Caistor-​by-​Norwich,
which revealed a possible theatre in the insula opposite the two Romano-​Celtic tem-
ples (Bowden and Bescoby 2008). However, some structures thought to have been
amphitheatres or theatres are unlikely. The so-​called amphitheatre at Caerwent illus-
trates some of these problems: this was an antiquarian excavation, poorly reported. The
walls do not fit well into the surrounding buildings if this was an amphitheatre, and, if it
was indeed one, it seems to have been a late construction and never completed (Ashby
et al. 1905).
The association between theatres and temples has already been discussed, but how do
these work as buildings in their own right? There is evidence for seven urban theatres
from Britain, from a total of six towns (this includes two at Colchester). However, the
theatre at Caistor-​by-​Norwich is known only from the recent geophysics, and that at
Brough-​on-​Humber is assumed from an inscription commemorating its construction
(RIB 707). Only Canterbury and Verulamium have produced any form of stratigraphy,
but both show the same sequence of development. The former was constructed in the
first century ad, and Wilmott (2007) has argued that this initially followed the plan of
the Gallo-​Roman hybrid theatre–​amphitheatre, where the seating extends beyond the
usual semi-circle, creating a larger orchestra that could then be used as the arena for
gladiatorial games. However, in reality it had an unusually elongated cavea, which does
not suggest its use as an amphitheatre (Frere and Simpson 1970). It was then substantially
780

780   Louise Revell

altered in the third century ad to create a more classical-​style theatre. This pattern is also
seen with the theatre at Verulamium (Kenyon 1935), which was constructed in the mid-​
second century ad, again as a Gallo-​Roman theatre. It was rebuilt within two or three
decades, and again the new layout resembled a more classical-​style theatre. In contrast,
the theatre at Gosbecks, three miles west of Colchester, was constructed in a classical
style from its earliest phase, although there is some uncertainty over the arrangement of
the stage (Dunnett 1971). The theatre within Colchester itself seems to have been of clas-
sical style, although the stratigraphy is unclear and the remains may not represent the
earliest phase (Crummy 1982). The Gallo-​Roman theatres have been the subject of much
debate (summarized in Sear 2006: 98–​101), and they have variously been argued to be
hybrid theatre–​amphitheatres because of the shape of the seating area, or not theatres at
all because of the lack of a substantial stage-​setting. The rebuilding of two of the British
examples as more classical-​style theatres suggests that they were considered theatres of
some type, but that, over time, different architectural arrangements were desired by the
townspeople.
Amphitheatres are more common than theatres throughout the province overall,
but in towns the proportions are more equal, with seven amphitheatres (excluding
Caerwent but including Aldborough) as opposed to seven theatres. Those that can be
dated were early constructions, with the earliest at Silchester constructed in timber dur-
ing the period c. ad 55–​75. This was followed by a group dated to the Flavian period:
the first timber phase at London, and the masonry amphitheatres at Chichester and
Dorchester, while the amphitheatre at Cirencester dates to the early second century ad.
Unlike theatres, these never formed part of the monumental centres of the town, and
were positioned either at the outer edge of the town or just outside. Nevertheless, their
positions near major routes allowed them to provide a dominant landmark within the
cityscape. The amphitheatre at Cirencester, for example, stood just outside the western
bath gate, on the main west–​east road through the town, which ran past the front of the
forum. The amphitheatre at Dorchester reused a Neolithic henge monument. There is
an obvious functional explanation, as this reuse would have reduced the effort needed
for its construction. However, this may also have served as a memorial to the past occu-
pation of the area, creating an idea of continuity which we find in the choice of location
of buildings in other towns. This reflects an unexpected pattern: Welch (1994, 2007) has
argued that the growth in amphitheatres during the mid to late Republic was due to their
association with the military and with veteran settlements. However, although we do
see non-​urban amphitheatres associated with fortresses in Britain (Wilmott 2007), the
urban examples do not show this connection. Only London has definite military con-
nections, with the Cripplegate fort and some form of military installation at Southwark,
while no colonia has produced evidence for an amphitheatre. However, other towns
with amphitheatres, such as Chichester, Cirencester, and Silchester, do not seem to have
had a military origin. This is not to deny the military association as a general phenom-
enon, but it does suggest that, in Britain, there were multiple routes to their adoption.
A type of building which can be included in this group is the monumental bathing
complexes. Discussing these in any depth is difficult due to the problematic nature of the
 Urban Monumentality in Roman Britain    781

evidence. In some towns, such as Colchester, there is no evidence for large public baths,
but this is attributable more to continuous occupation on the site than to certainty of
their absence. Furthermore, in most cases where a large set of public baths has been exca-
vated, this is only a partial excavation, in many cases revealing only part of the main suite,
or a series of snapshots difficult to piece together. However, in terms of their contribution
to the monumentality of the town, there are three important points to be made. The first
is that, while they were not necessarily the first public buildings the town possessed, they
were an early part of their monumentalization, and by the mid-second century all those
which we can date either were completed or were underway. The first phase of the baths
at Caerwent date to the late first century ad, at about the same time as the construction
of the forum (Nash-​Williams 1930). The baths at Chichester (Down 1978) and at Exeter
(Bidwell 1979) also date to the first century ad, while those at Canterbury (Blockley et al.
1995), Wroxeter (Ellis 2000), and Lincoln (Jones et al. 2003: 79–​80) date to the first half
of the second century. Others may have had similarly early dates, but lack published dat-
ing evidence. The second feature is that most of these bath buildings provide evidence
for their rebuilding or rearrangement. At Caerwent, the entire structure was completely
rebuilt seemingly within fifty years of its original construction, and then additional
rooms and a new wing added, apparently during the third and fourth centuries (Nash-​
Williams 1930). The baths at Wollaston House in Dorchester also seem to have been
rebuilt, which would have doubled the size of the complex, with further rooms added
at a later date. Other baths show less dramatic extensions: at Silchester the frigidarium
was extended and a second caldarium added in period 3, and the palaestra was widened
in period 4 (Boon 1974). At Canterbury and Chichester, rooms went through changes of
function, usually changing between heated and unheated.
The third point is that these were large buildings, which seem to have gone beyond the
functional row of cold, warm, and hot rooms. In many of the British urban bathhouses
there were associated steam rooms, plunge pools, palaestrae or basilicas, and swimming
pools (usually outdoor). Most of the bathhouses seem to show some evidence for an
entrance through either a peristyle/​palaestra or a basilica. Both Caerwent and Wroxeter
were entered through a large basilica, while at Caistor-​by-​Norwich and Silchester there
was evidence for a porticoed peristyle court. At Canterbury and Exeter these seem to
have contained open-​air swimming pools, and this also seems to have been the case in
the insula 19 baths at Verulamium. Similarly, we see multiple heated rooms at Lincoln,
Verulamium, and Leicester going beyond the single tepidarium and caldarium. Many
contain pools or plunge baths, whether heated or cold, and at Silchester there seems
to have been an additional sweat room. DeLaine (1992) has argued for a distinction
between bathing complexes primarily for hygiene in the Republic and the imperial
baths with an emphasis on pleasure and enjoyment, and the idea of variability in the
social role of bathing can be seen in the difference between the facilities and decoration
evident in the legionary and auxiliary bathhouses of Britain (Revell 2007b). These urban
bathhouses all adhere to the idea of bathing as a social experience, with the basilicas and
palaestrae providing space for exercise and chatting, while the multiple bathing facilities
allowed for a more leisurely visit.
782

782   Louise Revell

Often ignored or seen as frivolous, the baths and the buildings for games all share two
important features. The first is that they were usually constructed early in the history of
each town, either at its inception, or during the first fifty years. The second character-
istic is that they were subjected to continual rebuilding through to the fourth century.
Romano-​British urbanism has been categorized as a failed experiment, and the lack of
public facilities has been used to support this. However, it is an error to view a Roman
town as consisting of an essential suite of buildings, and there are examples of towns in
Gaul and Iberia which do not have baths, theatre, and amphitheatre. In Britain, we see
that most of the towns have a large public bathhouse with a range of bathing facilities.
Similarly, most also have a theatre or an amphitheatre, with Cirencester and Aldborough
possibly having both, and Colchester suffering an embarrassment of riches, with two
theatres and a circus. This was not true for all towns, and the recent publication of the
geophysics plan for Wroxeter demonstrates that, while it had a monumental bathhouse,
it lacked both theatre and amphitheatre (White et al. 2013). Similarly, there is no evi-
dence for buildings for spectacles at Lincoln, Exeter, Leicester, and York, although these
towns lack full excavation due to continued occupation. Nevertheless, the late second
century and early third century were a time of renovation for many of these buildings,
showing the ongoing reformulation of the idea of desired building type, and the devel-
opment of a discourse of architectural design within the province.

Space for Living and Working

A Roman town was more than a collection of public buildings; it was also a place for
living and working. This is an overlooked element of Romano-​British urbanism, and
contributes to the picture of towns as underpopulated spaces. While we have an image
of the elite (in spite of their lack of visibility epigraphically in contrast to elsewhere in the
empire), and can set them within the towns, there is less consideration of the non-​elite
within the towns. Urban populations are usually assumed to be low, although in reality
there has not been any systematic attempt to calculate the demography of the towns.
However, the archaeology suggests a more nuanced picture. If we take the example of
Wroxeter, the geophysics results show the centre of the town dominated by public build-
ings, and immediately around this a dense area of generally elite houses (White et al.
2013). The rest of the insulae do not necessarily show evidence for specific houses, but
the geophysical survey suggests industrial activity, and possibly timber housing. This
raises the possibility that the towns were more densely occupied than we have acknowl-
edged in the past.
Studies investigating the economic activities in individual towns have reinforced this
picture. Hall’s study (2005) of crafts in London represents a somewhat typical picture of
multiple, small-​scale trades. There is evidence for all stages of metalworking from the
smelting of ores through to the production of metal objects. Crucible fragments were
found on numerous sites, suggesting this was widespread throughout the city. Dumps
 Urban Monumentality in Roman Britain    783

of glass-​working debris indicate glass production, and similarly wooden offcuts indicate
carpentry and dumped leather shoe-​making and other leatherworking. There is also the
evidence for the production of foodstuffs. This picture is repeated at Verulamium, where
again there is evidence for varied forms of small-​scale industrial activity throughout the
town (Niblett and Thompson 2005). Metalworking and pottery production flourished
from the earliest phase of the town through to the mid-​third century. Those living in
the towns may also have been involved in low-​level agricultural production, with allot-
ments suggested from aerial photographs, and the possibility of back gardens. Some
made their living from the processing of agricultural produce for food, with remains
of corn-​drying ovens, bakeries, and brewing beer. As at London, there is evidence for
leather-​working and possibly tanning. There may also have been wool working or weav-
ing, with the possible remains of large wool carding combs, and possibly cloth prepara-
tion. Furthermore, at Colchester, ceramic, glass, and bone-​working industries have all
been identified, and, as well as general metalworking, there is both material and epi-
graphic evidence for a specialist coppersmith (Gascoyne 2013; RIB 194).
Those engaged in many of these industries would be living and working in the same
premises. A typical example is the timber-​built insula 14 buildings in Verulamium,
excavated and published by Frere (1972). Although there has been some debate over
the significance of the building technique (contrast Frere 1972 with Millett 1990), the
interpretation of the buildings’ function has not been debated. These combined living
quarters, workshop, and retail premises of metalsmiths, who worked mainly in bronze
but also in iron and gold, making and repairing metal items. Similar structures have
also been seen in Caerwent, although the sequence is difficult to interpret, and at No.
1 Poultry in London (Brewer 2006; Hill and Rowsome 2011). These buildings lack the
investment we see in the public buildings, being largely built of timber, not stone or
brick, with few rooms in addition to the workshop, and lacking decoration. It is only in
the late second century and the early third century that we see the replacement of timber
buildings with masonry with new internal arrangements and more elaborate decora-
tion. Faulkner’s analysis (2002) of the development of domestic architecture in Britain
shows an increase in the number of rooms occupied in town houses throughout the sec-
ond and early third centuries until c. 250, after which it remains broadly constant until a
decrease begins from c. 325, which continues throughout the fourth century.
This long sequence of construction and reconstruction can be seen at Dorchester.
At the Greyhound Yard site the initial first-​century sequence of buildings comprised
low-​density individual houses occupying street frontages, possibly with enclosures to
the rear (Woodward et al. 1993). Towards the end of the second century, buildings with
continuous stone footings began to be constructed, and by the early third century many
of the timber houses had been replaced, but on a roughly similar scale. In the late third
century these were then replaced with larger buildings, either through extensions, or
through demolition and rebuilding. Similarly at the County Hospital site, the earliest
buildings were of timber, dating possibly as early as ad 60–​65, and were replaced by a
new series of town houses, probably masonry built, with a layout of multiple rooms and
a portico, and decorated with mosaic floors. In the fourth century a large town house
784

784   Louise Revell

was built, set back from the street, and with evidence for elaborate polychrome mosa-
ics. This trajectory can also be seen in Silchester in the ongoing excavations in insula
9, where the early timber houses were replaced with two masonry houses during the
second century, and then, in the third century, these were amalgamated into a single
masonry house. This new house was multi-​roomed, arranged along a veranda (Fulford
and Clarke 2011). Wallace’s study (2014) of early Roman London reinforces this picture
of lack of investment in domestic structures, arguing for a lack of mosaic, marbles, large
receptions rooms, and other accoutrements associated with social display.
The development of non-​public structures stands in contrast to that of the public
buildings. Initially low key and timber built, they lack the evidence of investment seen
in the fora, temples, baths, theatres, and amphitheatres. Nevertheless, it is important
to recognize two key characteristics. The first is that, in spite of the comparative lack of
investment, these buildings are early constructions, and illustrate that Romano-​British
urbanism was more than the acquisition of a series of public buildings, but were places
where people lived and worked from the outset. Secondly, while these were initially
small scale, the townspeople were committed enough to the towns to continue investing
in urbanism, redeveloping as finance and opportunity arose.

Urban Variability

Thus far, I have argued that, in the first two centuries after the Claudian conquest, we
can see a pattern of urban monumentalization which focused on the public above the
domestic, and on spaces for public gatherings, whether political, religious, or for enter-
tainment, all of which broadly followed the ideal of urbanism set out in Rome itself.
Nevertheless, the one issue I have yet to raise is that of the variability between towns.
While all the chartered towns of Britain accord with this ideal to some extent, this does
not mean we see uniform patterns of urbanism, and it is this variability I will address in
this section.
While we lack complete plans for all towns, there are enough to demonstrate that
there are substantial differences between the towns. The first element of this is in the
size of the towns. The rank-​size graph of the area enclosed by defences for the towns
of Britain demonstrate that the public towns have a steep drop-​off, from the largest,
London, at over 130 hectares (Millett 1990: figure 62). Only a further three are over
60 hectares, with most clustering at 30–​50 hectares in size. The smallest public town
was Caistor-​by-​Norwich, at 14 hectares. There are obvious problems with this means
of quantification, as argued by Hurst (2005), because of the exclusion of suburbs from
these figures. Nevertheless, the symbolism of the city walls and their role in creating the
idea of inside and outside make this a factor in the creation of difference.
A second element is the distribution of public buildings, with some containing mul-
tiple examples, while others are more meanly furnished. Verulamium, for example, has
a monumental centre of forum, temple, and theatre, while elsewhere in the town we see
 Urban Monumentality in Roman Britain    785

further temples and the suggestion of public baths. The public buildings in Silchester are
more dispersed, but still include a forum, multiple temples, an amphitheatre, and large
town baths. Further west, Cirencester also seems to develop a monumental centre, with
a theatre and amphitheatre in addition to the forum, although no temples have been
securely identified. In contrast to these, there are towns with fewer public buildings,
such as Wroxeter, which survey evidence demonstrates to have been a densely occu-
pied town of 76.8 hectares, but which does not show the same level of monumentality
(see Figure 37.3): there was a large forum–​basilica with a substantial baths and macellum
occupying the whole of the opposite insula (White et al. 2013). However, we do not see
a theatre or amphitheatre forming part of the monumental core. While it is tempting to
relate this to size, Caistor-​by-​Norwich, much smaller than Wroxeter, was furnished with
a forum–​basilica, twin temples, a public baths, a theatre, and an amphitheatre (Frere
1971; Bowden and Bescoby 2008).
It is this variability in size and monumentality which has led to the categorization of
towns as more or less successful, as though there was an expectation of Roman towns
to be homogenous. However, various studies have pointed to the inherent variability of
urbanism within any single province of the empire (Keay 1995; Woolf 1997), and that, in
reality, urbanism was the creation of a hierarchy of towns. This was in part political, but
other factors came into play, such as economics, religion, pilgrimage, and so on. The lack
of textual evidence (including epigraphic) makes such factors difficult to understand,
but we can identify their influence on urbanism in Britain. In terms of political hier-
archies, the role of London as an administrative centre in the post-​Boudiccan period
creates a different form of monumentality from that seen in the first two decades of its
development (Wallace 2013). Cirencester and York may have similarly developed as pro-
vincial capitals as Britain was divided into two and then four provinces. Onto this, we
can add the possibility of a religious hierarchy, most notably Canterbury, with its ritual
centre of temple, baths, and theatre, and also Colchester as the centre for the provincial
imperial cult with the Temple to Deified Claudius. These different hierarchies and net-
works would have intersected, creating a complexity to urbanism within the province
which goes beyond a single criterion for their assessment.

Conclusion

There has been a tendency to view Romano-​British urbanism as something of a failed

experiment, reflected (and distorted) through a prism of the idealized Roman town. In
contrast, when towns are studied on their own terms, we see a different picture. The
emphasis on monumentality and public spectacle ties the towns of Britain into the
empire-​wide discourse of urbanism. The ongoing maintenance and reconstruction of
the buildings, as well as the gradual investment in commercial and domestic buildings,
demonstrate that this idea continued through to the late third and fourth centuries, by
which time a new concept of urbanism was emerging. At the same time, the differences
786

786   Louise Revell

between Britain and other provinces, as well as the differences between towns within
Britain, all point to an inherent variability. As Hurst has argued, the differences between
Gloucester and Cirencester were the result of ‘two communities expressing themselves,
through material culture we know as Romano-​British, in ways which convey their
identity and aspirations. They used that expression partly to be alike … but also … to
mark their differences’ (Hurst 2005: 303; see also Revell 2009). None of this is unique to
Britain, but in many ways echoes the picture in other provinces, warning against view-
ing the province in a vacuum. For example, urbanism in Britain is often argued to have
been a delayed development (Mattingly 2006), but Woolf (1997) has shown that the con-
struction of towns in Gaul could take up to two generations. Similarly, the majority of
towns in Iberia and Gaul do not differ substantially in size or monumentality from those
in Britain (Bedon et al. 1988; Bendala Galán 1993). Viewed within their wider context,
Romano-​British towns represent an important change in the lives of the inhabitants of
the provinces, and a key part of the political, social, economic, and cultural changes in
the province following the Roman conquest.

Abbreviation
RIB R. G. Collingwood and R. P. Wright, The Roman Inscriptions of Britain. Volume
I: Inscriptions on Stone. Oxford: Clarendon Press, 1965–​to date.

References
Primary Sources
Tacitus, Annals, trans. J. Jackson (London: Heinemann, 1925–37).

Secondary Soiurces
Alföldy, G. (1996). ‘Subject and Ruler: Subjects and Methods: An Attempt at a Conclusion’, in
A. Small (ed.), Subject and Ruler: The Cult of the Ruling Power in Classical Antiquity. Journal
of Roman Archaeology Supplementary Series 17. Ann Arbor, Michigan, 237–​259.
Ashby, T. (1906). ‘Excavations at Caerwent, Monmouthshire, on the Site of the Romano-​British
City of Venta Silurum, in the Year 1905’, Archaeologia, 60/​1: 111–​130.
Ashby, T., Hudd, A. E., and Martin, A. T. (1905). ‘Excavations on the Site of the Romano-​British
city of Venta Silurum’, Archaeologia, 50: 88–​115.
Atkinson, D. (1931). ‘Caistor Excavations, 1929’, Norfolk Archaeology, 26: 93–​139.
Balty, J. C. (1991). Curia ordinis: Recherches d’architecture et d’urbanisme antiques sur les curies
provinciales du monde romain. Brussels: Palais des Academies.
Bedon, R., Chevallier, R., and Pinon, P. (1988). Architecture et urbanisme en Gaule romaine.
Paris: Editions Errance.
Bendala Galán, M. (1993). The Hispano-​Roman Town. Barcelona, Ministerio of Cultura: Àmbit
Servicios Editoriales.
 Urban Monumentality in Roman Britain    787

Bennett, P. (1981). ‘68–​69A Stour Street’, Archaeologia Cantiaca, 97: 279–​281.

Bennett, P., and Nebiker, D. (1989). ‘No. 76 Castle Street’, Archaeologia Cantiaca, 107: 283–​286.
Bidwell, P. T. (1979). The Legionary Bath-​House and Basilica and Forum at Exeter. Exeter: Exeter
City Council, University of Exeter.
Blagg, T. F.  C. (1984). ‘Roman Architectural Ornament in Kent’, Archaeologia Cantiaca,
100: 65–​80.
Blagg, T. F. C. (1990). ‘Architectural Munificence on Britain: The Evidence of the Inscriptions’,
Britannia, 21: 13–​31.
Blockley, K., Blockley, M., Blockley, P., Frere, S., and Stow, S. (1995). The Archaeology
of Canterbury:  Excavations in the Marlowe Car Park and Surrounding Areas.
Canterbury: Canterbury Archaeological Trust.
Boon, G. C. (1973). ‘Serapis and Tutela: A Silchester Coincidence’, Britannia, 4: 107–​114.
Boon, G. C. (1974). Silchester, the Roman Town of Calleva. Newton Abott: David & Charles.
Bowden, W. (2013). ‘The Urban Plan of Venta Icenorum and its Relationship with the Boudican
Revolt’, Britannia, 44: 145–​169.
Bowden, W., and Bescoby, D. (2008). ‘The Plan of Venta Icenorum (Caistor-​by-​
Norwich): Interpreting a New Geophysical Survey’, Journal of Roman Archaeology, 21/​1: 324–​335.
Brewer, R. J. (1993). ‘Caerwent—​Venta Silurum: A Civitas Capital’, in S. J. Greep (ed.), Roman
Towns:  The Wheeler Inheritance. A  Review of Fifty Years Research. London:  Council for
British Archaeology Research Report 93: 56–​65.
Brewer, R. J. (2006). Caerwent Roman Town. Cardiff: Cadw.
Crawford, M. (1996). Roman Statutes. London: Bulletin of the Institute of Classical Studies 64.
Creighton, J. (2006). Britannia: The Creation of a Roman Province. London: Routledge.
Crompton, J. (1971). ‘Archaeology in Leicestershire and Rutland April 1968–​March 1970’,
Transactions of the Leicestershire Archaeological and Historical Society, 45: 74–​78.
Crummy, N. (2006). ‘Worshipping Mercury on the Balkerne Hill, Colchester’, in P. Ottaway
(ed.), A Victory Celebration: Papers on the Archaeology of Colchester and Late Iron Age Roman
Britain. Colchester: Friends of Colchester Archaeological Trust, 55–​68.
Crummy, P. (1982). ‘The Roman Theatre at Colchester’, Britannia, 13: 299–​302.
Cunliffe, B. W., and Fulford, M. (1982). Bath and the Rest of Wessex. Oxford:  Oxford
University Press.
DeLaine, J. (1992). ‘New Models, Old Modes: Continuity and Change in the Design of Public
Baths’, in H. J. Schalles and P. Zanker (eds), Der Romische Stadt im 2 Jahrhundert n. Chr. Der
Functionswandel des offentlichem Raumes. Cologne: Reinland-​Verlag GmbH, 257–​275.
Down, A. (1978). Chichester Excavations III. Chichester: Phillimore and Co.
Drury, P. J., Bayley, J., Blagg, T. F.  C., Evans, J., Going, C. J., Hassall, M. W.  C., Niblett, B.
R. K., and Wickenden, N. P. (1984). ‘The Temple of Claudius at Colchester Reconsidered’,
Britannia, 15: 7–​50.
Dunnett, R. (1971). ‘The Excavation of the Roman Theatre at Gosbecks’, Britannia, 2: 27–​47.
Durrani, N. (2004). ‘Tabard Square Excavations, Southwark’, Current Archaeology,
192: 540–​547.
Ellis, P. (2000). The Roman Baths and Macellum at Wroxeter: Excavations by Graham Webster,
1955–​85. London: English Heritage.
Esmonde Cleary, A. S. (2005). ‘Beating the Bounds: Ritual and the Articulation of Urban Space
in Roman Britain’, in A. MacMahon and J. Price (eds), Roman Working Lives and Urban
Living. Oxford: Oxbow Books, 1–​17.
788

788   Louise Revell

Euzennat, M., and Hallier, G. (1986). ‘Les Forums de Tingitane: Observations sur l’influence
de l’architecture militaire sur les constructions civiles de l’Occident romain’, Antiquités afric-
aines, 22/​1: 73–​103.
Faulkner, N. (2002). ‘The Debate about the End:  A  Review of Evidence and Methods’,
Archaeological Journal, 159: 59–​76.
Fox, G. E., and Hope, S. J. W. H. (1893). ‘Excavations on the Site of the Roman City at Silchester,
Hants, in 1982’, Archaeologia, 53/​2: 539–​573.
Frere, S. S. (1971). ‘The Forum and Baths at Caistor by Norwich’, Britannia, 2: 1–​26.
Frere, S. S. (1972). Verulamium Excavations. Volume I. London: Society of Antiquaries.
Frere, S. S. (1989). ‘Roman Britain in 1988. 1. Sites Explored’, Britannia, 20: 258–​326.
Frere, S., and Simpson, G. (1970). ‘The Roman Theatre at Canterbury’, Britannia, 1: 83–​113.
Fulford, M. G., and Clarke, A. (2011). Silchester: City in Transition. The Mid-​Roman Occupation
of Insula IX C. a.d. 125–​250/​300: A Report on Excavations Undertaken since 1997. London:
Society for the Promotion of Roman Studies.
Fulford, M., and Timby, J. (2000). Late Iron Age and Roman Silchester: Excavations on the
Site of the Forum–​Basilica 1977, 1980–​86. London: Society from the Promotion of Roman
Studies.
Gascoyne, A. (2013). Colchester, Fortress of the War God:  An Archaeological Assessment.
Oxford: Oxbow.
González, J. (1986). ‘The Lex Irnitana: A New Copy of the Flavian Municipal Law’, Journal of
Roman Studies, 76: 147–​243.
Goodman, P. J. (2007). The Roman City and its Periphery:  from Rome to Gaul.
London: Routledge.
Goodman, P. (2013). ‘Temple Architecture and Urban Boundaries in Gaul and Britain’, in
T. Kaizer, A. Leone, E. Thomas, and R. Witcher (eds), Cities and Gods: Religious Space in
Transition. Leuven: Peeters, 81–​96.
Gros, P. (1996). L’Architecture romaine. Paris: Les Manuels d’Art et d’Archeologie Antiques.
Hall, J. (2005). ‘The Shopkeepers and Craft Workers of Roman London’, in A. MacMahon and
J. Price (eds), Roman Working Lives and Urban Living. Oxford: Oxbow Books, 167–​190.
Haselgrove, C., and Millett, M. (1997). ‘Verlamion Reconsidered’, in A. Gwilt and C.
Haselgrove (eds), Reconstructing Iron Age Societies: New Approaches to the British Iron Age.
Oxford: Oxbow, 282–​296.
Haynes, I. (2000). ‘Religion in Roman London’, in I. Haynes, H. Sheldon, and L. Hannigan,
(eds), London Under Ground: The Archaeology of a City. Oxford, Oxbow Books, 85–​101.
Henig, M. (1993). Roman Sculpture from the Cotswold Region, with Devon and Cornwall.
Oxford: Oxford University Press.
Hill, J., and Rowsome, Pl (2011). Roman London and the Walbrook Stream Crossing: Excavations
at 1 Poultry and Vicinity, City of London. London: Museum of London Archaeology.
Holbrook, N. (1998). Cirencester:  The Roman Town Defences, Public Buildings and Shops.
Cirencester: Cotswold Archaeological Trust.
Hull, M. R. (1958). Roman Colchester. London: Society of Antiquaries.
Hurst, H. (1972). ‘Excavations at Gloucestershire, 1968–​7 1: First Interim Report’, Antiquaries
Journal, 52: 24–​69.
Hurst, H. (1999). ‘Civic Space at Glevum’, in H. Hurst (ed.), The Coloniae of Roman Britain: New
Studies and a Review. Ann Arbor: Journal of Roman Archaeology, 152–​160.
Hurst, H. (2005). ‘Roman Cirencester and Gloucester Compared’, Oxford Journal of
Archaeology, 24/3: 293–​305.
 Urban Monumentality in Roman Britain    789

Isserlin, R. M. J. (1998). ‘A Spirit of Improvement? Marble and the Culture of Roman Britain’, in
R. Laurence and J. Berry (eds), Cultural Identity in the Roman Empire. London: Routledge,
125–​155.
Jones, M. J. (2002). Roman Lincoln: Conquest, Colony and Capital. Stroud: Tempus.
Jones, M. J., Stocker, D., and Vince, A. (2003). The City by the Pool: Assessing the Archaeology of
the City of Lincoln. Oxford: Oxbow.
Joyce, J. G. (1873). ‘The Excavations at Silchester’, Archaeological Journal, 30: 10–​27.
Kaiser, A. (2000). The Urban Dialogue: An Analysis of the Use of Space in the Roman City
of Empuries, Spain. British Archaeology Reports International Series 901. Oxford:
Archaeopress.
Keay, S. J. (1995). ‘Innovation and Adaptation:  The Contribution of Rome to Urbanism in
Iberia’, in B. W. Cunliffe and S. J. Keay (eds), Social Complexity and the Development of Towns
in Iberia. London: Proceedings of the British Academy, 86: 291–​337.
Kenyon, K. M. (1935). ‘The Roman Theatre at Verulamium, St Albans’, Archaeologia, 84: 213–​261.
Laurence, R., Esmonde ​Cleary, S., and Sears, G. (2011). The City in the Roman West, c.250 bc–​c.
ad 250. Cambridge: Cambridge University Press.
Laurence, R., and Newsome, D. J. (2011). Rome, Ostia, Pompeii:  Movement and Space.
Oxford: Oxford University Press.
Lefebvre, S. (2004). ‘Espace et pouvoir local dans les provinces occidentales:  Quelques
remarques’, in C. Auliard and L. Bodiou (eds), Au jardin des Hespérides: Histoire, société
et épigraphie des mondes anciens. Mélanges offerts à Alain Tranoy. Rennes:  Presses
Universitaires de Rennes, 379–​406.
MacDonald, W. L. (1986). The Architecture of the Roman Empire. II. An Urban Appraisal. New
Haven: Yale University Press.
Martin, A. T. (1898). ‘Notes in the Roman Basilica at Cirencester Lately Delivered by Wilfred
J. Cripps Esq., C.B.’, Transactions of the Bristol and Gloucestershire Archaeological Society, 21:
70–​78.
Mattingly, D. (2006). Britain: An Imperial Possession. London: Allen Lane.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Millett, M. (2001). ‘Review of Fulfold and Timby’, Archaeological Journal, 158: 394–​395.
Millett, M. (2007). ‘Roman Kent’, in J. H. Williams (ed.), The Archeology of Kent to ad 800.
Woodbridge: Boydell Press, 135–​184.
Milne, G. (1992) (ed.). From Roman Basilica to Medieval Market: Archaeology in Action in the
City of London. London: HMSO.
Nash-​Williams, V. E. (1930). ‘IX.—​Further Excavations at Caerwent, Monmouthshire, 1923–​5’,
Archaeologia (Second Series), 80: 229–​288.
Niblett, R. (2001). Verulamium: The Roman city of St Albans. Stroud: Tempus.
Niblett, R., and Thompson, I. (2005). Alban’s Buried Towns: An Assessment of St Alban’s
Archaeology up to ad 1600. Oxford: Oxbow Books.
Ottaway, P. (1993). Roman York. London: Batsford/​English Heritage.
Revell, L. (1999). ‘Constructing Romanitas: Roman Public Architecture and the Archaeology
of Practice’, in P. Baker, C. Forcey, S. Jundi, and R. Witcher (eds), TRAC98. Proceedings
of the Eighth Annual Theoretical Roman Archaeology Conference. Leicester 1998.
Oxford: Oxbow, 52–​58.
Revell, L. (2007a). ‘Architecture, Power and Politics: The Forum-​Basilica in Roman Britain’, in J.
Sofaer (ed.), Projecting Identities. London: Blackwells, 127–​151.
790

790   Louise Revell

Revell, L. (2007b). ‘Military Bath-​Houses in Britain: A Comment.’ Britannia, 38: 230–​237.

Revell, L. (2009). Roman Imperialism and Local Identities. Cambridge:  Cambridge
University Press.
Revell, L. (2013). ‘The Written City:  Political Inscriptions from Roman Baetica’, in G.
Sears, P. Keegan, and R. Laurence (eds), Written Space in the Latin West: 200 bc–​ad 300.
London: Bloomsbury, 231–​245.
Rogers, A. (2013). Water and Roman Urbanism: Towns, Waterscapes, Land Transformation and
Experience in Roman Britain. Leiden: Brill.
Sauer, E. (2004). ‘Not just Small Change: C​ oins in Mithraea’, in M. Martens and G. de Boe (eds),
Roman Mithraism: The Evidence of the Small Finds. Papers of the International Conference,
Tienen, 7–​8 November 2001. Brussels, 327–​353.
Sear, F. (2006). Roman Theatres: An Architectural Study. Oxford: Oxford University Press.
Shepherd, J. D. (1998). The Temple of Mithras, London:  Excavations by W.  F. Grimes and
A. Williams at the Walbrook. London: English Heritage.
Wacher, J. S. (1995). The Towns of Roman Britain. 2nd edn. London: Batsford.
Wallace, L. (2013). ‘The Foundation of Roman London:  Examining the Claudian Fort
Hypothesis’, Oxford Journal of Archaeology, 32/​3: 275–​291.
Wallace, L. M. a. (2014). The Origin of Roman London. Cambridge: Cambridge University Press.
Wallace-​Hadrill, A. (2011). ‘The Monumental Centre of Herculaneum:  In Search of the
Identities of the Public Buildings’, Journal of Roman Archaeology, 24: 121–​160.
Welch, K. (1994). ‘The Roman Arena in Late-​Republican Italy: A New Interpretation’, Journal of
Roman Archaeology, 7: 59–​80.
Welch, K. E. (2007). The Roman Amphitheatre: From its Origins to the Colosseum. New York
and Cambridge: Cambridge University Press.
White, R. H. (2002). Wroxeter: Life & Death of a Roman City. 2nd edn. Stroud: Tempus.
White, R. H., Gaffney, C., and Gaffney, V. L. (2013). Wroxeter, the Cornovii and the Urban
Process: Final Report on the Wroxeter Hinterland Project 1994–​1997, ii. Characterizing the
City. Oxford: Archaeopress.
Wilmott, T. (2007). The Roman Amphitheatre in Britain. Stroud: Tempus.
Woodward, P. J., Davies, S. M., and Graham, A. H. (1993). Excavations at the Old Methodist
Chapel and Greyhound Yard, Dorchester, 1981–​1984. Dorchester: Dorset Natural History &
Archaeological Society.
Woolf, G. (1997). ‘The Roman Urbanization of the East’, in S. E. Alcock (ed.), The Early Roman
Empire in the East. Oxford: Oxbow Books.
Zanker, P. (1998). Pompeii: Public and Private Life. Cambridge, MA: Harvard University Press.
 Chapter 38

The Expl oitat i on of

An im als in Roma n Bri ta i n

Mark Maltby

Introduction: The History of
Romano-​British Zooarchaeology

The study of the exploitation of animals in Roman Britain using archaeological evi-
dence is a relatively recent development. Although classic studies of the province often
included sections on animal husbandry or diet, the evidence was based on a very limited
number of archaeological investigations and often relied heavily on historical inference
drawn from Roman agronomists (e.g. White 1970). The advent of historical zooar-
chaeology from the 1970s onwards has resulted in the study of assemblages from many
hundreds of Romano-​British sites, producing a burgeoning and increasingly unwieldy
archive. As the information has increased, several syntheses for the evidence have been
produced (e.g. Grant 1989, 2004; King 1991, 1999; van der Veen and O’Connor 1998; Cool
2006; Albarella et al. 2008; Hesse 2011). This chapter will both draw upon these previous
surveys and incorporate new evidence.

The Scope of This Review

Evidence for the exploitation of animals in Roman Britain is derived from many facets
of archaeological and, to a lesser extent, documentary sources. These will be mentioned
where relevant, but this chapter will focus primarily on the zooarchaeological evidence.
It will largely exclude discussion of the roles of animals in ritual and religion, as this is
covered in other chapters in this volume (e.g. Smith and Weekes, this volume). The main
792

792   Mark Maltby

theme that will be examined is that of innovation. Put simply, in what ways did human
exploitation of animals change in the Romano-​British period?
It should be noted from the outset that, although there are some common trends that
can be seen with regard to the exploitation of animals in Roman Britain, there are also
many variations. It can be shown that bone assemblages can vary significantly chrono-
logically, regionally, and intra-​regionally, between different types of settlement, between
settlements of the same type, and within individual settlements themselves. Arguably, it
is these differences that tell us more about the complexities of the pastoral economy and
the meat diet than the similarities. The review will focus initially on domestic mammals
and their exploitation. This will be followed by a shorter summary of the evidence for
the exploitation of other mammals, birds, and fish.

Cattle

Previous reviews of Romano-​British diet have concluded that beef was comfortably the
meat most commonly consumed. Cattle bones form the highest component of many
faunal assemblages (King 1984, 1991, 1999; Grant 2004), whereas Iron Age assemblages
are more commonly dominated by sheep/​goat (Hambleton 1999; Albarella 2007). Even
in cases where minimum number calculations indicate that more sheep were eaten than
cattle, the much heavier carcass weight of cattle would mean that beef still formed the
bulk of the meat diet (Grant 2000: 428).
King’s analyses (1984, 1999) of species representation in a wide range of assemblages
demonstrated that, in addition to the general dominance of cattle, there were variations
in the percentages of cattle, sheep/​goat, and pig on different types of site. This calcula-
tion masks a considerable amount of variation, which has been encountered in assem-
blages both within and between settlements, but King’s surveys showed that cattle tend
to be best represented on military sites and large towns—​settlements that would have
contained higher proportions of people not directly involved with food production.
The increased demand for food created by urban populations and military personnel
is likely to have been a major factor in the change of emphasis in Romano-​British meat
production.
The consensus view has been that the dominance of cattle tended to increase in the
later Roman period following King’s original results (1984) (Grant 2004; Albarella 2007),
with cattle forming a higher proportion of the bones on all types of settlement. Although
this generalization can still be supported by the evidence, there are a lot of variations and
numerous exceptions to this trend. For example, in a recent survey of Romano-​British
assemblages from twenty-​eight sites in fourteen large towns where material from dif-
ferent periods could be compared, cattle percentages (in relation to sheep/​goat and pig)
increased in the later Roman period in seventeen cases but decreased or remained static
in the other eleven (Maltby 2010: 266–​267). In addition to changes in the diet, there are
a number of factors that can account for variations in species in assemblages, including
 The Exploitation of Animals in Roman Britain    793

differential preservation, butchery, and disposal practices. More detailed and critically
evaluative reviews of species representation are required to provide a more comprehen-
sive understanding of the extent to which beef production became more important in
the later Roman period.
A major change that accompanied the increased emphasis on beef production con-
cerned how cattle carcasses were processed. It has long been recognized that butchery
methods involved the greater use of cleavers and heavy blades in carcass-​processing
than has been encountered in British Iron Age assemblages (Grant 1989; Maltby 1989).
Assemblages from large urban and military sites in particular have evidence for con-
sistent treatment of carcasses utilizing these implements, which have left characteristic
butchery marks (Maltby 1989, 2007; Seetah 2006). In particular, the appearance of dis-
tinctive indentations (scoops) on the shafts of upper limb bones made by heavy blades
during the removal of meat is testament to the processing of carcasses by specialist
butchers. While such butchery is found frequently in urban and military assemblages,
similar filleting marks have been found much less consistently on small nucleated settle-
ments and villas. They tend to be even less frequent and often absent on other rural set-
tlements (Maltby 2007, 2010). These methods, which may well have had military origins,
as they have been evidenced in some early forts such as Exeter (Maltby 1979), would
have increased the speed of processing and were thus favoured by butchers dealing with
large numbers of carcasses (Seetah 2006).
There is abundant evidence for large-​scale processing of cattle in towns, and exam-
ples of substantial accumulations of waste emanating from this can be found in all the
major towns that have received extensive excavations (Maltby 1984, 2007, 2010). This
is one of the major reasons why species representation varies so greatly within them.
Areas (both intra-​and extra-​mural) that include large dumps of processing waste have
invariably produced high percentages of cattle. Examples include Rack Street, Exeter
(Maltby 1979), Chester Street, Cirencester (Maltby 1998), and Balkerne Lane, Colchester
(Luff 1993).
Other common finds on urban and military sites are cattle scapulae with distinc-
tive blade marks on the edges of the articular surface (glenoid) and neck. These
scapulae were disarticulated from the humerus. Holes in the blade of some of the
scapulae suggest they were hung probably during preservation by smoking and salt-
ing. Well-​known examples have been illustrated from York (O’Connor 1988) and
Lincoln (Dobney et al. 1996), and perforated specimens have been recorded in several
other major towns (Maltby 2007). In addition, scapulae with similar blade marks on
the neck and the glenoid to those commonly found in towns have also been recovered
on some villas and other rural settlements such as Owslebury in Hampshire (Maltby
1989) and Marsh Leys Farm in Bedfordshire (Maltby 2011). This raises the possibility
that preserved shoulders of beef may have been traded. There is evidence for a sub-
stantial increase in salt production during the Romano-​British period in many areas
of England (de Brisay and Evans 1975; Bradley 1992; Lane and Morris 2001; Hathaway
2006; Rippon 2008), raising the probability that much of the meat eaten was in the form
of cured products.
794

794   Mark Maltby

Marrow and grease were other cattle products that were commonly utilized in Roman
Britain. In addition to evidence that many bones were routinely broken for pot-​boiling,
there is evidence that upper limb bones were sometimes collected for bulk process-
ing. In addition to the characteristic filleting marks already described, these bones
were split longitudinally to release marrow. Large accumulations of these split bones
have been found mainly, but not exclusively, on major urban sites such as Gloucester
and Winchester (Maltby 2007). Such accumulations have often been interpreted as evi-
dence for the production of stock or broth. However, marrow can also be utilized for
other products. Dobney et al. (1996: 25), for example, have suggested that marrow was
obtained from mandibles in Lincoln and may have been used in oil lamps.
Studies of mortality patterns based on tooth ageing evidence have shown that in
many settlements a high percentage of cattle were not slaughtered until adulthood
(Grant 2004: 373; Maltby 2010: 288). In most cases, however, relatively few of the adult
cattle mandibles were from very old animals. Therefore, although cattle were undoubt-
edly valued for dairy products and traction power, cattle were commonly slaughtered
for beef as young or mature adults (perhaps mainly between 4 and 8 years of age). Many
major towns have high percentages of cattle of this age. Examples include Winchester,
Dorchester (Maltby 1994), Caerwent (Maltby 2010: 288), Lincoln (Dobney et al. 1996),
Silchester (Grant 2000; Ingrem 2006), Exeter (Maltby 1979), and Colchester (Luff 1993).
Similar results have been obtained in assemblages from small towns such as Alcester,
Warwickshire (Maltby 2001), and Heybridge, Essex (Albarella et al. 2008: 1837). Military
sites producing assemblages with high percentages of adult cattle include South Shields
(Stokes 2000), Loughor (Sadler 1997), and Portchester Castle (Grant 1975).
Cattle mortality rates on villas and other rural settlements appear to be more variable.
Although many assemblages contain high percentages of adult cattle, the peak of slaugh-
ter of animals of this age is often less marked. Examples include Owslebury, Hampshire
(Maltby 1994), Odell (Grant 2000), Marsh Leys Farm and Newnham in Bedfordshire
(Maltby 2011), and Abingdon and Barton Court Farm villa in Oxfordshire (Grant 2000).
Most of these assemblages include fairly high percentages of cattle slaughtered between
1 and 3 years of age. These animals were not required for breeding, working, or sale. This
may indicate that the inhabitants at these settlements were sufficiently wealthy to afford
to cull or acquire potentially productive immature cattle for their meat. However, this
is not necessarily related to status, as the phenomenon seems to occur at both villa and
non-​villa sites.
Veal was seemingly eaten in relatively small amounts compared to mature beef. Teeth
and bones of young calves have been found on many sites but never in very high per-
centages. Some of these calves may have been natural mortalities, but others, particu-
larly those found in towns such as Dorchester (Maltby 1994), probably reflect deliberate
slaughter of animals acquired for their meat.
Another aspect of cattle in Roman towns is the bias towards smaller cattle. Most
breadth measurements of cattle limb bones in assemblages from Roman towns are posi-
tively skewed. Examples can be drawn from Cirencester (Maltby 1998: 362), Winchester
(Maltby 2010: 150), and Alcester (Maltby 2001). This probably indicates that the majority
 The Exploitation of Animals in Roman Britain    795

of the cattle were females, although this interpretation is complicated by the introduc-
tion of larger cattle in some parts of Roman Britain. However, the comparative lengths
and breadths of metacarpals display sexual dimorphism, and measurements of com-
plete bones consistently indicate that most of the adult cattle in urban assemblages
belonged to females. Good examples can be found in Colchester (Luff 1993: 61), Lincoln
(Dobney et al. 1996), and Winchester (Maltby 2010: 148). Unfortunately most meta-
carpals on Roman sites have been broken to release the marrow, but distal measure-
ments also demonstrate the bias towards smaller (female) cattle in urban assemblages.
Examples include York (O’Connor 1988:  55), Exeter (Maltby 1979), and Dorchester
(Maltby 1994). If this interpretation is correct, it seems that the procurers of cattle in
towns and forts were preferentially acquiring mature but not very elderly female cattle,
and/​or farmers were supplying cattle at an age when their productivity in breeding and
milking decreased.
Assemblages containing a more even distribution of distal metacarpal breadths,
which may indicate the presence of higher percentages of adult males, include the rural
sites at Owslebury, Hampshire (Maltby 2010:  148), Frocester villa, Gloucestershire
(Noddle 2000), and Marsh Leys Farm, Bedfordshire (Maltby 2011). This may indicate
that plough oxen were more often likely to have been retained for slaughter on farms
after they had ceased working. Evidence for the use of cattle for ploughing is also indi-
cated by the presence of pathological conditions on cattle foot bones. In several sam-
ples, including those from Wroxeter (Hammon 2005) and several rural settlements in
Bedfordshire, severe pathology tended to be more prevalent in larger specimens. This
suggests that more males were affected.
Metrical analysis has also shown that some Romano-​British cattle were substantially
larger than their Iron Age counterparts. Increases in the average size of cattle from early
in the Roman period have been demonstrated in, for example, Hampshire (Maltby 1981)
and Essex (Albarella et al. 2008). The latter have also shown that exceptionally large cat-
tle were present at Great Holts Farm, Essex, in the third century ad and have argued
convincingly that new stock were introduced to Britain from the continent from the
early Roman period.
However, there are regional and chronological variations in the presence of larger cat-
tle in Roman Britain. Most of the sites that have produced evidence for substantial num-
bers of larger cattle are located in the south-​east and the Midlands. Assemblages from
south Wales and the south-​west do not show much in the way of size improvements
throughout the Roman period (Maltby 1981, 2010; Hammon 2005).

Sheep and Goats

Apart from the exceptional case of Uley, Gloucestershire, where goats were selected
specifically for slaughter at the temple (Levitan 1993), sheep greatly outnumber goats in
assemblages where the two species have been differentiated. In major towns, they have
796

796   Mark Maltby

not provided more than 10 per cent of any assemblage, even including horn core counts,
and they usually contribute much less than that (Maltby 2010: 268).
Sheep are regarded as the mainstay of Iron Age animal husbandry (Albarella 2007),
although their percentages in faunal assemblages can vary quite substantially region-
ally (Hambleton 1999; Albarella 2007). As already noted, they tend to be better repre-
sented on non-​villa rural settlements than on other types of Romano-​British settlement
(King 1984, 1999) and on early Roman sites, although there is a wide range of variability.
Minimum number counts indicate that sheep continued to be the most common ani-
mal slaughtered on many sites, possibly including some towns, but, as already discussed,
when carcass weights are brought into consideration, lamb and mutton were much less
important than beef in the average diet.
This is not to say that meat production was not an important consideration in sheep
husbandry. Although mortality data show a lot of variation, many faunal assemblages
have substantial numbers of sheep killed between 6 months and 3 years of age. Often
there is a peak of slaughter of sheep aged between 18 and 36 months, indicating a focus
on the culling of sub-​adult and young adult animals once they had reached a good car-
cass size. Some assemblages, particularly in the later Roman period, have quite high per-
centages of adult animals, supporting the contention that woollen textile production
was a significant consideration in sheep husbandry for some Romano-​British farmers
(Wild 2002; Grant 2004: 378; Maltby 2010: 290). Young lambs were slaughtered in quite
large numbers on some sites. In some cases, particularly on rural sites, these may simply
have been stock surplus to breeding or market requirements. However, as in the case
of veal, lamb may have been regarded as a luxury meat. At Colchester, there were high
percentages of lambs in several assemblages, particularly from intra-​mural sites (Luff
1993: 73). Relatively high percentages of lambs have also been found in assemblages in
the centres of several other towns (Grant 2004: 378; Maltby 2010: 290). Sheep associ-
ated with temples and foundation deposits have tended to consist mainly of juvenile and
immature animals (King 2005; Maltby 2012).
Although sheep as small as the native Iron Age stock continued to be exploited
throughout the Roman period, particularly in western and northern areas (Maltby 2010:
294–​295), there is, as in the case of cattle, evidence for significant improvement in the
stature of the stock in some areas of the province. The average sizes of sheep were slightly
larger in the towns of Gloucester, Cirencester, and Alcester than in Exeter in the south-​
west (Maltby 2001; 2010: 294–​295). Large sheep have been found on several sites in Essex
dating from the second century ad onwards (Albarella et al. 2008). Some of these sheep
may have derived from stock imported from the continent. Some breeds of sheep have
horned skulls, others are hornless. Hornless sheep have been found on a number of
Romano-​British sites in areas where only horned sheep were found previously (Maltby
2010: 181). Hornless skulls appear in early Roman features in Winchester and Dorchester
(Maltby 1994), indicating that the inhabitants had access to new types of sheep from
early in the Roman occupation. Such skulls have not been found on several Iron Age
and early Roman rural settlements in their hinterland and appear only in late Roman
deposits at Owslebury, where they were associated with large metapodials, implying
 The Exploitation of Animals in Roman Britain    797

that hornless sheep were larger than the horned variety (Maltby 1994). Assemblages
from later Roman Winchester included a higher proportion of hornless sheep that at
Owslebury and they also produced more bones from large sheep (Maltby 2010: 182),
suggesting that the hornless type may have been of larger stature. However, it seems
unlikely that the two types were kept totally isolated, and the effects of potential hybridi-
zation are poorly understood.

Pigs

Pigs generally rank third behind cattle and sheep/​goat in most Romano-​British assem-
blages (King 1999; Grant 2004; Maltby 2010: 264–​265; Hesse 2011). However, there are
again some consistent variations in their relative abundance. They tend to be better
represented on most military sites, major towns, and some villas than in small towns
and other rural settlements (King 1984, 1999). The reasons for this could include dietary
preferences, status, and the usefulness of pigs as meat producers for settlements where
large amounts of meat were consumed. Taking carcass weights into account, pork prod-
ucts were probably more commonly eaten than lamb and mutton in most military and
large urban centres.
There are some interesting variations in pig abundance that transcend gross settle-
ment comparisons. For example, pig remains are consistently very well represented in
urban deposits in towns in south-​east England, such as Colchester and London, com-
pared with many other urban assemblages elsewhere in England (Maltby 2010: 269).
This may reflect that more pigs were kept in the region. It may also reflect a continu-
ation of a late Iron Age phenomenon that saw high levels of pigs on some of the opp-
ida and other trading centres in the region, such as Silchester and Braughing (King
1984; Hambleton 1999; Grant 2000; Albarella 2007). This could be related to well-​
documented continental trading and cultural influences that emerged in that region
during the late Iron Age. There is evidence to suggest that pig carcasses were prepared
for shipment during that period—​a trade probably facilitated by the increase in salt
production (Maltby 2006). Smoked and cured ham and bacon are likely to have been
commonly traded products throughout the Roman period. Large-​scale processing of
pigs is evidenced, for example, in a dump of pig foot bones discarded at Nazeingbury,
Essex (Huggins 1978). The carcasses processed there may well have been destined for
the urban market.
Another striking pattern is that pigs are often better represented on sites in centrally
located sites within towns than in their suburbs. Examples include Caerwent, Exeter,
Dorchester, Winchester, and Wroxeter (Maltby 2010: 264–​265). This may again be linked
to the greater demand for pork products by those of higher status within the towns. The
link between wealth and pig consumption is also evidenced by their high percentages
at high-​status sites such as Fishbourne (Grant 2004) and the Winchester Palace site in
Southwark (Reilly 2005). In contrast, less prestigious rural settlements generally have
798

798   Mark Maltby

fewer pigs than found on urban settlements in their vicinity, as can be demonstrated in
the Dorchester and Winchester hinterlands (Maltby 1994, 2010). It would appear that
the supply and demand of pork were more heavily focused on urban markets and on
some high-​status settlements as well as major military centres.
As expected, few Romano-​British sites have produced large percentages of adult ani-
mals. Most pigs are killed for their meat while immature. There is again a lot of variability
between assemblages, but many include high percentages of pigs killed in their second
and third years. Younger piglets have been found less consistently, but form a significant
component of some assemblages, again particularly on some (but not all) sites near the
centres of some towns such as Caerwent, Silchester, Leicester, and Dorchester (Grant
2004: 379; Maltby 2010: 291). It has been speculated (e.g. Maltby 1994), but never proven,
that some pigs were raised within towns.
Discussions about the sizes of pigs in Roman Britain are limited. At Heybridge, Essex,
Albarella et al. (2008) have demonstrated from measurements of teeth and bones that
there were small but significant improvements in the sizes of pigs in the second century
ad. Pig distal tibiae breadths in later Roman Winchester were nearly all larger than those
from the neighbouring rural settlement of Owslebury (Maltby 2010: 203). Although this
may reflect greater numbers of males being consumed in the town, it may also indi-
cate that Winchester had access to pigs (both male and female) of larger stature than
those eaten at Owslebury. Similarly large pigs were found in Dorchester (Maltby 1994).
MacKinnon (2006) has shown from zooarchaeological, documentary, and pictorial evi-
dence that two types of pig were commonly exploited in Roman Italy: one was hairy
and slender, often fattened in woodland; the second tended to be larger, hairless, and
raised in sties. Similar variations in husbandry methods and stock types may account
for the variations observed in Roman Britain, although much more research is needed
on this topic.

Horses and Other Equids

Horses of course played an important role within the Roman army (Hyland 1990), and
one assumes that many were imported to Britain during the occupation to supplement
native stock. At Heybridge, Albarella et al. (2008) have shown a significant increase in
the average size of horses from the second century ad onwards. These were comparable
in size to a New Forest pony. This is slightly higher than averages obtained from several
Roman towns (Maltby 2010: 297).
Horses tend to be better represented on rural settlements and suburbs of towns than
in their centres, where they rarely form more than 5 per cent of the total cattle and horse
assemblage (Maltby 2010: 269–​270). This is because they were much less important as a
source of food than cattle, sheep, and pig. They tend to be found in relatively higher fre-
quencies on all types of rural settlement, often forming over 10 per cent of the total cattle
and horse assemblage (Maltby 1994), but there is little evidence that they were exploited
 The Exploitation of Animals in Roman Britain    799

frequently for food. Although butchery marks have occasionally been found on horse
bones on Romano-​British sites, they were exploited for meat much less intensively than
cattle. Their bones tend to be less fragmented, and there are many more examples of
partial and complete horses being deposited than those of cattle. Horses were primar-
ily valued as a means of transport and as beasts of burden. There are very few examples
compared to the other domestic species of immature horse bones on Roman sites, indi-
cating that most could expect to be kept alive for as long as they were considered to be
useful unless they died of natural causes.
Donkeys and mules have been positively identified on a small number of Roman sites
(Johnstone 2010). The potential presence of mules has been overlooked by most zooar-
chaeologists, although neither mules nor donkeys were likely to have been present in
Britain in large numbers.

Dogs and Cats

Dogs were commonly kept by the inhabitants of Roman Britain. Their carcasses were
occasionally skinned but very rarely eaten. They were kept as pets, as farm animals, as
guards, and, judging from the evidence of several mosaics, for hunting. It has long been
recognized that the Roman period was one where there was a significant increase in the
diversity of dogs. Dogs ranged greatly in stature from animals the size of wolfhounds to
those as small as a Yorkshire terrier (Harcourt 1974; Clark 1995). Such diversity indicates
the importation of new types. From metrical analysis alone it is impossible to assign the
different sizes of dogs to breeds, but there must have been specialist breeding.
Dogs, including peri-​natal puppies, have been found commonly as complete or par-
tial skeletons. Sometimes these are isolated burials; in other cases large numbers of dogs
were deposited in the same feature (sometimes with other skeletons). Examples of mul-
tiple burials have been found in Dorchester, Winchester (Maltby 2010), and Springhead,
Kent (Grimm 2008). To what extent these can be regarded as ritual depositions beyond the
burial of carcasses not required for processing is the subject of debate that goes beyond the
remit of this chapter (see Smith, this volume, for discussion of animals in ritual deposition).
Cats are less commonly found than dogs, and, although they have been recorded on
many sites, it is only usually in small numbers. They first appeared in Britain in the Iron
Age and may have been kept to control vermin as well as for pets in Roman settlements
(Kitchener and O’Connor 2010).

Hunted Animals

Generally speaking, deer remains occur very infrequently on Romano-​ British

sites. Red and roe deer rarely provide over 1 per cent of the food mammal counts on
800

800   Mark Maltby

urban sites, and some of these counts include antler collected for working (Maltby
2010: 271). Deer bones appear more frequently (but rarely in high percentages) on
some sites of higher status (Allen, forthcoming), supporting the widely held belief
that deer-​hunting may have been mainly the prerogative of the upper echelons of
Romano-​British society. This is further supported by the introduction of fallow deer
to Fishbourne Palace and the subsequent establishment of a herd, presumably for the
pleasure of the owners and their guests (Sykes et al. 2011). Confirmed identifications
of fallow deer have been restricted to a handful of Roman sites mainly in southern
England (Sykes 2010a).
Hares have also been found in many Romano-​British faunal assemblages but again
usually only in small numbers. They rarely form over 1 per cent of the mammal assem-
blage, and, where hares are present, it is usually on sites of high status such as Winchester
Palace, Southwark (Reilly 2005) or Whitehall villa, Northamptonshire (Sykes 2010b).
Authenticated identifications of rabbits in Roman Britain are extremely rare, the most
convincing example coming from Beddingham villa, Essex (Sykes and Curl 2010). This
was possibly imported from the Mediterranean.
Identifications of wild boar are complicated by their similarity to large domestic pigs,
particularly if they are killed when immature. They are occasionally recorded, however.
One of the best examples comes from the legionary fortress at Caerleon, south Wales.
Here a wild boar was eaten with venison and cranes, possibly at a banquet hosted by a
high-​ranking officer (Hamilton-​Dyer 1993).

Birds

General reviews of birds in Roman Britain can be found in Parker (1988), Yalden and
Albarella (2009: 95–​153), and, for large towns, Maltby (2010: 272–​279). The main results
of these surveys have shown that domestic fowl (chickens) became significantly more
important during the Roman period, although they would not have formed a significant
part of the diet. Although introduced to Britain in the Iron Age, their distribution was
restricted and their use as a source of food debatable (Sykes 2012). In the Roman period,
the evidence suggests that chickens became more ubiquitous but not evenly exploited.
Higher percentages of chickens have generally been found in urban and military assem-
blages than in those from small towns and villas, and they tend to be most poorly repre-
sented in non-​villa rural assemblages (Maltby 1997). Again, this would appear to reflect
the dietary preferences of different sectors of the Romano-​British population. Chickens
were exploited for their eggs as well as their meat and probably used for cockfighting
(Sykes 2012). They have also been found in substantial numbers of human graves and
associated with other ritual sites such as Uley (Levitan 1993). There is some evidence to
suggest that chickens, particularly males, tended to increase in size during the Roman
period (Albarella et al. 2008).
 The Exploitation of Animals in Roman Britain    801

The domesticated status of ducks and geese in Roman Britain is less clear (Albarella
2005). Ducks and geese of various sizes are present on Roman sites in modest num-
bers (Yalden and Albarella 2009: 102–​105). The former tends to outnumber the lat-
ter, which usually is found more frequently in later Roman sites. Other avian species
likely to have been eaten that occur quite frequently on Roman sites include pigeons
(some of which may have been domestic), woodcock, and a range of other waders,
including crane, plus partridge, and black grouse (especially in northern England).
Non-​native pheasants have been recorded on several sites, and a peafowl bone was
identified at Portchester Castle (Yalden and Albarella 2009). Other rarities included
the butchered wing of a great auk on the Isle of Portland (Maltby and Hamilton-​D yer
2012). Ravens and small corvids are common, but are rarely claimed to have been
eaten. The former has often been suggested to be linked with ritual practices (e.g.
Fulford 2001).

Fish

Locker (2007) has produced a comprehensive survey of fish assemblages found in

Roman Britain, and this section largely summarizes those results. The first point to
make is that fish remains have been found extremely rarely on late prehistoric sites in
southern Britain, despite the increasing number of excavations where sieved sampling
has been carried out. This has led to the suggestion that there may have been a taboo on
the exploitation of marine resources (Dobney and Ervynck 2007).
There is no doubt that fish were more commonly eaten during the Romano-​British
period, although the evidence is patchy and handicapped by the inconsistency of siev-
ing. Eel is the species that that has been found most frequently. The distribution of most
species tends to reflect whether they were locally available. For example, York has pro-
duced relatively high frequencies of freshwater species such as cyprinids (carp family),
whereas cod is more common on London sites. Salmon were quite common on sites in
the north and the Midlands (Locker 2007). Bream, bass, gadids (cod family), and wrasse
caught in the inshore waters off Portland were the species most commonly exploited,
as well as eels in Dorchester and neighbouring sites (Maltby and Hamilton-​Dyer 2012).
There is little evidence for deep-​sea fishing.
Most commentaries on Roman food have highlighted the importance of garum and
other fish-​based sauces in the diet. Although they have been found on several Romano-​
British sites (including urban deposits in London, Lincoln, York, and Dorchester) pro-
cessing sources of small fish locally available (Locker 2007), the occurrences are rare and
the importance of this food resource may have been over-​emphasized. Indeed, although
isotopic studies attest to the variable presence of marine foods in the diet of some of the
inhabitants, these studies have shown that such sources remained a small component of
the diet (Redfern et al. 2010).
802

802   Mark Maltby

Summary

This brief review of the zooarchaeological evidence has confirmed that the Romano-​
British meat diet was dominated by the acquisition of beef and, to a lesser extent, pork,
lamb, and chicken. Other meats including goat, venison, other game, and fish supple-
mented some diets but rarely contributed significant portions of them. There is some
evidence that people of higher status enjoyed more varied diets and perhaps bestowed
largesse by hosting banquets that contained rarer meats. There are some indications that
those living in towns had access to, or preferred to eat, a wider range of meats than their
contemporaries in some rural farms.
There is clear evidence that the beef supply to fortresses and large towns in particu-
lar relied heavily on the work of specialist butchers. They mainly acquired, or were
supplied with, adult cattle, which they intensively processed for sale as both fresh and
preserved meat. They could also provide the raw material for large-​scale processing
of marrow, hides, horns, and bone, either by themselves or by other specialists. The
acquisition of cattle for the urban and military markets must have had a significant
impact on traditional means of slaughtering and the distribution of beef. It is less clear
to what extent specialist butchers were involved in processing pigs and sheep, and it is
feasible that there was a more diverse system of acquisition and distribution of their
products.
There were changes in general husbandry practices in the Romano-​British period.
More sheep were kept until their second and third years before slaughter for meat, and
it seems that increasingly more of them were allowed a longer life to produce wool for
the expanding textile market in the later Roman period. Some pigs may have become
largely confined to sties. Plough cattle possibly became increasingly important on
farms, and milk would certainly have been acquired from cattle, goats, and sheep,
although there is little evidence for intensive dairy production. Chickens would have
been a more common sight during the Roman period, perhaps particularly in and
around towns.
Through a combination of importation, more controlled breeding, and better hus-
bandry practices, larger cattle, sheep, pigs, horses, and even chickens have been found in
some parts of Roman Britain, allowing improvements in meat production in particular.
Conversely, Romano-​British people would have encountered a much greater range of
dogs than their ancestors.
There remains much further research to be carried out on human and animal rela-
tionships in the Roman period. The advent of isotopic analysis (e.g. Sykes et al 2006) and
genetic studies is already beginning to have a major impact on improving our knowl-
edge of the movements, diet, and ancestry of both animals and humans in this period.
Traditional zooarchaeological analysis can embrace these studies to develop deeper
understanding of how animals were exploited and perceived by different sections of the
community across Roman Britain.
 The Exploitation of Animals in Roman Britain    803

References
Albarella, U. (2005). ‘Alternate Fortunes? The Role of Domestic Ducks and Geese from
Roman to Medieval Times in Britain’, in G. Grupe and J. Peters (eds), Feathers, Grit and
Symbolism: Birds and Humans in the Ancient Old and New Worlds. Rahden: Verlag Marie
Leidorf, 249–​258
Albarella, U. (2007). ‘The End of the “Sheep Age”: People and Animals in the Late Iron Age’, in
C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow
Books, 389–​402.
Albarella, U., Johnstone, C., and Vickers, K. (2008). ‘The Development of Animal Husbandry
from the Late Iron Age to the End of the Roman Period: A Case Study from South-​East
Britain’, Journal of Archaeological Science, 35: 1828–​1848.
Allen, M. (forthcoming). ‘Chasing Sylvia’s Stag: Placing Deer in the Countryside of Roman
Britain’, in N. Sykes, K. Baker, R. Carden, and R. Madgwick (eds), Deer and People: Past,
Present and Future. Oxford: Windgather Press.
Bradley, R. (1992). ‘Roman Salt Production in Chichester Harbour:  Rescue Excavations at
Chidham, West Sussex’, Britannia, 23: 27–​44.
Clark, K. (1995). ‘The Later Prehistoric and Proto-​Historic Dog: The Emergence of Canine
Diversity’, Archaeozoologia, 7: 9–​32.
Cool, H. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge
University Press.
de Brisay, K., and Evans, K. (1975) (eds). Salt:  The Study of an Ancient Industry.
Colchester: Colchester Archaeology Group.
Dobney, K., and Ervynck, A. (2007). ‘To Fish or not to Fish? Evidence for the Possible Avoidance
of Fish Consumption during the Iron Age around the North Sea’, in C. Haselgrove and T.
Moore (eds), The Later Iron Age in Britain and Beyond. Oxford: Oxbow Books, 403–​418.
Dobney, K., Jacques, D., and Irving, B. (1996). Of Butchers and Breeds: Report on Vertebrate
Remains from Various Sites in the City of Lincoln. Lincoln: Lincoln Archaeological Studies.
Fulford, M. (2001). ‘Links with the Past:  Pervasive “Ritual” Behaviour in Roman Britain’,
Britannia, 32: 199–​218.
Grant, A. (1975). ‘The Animal Bones’, in B. Cunliffe, Excavations at Portchester Castle. Volume
1: Roman. Society of Antiquaries Research Report 32. London: The Society of Antiquaries,
378–​408.
Grant, A. (1989). ‘Animals in Roman Britain’, in M. Todd (ed.), Research on Roman Britain:
1960–​1989. Britannia Monograph Series 11. London: Society for the Promotion of Roman
Studies, 135–​146.
Grant, A. (2000). ‘Diet, Economy and Ritual Evidence from the Faunal Remains’, in M. Fulford
and J. Timby (eds), Late Iron Age and Roman Silchester: Excavations on the Site of the Forum-​
Basilica 1977, 1980–​86. Britannia Monograph Series 15. London: Society for the Promotion of
Roman Studies, 425–​482.
Grant, A. (2004). ‘Domestic Animals and their Uses’, in M. Todd (ed.) A Companion to Roman
Britain. Oxford: Blackwell, 371–​392.
Grimm, J. (2008). ‘A Dog’s Life: Animal Bone from a Romano-​ British Ritual Shaft at
Springhead, Kent’, in N. Benecke (ed.), Beiträge zur Archäozoologie und Prähistorischen
Anthropologie 6. Langenweißbach: Gesellschaft für Archäozoologie und Prähistorischen
Anthropologie, 54–​75.
804

804   Mark Maltby

Hambleton, E. (1999). Animal Husbandry Regimes in Iron Age Britain. BAR British Series 282.
Oxford: Archaeopress.
Hamilton-​Dyer, S. (1993). ‘The Animal Bones’, in V. Zienkiewicz, ‘Excavations in the Scamnum
Tribunorum at Caerleon’, Britannia, 24: 132–​136.
Hammon, A. (2005). ‘Late Romano-​British–​Early Medieval Socio-​Economic and Cultural
Change:  Analysis of the Mammal and Bird Bone Assemblages from the Roman City of
Viroconium Cornoviorum, Shropshire’. Unpublished Ph.D. thesis, University of Sheffield.
Harcourt, R. (1974). ‘The Dog in Prehistoric and Early Historic Britain’, Journal of Archaeological
Science, 1: 151–​176.
Hathaway, S.-​J. (2006). ‘Poole Harbour: A Review of Early and More Recent Archaeological
Investigations with Evidence for Iron Age and Romano-​British Salt Production’, Proceeding
of the Dorset Natural History and Archaeological Society, 127 (for 2005), 53–​58.
Hesse, R. (2011). ‘Reconsidering Animal Husbandry and Diet in the North-​West Provinces’,
Journal of Roman Archaeology, 24: 215–​248.
Huggins, P. (1978). ‘Excavation of Belgic and Romano-​British Farm with Middle Saxon
Cemetery and Churches at Nazeingbury, Essex, 1975–​6’, Essex Archaeology and History,
10: 29–​115.
Hyland, A, (1990). Equus: The Horse in the Roman World. London: Batsford.
Ingrem, C. (2006). ‘The Animal Bone’, in M. Fulford, A. Clarke, and H. Eckardt, Life and
Labour in Late Roman Silchester. Excavations in Insula IX since 1997. Britannia Monograph
Series 22. London: Society for the Promotion of Roman Studies, 167–​88.
Johnstone, C. (2010). ‘Donkeys and Mules’, in T. O’Connor and N. Sykes (eds), Extinctions and
Invasions: A Social History of British Fauna. Oxford: Windgather Press, 17–​25.
King, A. (1984). ‘Animal Bones and the Dietary Identity of Military and Civilian Groups in
Roman Britain, Germany and Gaul’, in T. Blagg and A. King (eds), Military and Civilian in
Roman Britain. BAR British Series 137. Oxford: Archaeopress, 187–​217.
King, A. (1991). ‘Food Production and Consumption—​Meat’, in R. Jones (ed.), Britain in the
Roman Period: Recent Trends. Sheffield: J. R. Collis, 15–​20.
King, A. (1999). ‘Meat Diet in the Roman World: A Regional Inter-​Site Comparison of the
Mammal Bones’, Journal of Roman Archaeology, 12: 168–​202.
King, A. (2005). ‘Animal Remains from Temples in Roman Britain’, Britannia, 36: 329–​369.
Kitchener, A., and O’Connor, T. (2010). ‘Wildcats, Domestic and Feral Cats’, in T. O’Connor
and N. Sykes (eds), Extinctions and Invasions:  A  Social History of British Fauna.
Oxford: Windgather Press, 83–​94.
Lane, T., and Morris, E. (2001). A Millennium of Saltmaking:  Prehistoric and Romano-​
British Salt Production in the Fenland. Lincolnshire Archaeology and Heritage Report
4. Sleaford: Lincolnshire Archaeology.
Levitan, B. (1993). ‘The Vertebrate Remains’, in A. Woodward and P. Leach, The Uley Shrines.
English Heritage Archaeological Report 17. London: English Heritage, 257–​345.
Locker, A. (2007). ‘In piscibus diversis: The Bone Evidence for Fish Consumption in Roman
Britain’, Britannia, 38: 141–​180.
Luff, R. (1993). Animal Bones from Excavations in Colchester 1971–​85. Colchester Archaeological
Report 12. Colchester: Colchester Archaeological Trust.
MacKinnon, M. (2006). ‘Bones, Text and Art in Roman Italy’, in M. Maltby (ed.), Integrating
Zooarchaeology. Oxford: Oxbow Books, 52–​59.
Maltby, M. (1979). Faunal Studies on Urban Sites:  The Animal Bones from Exeter.
Sheffield: University of Sheffield.
 The Exploitation of Animals in Roman Britain    805

Maltby, M. (1981). ‘Iron Age, Romano-​British and Anglo-​Saxon Animal Husbandry: A Review

of the Faunal Evidence’, in M. Jones and G. Dimbleby (eds), The Environment of Man: The
Iron Age to the Anglo-​Saxon Period. BAR British Series 87. Oxford: Archaeopress, 155–​204.
Maltby, M. (1984). ‘Animal Bones and the Romano-​British Economy’, in J. Clutton-​Brock
and C. Grigson (eds), Animals and Archaeology. Volume 4:  Husbandry in Europe. BAR
International Series 227. Oxford: Archaeopress, 125–​138.
Maltby, M. (1989). ‘Urban–​Rural Variation in the Butchering of Cattle in Romano-​British
Hampshire’, in D. Serjeantson and T. Waldron (eds), Diets and Crafts in Towns. BAR British
Series 199. Oxford: Archaeopress, 75–​106.
Maltby, M. (1994). ‘The Meat Supply in Roman Dorchester and Winchester’, in A. Hall and
H. Kenward (eds), Urban–​Rural Connexions: Perspectives from Environmental Archaeology.
Oxbow Monograph 47/​Symposia of the Association of Environmental Archaeologists 12.
Oxford: Oxbow Books, 85–​102.
Maltby, M. (1997). ‘Domestic Fowl on Romano-​British Sites:  Inter-​Site Comparisons of
Abundance’, International Journal of Osteoarchaeology, 7: 402–​414.
Maltby, M. (1998). ‘Animal Bones from Romano-​British Deposits in Cirencester’, in N.
Holbrook (ed.), Cirencester Excavations V: The Roman Town Defences, Public Buildings and
Shops. Cirencester: Cotswold Archaeological Trust, 352–​370.
Maltby, M. (2001). ‘Faunal Remains (AES76–​ 7)’, in P. Booth and J. Evans, Roman
Alcester: Northern Extramural Area. CBA Research Report 127. London: Council for British
Archaeology, 265–​290.
Maltby, M. (2006). ‘Salt and Animal Products: Linking Production and Use in Iron Age Britain’,
in M. Maltby (ed.), Integrating Zooarchaeology: Proceedings of the Ninth ICAZ Conference,
Durham 2002. Oxford: Oxbow Books, 119–​124.
Maltby, M. (2007). ‘Chop and Change: Specialist Cattle Carcass Processing in Roman Britain’,
in B. Croxford, N. Ray, R. Roth, and N. White (eds), TRAC 2006: Proceedings of the Sixteenth
Annual Theoretical Roman Archaeology Conference, Cambridge 2006. Oxford:  Oxbow
Books, 59–​76.
Maltby, M. (2010). Feeding a Roman Town:  Environmental Evidence from Excavations in
Winchester, 1972–​1985. Winchester: Winchester Museums Service.
Maltby, M. (2011). ‘Animal Bone’, in M. Luke and T. Preece, Farm and Forge: Late Iron Age/​
Romano-​ British Farmsteads at Marsh Leys, Kempston, Bedfordshire Bedford:  Albion
Archaeology, 123–​128.
Maltby, M. (2012). ‘Sheep Foundation Burials in Roman Winchester’, in A. Pluskowski (ed.), The
Ritual Killing and Burial of Animals: European Perspectives. Oxford: Oxbow Books, 152–​163.
Maltby, M., and Hamilton-​Dyer, S. (2012). ‘Big Fish and Great Auks: Exploitation of Birds and
Fish on the Isle of Portland, Dorset, during the Romano-​British Period’, Environmental
Archaeology, 17: 168–​176.
Noddle, B. (2000). ‘Large Vertebrate Remains’, in E. Price, Frocester:  A  Romano-​British
Settlement, its Antecedents and Successors. Volume 2: The Finds. Stonehouse: Gloucester and
District Archaeological Research Group, 217–​244.
O’Connor, T. (1988). Bones from the General Accident Site, Tanner Row. The Archaeology of
York: Volume 15, Fascicule 2. York: Council for British Archaeology.
Parker, A. (1988). ‘The Birds of Roman Britain’, Oxford Journal of Archaeology, 7: 197–​226.
Redfern, R., Hamlin, C., and Athfield, N. (2010). ‘Temporal Changes in Diet: A Stable Isotope
Analysis of Late Iron Age and Roman Dorset, Britain’, Journal of Archaeological Science,
37: 1149–​1160.
806

806   Mark Maltby

Reilly, K. (2005). ‘Animal Bones’, in B. Yule, A Prestigious Roman Building Complex on the
Southwark Waterfront: Excavations at Winchester Palace, London, 1983–​90. London: MoLAS
Monograph 23. London: Museum of London Archaeology Service, 158–​166.
Rippon, S. (2008). ‘Coastal Trade in Roman Britain:  The Investigation of Crandon Bridge,
Somerset, a Romano-​British Trans-​Shipment Port beside the Severn Estuary’, Britannia,
39: 85–​144.
Sadler, P. (1997). ‘Faunal Remains’, in A. Marvell and H. Owen-​John, Leucarum: Excavations
at the Roman Auxiliary Fort at Loughor, West Glamorgan 1982–​4 and 1987–​8. Britannia
Monograph Series 12. London: Society for the Promotion of Roman Studies, 396–​409.
Seetah, K. (2006). ‘Multidisciplinary Approach to Romano-​British Cattle Butchery’, in M.
Maltby (ed.), Integrating Zooarchaeology. Oxford: Oxbow Books, 109–​116.
Stokes, P., (2000). ‘A Cut above the Rest? Officers and Men at South Shields Roman Fort’, in P.
Rowley-​Conwy (ed.), Animal Bones, Human Societies. Oxford: Oxbow Books, 145–​151.
Sykes, N. (2010a). ‘European Fallow Deer’, in T. O’Connor and N. Sykes (eds), Extinctions and
Invasions: A Social History of British Fauna. Oxford: Windgather Press, 51–​58.
Sykes, N. (2010b) ‘Worldviews in Transition: The Impact of Exotic Plants and Animals on Iron
Age/​Romano-​British Landscapes’, Landscapes, 10: 19–​36.
Sykes, N. (2012). ‘A Social Perspective on the Introduction of Exotic Animals: The Case of the
Chicken’, World Archaeology, 44: 158–​169.
Sykes, N., and Curl, J. (2010). ‘The Rabbit’, in T. O’Connor and N. Sykes (eds), Extinctions and
Invasions: A Social History of British Fauna. Oxford: Windgather Press, 116–​126.
Sykes, N., White, J., Hayes, T., and Palmer, M. (2006). ‘Tracking Animals Using Strontium
Isotopes in Teeth:  The Role of Fallow Deer (Dama dama) in Roman Britain’, Antiquity,
80: 948–​959.
Sykes, N., Baker, K., Carden, R., Higham, T., Hoelzel, R., and Stevens, R. (2011). ‘New Evidence
for the Establishment and Management of European Fallow Deer (Dama dama) in Roman
Britain’, Journal of Archaeological Science, 38: 156–​165.
Van der Veen, M., and O’Connor, T. (1998). ‘The Expansion of Agricultural Production in Later
Iron Age and Roman Britain’, in J. Bayley (ed.), Science in Archaeology: An Agenda for the
Future. London: English Heritage, 127–​143.
White, K. (1970). Roman Farming. London: Thames and Hudson.
Wild, J. (2002). ‘The Textile Industries of Roman Britain’, Britannia, 33: 1–​42.
Yalden, D., and Albarella, U. (2009). The History of British Birds. Oxford:  Oxford
University Press.
 Chapter 39

A rable Fa rmi ng ,
Horticu ltu re, a nd  Fo od
Expansion, Innovation, and Diversity

Marijke Van der Veen

Introduction

The production and acquisition of enough food were central to most people’s existence
in late Iron Age and Roman Britain, and a consideration of continuity and change in
agriculture and food following the Roman conquest thus offers an interesting angle to
studying Rome’s impact on life in Britain. Not all aspects of food and farming can be
covered in this short chapter, and the reader is referred to the previous reviews, men-
tioned in the next paragraph, for a full treatment of the topic; here some of the latest
evidence will be reviewed. Specifically, this chapter will consider three aspects—​arable
production, the spread of horticulture, and the diversification of consumption pat-
terns—​followed by a consideration of the impact of these changes in terms of grow-
ing regionality, diversity, and new social realities. Evidence for changes in agricultural
production and food consumption can, of course, be observed through a wide variety
of sources. Here the focus lies on the archaeobotanical evidence—​that is, on the botani-
cal remains of crops and other foods retrieved from archaeological excavations—​but
other evidence will be briefly considered where appropriate (for animal husbandry, see
Maltby, this volume).
Any discussion of these issues is only as good as the data on which it is based. While
a huge amount of archaeobotanical work has been carried out on late Iron Age and
Roman period sites in Britain, there are marked regional differences in availability of
such data, with western and northern parts of the country particularly poorly covered
(reflecting the emphasis of archaeological work carried out in Britain since the 1960s
808

808   Marijke van der Veen

which, in turn, is a reflection of modern settlement density and development (Van

der Veen et al. 2007). Moreover, most data are found in individual site reports, with
few regional syntheses available. Previous reviews of Roman and Iron Age agriculture
have been published by Jones (1981, 1989, 1991) and Van der Veen and O’Connor (1998);
while Morris (1979), Dark and Dark (1997), Fowler (2002), and Taylor (2007) provide
additional discussions of landscape, rural settlement, and farming more generally. The
food supply to the army is considered in Hall and Huntley (2007) and Stallibrass and
Thomas (2008). Finally, a particular group of foods and crops—​that is, those brought
to Britain by the Romans—​is reviewed in Van der Veen (2008) and Van der Veen et al.
(2008).
Throughout, the term ‘Roman’ or ‘Mediterranean’ is used to refer to foods and food-
ways that were introduced during the Roman period or brought from the Mediterranean
region and/​or Gaul just prior to the Roman conquest. It does not suggest that these
foods originated from that region (after all, black pepper is native to India, and peaches
originally come from China and South Asia). Even so, the term remains problematic, as
it implies some degree of uniformity in such food ways, while it is clear that the empire
incorporated a mix of different food traditions, though some communality can be
observed.
A number of trends can be identified within the three aspects considered here, but
these trends are largely relevant to central-​southern and eastern Britain. Variability in
practice and response is evident at all levels of analysis: countrywide, within regions,
within types of settlement, within particular groups of sites, and within sites, and future
studies may usefully focus on this variability, to reveal more clearly the immense com-
plexity of life in Roman Britain.

Arable Farming: A Change of Scale

Agricultural responses to the Roman conquest may include changes to the crop rep-
ertoire, to the balance between animals and crops and land use, to the scale and mode
of production, and regional variation in these. Concentrating here on arable farm-
ing and field crops, there is remarkably little change in the range of crops cultivated.
Arable farming during the late Iron Age and Roman period was dominated by the cul-
tivation of two crops: spelt wheat (Triticum spelta) and barley (six-​row hulled barley,
Hordeum vulgare). Four further cereal crops are known from this period: emmer wheat
(Triticum dicoccum), bread wheat (Triticum aestivum), rye (Secale cereale), and culti-
vated oats (Avena sativa/​strigosa), but these all appear to be minor crops, found on a
few sites and in small quantities, though there are regional differences (see later in this
section). In some parts of Britain there is additional evidence for the cultivation of peas
(Pisum sativum), field/​broad bean (Vicia faba), opium poppy (Papaver somniferum),
brassicas (cabbage, rape, turnip, black mustard; Brassica spp.) and flax/​linseed (Linum
usitatissimum).
 Arable Farming, Horticulture, and Food    809

The real change in arable farming noticeable in this period concerns not so much
crop choice, but the scale of production, and this is evident from a range of different
sources. In archaeobotanical assemblages, changes in scale can be observed by moni-
toring regional and/​or temporal variations in the bulk deposition of charred grain or
chaff (Van der Veen 2007; Van der Veen and Jones 2007). Such remains are generally
found in charred form (see later in this section), and, with the exception of deliber-
ate offerings or destruction owing to conflict, the burning of botanical remains usu-
ally occurs in one of three circumstances: (1) when the by-​products of grain dehusking
and cleaning are deliberately burnt as either fuel or waste; (2) when an accident occurs
during some process involving fire (for example, during parching, drying, or cooking)
and (3) when a building containing stored produce catches fire. Most archaeobotanical
material originates from (1), the routine practice of preparing grain for consumption
(dehusking, cleaning), carried out on a day-​to-​day basis. This practice tends to result
in the regular and frequent deposition of small amounts (that is, low densities) of chaff
and weed seeds, with only little grain (the grain being consumed). In contrast, the acci-
dents described under (2) and (3) occur much more rarely, but result in sporadic depo-
sition of large amounts of grain (that is, high densities). A rise in the frequency of such
high-​density deposits over time suggests a change in the scale of production—​that is,
an increase in the bulk handling of grain (processing, storing), which has resulted in an
increased likelihood of accidents. Likewise, an increase in the density of chaff remains
suggests a move from day-​to-​day processing (to feed the family) to bulk processing (in
advance of bulk transport or storage). Thus, a marked rise in either grain-​rich or chaff-​
rich samples points towards a move from a farming system primarily focused on sub-
sistence to one focusing on production for exchange or taxation.
Examples of both have been identified in Wessex and the east of England. Among
the sites of the Danebury Environs project, the frequency of samples representing acci-
dentally burnt grain increases markedly from the Iron Age to the Roman period (from
6 per cent in the early Iron Age, to 13 per cent in the mid/​late Iron Age, 38 per cent in
the latest Iron Age, and 47 per cent in the Roman period (Campbell 2008)). Similarly,
Parks (2012: sections 4.6 and 4.7) has identified an increase in samples with dense plant
remains (containing > 25 items/​litre of deposit) in the east of England, with the highest
levels reached in the middle Roman period.
An additional indicator of change is the rise in frequency of germinated grain.
Germination may be accidental (poor storage conditions) or deliberate (part of the
brewing process), and both signal an increase in the scale of cereal production. Parks
(2012: section 4.8, fig. 4.34) has calculated the proportion of samples showing signifi-
cant evidence of germination (that is, ≥ 20 per cent of grain is germinated) and found
that, in the east of England, this rose from 1 per cent in the Iron Age, to 6 per cent in the
early Roman, 15 per cent in the mid-​Roman period, then dropping again to 8 per cent.
Many other authors report the presence of dense deposits of germinated spelt grain—​
for example, at the nucleated settlement of Catsgore, Somerset (Hillman 1982), the road-
side settlement of Springhead, Kent (Stevens 2011), and the Roman villa at Northfleet,
Kent (W. Smith 2011).
810

810   Marijke van der Veen

Starting with storage, there is now convincing evidence that grain spoiled by inad-
equate storage was a serious problem in Roman Britain (Figure 39.1(a–​b)). Grain pests
(Coleoptera), which thrive in poorly ventilated storage buildings and in grain that is not
fully dry when put into storage, make their first appearance in Britain during the Roman
period (Smith and Kenward 2011). These grain beetles have not been recorded on Iron
Age or earlier sites and are not thought to be native to Britain. They appear from the very
start of the Roman conquest, probably as adventitious inclusions in grain brought into
Britain by the Roman army during its campaigns. For example, a grain weevil (Sitophilus
granarius) was found in a ditch of the annexe to the vexillation fortress at Alchester,
dating to ad 44–​45, together with other imports, such as millet and coriander. Grain
pests were also discovered in deposits dating to between ad 47–​60 at the 1 Poultry site
in central London and at York (Kenward and Williams 1979; Booth et al. 2007: 24, 281;
D. Smith 2011). The sudden appearance of these grain pests is probably related to the
increased use of large, open grain stores containing bulk quantities of grain, which cre-
ated environments in which these grain pests could thrive, while the large-​scale trade
and movement of grain—​both internal and across the Channel—​facilitated their rapid
spread (Smith and Kenward 2011). This would appear to be corroborated by the near
absence of these pests from early medieval sites and their reappearance with the return
of long-​distance grain transport during the later medieval period—​though, interest-
ingly, never to the levels recorded for the Roman period.
Germinated grain is also associated with beer-​brewing. Here the grain is deliber-
ately germinated as part of the ‘malting’ process—​an essential step in the production
of beer. Such germinated grain, and the associated detached sprouts (coleoptiles), are
often found associated with so-​called corn-​driers, as at the Roman villa at Northfleet,
Kent (Figure 39.1(c–​d)) (W. Smith 2011). These structures are primarily a mid-​to-​late
Roman phenomenon, but start appearing as early as the first century ad (Morris 1979).
Early experiments to determine their function suggested that, rather than drying grain,
they may have been more suited to produce malt (Reynolds and Langley 1979; Reynolds
1981). A review of the archaeobotanical assemblages found associated with these struc-
tures highlights that they were probably multi-​functional, with evidence for both the
drying of grain and the production of malt, and that cereal chaff—​especially spelt glume
bases—​was frequently used as fuel in these ovens (Van der Veen 1989). Since that publi-
cation, a wealth of further information has become available, with finds of germinated
grain and large quantities of burnt spelt chaff confirming the earlier interpretations (e.g.
Arrow Valley, Warwickshire (Moffett and Ciaraldi 1999); Showell Farm, Chippenham,
Wiltshire (Carruthers 2006); Grateley, Hampshire (Campbell 2008); Springhead, Kent
(Stevens 2011) and Northfleet, Kent (W. Smith 2011). At Grateley, three pairs of corn-​
driers were discovered, and in each pair the right-​hand oven showed evidence of more
intensive burning than the other, suggesting differential usage. The more intensely
heated ovens are thought to have been used to dry spelt grain and the more moderately
heated ones to germinate grain and produce malt (Cunliffe 2009).
As with grain storage, beer-​brewing is not a Roman invention. Beer is likely to have
been produced throughout prehistory but on a household scale, using ordinary vessels
 (a)

(b)

(c)

Figure 39.1  (a) Charred spelt wheat grain, showing infestation by a grain weevil from a third/​
fourth century corn drier at Grateley South, Hampshire
Photo: Gill Campbell; © English Heritage; reproduced by kind permission of English Heritage;
(b) Charred remains of the granary weevil (Sitophilus granarius L.) from a deposit filled with
charred ‘rubbings’ of malted spelt in an early/​mid Roman ditch at Northfleet villa, Kent ([D.]
Smith 2011)
Photo: David Smith;
(c) Charred germinated spelt grains from a late Roman corn drier at Northfleet villa, Kent ([W.]
Smith 2011: Plate 9)
© High Speed 1 Ltd; image reproduced with the kind permission of High Speed 1 Ltd;
812

812   Marijke van der Veen

(d) (e)

0 5mm
0.5 mm

(f)

Figure 39.1  Continued(d) Charred detached spelt grain sprouts from a late Roman ditch at
Northfleet villa, Kent ([W.] Smith 2011: Plate 10)
(© High Speed 1 Ltd; image reproduced with the kind permission of High Speed 1 Ltd);
(e) Charred coriander; one of a cache of more than 1,000 coriander fruits found on the floor of a
shop in Colchester that was burnt down during the Boudiccan Revolt of ad 60/61. The material
was recovered as part of the excavations carried out at 45-​6 High Street (Murphy 1977).
(Photo: Gill Campbell; © English Heritage; image reproduced by kind permission of English Heritage);
(f) Waterlogged olives from a latrine block at the rear of a first century tavern at 1 Poultry, London
(Davis 2011: figure 39.275).
(Photo: Andy Chopping; image reproduced by kind permission of Museum of London Archaeology).

and ovens, and thus not easily detectable in the archaeological record. The use of spe-
cialized structures and the frequency at which these are found in Roman Britain is
indicative of a rural industry. Beer may have represented a ‘cash crop’, where a surplus
of grain could be turned into a product that had ‘added value’ and thus could be sold at
a profit (Jones 1981). Besides, the quantities of spelt chaff present at these sites highlight
 Arable Farming, Horticulture, and Food    813

that the spelt grain was no longer dehusked on a piecemeal, day-​to-​day basis but,
instead, bulk-​processed in advance of bulk grain storage or shipment of grain, and the
chaff used to fuel the drying/​malting ovens. Thus, the very appearance of corn-​driers—​
specialized structures—​associated with germinated grain and dense spelt chaff depos-
its indicates that routine household activities such as storing grain and beer-​brewing
had—​at some sites at least—​been replaced by large-​scale agricultural operations, tied
into bulk storage and long-​distance transport of surplus grain and the production of a
cash crop.
The appearance of large barns and mills also underlines a growing emphasis on the
efficient and bulk processing of cereal crops. For example, at Grateley (Hampshire), two
large rectangular buildings were clearly used for crop-​processing, with each containing
a pair of corn-​driers, as well as evidence for grain storage, milling, and bread baking
(similarly at Darenth, Horton Kirby, and Littlehampton). Examples of watermills have
been found at Fullerton (Hampshire), Ickham (Kent), Stanwick (Northamptonshire),
and Orton Hall Farm (Cambridgeshire), and animal-​or water-​driven mills (in the
form of large mill stones) at Chew Park (Somerset), Heronbridge (Cheshire), Littlecote
Park villa at Ramsbury (Wiltshire), and Barton Court Farm (Oxfordshire) (Spain 1984;
Fowler 2002: 175; Booth et al. 2007: 298–​303; Cunliffe 2009; Taylor 2011). Technological
advances include improvements to ploughs and other tools, in the form of iron coul-
ters, shares and share tips, as well as large iron scythes—​the latter pointing to a greater
emphasis on hay-​making, though they may also have been used to harvest other crops
(Fowler 2002: ch. 8; Booth et al. 2007: 294). This investment in the agricultural infra-
structure does not just point to economic change, but also relates to a wider emphasis on
display—​in this case the display of agricultural wealth (Taylor 2011).
Additional support for agricultural expansion comes from the Thames Valley, where
the higher areas of the floodplain were turned into arable fields, while grassland on
the second terrace was ploughed up and parts of the floodplain were converted to hay
meadows (Booth et al. 2007: 21–​29, 280, 284). A change in the weed flora associated
with the crops has been interpreted as a move onto the heavier clay soils during the late
Roman period; something first identified by Jones (1981). At Yarnton (Oxfordshire),
the seeds of stinking mayweed (Anthemis cotula)—​a plant that can thrive on calcare-
ous clay soils and heavy loams—​increase from 2 per cent to 8 per cent from the early to
the late Roman period, while in the east of England this frequency rises from less than
1 per cent to 9 per cent over the same time period (Parks 2012: figure 6.4). This species
has also been found at several other Roman sites (Stevens 2006; Booth et al. 2007: 27;
Campbell 2008; W. Smith 2011). While Robinson (1981) suggests this species may repre-
sent a Roman import, it has been found at a few Iron Age sites (Parks 2012: section 6.4.3),
which highlights the need for a full appraisal of the evidence.
Here the importance of studying the weeds becomes clear. While the lack of signifi-
cant change in the principal cereal crops may suggest continuity, the arable weeds could
suggest otherwise. Their life form (annual/​perennial) and ecology (preference for nutri-
ent rich/​poor, acid/​neutral and wet/​dry or heavy/​light soils) help identify the condi-
tions in the arable fields, from which changes in cultivation techniques, arable strategies
814

814   Marijke van der Veen

(ploughing, manuring, weeding, and so on), and scales of production can be inferred.
Besides, an expansion of agriculture onto heavier clay soils will have demanded a greater
use of cattle for traction (and manure), and the heavy work may have increased the
occurrence of pathologies on cattle foot bones. The greater importance of cattle in the
overall farming system, compared to the Iron Age, is visible in the faunal assemblages
(Van der Veen and O’Connor 1998; Grant 2004). There is, currently, little evidence that
the mouldboard plough—​which inverts the soil rather than scratch or cut the surface as
the ard does—​was introduced prior to the tenth century ad (Fowler 2002: ch. 9; Booth
et al. 2007: 286–​288). Systematic regional reviews of changes in arable weed floras and
cattle pathology are needed before it can be established how widespread the use of clay
soils was during the later Roman period.
All these developments emphasize a marked increase in arable production during the
Roman period, but it would be wrong to assume that all regions of Britain participated
in this process, that all settlements within a region participated, or that this process was
continuous. Here the lack of detailed, quantitative, regional studies hampers our abil-
ity to offer precise regional definition to the trends observed, but most of the trends
identified above concern central-​southern and eastern England, rather than western,
south-​western, and northern England, Wales, or Scotland. In these latter regions there
is a higher degree of continuity in the nature of arable farming. Besides, in northern
England and Scotland barley is the more important crop, while emmer rather than
spelt takes second place (Van der Veen 1992; Huntley 2000; Hall and Huntley 2007).
Additionally, there are a few areas of England where emmer remains important or
reappears—​though no clear pattern or reason has yet been identified (Pelling 2000;
Booth et al. 2007: 293; Campbell 2008; Stevens 2011). The role of bread wheat needs fur-
ther study (a review is in preparation). Samples dominated by bread wheat start to occur
from the first century ad onwards but may represent imports rather than British pro-
duce. There is a slight increase by the late Roman period, but bread wheat tends to be
associated with selected—​possibly mainly high-​status—​sites only (Campbell 2008).

Horticulture: An Innovative
New Practice

A significant innovation in Romano-​British farming concerns the introduction and

development of horticulture—​that is, the cultivation of fruits, vegetables, and herbs in
small garden plots and/​or orchards, grown largely for nearby markets rather than just
personal consumption. Wild fruits, greens, and roots must have been consumed in pre-
historic Britain—​they provide essential vitamins and minerals—​but the archaeobotani-
cal evidence for this remains scarce. The actual cultivation of such foods became—​based
on evidence available at the time of writing—​established during the Roman period, as
part of a wider introduction of some fifty new food plants. The import and introduction
 Arable Farming, Horticulture, and Food    815

of Mediterranean foodstuffs and new food ways into Roman Britain started during the
late Iron Age (Hill 2007; Lodwick 2013; see also the section ‘Food: Growing variety and
diversity’ below), but saw a significant increase in both range and scale after the con-
quest. Examples of these newly introduced crops are apple, pear, plum, cherry, walnut,
cabbage, turnip, leaf beet, carrot, coriander, dill, and celery. While initially imports—​
that is, foodstuffs brought into the country from elsewhere—​these are so-​called intro-
duced crops because they become incorporated into British agriculture, unlike the
imports (which are considered in the section ‘Food: Growing variety and diversity’
below). Detailed discussions of the social and regional uptake of these foods and prac-
tices have been published in Van der Veen (2008) and Van der Veen et al. (2008).
Before I discuss the evidence, it is important to explain some basic principles of pres-
ervation. Dead plant tissues (seeds, grains, chaff, fruit stones) normally decay after a
number of years, and some special mechanism of preservation is thus needed to ensure
their recovery on archaeological sites. The three most common modes of preserva-
tion in Britain are carbonization or charring, waterlogging, and mineral replacement.
Crucially, different categories of plant foods tend to be associated primarily with one or
other mode of preservation (Figure 39.2). Remains of fruits, vegetables, herbs, and spices
are usually recovered in waterlogged form, and any analysis of horticulture and the con-
sumption of fruits, herbs, and vegetables thus tends to rely largely on assemblages of

100

80

Mineralized
60 (N = 535)
Percentage

Waterlogged
(N = 1,523)
40 Carbonized
(N = 2,369)

20

0
Cereals and Fruits, vegetables, Nuts and oil/fibre
pulses and herbs plants

Figure 39.2  Predominant mode of preservation for the three main categories of food and
fibre crops, based on the total number of occurrences of each of these plants in archaeobotanical
assemblages from Roman Britain.
Source: after Van der Veen (2008); Van der Veen et al. (2008).
816

816   Marijke van der Veen

waterlogged plant remains. While such preservation is commonly found on urban—​

and some military—​sites, it is rarely encountered on rural ones; r​ esulting in skewed
distributions of sites with these foods, both in terms of site type, geographical area of
the country, and chronological period. The rural character of pre-​Roman settlement in
Britain, combined with the general lack of waterlogged deposits at rural sites, make it
particularly difficult to determine exactly when these new foods first start to appear (see
also the section ‘Food: Growing variety and diversity’). Consequently, any analysis of
the spread of horticulture and consumption of these types of food does need to take
these biases into account (Van der Veen et al. 2007).
It is difficult to determine with any certainty exactly when the practice of horticulture
starts in Britain; the occurrence of seeds of herbs, stones of fruits, and shells of nuts
on archaeological sites may signify either long-​distance trade or horticulture. However,
their increasing presence on rural, rather than just urban and military, sites is a good
indicator that horticulture was practised. This is particularly the case when seeds of veg-
etables such as leaf beet, leek, lettuce, or carrot are found, as we eat the leaves or roots
rather than the seeds. The seeds of these crops tend to be found at locations of cultiva-
tion. This distinction is not clear-​cut, of course. Herbs such as celery, dill, and fennel are
utilized for their seeds as well as their leaves, leaf petioles, and stem-​bases, and the seeds
of cabbage and turnip (the brassicas) may be used as flavourings or to extract oil, while
the leaves and roots of these crops are consumed as vegetables.
Many of the newly introduced foods concern plants exotic to Britain (for example,
pear, plum, walnut, coriander, dill, leek, onion, cucumber, and lettuce), thus indicat-
ing that the Romans brought actual cultivars with them. However, some—​such as cel-
ery and apple—​do occur wild in Britain, and here it is theoretically possible (though
improbable) that British plants were newly domesticated. For example, celery occurs
only in some British salt marshes and thus was most likely grown from imported seeds
(Booth et al. 2007: 282). Significantly, the DNA of modern apple cultivars suggests that
their wild progenitor is Malus sieversii, a native of the mountain region of Kyrgystan and
north-​west China, rather than the British/​European crab apple (Malus sylvestris). This
reveals that the Romans brought the cultivated apple to Britain and did not use the local
wild variety (Harris et al. 2002). Future DNA analysis of other crops will help clarify the
situation further.
Most of these new crops are found from the early Roman period onwards (see the sec-
tion ‘Food: Growing variety and diversity’ for those that occur from the late Iron Age).
At rural sites we see an increase over time in the frequencies of fruits such as apple/​
pear (archaeobotanically it is often difficult to distinguish between the two), cherries,
plum, damson, and walnut, as well as coriander, celery, mint, carrot, and poppy seed.
This suggests that their cultivation was rapidly adopted by local farmers, though evi-
dence for gardening is also found at several military sites (Van der Veen 2008). In terms
of geographical distribution there are differences between the fruits and herbs on rural
sites: while most herbs, vegetables, and fruits are found across central-​southern and
eastern England, this is not the case further north and west, where—​to date—​no herbs
have been found and only a few fruits and vegetables (e.g. Van der Veen 2008: figure 11),
 Arable Farming, Horticulture, and Food    817

although the meagre database of rural sites with waterlogged preservation makes any
such generalizations tentative.
Many of the newly introduced herbs—​such as coriander, celery, dill, and summer
savoury—​seem to mirror the fortunes of the Roman Empire in Britain: early attrac-
tion is witnessed by rare late Iron Age imports; essential requirement is seen in their
early Roman presence at military and some urban sites; their growing popularity is
evident through increased frequency in the middle Roman period (including on rural
sites); while the waning of their appeal is seen in their decline during the late Roman
period (Figure 39.3(a)). In contrast, several fruits and vegetables (for example, cherry,
plum, walnut, carrot, and leaf beet) display an upward frequency curve (Figure
39.3(b)), becoming increasingly more popular during the course of the Roman period,
suggesting they begin to be cultivated in Britain (Van der Veen 2008; Van der Veen et
al. 2008).
A brief mention of viticulture seems appropriate here. There is one convincing exam-
ple of a vineyard: at Wollaston in the Nene Valley. Here a complex of parallel trenches
was discovered, with low but consistent occurrences of grape (Vitis) pollen in these
trenches and with stake holes suggestive of a pastinatio-​type vineyard (Brown et al.
2001). Vineyards have been tentatively identified at a further seven locations (Van der
Veen 2008: figure 9(b)), but Wollaston is the only one where remains of grape pollen
have been found. How successful this and (the possible) other vineyards were, and
whether the fruits were used as table grapes or for wine production, remains uncertain.
The prevalence of these crops in central-​southern and eastern England can be partly
attributed to the greater volume of archaeological work carried out here and the greater
number of sites with waterlogged preservation (Van der Veen et al. 2007: figure 11), but
an additional, if not the main, explanation may relate more to the degree of socio-​eco-
nomic development in this region. Both fruits and vegetables are seasonal, perishable,
and not easily stored for long periods of time or transported over large distances (unlike
cereal grain and pulses). Large-​scale horticulture or market gardening is, consequently,
often found concentrated in or near large centres of population and on important trans-
port routes. The current concentration of evidence for horticulture in central-​southern
and eastern parts of the country matches the distribution of towns and evidence for rural
development and wealth (villas, corn-​driers, large barns, and so on), suggesting that the
logistics of this type of food production was a key factor in its geographical distribution.

Food: Growing Variety and Diversity

Food consumption is another area of life where we see significant changes in parts of
Roman Britain. A dramatic increase in the choice and variety of available plant foods
has been identified, and this is accompanied by marked changes in the range of ceram-
ics; in particular, we see an increase in imported ceramics and the introduction of table
wares (e.g. Cool 2006; Pitts 2008). Both are indicative of quite significant changes in
818

(a)
45
Coriander
40
Celery
% of all waterlogged records 35
Dill
30

25

20

15

10

(b)
20
Plum
18
Walnut
% of all waterlogged records

16
14 Leaf beet
12
10
8
6
4
2
0
Early Roman Middle Roman Late Roman
(N = 71) (N = 64) (N = 63)

(c)
45
Fig
40
Grape
35
% of all waterlogged records

Mulberry
30

25

20

15

10

Figure 39.3  Frequency patterns of selected foods, waterlogged records only. (a) foods that ini-
tially increase but then decline; (b) foods that increase over time; (c) foods that decline over time.
N = number of records with waterlogged remains.
Source: after Van der Veen (2008); Van der Veen et al. (2008).
 Arable Farming, Horticulture, and Food    819

local food ways, although, as with the other changes identified in the previous sections,
there are marked regional and social differences in their uptake.
Prior to the conquest, the diet of the majority of people in Britain consisted of cereal
grains (wheat, barley), some pulses (peas, beans), a few oil seeds (flax, poppy), and
a restricted number of wild foods (for example, hazelnut, blackberries, and proba-
bly some green leaves), as well as animal protein. By the early Roman period, this
repertoire had been increased by some fifty new food plants, many of them imports
and scarce, but others—​the horticultural crops discussed above—​rapidly becom-
ing incorporated into British agriculture. On current evidence, the majority of
these foods are first encountered in the Roman period, but some food imports start
to arrive during the late Iron Age (the general lack of waterlogged deposits at rural
Iron Age sites makes exact monitoring difficult). For example, wine, olive oil, and
fish sauce are manifested by Mediterranean amphorae found in south-​east England,
primarily in the graves of wealthy individuals (Hill 2007), while a fig seed is known
from the trading centre at Hengistbury Head (Cunliffe 2000: 191–​192). Significantly,
recent archaeobotanical research at the late Iron Age oppidum at Silchester has pro-
duced olive, celery, coriander, and dill seeds from well deposits dating to c. ad 25–​
50 (Lodwick 2013). This new evidence mirrors the evidence from Germany, France,
and Switzerland for the importation of ‘processed’ (wine, oil) and ‘raw’ (fruits, seeds)
foods, prior to the incorporation of the region into the Roman Empire (e.g. Kreuz
2004; Zech-​Matterne et al. 2009). Current evidence suggests that these late Iron Age
imports are going to elite locations, in particular to the oppida; highlighting the desire
of the late Iron Age elite—​or at least some of them—​to participate in Roman-​style
dining practices (Lodwick 2013). Thus, the desire for and acquisition of some exotic
‘Roman-​style’ foods started during the late Iron Age, but a marked increase in both
scale and range of imports is noticeable immediately after the conquest; as can be seen
in other parts of north-​west Europe.
That these foods were brought in from the very start of the conquest is evident from
several early Roman sites: millet, coriander, and celery have been recovered at the vexil-
lation fortress at Alchester, dating to ad 44–​45 (Booth et al. 2007: 24, 281); coriander,
fig, and walnut, at 1 Poultry, London, dating to ad 45–​53 (Davis 2011); fig, mulberry,
and olive at Southwark, London, dating to ad 50–​70 (Giorgi 2009); coriander (Figure
39.1(e)), fig, pine nut, aniseed, lentil, dill, and celery at Colchester dating to ad 60–​61
(Murphy 1977), and fig, grape, walnut, dill, and coriander in deposits at Carlisle dat-
ing to ad 72–​105 (Huckerby and Graham 2009). Moreover, many of these foods were
not restricted to use by the army; they were quickly available for sale, as is evident from
the presence of coriander, anise, fennel, dill, mustard, and poppy seeds in shops at
Colchester and London (both burnt down during the Boudiccan revolt of ad 60/​61) and
olives (Figure 39.1(f)) and grapes in a latrine at the back of a first-​century ad tavern in
London (Murphy 1977; Davis 2011).
Fig is particularly common in Roman Britain, being present at 39 per cent of all early
Roman sites where waterlogged plants were recovered (Figure 39.3(c)). In addition to
these exotics, staples such as grain were also imported. The finds of grain weevils at
820

820   Marijke van der Veen

mid-​first-​century Alchester (see section: ‘Arable Farming: A change of scale’ above) and
the first-​century legionary fortress at Exeter (Straker et al. 1984), among others, point
to the importation of grain from the continent; while the occurrence of some seeds of
non-​British species such as einkorn, lentils, and bitter vetch, in a batch of first-​century
charred grain in London, also suggests the presence of imported grain, this time prob-
ably from the Mediterranean region (Straker 1984).
As mentioned in the section ‘Horticulture: An innovative new practice’, the
temporal uptake of these exotics varies significantly. Some imports—​for example,
black pepper, pomegranate, peach, sesame, lettuce, and black cumin—​remain rare
throughout the Roman period and are found almost exclusively in London and at
military sites. Those that are more common but cannot easily be grown in Britain—​
for example, fig, grape, mulberry, lentil—​show a downward trend—​that is, their
frequencies decline, suggesting that many were popular with the invading army,
but their importation did not increase over time (Figure 39.3(c)). Yet, one of these
foods, fig, is the most common import found—​possibly because of its ease of trans-
port (in dried form) and the numerous seeds (1,000+ per fruit)—​and, although
down from a peak of 39 per cent, it was still present at about 25 per cent of sites by
the late Roman period. Other foods increase in popularity after their initial intro-
duction but drop back to early Roman levels by the late Roman period. Examples are
many of the herbs (coriander, celery, dill, mint, summer savory (Figure 39.3(a)), as
well as the brassicas and parsnip. A final group of foods—​mostly those that can eas-
ily be grown in Britain and indeed became incorporated into British farming, such
as cherry, plum, walnut, leaf beet, black mustard—​increase steadily throughout the
period, though their frequencies remain low (below 15 per cent (Figure 39.3(b)). See
Van der Veen (2008) and Van der Veen et al. (2008) for a more detailed discussion of
the patterns found.
There are regional and social variations as well. To date, most foods have been found
on settlements in central-​southern and eastern England, but imports and herbs are
absent from rural sites further north, though they are present at the northern military
sites. Only certain foods—​apple, plum, cherry, walnut, brassicas, carrot, and parsnip—​
that is, all foods that became incorporated into British farming practice—​have been
found on rural sites in northern Britain (Van der Veen 2008: figures 11c–​d).
Of note is the social uptake of these foods. Several different consumer groups can be
identified. London stands out as a community with access to the widest range of new
foods and the greatest variability—​not surprisingly given its special position as a trad-
ing centre and its metropolitan character, with a large immigrant population and social
mix of people. The other major towns show similar consumption patterns, although
with more limited access to the rare imports (which may be a sampling issue). The small
towns are very different, however: they resemble the rural sites rather than the major
towns as far as food consumption is concerned (Van der Veen 2008: table 5)—​a conclu-
sion also reached by studies of other material culture, such as nail-​cleaners and samian
ware (Eckardt 2005; Willis 2005). The military sites form a separate group, with access
to most new foods and, in particular, a high proportion of herbs (especially coriander,
 Arable Farming, Horticulture, and Food    821

dill, and celery). The abundance of herbs is especially suggestive of the adoption of new
practices (see later in this section).
The rural inhabitants of Britain can be divided into at least three separate consumer
groups. First, there are those people living in south-​east England with access to many of
the new foods, including ‘exotics’ such as figs, grapes, pine nuts, and lentils. Strikingly,
this group does not consist exclusively of rural elite sites (villas) but includes some nucle-
ated rural sites and small farmsteads, as well as a few minor towns (Van der Veen et al.
2008: figure 7). A second group consists of rural sites, mostly located in the south-​east
and again including elite and non-​elite sites, that had access to a selection of the new
foods, especially those that later became part of British farming practice (namely, apple,
pear, plum, cherry, turnip, leaf beet, parsnip, and carrot). Finally, the third rural group
is made up of settlements where no evidence of the new foods has been found, and this
group concerns most of those living in northern and western Britain, and many living in
central-​southern and eastern Britain. Yet, this group nevertheless made some changes in
their diet in that the frequency of native wild foods (for example, hazelnut, blackberry,
raspberry, elderberry, sloe) on rural sites increases from the Iron Age to the Roman
period, suggesting that they too started eating more fruits and nuts, though in this case
local wild ones, rather than introduced cultivated ones (Van der Veen 2008: table 6).
The formation of different consumer groups is also apparent from ceramic assem-
blages and kitchen/​dining utensils more generally (e.g. Cool 2006; Hill 2007; Pitts 2008;
Cramp et al. 2011), and from the stable isotopes surviving in human bone (e.g. Richards
and Hedges 1998; Redfern et al. 2010; Müldner 2013). This latter evidence supports the
conclusions reached above for a significant change in diet, with an increase in dietary
breadth, greater diversity, and more status-​related differences in the Roman compared
to the Iron Age period. There is some evidence that marine foods (for example, fish
sauce, fish, oysters) may represent high-​status foods. This, together with the evidence
of selective access to exotic plant foods, brings to light the rise of dietary inequality dur-
ing the Roman period. Some degree of differentiation between the sexes has also been
identified, as has the presence of greater nutritional variety in the towns; the latter also
evident from the botanical data (Van der Veen 2008; Müldner 2013).
This growing diversity emanates not just from the introduction of new foods, but
also from the arrival of migrants from other parts of the Roman Empire (see Ivleva and
Eckardt, this volume, for discussions on mobility within Britain and between prov-
inces). At York, for example, the atypical diets of some individuals suggests they origi-
nated from North Africa; while one or more of the so-​called Headless Romans at York
may have originated from central Europe, as indicated by their high carbon ratio, which
suggests the consumption of C4 plants such as millet (Leach et al. 2009; Müldner et al.
2011; Müldner 2013). Millet has never been part of the cereal repertoire in Britain, but a
few grains have been found at early Roman deposits at military sites such as Alchester
and Carlisle, and in London, always associated with other food imports (Van der Veen
et al. 2008; Müldner et al. 2011). Thus some migrants can be identified as such by evi-
dence for a different childhood diet, whereas others brought familiar foods such as mil-
let with them, before switching to foods more widely available in Britain.
822

822   Marijke van der Veen

Not all foodstuffs exclusively represent food, of course. Several, especially the herbs
and spices, were in antiquity (as today) regularly used in medicine, as herbal remedies
and food supplements. Importantly, in antiquity the distinction between food, remedy,
and ritual was much less marked than today, if it existed at all. Unfortunately, it is virtu-
ally impossible to determine from the archaeobotanical evidence whether or not these
plants were used as part of treatments, though there is one instance where the archaeo-
logical context hints at a medicinal use. A number of medicinal herbs, including dill
(Anethum graveolens), henbane (Hyoscyamus niger), vervain (Verbena officinalis), St
John’s wort (Hypericum perforatum), and fenugreek (Trigonella foenum-​graecum), were
found in an early first-​century ad context in the hospital of the Roman fort at Neuss,
Germany. Interestingly, they were found together with charred grains of rice (Oryza
sativa) (Knörzer 1967). Rice is frequently mentioned in ancient sources as having bene-
ficial properties for patients—​being easily digested—​and it was recommended for upset
stomachs and intestinal problems, and doctors prescribed a medicinal gruel contain-
ing rice to their wealthier patients (Konen 1999; Dalby 2000: 197). Many of these herbs
and spices will have offered benefits to the consumer, regardless of whether they were
ingested as a remedy or as part of the flavouring of dishes (see Baker, this volume, for a
further discussion of medicine).
The use of plants in ritual and ceremonial activities is another important facet of
their occurrence and dispersal in Britain and north-​west Europe more widely, and
botanical remains have at present been recovered from at least twelve cemetery
sites and eight shrine/​temple sites in Britain (e.g. Davis 2000; Giorgi 2000; Van der
Veen et  al. 2008). The choice of food consumed during burial rituals, donated to
the deceased, and/​or burnt as offerings in shrines and temples, concerns not solely
imported or newly introduced foodstuffs but also local staples; complex ties between
foods and identities are in evidence here (e.g. Bouby and Marinval 2004). For exam-
ple, pine nuts have often been found associated with ceremonial sites, but are equally
and possibly more frequently found in domestic contexts; while, in contrast, date
stones are almost exclusively found in high-​status ceremonial contexts and linked to
particular mystic cults (Kislev 1988; Livarda 2013). Here, even more clearly than else-
where, food can be seen to represent a form of material culture, rather than simply
essential fuel for the body.
Nutrition is important too, of course, and the influx of new foods broadened the nutri-
tional base of the British diet. While wheat and barley provided important carbohydrates
and thus energy, the new fruits, vegetables, and herbs contributed vitamin C, beta-​caro-
tene, folic acid, iron, as well as other essential minerals and antioxidants—​all vital to health
and nutrition. This supplemented the wild fruits, nuts, and leaves already consumed dur-
ing the Iron Age, and, intriguingly, there is some evidence that the use of these wild plants
increased during the Roman period (Van der Veen 2008). Additionally, the herbs and other
foods offered new ways to flavour basic staples such as gruel, stews, and bread. More work
is needed to determine in what way all these foods were consumed and how this may have
varied across Britain and between and within different social groups; here integrating the
plants with the faunal remains and ceramics will be beneficial (see Cool 2006).
 Arable Farming, Horticulture, and Food    823

What an analysis of these foods in Roman Britain emphasizes is that simple distinc-
tions between Roman and native, elite and non-​elite, or urban and rural are not helpful
in understanding changing food patterns: the situation was far more complex than that.
But what is clear is that, broadly speaking, selected groups of people in central-​south-
ern and eastern Britain started adopting new food ways, while there is currently little
evidence of this in rural parts of northern and western Britain (see discussion in the
next section below). It is also worth stressing here that different degrees of archaeologi-
cal intervention and different preservation conditions strongly influence these patterns,
and that future work may reveal more complex patterns and/​or alter our current view.
A full discussion of the complexities is given in Van der Veen (2008) and Van der Veen
et al. (2008), while the European evidence is discussed in Bakels and Jacomet (2003),
Kreuz (2004), Livarda and Van der Veen (2008), and Livarda (2011, 2013).

Regionality, Diversity,
and Materiality

This review of the botanical evidence for arable agriculture, horticulture, and food con-
sumption has made it clear that there were many and diverse responses to the Roman
occupation and that, in part, these responses have a regional basis, but are additionally
linked to social and economic realities within each area. On a very basic level we can see
two broad regions, central-​southern and eastern England, where significant changes can
be observed, and western and northern Britain, where no such changes are in evidence
(though the much reduced level of archaeological intervention here is likely to give a false
impression of continuity). It would be wrong, however, to assume uniformity within
these regions. In northern and western regions the data set is too small for reliable obser-
vations of small-​scale differences in response. In contrast, central-​southern and eastern
England has rich archaeological data sets that can highlight complex patterns in both the
temporality and social participation of change. For example, the evidence for an increase
in the scale of arable production and the creation of market gardening is not restricted
to elite rural sites, but neither do all elite sites show evidence for such changes. Similarly,
access to the newly imported and introduced foods is not limited to elite sites. In fact,
exotic foods and introduced foods are found on some small rural settlements, as well as
on nucleated settlements, villas, and small towns. Notably, the villas do not stand out as a
distinct consumer group, while most small towns appear more similar to rural sites than
major towns in terms of food consumption. Differential sampling and preservation (few/​
many samples, presence/​absence of deposits with waterlogged/​mineral-​replaced pres-
ervation) is obviously a major concern here, and future work may change the patterns
found; nevertheless, the diversity of practices observed will undoubtedly stand.
Explanations for the patterns found are equally complex and diverse. Some relate
to the wider economic and political realities of the empire, while others concern more
824

824   Marijke van der Veen

local issues. For example, structural changes in the empire’s economy, interruptions in
supply, and a reduction in the size of the British garrison during the third century ad,
are thought to have affected the volume of trade (Millett 1992: 162–​173; Fulford 2004;
Mattingly 2006: 501, 514). Imports (as reflected by amphorae) dropped in this period, as
did the occurrence of imported foods (for example, reduced frequencies of figs, grapes,
lentils, mulberries, pine nuts, and fennel). The supply did not cease, however; as is clear
from the finds of fig, mulberry, pine nut, and olive in late Roman London, fig and lentil
in fourth/​early fifth-​century ad Silchester (Robinson 2006), and a near-​complete pine
cone (Figure 39.4) from a late-​third/​early fourth-​century ad ditch fill at Clatterford, Isle
of Wight (McPhillips 2001).
The number of people using the new foods to express a new cultural identity, or
aspiration to such an identity, is likely to have changed over time, which may possibly
explain the frequency pattern of some of the herbs. These foods are strongly associated
with military sites, but their frequencies drop off after the mid-​Roman period. Many of
the herbs were probably grown in Britain by that time, and their frequency is thus not
affected by changes in trade. While the general pattern is that herbs decline by the late

0 1 2 3 4 5 cm
Figure 39.4  A near complete waterlogged imported pine cone (Pinus pinea) from a late third/​
early fourth century ditch fill containing possible other votive objects at Clatterford Roman villa
(McPhillips 2001).
© English Heritage; image reproduced by kind permission of English Heritage.
 Arable Farming, Horticulture, and Food    825

Roman period, they actually increase at rural sites by that time, but this may perhaps be
explained by the move of the elite from the towns to the countryside and the associated
rise in rural villas. Thus, here a combination of empire-​wide strategies and local situa-
tions influences the pattern we see.
Geographical location of individual settlements and general prosperity of a region
also influenced people’s chances of participating in the new developments. It is gen-
erally agreed that the heavy military presence in northern Britain hindered this
region’s economic and social development (Mattingly 2006: 174), which may, par-
tially, explain why we see little evidence of significant agricultural change or adop-
tion of new food ways. Even in central-​southern and western England—​the area with
the highest degree of prosperity and economic development—​not all settlements
show evidence of significant change. Those that do tend to be located in proxim-
ity to towns and/​or major transport routes, which offered easier access to the new
foods and awarded economic opportunity (ready contact with individuals requiring
or desiring grain, vegetables, and fruits). What is not currently clear is the extent to
which these changes were shaped by state-​driven supply networks or market forces
(Pitts 2008).
As far as the actual foods are concerned, these changes accentuated, in greater meas-
ure than before, the role foods play in the creation and maintenance of social identity.
The availability of foods strongly associated with the new Roman elite offered new ways
of displaying one’s wealth and status. But the wide range of foods now available (true
exotics as well as foods quickly incorporated into local agriculture) also meant that a
much larger section of British society was able to participate in some degree, and we see
this, for example, in the presence of some of these foods on small farmsteads. The rapid
spread of certain foods may, in part, be explained by emulation—​the adoption of aspects
of Roman culture by the local elite in order to belong to the new political order. But
other explanations include ones more focused on economic and social opportunity and
an ability to take advantage of the new situation. For example, not everyone will have
been able—​financially or psychologically—​to make long-​term investments such as the
construction of corn-​driers, storage barns, watermills or the planting of orchards (apple,
cherry, plum, and walnut trees, for example, typically take five or more years before they
start bearing fruit) or divert labour away from the staple cereals to the production of
fruits, herbs, and vegetables. The factors involved in the successful adoption of agricul-
tural innovation are complex (see Van der Veen 2010 for a fuller discussion). The chance
of innovations being fruitful will have been higher in the more developed and prosper-
ous south and east of the country than in the north and west, owing to differential access
to markets, transport infrastructure, and demand. Thus, increasing levels of diversity
and inequality are seen in Roman Britain from a variety of archaeological evidence, and
plant foods are no exception in this respect.
It is more difficult to establish how the new foods were utilized in cuisine. Here the
analysis of the actual foods needs to be combined with studies of kitchen ware, fine
wares, and other material culture (e.g. Cool 2006). Were the new foods consumed
because people enjoyed the taste or primarily because their consumption signalled
826

826   Marijke van der Veen

membership of a particular group? And were these foods incorporated into local cui-
sine and consumed from local table wares—​that is, a form of creolization—​or were
they consumed as part of a new cuisine and using new, Roman fine wares? The vary-
ing ways in which Roman ceramics are utilized in different parts of Britain suggests
that we should expect not a uniform adoption of ‘Roman’ styles of eating, but rather
variable adaptations to suit local circumstances (e.g. Evans 1993; Pitts 2005, 2008;
Willis 2011).
The very presence of these new foods must have affected how people thought about
their food. Were these rejected by some, but perceived as ‘must-​have’ ingredients,
regardless of their taste, by others? The arrival of these foods demanded a response and,
just as any other type of material culture (fine wares, glass, personal adornments, and
so on), created mechanisms through which new social groups were created: by either
adopting the new culture or incorporating certain aspects into existing practices.
Studying the trajectories of these foods over a longer time span—​that is, into the medi-
eval period—​may help narrow down the specific roles particular foods served in dif-
ferent social agendas and different cultural settings (Livarda and Van der Veen 2008;
Livarda 2011; Van der Veen, forthcoming).
All these changes will have had an impact on the social fabric of Roman Britain, and
social changes may have been located at the household and local community level.
The switch from subsistence farming to that of growing a surplus for market exchange
(through the sale of surplus grain or beer) impacted on time commitments of individ-
ual farmers, as more time was needed (as well as more fields, tools, manure, traction
animals, ploughs, barns, corn-​driers, and so on) to plough the fields, manure the soils,
sow and weed the crops, harvest and process them (reaping, threshing, dehusking, siev-
ing) and bring them to market, or store them and convert them to beer later in the year.
These new day-​to-​day realities, in terms of both time commitments and material set-
tings, will have prompted new social and labour relationships. It meant developing more
community-​oriented approaches to the agricultural process, or enlisting more children
(larger families?), or taking on farm labourers. In each of these scenarios the social
relations of the farmers and their families changed significantly; for example, the lat-
ter case meant the adoption of unequal relationships within the farm itself. New social
relationships were formed at the point of sale as well, with farmers gaining contacts and
contracts with new social groups—​that is, with merchants shipping the grain and other
foods or with townspeople if they brought the produce to market themselves. In con-
trast, the social fabric of society may have changed much less markedly in the north and
west of the country.
Gender relations are likely to have changed too, especially with the introduction of
market gardening. Traditionally, gardens have been located closer to the homestead
than arable fields—​to enable more intensive cultivation—​and this proximity to the
house may help explain why gardening is perceived by some as a ‘domestic’ activity, car-
ried out foremost by women (Leach 1997). Several authors have identified apparent gen-
der differences between plough agriculture and hoe-​based cultivation (see Van der Veen
2005 for references and discussion of these different practices), and the move towards
 Arable Farming, Horticulture, and Food    827

horticulture may, therefore, have meant women becoming involved with the production
of fruits, vegetables, and herbs, and thus acquiring new roles—​roles that may also have
brought them into contact with the consumers of their produce—​that is, traders passing
by the farm, nearby townspeople, or military personnel. The impact of these changes to
the native economy was complex and varied, and this may have either benefited or dis-
advantaged women, depending on the local situation (e.g. Devens 1991). A challenge for
the future will be how to start identifying these new relationships in the archaeological
record.

Future Work

Despite all the work carried out over the last few decades, much remains to be done.
The sheer volume of data now available makes new regional surveys essential, and
quantification by site type and region of archaeobotanical data, corn-​driers, mills, large
barns, tools, ceramic types, and so on, will undoubtedly reveal more complex patterns
of change than those visible at the time of writing this chapter. Future work on the ecol-
ogy of the weeds associated with staple crops—​especially when combined with stable
isotope evidence (carbon and nitrogen ratios in the grains themselves)—​will reveal how
farmers treated individual crops (degree of manuring, ploughing, weeding, and so on)
and how this changed over time in each region. More radio-​carbon dates are needed on
‘late’ occurrences of emmer and ‘early’ ones of bread wheat, rye, and oats. DNA analy-
sis on selected crops will help determine from where these originated and whether any
were, instead, developed from local varieties. Once these more detailed studies are inte-
grated with other lines of evidence (field systems, settlement types, transport routes)
and other forms of material culture (for example, ceramics and faunal remains), pos-
sibly initially on a site-​by-​site basis, we will be able to gain a deeper understanding of
Rome’s impact on life in Britain.

Conclusion

To conclude, much continuity of practice can be seen in the range of staple crops grown,
and in the actual diet of large parts of the population, though this diet was augmented
with a greater consumption of wild fruits and nuts. Significant change can be observed
in the scale of cereal production, though only in parts of central-​southern and eastern
England. Some innovation took place within this field too; in the form of structures and
tools that facilitated the increased scale of production and helped display agricultural
wealth (better ploughs, scythes, corn-​driers, barns, watermills, hay production, and so
on). Significant innovation can be seen in the introduction of horticulture, the culti-
vation of fruits, nuts, vegetables, and herbs for market. This was made possible by the
828

828   Marijke van der Veen

introduction of many new cultivars, plus the know-​how to cultivate and reproduce these
crops. Again, this practice is found largely, though not exclusively, in central-​southern
and eastern England—​the area where demand, transport facilities, and prosperity coin-
cided. The fact that some fruits and vegetables are found in northern Britain, and the
fact that these concern crops that continue to be cultivated in Britain after the with-
drawal of the Roman army, brings into focus the fact that these foods were not exclu-
sively desired for their association with Roman elite culture. Innovation is also seen in
the diet of some new consumer groups, with the introduction of a vast range of new
fruits, nuts, vegetables, and herbs, which brought significant nutritional and health ben-
efits, but also greater variety of taste and cuisine, as well as new social relations. Again we
see marked regional and social differences, and at all levels (farming, market-​gardening,
and food consumption) we see greater regionality, diversity, and inequality in the
Roman period compared to the Iron Age. In material terms this meant greater social
and economic diversity, new social relationships, and differentiation between those who
took advantage of the new realities and those who did not. Whether the degree of uptake
was caused entirely by differences in locally available social and economic opportunities
or related more to different degrees of aspiration to ‘become Roman’ remains a matter
for future research.

Acknowledgements

I would like to thank the Leverhulme Trust for a Major Research Fellowship and the
Netherlands Institute for Advanced Study for a NIAS Fellowship during which this
chapter was researched. I am grateful to Julie-​Anne Bouchard Perron, Martin Millett,
Alison Moore, and Jeremy Taylor for helpful discussions and/​or comments on an earlier
draft. Edward Biddulph, Gill Campbell, Andy Chopping, Anne Davis, Rob Goller, David
Smith, and Wendy Smith generously offered images for Figure 39.1 and/​or Figure 39.4.

References
Bakels, C., and Jacomet, S. (2003). ‘Access to Luxury Foods in Central Europe during the
Roman Period: The Archaeobotanical Evidence’, World Archaeology, 34/​1: 542–​557.
Booth, P., Dodd, A., Robinson, M., and Smith, A. (2007). Thames through Time: The
Archaeology of the Gravel Terraces of the Upper and Middle Thames. The Early Historical
Period: ad 1–​1000. Thames Valley Landscapes Monograph 27. Oxford: Oxford Archaeology.
Bouby, L., and Marinval, P. (2004). ‘Fruits and Seeds from Roman Cremations in Limagne
(Massif Central) and the Spatial Variability of Plant Offerings in France’, Journal of
Archaeological Science, 31: 77–​86.
Brown, A. G., Meadows, I., Turner, S. D., and Mattingly, D. J. (2001). ‘Roman Vineyards in
Britain: Stratigraphic and Palynological Data from Wollaston in the Nene Valley, England’,
Antiquity, 75: 275–​287.
 Arable Farming, Horticulture, and Food    829

Campbell, G. (2008). ‘Plant Utilization in the Countryside around Danebury:  A  Roman

Perspective’, in B. Cunliffe (ed.), The Danebury Environs Programme. A  Wessex
Landscape during the Roman Era. Volume 1:  Overview. Oxford University School of
Archaeology Monograph 70. Oxford:  English Heritage and Oxford University School of
Archaeology, 53–​74.
Carruthers, W. (2006). ‘Charred Plant Remains’, in R. Young and A. Hancocks, ‘Early Bronze
Age Ring Ditches and Romano-​ British Agriculture at Showell Farm, Chippenham.
Excavations in 1999’, Wiltshire Archaeological and Natural History Magazine, 99: 38–​43.
Cool, H. E.  M. (2006). Eating and Drinking in Roman Britain. Cambridge:  Cambridge
University Press.
Cramp, L. J.  E., Evershed, R. P., and Eckardt, H. (2011). ‘What was a Mortarium Used for?
Organic Residues and Cultural Change in Iron Age and Roman Britain’, Antiquity,
85: 1339–​1352.
Cunliffe, B. (2000). The Danebury Environs Programme. The Prehistory of a Wessex Landscape.
Volume I:  Introduction. Oxford University Committee for Archaeology Monograph 48.
London: English Heritage and Oxford University Committee for Archaeology.
Cunliffe, B. (2009). ‘Continuity and Change in a Wessex Landscape’, Proceedings of the British
Academy, 162: 161–​210.
Dalby, A. (2000). Empire of Pleasures:  Luxury and Indulgence in the Roman World.
London: Routledge.
Dark, K., and Dark, P. (1997). The Landscape of Roman Britain. Stroud: Sutton.
Davis, A. (2000). ‘Plant Remains’, in B. Barber and D. Bowsher, The Eastern Cemetery of Roman
London:  Excavations 1983–​1990. MoLAS Monograph 4.  London:  Museum of London
Archaeology Service, 368–​378.
Davis, A. (2011). ‘Plant Foods’, in J. Hill and P. Rowsome, Roman London and the Walbrook
Stream Crossing: Exacavations at 1 Poultry and Vicinity, City of London. Volume 2. MoLAS
Monograph 37. London: Museum of London Archaeology Service, 290–​304.
Devens, C. (1991). ‘Gender and Colonization in Native Canadian Communities: Examining the
Historical Record’, in D. Walde (ed.), The Archaeology of Gender. Proceedings of the Twenty-​
Second Annual Conference of the Archaeological Association of the University of Calgary.
Calgary: The University of Calgary Archaeological Association, 510–​515.
Eckardt, H. (2005). ‘The Social Distribution of Roman Artefacts: The Case of Nail-​Cleaners
and Brooches in Britain’, Journal of Roman Archaeology, 18: 139–​160.
Evans, J. (1993). ‘Pottery Function and Fine Wares in the Roman North’, Journal of Roman
Pottery Studies, 6: 95–​118.
Fowler, P. (2002). Farming in the First Millennium ad. Cambridge: Cambridge University Press.
Fulford, M. (2004). ‘Economic Structures’, in M. Todd (ed.), A Companion to Roman Britain.
Oxford: Blackwell, 309–​326.
Giorgi, J. (2000). ‘The Plant Remains’, in A. Mackinder, A Romano-​British Cemetery on Watling
Street: Excavations at 156 Great Dover Street, Southwark, London. MoLAS Archaeological
Study Series 4. London: Museum of London Archaeology Service, 65–​66.
Giorgi, J. (2009). ‘Plant Remains’, in C. Cowan, F. Seeley, A. Wardle, A. Westman, and L.
Wheeler (eds), Roman Southwark, Settlement and Economy:  Excavations in Southwark,
1973–​91. MoLAS Monograph 42. London: Museum of London Archaeology Service, 101–​
104, 112–​115, 242–​243 + CD.
Grant, A. (2004). ‘Animals and the Economy and Ideology of Roman Britain’, in M. Todd (ed.),
A Companion to Roman Britain. Oxford: Blackwell, 371–​392.
830

830   Marijke van der Veen

Hall, A. R., and Huntley, J. P. (2007). A Review of the Evidence for Macrofossil Plant Remains
from Archaeological Deposits in Northern England. English Heritage Research Department
Report Series 87/​2007. London: English Heritage.
Harris, S. A., Robinson, J. P., and Juniper, B. E. (2002). ‘Genetic Clues to the Origin of the Apple’,
Trends in Genetics, 18: 426–​430.
Hill, J. D. (2007). ‘The Dynamics of Social Change in Late Iron Age Eastern and South-​Eastern
England c.300 bc–​ad 43’, in C. Haselgrove and T. Moore (eds), The Later Iron Age in Britain
and Beyond. Oxford: Oxbow Books, 16–​40.
Hillman, G. (1982). ‘Evidence for Spelt Malt’, in R. Leech (ed.), Excavations at Catsgore 1970–​
73. A  Romano-​ British Village. Western Archaeological Trust Excavation Monograph
2. Bristol: Western Archaeological Trust, 137–​141.
Huckerby, E., and Graham, F. (2009). ‘Waterlogged and Charred Plant Remains’, in C. Howard-​
Davis (ed.), The Carlisle Millennium Project, Excavations in Carlisle, 1998–​2001. Volume
2: The Finds. Lancaster: Oxford Archaeology North, 926–​936 + appendix 14.
Huntley, J. (2000). ‘The Charred and Waterlogged Remains’, in C. Haselgrove and R.
McCullagh (eds), An Iron Age Coastal Community in East Lothian: The Excavation of Two
Later Prehistoric Enclosure Complexes at Fisher Road, Port Seton, 1994–​95. Scottish Trust
for Archaeological Research. Edinburgh:  Scottish Trust for Archaeological Research,
157–​169.
Jones, M. (1981). ‘The Development of Crop Husbandry’, in M. Jones and G. Dimbleby (eds),
The Environment of Man: The Iron Age to the Anglo-​Saxon Period. BAR British Series 87.
Oxford: Archaeopress, 95–​127.
Jones, M. (1989). ‘Agriculture in Roman Britain: The Dynamics of Change’, in M. Todd (ed.),
Research on Roman Britain: 1960–​1989. Britannia Monograph Series 11. London: Britannia,
117–​134.
Jones, M. (1991). ‘Food Production and Consumption—​Plants’, in R. F. J. Jones (ed.), Britain in
the Roman Period: Recent Trends. Sheffield: J. R. Collis Publications, 21–​27.
Kenward, H. K., and Williams, D. (1979). ‘Biological Evidence from the Roman Warehouses at
Coney Street’, The Archaeology of York, The Past Environment of York, 14/​2: 44–​100.
Kislev, M. (1988). ‘Pinus pinea in Agriculture, Culture and Cult’, in H. J. Küster (ed.),
Der Prähistorische Mensch und seine Umwelt:  Festschrift für Udelgard Körber-​Grohne.
Stuttgart: Konrad Theiss, 73–​79.
Knörzer, K.-​ H. (1967). ‘Der Römerzeitliche Heilkräuterfund aus Neuss/​ RH’, Bonner
Jahrbücher, 23/​2: 65–​75.
Konen, H. (1999). ‘Reis im Imperium Romanum: Bemerkungen zu seinem Anbau und seiner
Stellung als Bedarfs–​und Handelsartikel in der Römischer Kaiserzeit’, Münstersche Beiträge
zur antiken Handelsgeschichte, 18: 23–​47.
Kreuz, A. (2004). ‘Landwirtschaft im Umbruch? Archäobotanische Untersuchungen zu
den Jahrhunderten um Christi Geburt in Hessen und Mainfranken’, Bericht der Römisch -​
Germanischen Kommission, 85, 97–​104.
Leach, H. M. (1997). ‘The Terminology of Agricultural Origins and Food Production Systems:
A Horticultural Perspective’, Antiquity, 71/​271: 135–​148.
Leach, S., Lewis, M., Chenery, C. Müldner, G., and Eckardt, H. (2009). ‘Migration and Diversity
in Roman Britain:  A  Multidisciplinary Approach to the Identification of Immigrants in
Roman York, England’, American Journal of Physical Anthropology, 140: 546–​561.
Livarda, A. (2011). ‘Spicing up Life in North-​Western Europe: Exotic Food Plant Imports in the
Roman and Medieval World’, Vegetation History and Archaeobotany, 20: 143–​164.
 Arable Farming, Horticulture, and Food    831

Livarda, A. (2013). ‘Date, Rituals and Socio-​Cultural Identity in the North-​Western Roman
Provinces’, Oxford Journal of Archaeology, 32/​1: 101–​117.
Livarda, A., and Van der Veen, M. (2008). ‘Social Access and Dispersal of Condiments in
North-​West Europe from the Roman to the Medieval Period’, Vegetation History and
Archaeobotany, 17 (Supplement 1), 201–​209.
Lodwick, L. (2013). ‘Condiments before Claudius: New Plant Foods at the Late Iron Age oppi-
dum at Silchester, UK’, Vegetation History and Archaeobotany. Published on-​line 9 June 2013.
DOI 10.1007/​s00334-​013-​0407-​1.
McPhillips, S. (2001). ‘The Objects’, in P. Busby, D. de Moulins, M. Lyne, S. McPhillips, and R.
Scaife, ‘Excavations at Clatterford Roman Villa, Isle of Wight’, Proceedings of the Hampshire
Field Club and Archaeological Society, 56: 95–​128.
Mattingly, D. (2006). An Imperial Possession: Britain in the Roman Empire. London: Allen Lane.
Millett, M. (1992). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Moffett, L., and Ciaraldi M. (1999). ‘Charred Plant Remains’, in S. C. Palmer, ‘Archaeological
Excavations in the Arrow Valley, Warwickshire’, Birmingham Warwickshire Archaeological
Society Transactions, 103: 207–​208.
Morris, P. (1979). Agricultural Buildings in Roman Britain. BAR British Series 70.
Oxford: Archaeopress.
Müldner, G. (2013). ‘Stable Isotopes and Diet: Their Contribution to Romano-​British Research’,
Antiquity, 87: 137–​149.
Müldner, G., Chenery, C., and Eckardt, H. (2011). ‘The “Headless Romans”:  Multi-​isotope
Investigations of an Unusual Burial Ground from Roman Britain’, Journal of Archaeological
Science, 38: 280–​290.
Murphy, P. (1977). ‘45–​46 High Street, Colchester: Carbonised Fruits and Seeds’, AML Reports
(Old Series) 4018, English Heritage Research Department Reports <http://​research.english-​
heritage.org.uk/​results/​&number=4018> (accessed May 2013).
Parks, K. (2012). ‘Arable Practice in the Iron Age and Roman East of England’. Unpublished
Ph.D. thesis, University of Leicester <https://​lra.le.ac.uk/​handle/​2381/​27951> (accessed May
2013).
Pelling, R. (2000). ‘Charred Plant Remains’, in P. Booth and C. Hayden, ‘A Roman Settlement at
Mansfield College, Oxford’, Oxoniensia, 65: 324–​328.
Pitts, M. (2005). ‘Pots and Pits: Drinking and Deposition in Late Iron Age South-​East Britain’,
Oxford Journal of Archaeology, 24/​2: 143–​161.
Pitts, M. (2008). ‘Globalizing the Local in Roman Britain: An Anthropological Approach to
Social Change’, Journal of Anthropological Archaeology, 27: 493–​506.
Redfern, R. C., Hamlin, C., and Beavan-​ Athfield, N. (2010). ‘Temporal Changes in
Diet:  A  Stable Isotope Analysis of Late Iron Age and Roman Dorset, Britain’, Journal of
Archaeological Science, 37/​6: 1149–​1160.
Reynolds, P. (1981). ‘New Approaches to Familiar Problems’, in M. Jones and G. Dimbleby
(eds), The Environment of Man: The Iron Age to the Anglo-​Saxon Period. BAR British Series
87. Oxford: Archaeopress, 19–​49.
Reynolds, P., and Langley, J. K. (1979). ‘Romano-​British Corn-​Drying Oven: An Experiment’,
Archaeological Journal, 136: 27–​42.
Richards, M. P., and Hedges, R. E. M. (1998). ‘Stable Isotope Analysis Reveals Variations in
Human Diet at the Poundbury Camp Cemetery Site’, Journal of Archaeological Science,
25: 1247–​1252.
832

832   Marijke van der Veen

Robinson, M. (1981). ‘The Iron Age to Early Saxon Environment of the Upper Thames Terraces’,
in M. Jones and G. Dimbleby (eds), The Environment of Man: The Iron Age to the Anglo-​
Saxon Period. BAR British Series 87. Oxford: Archaeopress, 251–​277.
Robinson, M. [with N. Fulford and K. Tootell] (2006). ‘Chapter 5:  The Macroscopic
Plant Remains’, in M. Fulford, A. Clarke, and H. Eckardt (eds), Life and Labour in
Roman Silchester:  Excavations in Insula IX since 1997. Britannia Monograph Series 22.
London: Britannia, 206–​218, 374–​379.
Smith, D. (2011). ‘Insects from Northfleet’, in C. Barnett, J. I. McKinley, E. Stafford, J. M.
Grimm, and C. J. Stevens (eds), Settling the Ebbsfleet Valley: High Speed I Excavations at
Springhead and Northfleet, Kent. The Late Iron Age, Roman, Saxon and Medieval Landscape.
Volume 3: Late Iron Age to Roman Human Remains and Environmental Reports. Oxford/​
Salisbury: Oxford/​Wessex Archaeology, 88–​91.
Smith, D., and Kenward, H. (2011). ‘Roman Grain Pests in Britain:  Implications for Grain
Supply and Agricultural Production’, Britannia, 42: 243–​262.
Smith, W. (2011). ‘Charred Plant Remains from Northfleet’, in C. Barnett, J. I. McKinley,
E. Stafford, J. M. Grimm, and C. J. Stevens (eds), Settling the Ebbsfleet Valley: High Speed
I Excavations at Springhead and Northfleet, Kent. The Late Iron Age, Roman, Saxon and
Medieval Landscape. Volume 3: Late Iron Age to Roman Human Remains and Environmental
Reports. Oxford/​Salisbury: Oxford/​Wessex Archaeology, 105–​113.
Spain, R. J. (1984). ‘Romano-​British Watermills’, Archaeologia Cantiana, 100: 101–​128.
Stallibrass, S., and Thomas, R. (2008). Feeding the Roman Army: The Archaeology of Production
and Supply in North-​west Europe. Oxford: Oxbow Books.
Stevens, C. J. (2006). ‘Charred Plant Remains’, in M. G. Fulford, A. B. Powell, R. Entwistle, and
F. Raymond, Iron Age and Romano-​British Settlements and Landscapes of the Salisbury Plain.
Wessex Archaeological Reports 20. Salisbury: Wessex Archaeology, 152–​158.
Stevens, C. J. (2011). ‘Charred Plant Remains from Springhead’, in C. Barnett, J. I. McKinley,
E. Stafford, J. M. Grimm, and C. J. Stevens, Settling the Ebbsfleet Valley:  High Speed
I  Excavations at Springhead and Northfleet, Kent. The Late Iron Age, Roman, Saxon and
Medieval Landscape. Volume 3: Late Iron Age to Roman Human Remains and Environmental
Reports. Oxford/​Salisbury: Oxford/​Wessex Archaeology, 95–​105.
Straker, V. (1984). ‘First and Second Century Carbonized Cereal Grain from Roman London’,
in W. van Zeist and W. A. Casparie (eds), Plants and Ancient Man. Rotterdam: Balkema,
323–​329.
Straker, V., Robinson, M., and Robinson, E. (1984). ‘Biological Investigations of Waterlogged
Deposits in the Roman Fortress Ditch at Exeter’, Proceedings of the Devon Archaeological
Society, 42: 59–​69.
Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151.
London: Council for British Archaeology.
Taylor, J. (2011). ‘The Idea of the Villa: Reassessing Villa Development in South-​East Britain’, in
N. Roymans and T. Derks (eds), Villa Landscapes in the Roman North: Economy, Culture and
Lifestyles. Amsterdam: Amsterdam University Press, 179–​194.
Van der Veen, M. (1989). ‘Charred Grain Assemblages from Roman-​Period Corn Driers in
Britain’, Archaeological Journal, 146: 302–​319.
Van der Veen, M. (1992). Crop Husbandry Regimes: An Archaeobotanical Study of Farming in
Northern England, 1000 bc–​ad 500. Sheffield: J. R. Collis.
Van der Veen, M. (2005). ‘Gardens and Fields: The Intensity and Scale of Food Production’,
World Archaeology, 37/​2: 157–​163.
 Arable Farming, Horticulture, and Food    833

Van der Veen, M. (2007). ‘Formation Processes of Desiccated and Carbonised Plant
Remains:  The Identification of Routine Practice’, Journal of Archaeological Science,
34: 968–​990.
Van der Veen, M. (2008). ‘Food as Embodied Material Culture: Diversity and Change in Plant
Food Consumption in Roman Britain’, Journal of Roman Archaeology, 21: 83–​110.
Van der Veen, M. (2010). ‘Agricultural Innovation: Invention and Adoption or Change and
Adaptation?’, World Archaeology 42/​1: 1–​12.
Van der Veen, M. (forthcoming). ‘The Dispersal of Cultivated Fruits, Vegetables and Herbs in
Roman and Medieval Britain’.
Van der Veen, M., and Jones, G. (2007). ‘The Production and Consumption of
Cereals: A Question of Scale’, in C. Haselgrove and T. Moore (eds), The Later Iron Age of
Britain and Beyond. Oxford: Oxbow Books, 419–​429.
Van der Veen, M., and O’Connor, T. (1998). ‘The Expansion of Agricultural Production in Later
Iron Age and Roman Britain’, in J. Bayley (ed.), Science in Archaeology: An Agenda for the
Future. London: English Heritage, 127–​143.
Van der Veen, M. Livarda, A., and Hill, A. (2007). ‘The Archaeobotany of Roman
Britain: Current State and Identification of Research Priorities’, Britannia, 38: 181–​210.
Van der Veen, M., Livarda A., and Hill A. (2008). ‘New Food Plants in Roman Britain: Dispersal
and Social Access’, Environmental Archaeology, 13/​1: 11–​36.
Willis, S. (2005). ‘Samian Pottery, a Resource for the Study of Roman Britain and Beyond:
The Results of the English Heritage-​Funded Samian Project. An e-​monograph. Internet
Archaeology 17 <http://​intarch.ac.uk/​journal/​issue17/​willis_​index.html> (accessed May
2013).
Willis, S. (2011). ‘Samian Ware and Society in Roman Britain and Beyond’, Britannia,
42: 167–​242.
Zech-​Matterne, V., Bouby, L., Bouchette, A., Cabanis, M., Derreumaux, M., Durand, F.,
Marinval, P., Pradat, B., Sellami, M.-​F. and Wiethold, J. (2009). ‘L’Agriculture du VIe au Ier
siècle avant J.-​C. en France: État des recherches carpologiques sur les établissements ruraux’,
in I. Bertrand, A. Duval, J. Gomez de Soto, and P. Maguer (eds), Habitats et paysages ruraux
en Gaule et regards sur d’autres régions du monde celtique. Tome II. Chauvigny: Association
des Publications Chauvinoises (Mémoire XXXV), 383–​416.
834

Chapter 40

C oinag e a nd
the Ec onomy

Philippa Walton and Sam Moorhead

Introduction

Scholars of Roman coinage in Britain are blessed with a numismatic record that is
unlikely to be rivalled by any other province of the Roman Empire. Over the past 200
years, more than 2,570 coin hoards have been found, and the numbers continue to rise,
particularly since the introduction of Treasure legislation in 1997. These hoards are com-
plemented by records of more than 180,000 stray losses and site finds collected through-
out England and Wales by the Portable Antiquities Scheme (www.finds.org.uk) and the
Iron Age and Roman Coin Finds from Wales Project (Guest and Wells 2007) (see Figure
40.1) as well as more than 450 assemblages found during research excavations and devel-
opment control projects published both individually and as summaries (cf. Reece 1991;
Walton 2012).
Most research undertaken with these numismatic data has focused on the quantifi-
cation, classification, and publication of individual hoards and assemblages. This has
led to the compilation of a range of invaluable reference works for the identification of
Roman coins; most notably the Cunetio and Normanby hoard catalogues (Besly and
Bland 1983; Bland and Burnett 1988) and Robertson’s Inventory of Romano-​British Coin
Hoards (2000). In recent years, there has also been a growing interest in the numeri-
cal and statistical analyses of coin assemblages from archaeological sites, known col-
lectively as ‘applied numismatics’ (Casey and Reece 1974; Lockyear 2000; Walton 2012).
These analyses have led to a recognition that there are variations in distribution patterns
of coin loss and have engendered a new appreciation of the potential of numismatic data
for exploring the socio-​economic development of the province (Davies and Gregory
1991; Moorhead 2001; Guest 2008b; Walton 2012).
However, the focus on the publication of hoards and site assemblages has led to the
critical exploration of the use of coinage in Roman Britain being somewhat neglected.
Figure 40.1  Roman coins recorded by the Portable Antiquities Scheme.
Source: © Philippa Walton.
836

836    Philippa Walton and Sam Moorhead

Although there have been some detailed discussions regarding the role of money in
Romano-​British society (cf. Walker 1988: 301 and passim; Reece 2002: 107 and passim),
there has been a tendency to concentrate on the development of a ‘monetary economy’
rather than on a consideration of the non-​monetary roles it might acquire or represent.
An ‘evolutionist’ model continues to prevail (Aarts 2005: 41), situating monetization as
one aspect of a wider process of acculturation or ‘Romanization’. As such, native socie-
ties throughout the province moved from pre-​monetary systems of barter and social
exchange to a single monetized economy, as a direct result of integration within the
Roman Empire (Aarts 2005: 8). Hence, at the beginning of the Roman period, the use
of coinage was intimately linked to the incoming army and provincial administration
(Greene 1986: 61; Boon 1988: 118; Esmonde Cleary 1989: 8; Guest 2008a: 139), with mili-
tary installations and the nascent urban centres representing ‘coin using islands… in
an overwhelming sea of virtually coinless peasants’ (Abdy 2002b: 14). However, by the
mid-​second century other sectors of society had begun to employ coinage and it was
embraced even in rural areas (Reece 1988c: 37; Walker 1988: 288 ff.; Moorhead 2010: 157).
By the late third century ad, the transition from pre-​monetary to monetary economy
was complete and, until the collapse of Roman Britain in the early fifth century ad, func-
tioned in a manner similar to a modern capitalist economy (Reece 1988c: 102; Esmonde
Cleary 1989: 96; Millett 1990: 169; Mattingly 2006: 497).
The apparent ubiquity of coinage throughout the province has also led to the assump-
tion that processes of monetization and coin use were uniform and should therefore be
analysed and interpreted at this level. Hence, scholars have worked to produce average
profiles for coin loss in Britain such as ‘The British Mean’ and ‘The PAS Mean’ (Reece
1995; Walton 2012), and inter-​provincial comparisons have been made (Reece 1987: 91).
Although undoubtedly useful in some respects, this approach obscures the sheer varia-
tion in patterns of coin loss throughout the province, which some regional studies have
highlighted (Davies and Gregory 1991; Lockyear 2000; Moorhead 2001; Guest 2008b;
Walton 2012). More emphasis needs to be placed on an investigation of patterns of coin
loss at a regional level, as they are likely to indicate a new layer of complexity.
This chapter will attempt to convey some of this complexity. Although employing
a chronological framework, it is not intended to rival the exhaustive summaries of all
aspects of coinage used in Roman Britain provided by Reece (1987) and Moorhead (2013).
Instead, it will outline some of the prevailing theories concerning the monetization of the
province and will use new data recorded by the Portable Antiquities Scheme to re-​evalu-
ate these theories and to explore the role of Roman coinage in Romano-​British society
more broadly. Furthermore, it will explore the applicability of the prevailing ‘evolutionist’
model and the relevance of studying coinage from Britannia at a provincial level.

Roman Coin Use in the Late Iron Age

The earliest Roman coins found in Britain are Republican silver denarii, of which a large
number have been recorded, both in hoards and as site finds (Robertson 2000; Walton
 Coinage and the Economy    837

2012). Owing to their longevity of circulation and the tendency to overplay the impact
of the Roman invasion in ad 43, these denarii have invariably been interpreted as rep-
resenting military pay. Their distribution patterns are used to trace the manœuvres of
the army throughout the mid to late first century ad, even contributing to discussions
regarding the route taken by the invading force (cf. Orna-​Ornstein 1997; Frere and
Fulford 2001). While it is certainly true that a large proportion of Republican coinage
arrived and was lost after ad 43, there is some evidence that Iron Age societies, particu-
larly in southern Britain, had familiarity with, and access to, these coins during the first
and possibly even the second century bc.
This evidence is archaeological, iconographic, and metallurgical in nature. Excavations
of the temple complex on Hayling Island, Hampshire (Haselgrove 2005: 386 ff.), and Elm’s
Farm, Humberstone, Leicester (Haselgrove 2009), offer examples of Republican coins
probably deposited before ad 43, while PAS data provide tantalizing hints that Republican
issues may have been deposited within the prehistoric ritual landscape of the Witham
Valley, Lincolnshire (Walton 2012: 68), throughout the Isle of Wight (Walton 2012: 128) and
in Berkshire (Portable Antiquities Scheme 2014, BERK-​65D307) in the late pre-​Roman
Iron Age (Figure 40.2). Furthermore, numerous issues of the late first century bc struck
in the south-​east, including those of Cunobelin and Tincommius, incorporate icono-
graphic themes of contemporary Roman coinage (Scheers 1992: 34), and Republican dena-
rii appear to have been the primary source of metal used in the manufacture of the silver
denominations of several regions, including those attributed to the Atrebates (Northover
1992: 257; Fulford 1989: 178) and the Corieltavi (Julia Farley, pers. comm.).
Although it is possible to establish that some Republican denarii reached late Iron
Age Britain, it is more difficult to ascertain the mechanism(s) by which these coins
arrived and how they were subsequently employed within Iron Age societies. Strabo

Figure 40.2 BERK-​65D307—​a Roman Republican denarius issued in c. 207 bc found in

Berkshire.
Source: © Portable Antiquities Scheme.
838

838    Philippa Walton and Sam Moorhead

states that Britain was known for its exports of corn, cattle, hides, slaves, gold, silver,
iron, and hunting dogs (Strabo IV.5.3), and so it is possible that these coins represent
payment for exported commodities. However, recent research has also stressed the
importance of gift-​giving in the creation of political alliances between Rome and native
elites (Creighton 2006: 42; Hunter 2007: 218), and therefore their presence may repre-
sent not only trade and exchange, but also the articulation of Roman diplomatic policy.
Once they were in ‘circulation’, it is possible that these coins served a similar function to
contemporary Iron Age coinage, which may have been used in ‘non-​commercial’ pay-
ments, to emphasize ties of clientage and other social obligations, while also acting as
convenient stores of wealth (Haselgrove 1987; Hunter 2007: 221).

Roman Coin Use in the First Century ad

Although the invasion of ad 43 did not mark the beginning of Roman coin use in Britain,
the arrival of the Roman army en masse and the establishment of the provincial admin-
istration significantly increased the quantity of Roman coinage in the province dur-
ing the first century ad. It is estimated that the campaigning army required 8,500,000
asses (or 531,250 denarii) every four months to pay its soldiers (Boon 1988: 118), and,
therefore, it comes as no surprise that so much first-​and early second-​century coinage
has been found within military installations, associated urban foundations, and their
immediate hinterlands (Lockyear 2000: 403, 413; Guest 2008a: 139; Walton 2012: 62 and
passim). Military sites as remote from one another as Swanton Morley, in Norfolk, and
Dodderhill, in Herefordshire, possess comparable coin profiles (Davies and Gregory
1991: 71), while the wide geographical spread of Republican to Flavian coinage in Wales
(Guest 2008b) and the west Midlands (Walton 2012: 62 and passim) are likely to reflect
the militarization of the landscape during first-​century campaigning.
Once these coins had been distributed as military pay, establishing their ‘biographies’
until they were lost or deposited is no simple task. The concentrations of finds within
military installations suggest that soldiers spent at least some of their income within a
network of relatively closed monetary economies provided by forts—​presumably using
bronze denominations in lower-​value transactions for goods and services such as food,
alcohol, and sex. Indeed, spatial analysis of the distribution of coins from second-​cen-
tury forts on the Antonine Wall has demonstrated that low-​value bronze denominations
were lost by soldiers within the bathhouse, presumably while getting changed, eating,
drinking, and gambling, while silver denarii were lost around the headquarters’ build-
ing where military pay and savings were kept (Abdy 2002a). However, as early Roman
coinage is not recovered solely from military contexts (Walton 2012: 71), some monetary
interactions with the native population obviously took place, particularly after the initial
phase of invasion. Whether these interactions included the use of coinage as a medium
for exchange in everyday transactions at the level of the individual or army unit, the
retention of denarii as stores of precious metal, or the payment of taxes within the new
 Coinage and the Economy    839

Roman taxation system is open to interpretation. It is not clear, for example, how often
resources were requisitioned, compulsorily purchased, or just taken by the army in the
immediate post-​conquest period.
Imitations of Claudian bronze dupondii and asses, known as ‘Claudian copies’, are fre-
quently cited as evidence of early native adoption of Roman coinage and, by implication,
monetary exchange with the army. These were presumably minted in the mid-​first cen-
tury, and the relatively crude execution of the copies is used, along with the recitation of
punitive legislation against the counterfeiting of coin, to suggest that they were of native,
rather than Roman, manufacture (Sutherland 1935: 25; Boon 1988: 118). While further
research is needed to ascertain the exact function of these copies, this interpretation is
a clear example of the assumption of a colonial perspective, based on aesthetic judge-
ment rather than on distribution patterns. Indeed, huge numbers of Claudian copies
are found at first-​century forts and fortresses such as Richborough, Colchester, and Usk
(Kenyon 1987, 1991; Boon 1988: 118 and passim; Harper 2010; Walton 2012: 79 and pas-
sim). Coupled with a clear justification for military need following the cessation of pro-
duction of bronze coinage at the mint of Lyons in ad 42 (Boon 1988: 119), all this suggests
military manufacture.
More compelling are the epigraphic data provided by the Vindolanda tablets. These
documents, recovered from excavations of a fort in Northumberland, detail cash pay-
ments for foodstuffs and commodities purchased by military personnel on the north-
ern frontier in the late first and early second century ad (Bowman 2003). As a result,
they are frequently cited as evidence of a swift native transition from barter to mon-
etary exchange throughout the province, even in a frontier region (Bowman 2003: 66).
However, it is important both to emphasize the dangers of using information from a
single site to generalize at a provincial level and to consider other ways in which this
epigraphic material could be interpreted. For example, it is striking that the Vindolanda
tablets use the denarius (and fractions thereof) as the main unit of accounting, whereas
elsewhere in the empire it appears that the sestertius was commonly employed (Reece
1987: 32; van Heesch 2007: 80). This quirk in accounting practice may be related to the
fact that military salaries were paid and deductions taken from their salaries in denarii
(cf. Speidel 1992: 90) and that this was simply what was available. Alternatively, it might
reflect the demands of a local population who viewed denarii not as one denomination
in a monetary system but as convenient stores of precious metal. Tacitus implies that
Roman coinage was used in much this way on the Germanic frontier, where the native
population was reputed to pick out older, silver Roman coinage for use in trade (Tacitus,
Germania 5, 3–​5), and this is amply confirmed by evidence from the area outside the
empire. For example, finds from native sites in Scotland show a strong preference for
silver denarii over bronze small change (Hunter 2007: 218). Denarii were regarded as
one of many prestige goods by the native population in Scotland (Hunter 2007: 221) and
acted as tokens for specific specialized transactions, for storing wealth, and for display-
ing status (Hunter 2007: 218).
Although the situation remains ambiguous, what is clear is that any generalization
is unwise and that patterns of coin use are likely to have varied ‘from place to place and
840

840    Philippa Walton and Sam Moorhead

from time to time’ (Reece 1988b: 58). Indeed, there are marked regional variations in
both the quantity of first-​century coins found and the denominations represented.
Perhaps most striking is the dichotomy between north and south. In the north, the
presence of coinage appears inextricably linked with the presence of the Roman army.
Very few coins are recorded at any distance from forts, and the record is dominated by
the silver denarius (Walton 2012: 50). In the south, the quantity of coins recorded is far
larger, with the full range of denominations represented. Do these differences reflect the
separation of the province into ‘military’ and ‘civilian’ zones and the implementation of
a monetary taxation system in the ‘civilian’ zone or merely indicate a pre-​existing famili-
arity with, and willingness to use, coinage in southern Britain? It is impossible to be cer-
tain. However, similar denominational patterning in coin assemblages from ‘military’
and ‘civilian’ zones has been noted elsewhere. For example, long-​established ‘civilian’
provinces such as Gaul, Belgica, and Italy received more bronze coinage than ‘military’
ones, such as Upper and Lower Germany, Raetia, and Pannonia (Hobley 1998: 128).

Roman Coin Use in the Second and

Early Third Century ad

The second century ad is often regarded as a period of monetary consolidation, in

which further sectors of Romano-​British society were gradually welcomed into the
monetary fold. The increase in the size of the coinage circulation pool is argued to have
brought coinage within the reach of more people, while the appearance of Antonine
bronze coinage on rural sites is interpreted as reflecting a tentative expansion of the
monetary economy into the countryside (Moorhead 2010: 157). However, the stability of
Roman coinage in the mid-​Roman period is a major impediment to any investigation of
the processes of monetization at this time. Hoards demonstrate the longevity of circula-
tion of all denominations, with denarii of Mark Antony and bronze coins of Vespasian
available for deposition well into the second and as late as the third century ad (Reece
2002: 43). For example, a purse hoard from Birdoswald with a terminus post quem of ad
121 is dominated by Republican, rather than contemporary, issues (Robertson 2000: 24).
As a result, any observations regarding where and why coinage was being used in the
second century ad must be approached with a degree of caution.
Furthermore, there is some indication that the apparent ubiquity of coinage in the
second century ad is illusory. There was certainly a greater quantity of coinage in Britain
than in the previous century; but, even so, it has been calculated, using the assemblage
from the Sacred Spring at Bath, that by the ad 150s each person in the province might
have had access only to approximately two sestertii (Walker 1988). Richard Reece esti-
mated a range of between 0 and 30 sestertii (Reece 2002: 115). Either way, these totals
do not represent anything resembling monetization or everyday coin use. As Reece
(2002: 115) commented: ‘Coins and coin use were not universal… the coins are not there
 Coinage and the Economy    841

and their values are skewed way above the needs of the farm labourer wanting basic
food… ’.
Bearing this in mind, the large quantities of coins recovered from Romano-​British
temple sites and from votive deposits have a huge significance for understanding the
function of coinage. Indeed, at sites such as the Sacred Spring, Bath (Walker 1988),
Coventina’s Well, Northumberland (Allason-​Jones and Mackay 1985), and Piercebridge,
County Durham (Walton 2008), offerings of coinage represent a significant proportion
of the objects recovered and appear in far greater numbers than at contemporary settle-
ment sites. This may suggest that the act of votive deposition was restricted to certain
sectors of the population with access to coins, or that coinage was regarded as having
the same specialized religious functions it appears to have done in the late Iron Age.
Although this function is often described as being ‘non-​economic’ in nature, this is cer-
tainly not the case, as votive offerings represent economic transactions with the gods,
rather than with men.
The composition of these assemblages may also give us an insight into the nature of
these transactions. At Bath, large numbers of lower-​value bronze denominations were
chosen for deposition, as well as a number of contemporary copies. These copies, which
imitate Neronian to Hadrianic asses, form a coherent group thought to have been made
by a ‘forging mint’ (Walker 1988: 291). Given their context, it is possible that these bronze
pieces were made specifically for deposition as votive offerings and may have been
exchanged with pilgrims for genuine issues of a higher value. This theory is supported
by evidence from the riverine votive deposit at Piercebridge, County Durham, where
numerous lead–​alloy copies of Severan denarii have been found, of which several have
been folded and squeezed with tongs, so that they resemble miniature curse tablets. It is
possible that these copies also represent a specific votive product (Walton 2008). If this
is the case, a major reinterpretation of the function of other contemporary copies in all
periods is needed. Rather than always representing local demand for coinage, copies
may have served specialized religious functions associated with votive deposition.

Roman Coin Use in the Late Third

Century ad

While the mid-​Roman period is generally regarded as a period of monetary con-

solidation, the late third century ad is almost unanimously seen as the beginning of
an explosion of coin use that continued until the end of the Roman period (Walker
1988: 306). With the demise of the Augustan coinage system and the introduction of
the radiate coinage, there was a massive increase in the volume of coins produced. As a
result, debased coinage of the period ad 260–​74 is extremely common in Britain, both
in hoards such as Cunetio, Normanby, and Frome (Figure 40.3) (Besly and Bland 1983;
Bland and Burnett 1988; Moorhead et al. 2010) and as site-​finds. Indeed, 14.4 per cent of
842

842    Philippa Walton and Sam Moorhead

Figure 40.3  The Frome hoard being excavated.

Source: © Somerset County Museums Service.

all coins recorded in Reece (1991) and 12.2 per cent of all coins recorded by Walton (2012)
dated to the period ad 260–​75. This plentiful supply is argued to have brought coinage
within the reach of all rural populations. Instead of functioning as a unit of taxation
and a mechanism for paying the army and administration, coinage was at last embraced
by the rural marketplace and used in everyday exchange (Reece 1988b: 102; Esmonde
Cleary 1989: 96; Millett 1990: 169; Mattingly 2006: 497). However, although radiates are
found in the countryside, they are proportionally more common on urban sites than
rural ones (Reece 1987: 91 ff.; 1988b: 103; Davies and Gregory 1991: 76). This does suggest
that their primary function was associated with some facet of urban life, whether that be
related to the collection of taxes or to the processing of commercial transactions in the
marketplace, or some combination of the two.
Indeed, the demand for coinage for use in commercial transactions, coupled with a
sudden reduction in supply after the demise of the Gallic Empire in ad 274, is usually
deemed to have been responsible for the enormous production of locally produced cop-
ies, which are universally referred to as ‘barbarous radiates’. These coins are uniform nei-
ther in design nor in size, and, as a result, they have tended to be interpreted as wholly
unofficial issues produced locally by opportunist forgers. However, as with Claudian
copies, do subjective judgements regarding their aesthetic value, exemplified by the use
of the label ‘barbarous’, cloud the ability to see radiate copies as official or quasi-​official
 Coinage and the Economy    843

products? If ‘barbarous radiates’ were an unofficial response to the lack of coin sup-
ply from the continent, how did the army and provincial administration respond? By
what mechanisms did soldiers and civil servants continue to be paid efficiently and
effectively?
The phenomenon of hoarding in the late third century ad may also provide further
comment on how coinage was being used at this time. More than 650 hoards of radi-
ates are known from Britain, more than for the rest of the empire put together. Hoards
such as Cunetio with 54,951 coins (Besly and Bland 1983) and Frome with 52,503 coins
(Moorhead et al. 2010) illustrate the enormous number of radiates circulating in the
province. Primarily because of the size of these hoards, scholars have interpreted them
as representing ‘pay-​chests’ or savings hoards (Johns 1996: 6 and passim). Conversely,
the willingness to deposit such large quantities of coins may indicate that they had lit-
tle or no monetary value (at least outside of the religious sphere) and should be com-
pared with earlier votive deposits of coins such as those from the Sacred Spring, Bath,
Coventina’s Well, and Piercebridge. Indeed, there are some strong indications that the
Frome hoard was votive in nature, buried in a single episode near an ancient water
source (Moorhead et al. 2010). Therefore, it is possible that, as radiates had little value
in the rural marketplace, they were offered to the gods as part of numinous transactions
(Walton, forthcoming).

Coin Use in the Fourth Century ad

Following an explosion in the production of coinage in the late third century, the full
integration of the Romano-​British countryside into the ‘monetary economy’ is gener-
ally agreed to have occurred in the fourth century ad (Reece 1972, 1988c, 1995; Walker
1988: 304; Lockyear 2000: 403). Nearly every rural site with coinage produces a num-
mus dating to the period ad 330 to 348. However, an increase in coin production need
not necessarily equate with an increase in the number of users. Indeed, it is becoming
increasingly clear that, even in the fourth century, coin use was not a province-​wide phe-
nomenon. As in earlier centuries, sites where coinage was used and lost were situated,
for the most part, in southern and eastern Britain (Walton, forthcoming). Nor was coin
use a long-​lived phenomenon. As the fourth century progressed, although people may
have continued to use earlier coinage, the area that maintained its coin supply shrank to
a narrow band of land stretching from the West Country through Northamptonshire to
the western edge of East Anglia, with a few outliers to the north (Walton 2012: 105; forth-
coming). This pattern certainly does not accord with an evolutionist model of coin use
in Roman Britain.
While the appearance of large quantities of coinage in rural areas in the early fourth
century ad may reflect the embrace of a ‘monetary economy’, it may also be a reflec-
tion of the de-​centralization of military and administrative structures. Towns appear to
have been in decline, and officials and military units may instead have been stationed
844

844    Philippa Walton and Sam Moorhead

throughout the countryside to extract taxes, such as the annona militaris (Moorhead
2001: 94 ff; Moorhead and Stuttard 2012: 226 and passim). Although the archaeological
evidence for instability on the continent as the result of barbarian invasion is ambigu-
ous, contemporary literary sources strongly indicate an increased interest in the export
of Britain’s agricultural resources (Moorhead and Stuttard 2012: 171 ff.). With this in
mind, it is worth noting the striking similarity in the distribution of rural sites using
coinage in the late Roman period and the find-​spots of late Roman belt fittings (Laycock
2008: 115; Worrell 2012: 390). These fittings have been interpreted as the insignia of civil-
ian administrators and individual soldiers (Leahy 2007).

Coinage and the End of Roman Britain

The study of coinage in the late fourth and early fifth century ad is inextricably linked
with an exploration of the fate of the Diocese—​the group of late Roman provinces into
which Britannia was divided at this time. As there is so little material culture specifically
dated to the period, coinage has often been relied upon to provide reliable dating evi-
dence for events or developments, with a particular emphasis on the chronology of the
collapse of Roman Britain.
Despite the brief expansion of coinage into rural areas in the fourth century ad, by the
Theodosian period (ad 388–​402) coin-​using activity had overwhelmingly contracted
back to urban centres, military installations, and settlements situated at nodal points on
major arteries (Walton 2012: 106)—​in fact, the very types of site that had used coinage at
the beginning of the Roman period. This might suggest that the experiment in de-​cen-
tralization had failed and that safety was sought in accessible, yet defensible locations. In
a further echo of the early Roman period, variation in the denominations of coins being
used at sites with different functions becomes apparent. While late Roman coin assem-
blages from urban and military sites such as Canterbury, Caerwent, and Richborough
are dominated by bronze nummi, silver siliquae are proportionally more common on
rural settlement sites (Bland et al. 2013; Moorhead and Walton 2014). It is possible that,
although something approaching a tri-​metallic ‘monetary economy’ existed within
towns and forts, the rural population continued to select and retain precious metal
coinage for its intrinsic value, as they had several centuries before, rather than to use as
money.
Bronze coinage stopped arriving in Britain soon after ad 395 following the closure of
the western mints, although occasional later bronze nummi have been found in Britain,
including pieces from Hadrian’s Wall (Collins 2008), St Albans, Wroxeter, Dunstable,
and Richborough (Abdy and Williams 2006: 31 and passim). In ad 402 and ad 406
respectively, the silver and gold supply diminished to a trickle (Bland and Loriot 2010;
Moorhead and Walton 2014). The muted response to the cessation of coin supply is
illuminating. Although some copies of nummi were made, such as those found in the
Bishop’s Canning hoard (Guest et al. 1997), there was not the spate of copying seen in
 Coinage and the Economy    845

Figure 40.4  The Coleraine Hoard (County Antrim) deposited in the first half of the fifth cen-
tury ad.
Source: © The Trustees of the British Museum.

earlier periods, which may suggest there was little demand for or use of bronze denomi-
nations. However, the widespread practice of clipping the circumference of silver siliquae
does suggest an attempt to make the best out of the dwindling supply of silver, with the
clippings being used to strike siliqua copies, or melted down to form ingots. Who was
responsible for clipping these coins and for how long is a matter of some debate (Burnett
1984; Guest 2005; Abdy 2009; Walton 2012; Bland et al. 2013), although again it is likely
that there was a significant official element in the process. By ad 430, it is likely that any
surviving vestige of a monetary economy in Roman Britain, such as there was, had col-
lapsed. It is likely that any Roman coinage remaining in Britain was regarded as just
another form of bullion, in much the same way as the siliquae in contemporary hoards
from Barbaricum, such as Coleraine (Figure 40.4) and Traprain Law, appear to have been
(Kent and Painter 1977: 123; Robertson 2000: 405 and passim; Hunter and Painter 2013).

Conclusion

Many questions regarding the history of coin use in Roman Britain remain unan-
swered. However, it is clear that the Roman occupation did not provide the impetus for
846

846    Philippa Walton and Sam Moorhead

a province-​wide conversion from systems of barter and social exchange to a single mon-
etary economy. Regional and site-​specific variation in patterns of coin loss clearly illus-
trate that the ‘evolutionist’ model of monetization should not hold sway. Instead, there
appears to be a strong case for the military, administrative, and religious functions of
Roman coinage in Britain remaining essentially unchanged and unchallenged for more
than 400 years of Roman rule (Walton 2012: 169).

Acknowledgements

The authors’ thanks are extended to Richard Abdy, Roger Bland, Andrew Burnett, Kris
Lockyear, Ian Leins, Eleanor Ghey, Julia Farley, and Julian Bowsher for their informa-
tion and advice.

References
Primary Sources
Strabo, Geography, trans. H. L. Jones. Loeb Classical Library. London: Heinemann, 1923.
Tacitus, Agricola, Germania and Dialogus de Oratoribus, trans. W. Peterson and M. Hutton.
Loeb Classical Library. London: Heinemann, 1914.

Secondary Sources
Aarts, I. (2005). ‘Coins, Money and Exchange in the Roman World: A Cultural–​Economic
Perspective’, Archaeological Dialogues, 12/​1: 1–​28.
Abdy, R. (2002a). ‘A Survey of the Coin Finds from the Antonine Wall’, Britannia, 33: 189–​217.
Abdy, R. (2002b). Romano-​British Coin Hoards. Princes Risborough: Shire Publications.
Abdy, R. (2009). ‘Stanchester, Wiltshire’, in R. Abdy, E. Ghey, C. Hughes, and I. Leins (eds),
Coin Hoards from Roman Britain XII. Wetteren: Moneta, 366–​374.
Abdy, R. A., and Williams, G. (2006). ‘A Catalogue of Hoards and Single Finds from the
British Isles, ad 410–​675’, in B. Cook and G. Williams (eds), Coinage and History in the
North Sea World, c.500–​1250: Essays in Honour of Marion Archibald. Northern World 19.
Leiden: Koninklijke Brill, 11–​73.
Abdy, R., and Minnit, S. (2002). ‘Shapwick Villa, Somerset’, in R. Abdy, I. Leins, and J. Williams
(eds), Coin Hoards from Roman Britain XI. London: Royal Numismatic Society Special
Publication 36, 169–​234.
Allason-​Jones, L., and Mackay, B. (1985). Coventina’s Well:  A  Shrine on Hadrian’s Wall.
Hexham: Trustees of the Clayton Collection.
Besly, E., and Bland, R. (1983). The Cunetio Treasure. London: British Museum Press.
Birley, A. R. (2005). The Roman Government of Britain. Oxford and New  York:  Oxford
University Press.
Bland, R., and Burnett, A. (1988). The Normanby Hoard and other Coin Hoards: Coin Hoards
from Roman Britain VIII. London: British Museum Press.
 Coinage and the Economy    847

Bland, R., and Loriot, X. (2010). Roman and Early Byzantine Gold Coins Found in Britain
and Ireland with an Appendix of New Finds from Gaul. Royal Numismatic Society Special
Publication 46. London: Royal Numismatic Society.
Bland R., Moorhead, T. S. N., and Walton, P. J. (2013). ‘Finds of Late Roman Silver Coins from
Britain: The Contribution of the Portable Antiquities Scheme’, in F. Hunter and K. Painter
(eds), Late Roman Silver and the End of Empire:  The Traprain Law Treasure in Context.
Edinburgh: Society of Antiquaries of Scotland, 117–​166.
Boon, G. C. (1988). ‘Counterfeit Coins in Roman Britain’, in J. Casey and R. Reece (eds), Coins
and the Archaeologist. London: Seaby, 102–​188.
Bowman, A. K. (2003). Life and Letters on the Roman Frontier: Vindolanda and its People. 3rd
edn. London: British Museum Press.
Burnett, A. M. (1984). ‘Clipped siliquae and the End of Roman Britain’, Britannia, 15: 163–​168.
Casey, J., and Reece, R. (1974). Coins and the Archaeologist. BAR British Series
4. Oxford: Archaeopress.
Collins, R. (2008). ‘The Latest Roman Coin from Hadrian’s Wall: A Small Fifth-​Century Purse
Group’, Britannia, 39: 356–​361.
Creighton, J. (2006) Britannia: The Creation of a Roman Province. Abingdon: Routledge.
Davies, J. A., and Gregory, T. (1991). ‘Coinage from a “civitas”: A Survey of the Roman Coins
found in Norfolk and their Contribution to the Archaeology of the ‘Civitas Icenorum’,
Britannia, 22: 65–​101.
Esmonde Cleary, S. (1989). The Ending of Roman Britain. London: B. T. Batsford.
Frere, S., and Fulford, M. (2001). ‘The Roman Invasion of ad 43’, Britannia, 32: 45–​55.
Fulford. M. (1989). ‘The Economy of Roman Britain’, in M. Todd (ed.), Research on Roman
Britain 1960–​89. Britannia Monograph Series 11. London: Society for Promotion of Roman
Studies, 195–​211.
Greene, K. (1986). The Archaeology of the Roman Economy. London: B. T. Batsford.
Guest, P. (2005). The Late Roman Gold and Silver Coins from the Hoxne Treasure.
London: British Museum Press.
Guest, P. (2008a). ‘Coins’, in P. Booth, A-​M. Bingham, and S. Lawrence (eds), The Roman
Roadside Settlement at Westhawk Farm, Ashford, Kent, Excavations 1998–​ 9. Oxford
Archaeology Monograph Series 2. Oxford: Oxford Archaeology, 35–​139.
Guest, P. (2008b). ‘The Early Monetary History of Roman Wales:  Identity, Conquest and
Acculturation on The Imperial Fringe’, Britannia, 39: 33–​58.
Guest, P., and Wells, N. (2007). Iron Age and Roman Coins from Wales. Collection Moneta 66.
Wetteren: Moneta.
Guest, P., Bland, R., Orna-​Ornstein, J., and Robinson, P. (1997). ‘Bishops Cannings (Blagan Hill),
Wiltshire’, Coin Hoards from Roman Britain X. London: British Museum Press, 426–​462.
Harper, P. (2010). ‘Stylistic Links among Copies of Bronze Coins of Claudius I Found in Britain
and Spain’, Numismatic Chronicle, 170: 105–​114.
Haselgrove, C. (1987). Iron Age Coinage in South-​East England: The Archaeological Context.
BAR British Series 174. Oxford: Archaeopress.
Haselgrove, C. (2005). ‘A Trio of Temples: A Reassessment of Iron Age Coin Deposition at
Harlow, Hayling Island and Wanborough’, in C. Hazelgrove and D. Wigg-​Wolf (eds), Iron
Age Coinage and Ritual Practices. Studien zu Fundmünzen der Antike 20. Mainz: Verlag
Philipp von Zabern, 381–​418.
Haselgrove, C. (2009). ‘Small Change:  Two Recent Coin Finds from Humberstone
(Leicestershire) and Long Distance Contacts in Later Iron Age Britain’, in G. Cooney, K.
848

848    Philippa Walton and Sam Moorhead

Becker, J. Coles, M. Ryan, and S. Sievers (eds), Relics of Old Decency: Archaeological Studies
in Later Prehistory. Festschrift for Barry Raftery. Dublin: Wordwell, 249–​258.
Heesch, J. van (2007). ‘Some Aspects of Wage Payments and Coinage in Ancient Rome, First
to Third Centuries ce’, in J. Lucassen (ed.), Wages and Currency: Global Comparisons from
Antiquity to the Twentieth Century. International and Comparative Social History 10.
Bern: Peter Lang, 77–​96.
Hobley, A. S. (1998). An Examination of Roman Bronze Coin Distribution in the Western Empire
ad 81–​192. BAR International Series 688. Oxford: Archaeopress.
Hunter, F. (2007). ‘Silver for the Barbarians: Interpreting denarii Hoards in Northern Britain
and Beyond’, in R. Hingley and S. Willis (eds), Roman Finds: Context and Theory. Proceedings
of a Conference held at the University of Durham. Oxford: Oxbow Books, 214–​224.
Hunter, F., and Painter, K. (2013). Late Roman Silver:  The Traprain Treasure in Context.
Edinburgh: Society of Antiquaries Scotland.
Johns, C. (1996). ‘The Classification and Interpretation of Romano-​ British Treasures’,
Britannia, 27: 1–​16.
Kent, J. P. C., and Painter, K. S. (1977). Wealth of the Roman World ad 300–​700. London: British
Museum Press.
Kenyon, R. (1987). ‘The Claudian Coinage’, in N. Crummy (ed.), Colchester Archaeological
Report 4:  The Coins from Excavations in Colchester 1971–​ 9. Colchester:  Colchester
Archaeological Trust, 24–​41.
Kenyon, R. (1991). ‘The Copying of Bronze Coins of Claudius I in Roman Britain’. Unpublished
Ph.D. thesis, University College London.
Laycock, S. (2008). Britannia—​the Failed State: Ethnic Conflict and the End of Roman Britain.
Stroud: The History Press.
Leahy, K. (2007). ‘Soldiers and Settlers in Britain, Fourth to Fifth Century—​Revisited’, in M.
Henig and T. J. Smith (eds), Collecteana Antiqua: Essays in Memory of Sonia Hawkes. BAR
International Series 1673. Oxford: Archaeopress.
Leins, I. (2011). ‘The Coins’, in V. Score (ed.), Hoards, Hounds and Helmets:  A  Conquest-​
Period Ritual Site at Hallaton, Leicestershire. Leicester Archaeology Monograph 21.
Leicester: University of Leicester, 39–​60, 175–​277.
Lockyear, K. (2000). ‘Site Finds in Roman Britain:  A  Comparison of Techniques’, Oxford
Journal of Archaeology 19(4), 397–​423.
Mattingly, D. J. (2006). An Imperial Possession: Britain in the Roman Empire. London: Penguin.
Metcalf, W. (2012) (ed.). The Oxford Handbook of Greek and Roman Coinage. New York: Oxford
University Press.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Moorhead, T. S. N. (2001). ‘Roman Coin Finds from Wiltshire’, in P. Ellis (ed.), Roman Wiltshire
and After. Devizes: Wiltshire Archaeological and Natural History Society, 85–​105.
Moorhead, T. S. N. (2010). ‘Expanding the Frontiers: How the Portable Antiquities Scheme
Database Increases Knowledge of Roman Coin Use in England’, in S. Worrell, G. Egan, J.
Naylor, K. Leahy, and M. Lewis (eds), A Decade of Discovery: Proceedings of the Portable
Antiquities Conference 2007. BAR British Series 520. Oxford: Archaeopress, 138–​155.
Moorhead, T. S.  N. (2013). Roman Coinage in Britain:  The Contribution of the Portable
Antiquities Scheme. Witham: Greenlight Publishing.
Moorhead, T. S. N., and Stuttard, D. (2012). The Romans who Shaped Britain. London: Thames
and Hudson.
Moorhead, T. S. N., and Walton, P. J. (2014). ‘Coinage at the End of Roman Britain’, in F. Haarer
with R. Collins, K. Matthews, S. Moorhead and P. Walton (eds), AD 410: The History and
 Coinage and the Economy    849

Archaeology of Late Roman and Post-​Roman Britain. London:  Society for Promotion of
Roman Studies.
Moorhead, T. S.  N., Booth, A., and Bland, R. (2010). The Frome Hoard. London:  British
Museum Press.
Northover, J. P. (1992). ‘Materials Issues in the Celtic Coinage’, in M. Mays (ed.), Celtic
Coinage: Britain and Beyond. The Eleventh Oxford Symposium on Coinage and Monetary
History. BAR British Series 222. Oxford: Tempus Reparatum, 253–​299.
Orna-​Ornstein, J. (1997). ‘Early Hoards of denarii from Britain’, in R. Bland and J. Orna-​
Ornstein (eds), Coin Hoards from Roman Britain X. London: British Museum Press, 23–​29.
Portable Antiquities Scheme (2014). <http://​finds.org.uk/​database/​artefacts/​record/​id/​
260139> (accessed 1 March 2014).
Reece, R. (1972). ‘A Short Survey of the Roman Coins Found on Fourteen Sites in Britain’,
Britannia, 3: 269–​276.
Reece, R. (1987). Coinage in Roman Britain. London: Seaby.
Reece, R. (1988a). ‘Numerical Aspects of Roman Coin Hoards in Britain’, in J. Casey and R.
Reece (eds), Coins and the Archaeologist. London: Seaby, 86–​101.
Reece, R. (1988b). My Roman Britain. Cirencester: Cotswold Studies.
Reece, R. (1988c). ‘Coins and Villas’, in K. Branigan and D. Miles (eds), The Economy of Romano-​
British Villas. Sheffield: Sheffield University Press, 34–​41; repr. in Reece (2003), 109–​117.
Reece, R. (1991). Roman Coins from 140 Sites in Britain. Cotswold Studies 4. Cirencester:
Cotswold Studies.
Reece, R. (1995). ‘Site Finds in Roman Britain’, Britannia, 26: 179–​206.
Reece, R. (2002). The Coinage of Roman Britain. Stroud: Tempus.
Reece, R. (2003). Roman Coins and Archaeology: C ​ ollected Papers. Wetteren: Moneta.
Robertson, A. S., with Hobbs, R., and Buttrey, T. V. (2000) (eds). An Inventory of Romano-​
British Coin Hoards. Royal Numismatic Society Special Publication 20. London:  Royal
Numismatic Society.
Scheers, S. (1992). ‘Celtic Coin Types in Britain and their Mediterranean Origins’, in M. Mays
(ed.), Celtic Coinage: Britain and Beyond: The Eleventh Oxford Symposium on Coinage and
Monetary History. BAR British Series 222. Oxford: Tempus Reparatum, 235–​299.
Speidel, M. A. (1992). ‘Roman Army Pay Scales’, Journal of Roman Studies, 82: 87–​106.
Sutherland, C. H.  V. (1935). Romano-​British Imitations of Coins of Claudius I. American
Numismatic Notes and Monographs 65. New York: American Numismatic Society.
Walker, D. R. (1988). ‘The Roman Coins’, in B. Cunliffe (ed.), The Temple of Sulis Minerva at
Bath. Volume 2: The Finds from the Sacred Spring. OUCA Monograph 16. Oxford: Oxford
University Committee for Archaeology, 281–​358.
Walton, P. J. (2008). ‘The Finds from the River’, in H. E. M. Cool and D. Mason (eds), Roman
Piercebridge:  Excavations by D.  W. Harding and Peter Scott 1969–​ 1981. Durham:  The
Architectural and Archaeological Society of Durham and Northumberland, 286–​293.
Walton, P. J. (2012). Rethinking Roman Britain: Coinage and Archaeology. Collection Moneta
136. Wetteren: Moneta.
Walton, P. J. (2014). ‘Where, when and what for? Coin Use in the Romano-​British Countryside’,
in M. Henig and B. Soffe (eds), Roman Villas: A Retrospective Review. Oxford: Oxbow Books.
Worrell, S. (2012). ‘Finds Reported under the Portable Antiquities Scheme’, Britannia,
43: 355–​393.
850

Chapter 41

Ec onom y and P ow e r i n
L ate Rom an Bri ta i n

James Gerrard

Economics and power are two sides of the same coin. Their inter-​relationship is self-​
evident in the contemporary world, with its banks that are too big to fail and now jail
(e.g. Sorkin 2009). The ability to mobilize and deploy economic resources also played an
important part in creating and maintaining power relationships in the Roman Empire.
At an imperial scale, this connection is perhaps most clear-​cut when considering the fall
of the western empire. Many scholars would argue that the collapse of Roman power in
the west was not caused by conflict in Gaul or court intrigue in Ravenna, but by the loss
of tax revenue from North Africa, following its capture by the Vandals (Heather 2005;
Ward-​Perkins 2005).
What follows is a review of how the economics of Roman Britain have been char-
acterized. It is argued that many of these characterizations have failed to problematize
Romano-​British economics. As a consequence, the economics of Roman Britain have
been generally understood as a set of proto-​capitalist systems. This view of economic
activity has skewed our understanding of the foundations of late Roman power. By
rethinking these issues and mobilizing new models, an alternative view can be presented
that has the potential to recast our understanding of late Roman social, economic, and
political dynamics.
The study of the past can often be characterized in one of two ways. Ancient socie-
ties were either readily explicable and populated by rational beings, motivated by the
same desires, fears, and emotions as us; or they were alien places inhabited by irrational
people with alien desires and fears. Such visions of the past are probably deeply rooted
philosophically and psychologically. Was the burial of a cow due to a farmer needing
to dispose of the infectious carcass of some bovine plague, or part of a complex ritual
involving ancestor worship and chthonic deities? Was the settlement built on top of
a hill to place it beyond the reach of a flooding river, or so that the inhabitants could
be perceived as being closer to the gods? Questions such as these, absurd though they
sound, can be found answered one way or the other in a host of excavation reports.
 Economy and Power in Late Roman Britain    851

The division of the past into periods that are perceived by modern viewers as being
either rational or irrational has traditionally cleaved along a line determined by the pres-
ence or absence of textual sources. Irrationality is commonplace in the interpretation
of temporally distant prehistoric societies, where exotic ethnographic analogies can be
used to determine just how bizarre and alien the diverse societies of Homo sapiens sapi-
ens are (and were) to modern western eyes (e.g. Parker Pearson and Ramilisonina 1998).
However, the historic periods, with their textual sources and long-​standing positions in
the origin myths of our modern nation states, always appeared more rational. The early
Middle Ages (and particularly the fifth, sixth, and seventh centuries) were an exception
to this rule, and this can arguably be linked to the near absence of contemporary textual
sources (Moreland 2001, 2006).
Roman Britain has often been seen as a kind of less sophisticated version of the present
(e.g. Frere 1967; Salway 1981; de la Bédoyère 2006). The trappings of classical antiquity—​
towns, money, an army, taxation, government, and infrastructure—​are all familiar to
modern eyes. Furthermore, our own world is founded on the intellectual principles of
the Enlightenment and Graeco-​Roman revival. Like it or not, Rome remains one of the
fundamental ideological touchstones of the western world (e.g. Dyson 2006).
The study of Romano-​British economics has largely perpetuated this vision of a
Roman past that can be intuitively understood. The historiography of this approach is
complex and cannot be dealt with in detail here. However, in broad terms the economics
of Roman Britain are usually presented as a development from a socially embedded Iron
Age system. Such a system would be reliant on the social context of exchange, where
personal and community obligations and reciprocity were paramount. The Roman
conquest introduced towns, markets, and money, and, by the fourth century, this devel-
opment had led to the emergence of socially disembedded economics and a fully mon-
etized economy (Frere 1967; Fulford 1979) (see Walton and Moorhead, this volume, for
a discussion on coinage). The problems with such an approach are legion, and scholars
have tinkered with it in various ways. State intervention to explain peculiar or irrational
distributions of material culture has been suggested (Allen and Fulford 1996: 269). The
survival of Iron Age social systems has been argued to influence the distribution of late
Roman pottery kilns (Millett 1990: 165–​169). Recently, Mattingly (2007) has invited us to
accept a Roman Britain with many different kinds of economy. Nevertheless, when one
studies the detail of material culture, the models for its distribution are often implicit.
In the 1970s Hodder (1974) studied the distribution of various kinds of coarse pottery
in south-​central England. He was explicitly looking for ‘marketing patterns’ in which
pottery was manufactured, transported, and sold in local market centres. Applying
the principles borrowed from geography, he mapped out ceramic distributions and
explained the success or failure of one type of pottery over another by cost–​benefit
analysis. Oxfordshire red-​slipped finewares were more successful than their New Forest
rivals because they were functionally better and there was access to the cheap river-​
borne transport offered by the Thames (Fulford and Hodder 1975). Implicit within these
interpretations is an economically rational world driven by concepts of profit, loss, and
margins.
852

852   James Gerrard

This uncritical acceptance of a vision of the past is not just confined to pottery stud-
ies. The large quantities of Roman period animal bones that have been excavated since
the 1960s have been treated in a largely similar fashion (e.g. Maltby 2010). Animals were
either raised for household consumption or driven to market and sold for consumption
by townsfolk. There seem to be few alternatives, and what is worse is the apparent lack of
any evaluation of why this type of model is thought to be appropriate (see also Maltby,
this volume, for further discussion of the role of animals in the economy).
Since 1990 the number and type of theoretical approaches applied to Roman Britain
has blossomed. This rich diversity of approach, which is best encapsulated within the
pages of the Theoretical Roman Archaeology Conference proceedings, show cases of just
what a different, alien, and irrational place Roman Britain was. Yet this wider sea-​change
in interpretation has rarely been equalled by a concomitant improvement in theoretical
approaches applied to Romano-​British economics (Greene 1998). As an example of this,
a recent study of pottery in the Midlands has followed the path blazed thirty years earlier
by Hodder (Taylor 2004). There may be a simple, if somewhat depressing, explanation
for this phenomenon. Hodder successfully navigated from working on Romano-​British
coarse wares to being internationally renowned as an archaeological theoretician. There
are few who can equal this career trajectory. Most people working on the nuts and bolts
of the archaeological evidence for late Roman economic activity are specialists of finds
and faunal remains. They are beavering away, dealing with things, not theories, and they
are often working in environments, such as commercial archaeology, where engaging in
theoretical musing is a luxury they can ill afford.
One recent study on ancient economics has argued that the Roman world can be con-
ceptualized economically as an agrarian society and tributary empire (Bang 2007). This
model, which draws heavily on ethnohistoric parallels from the Moghul period, was
developed explicitly for the early Roman period in the Mediterranean. However, it can
also be utilized to great effect in the study of late Roman Britain. It emphasizes the coex-
istence of an agrarian society with urbanism and literary ‘high culture’, while address-
ing the importance of imperial surplus extraction from localized economic and social
systems. Finally, the concept of the bazaar is introduced to describe an imperfect trading
world ‘characterised by chronic imbalances and asymmetries in the supply of available
information and goods’ (Bang 2007: 139).
The application of this model to fourth-​century Britain would suggest that the foun-
dation of the economy was not the archaeologically visible material culture like pottery,
but agricultural commodities such as crops and animals. One way of understanding
this economy is to consider the size of Britain’s population in the late Roman period. In
the early 1990s a figure of approximately 3.67 million people (Millett 1990: 181–​186) was
suggested. Such estimates are fraught with difficulties, but a number lying between the
Domesday estimate of 1.76 million and the medieval maximum of the mid-​fourteenth
century (4.81 million) (Broadberry et al. 2010) seems appropriate. Certainly, the number
of Romano-​British rural sites discovered by commercially funded excavation since the
early 1990s (Taylor 2007) would seem to indicate that Millett’s estimate is unlikely to be
 Economy and Power in Late Roman Britain    853

too high. If we use the hypothetical figure of 3.67 million people, the economic impact of
this population can be considered.
Modern estimates for the bare minimum calorific requirements needed by a peasant
cultivator suggest that each individual would require somewhere in the region of 250kg
of wheat-​equivalent per annum (Clark and Haswell 1970: 57–​73). Roman-​period grain
yields are uncertain and would, in any case, be variable according to climate, weather,
and soil type. However, a ratio of one pound of seed to four pounds of harvested grain
may be appropriate on slim Roman period evidence (Evans 1981; Mayerson 1984; Bang
2007: 89). This would add another 83kg of wheat, equivalent to the bare subsistence fig-
ure, as seed required for the next season’s planting. Thus, just to feed the population at
subsistence level, over 1.2 million tonnes of grain had to be grown every year.
This figure is admittedly uncertain, but it is also thought provoking. This is especially
true when it is emphasized that this is a minimum food intake. The production of other
commodities to provide further foodstuffs and clothing, as well as a surplus that could
be used to meet the demands of taxation and exchange, ensures that this figure must
be a mere percentage of total agricultural production. Archaeologically this element of
the late Roman economy is both immediately obvious and frustratingly obscure. Field
systems (Taylor 2007: 55–​7 1), rural farmsteads, palynological evidence (e.g. Dark 2000;
Fyfe and Rippon 2004), and masses of animal bone (e.g. Maltby 2010) all exist in relative
abundance. Yet the archaeobotanical evidence for the crops and cereals grown is still
rare and depends on suitable preservational circumstances and recovery methods (van
der Veen et al. 2007; see also van der Veen, this volume, for further discussion of crops).
The use of the modern economic concept of gross domestic product (GDP) as a term
is perhaps best avoided, owing to its modern and therefore anachronistic overtones.
However, against the mountain of crops, cereals, meat, cheese, wool, and leather pro-
duced in fourth-​century Britain, the production, distribution, and exchange of material
culture—​whether that be pots, building stone, mosaics, or items of jewellery—​must rep-
resent a minor element in late Romano-​British economic output. The recognition of this
fact allows the model of an agrarianate society and a tributary empire to be deployed to
its full effect.
During the 1970s a nucleated rural settlement was excavated just north of Somerton
in Somerset. This site, called Castgore (Leech 1982), has become something of a type
site for what is sometimes referred to anachronistically as the Romano-​British ‘village’.
It was formed of a series of rectangular buildings, often divided into three rooms, set
within ditched and later walled plots alongside a minor road. The finds included all of
the paraphernalia typical of late Roman farming community: coins, knives, agricultural
implements, querns, assemblages of faunal and archaeobotanical remains. The pottery
points to links with south-​east Dorset, the New Forest, Oxfordshire, and other regions.
Over recent years similar sites have been dug elsewhere in the lowlands of Britain; at
Kingscote in Gloucestershire (Timby 1998) and Higham Ferrers in Northamptonshire
(Lawrence and Smith 2009). It is wrong to describe these settlements as typical, but they
do provide a snapshot of fourth-​century life for rural communities (Figure 41.1).
854

854   James Gerrard

ace
surf
oad
vel r
Gra

0 100 m

Figure 41.1  Plan of the roadside settlement at Higham Ferrers.

Source: after Lawrence and Smith (2009: figure 2.18). © James Gerrard/​Andrew Agate.

Such settlements are usually characterized as being based on an agricultural economy,

and this is surely true. However, this type of site and others like them are often uncon-
sciously placed within an interpretative framework that argues that the inhabitants were
producing a surplus to be marketed through a nearby urban centre. The publication of
the finds often highlights the distance that goods, especially pottery, travelled before
being consumed, for instance, in Somerset or Gloucester (Leech 1982: 159). This is cer-
tainly one way of interpreting the evidence. However, an alternative can be suggested.
In a well-​known study of Hawaiian villages, Kolb and Snead (1997) used the concept
of ‘small worlds’ to characterize their social and economic structure. This model can be
applied to the fourth-​century rural situation. Did the inhabitants of Catsgore (Leech
1982), Kingscote (Timby 1998), or Cefn Cwmwd and Melin y Plas in Anglesey (Cuttler
 Economy and Power in Late Roman Britain    855

et al. 2012) consider themselves to be part of a globalized network that ultimately con-
nected them to Rome, Ravenna, Constantinople, and beyond? Or did they live in a
‘small world’ characterized by the rhythms of the agricultural season, occasional trips to
the local ‘small town’, and episodic visits by peddlers and gatherings at fayres? Was the
red-​slipped and black burnished pottery on their tables a constant feature on the market
stalls in the local forum? Or did it appear only occasionally—​perhaps seasonally—​as a
sudden glut, which was inevitably followed by a long period when no vessels were avail-
able (Bang 2007: 139)? Here a lone comment contained within an early Roman period
letter found at Vindolanda is a powerful reminder that the road network was probably
impassable for much of the autumn and winter:

The hides which you write are at Cataractonium [Catterick]—​write that they be
given to me and the wagon about which you write. And write to me what is with that
wagon. I would have already been to collect them except that I did not care to injure
the animals while the roads are bad. (Tab. Vindol. II. 343)

A chance observation like this indicates just how small and unpredictable a Roman rural
community’s world could become.
One element that has not yet been explored in this discussion is the relationship
between the agrarianate society and the tributary empire. For these rural dwellers with
their limited horizons these elements of social organization must have been articulated
through the local elites. This social class and their control over the economy bring this
discussion to the crux of the relationship between economics and power in the late
Roman world.
One of the defining aspects of the Roman Empire was the state’s ability to demand and
raise revenues through taxation. Late Roman legal texts, such as the Theodosian Code
(Pharr 1952), contain many laws relating to taxation, yet the intricacies of the tax system
remain unclear (e.g. Jones 1964: 462–​469, 820–​823). Even such seemingly straightfor-
ward issues such as the scale of taxation are contentious topics. There is also plenty of
room for temporal and geographical variation in how taxes were assessed and levied.
These unquantifiable factors make any detailed discussion of the late Roman taxation
system in Britain exceedingly difficult. However, this does not need to be concerning.
The key issue regarding taxation is that it was the economic link between the small
world of agricultural communities, like those at Catsgore and Kingscote, and the glo-
balized empire.
It is generally accepted that the collection of taxation was devolved upon regional
administrative units known as the civitates. Within each individual civitas the town
council or ordo was tasked with fulfilling the demands of the state and personally
responsible for making good any shortfall in revenue (Jones 1964: 457; Wickham 1984).
Much has been written about the negative impact on the curial class of this system.
However, it also offered provincial elites the opportunity for self-​aggrandizement.
The social structure of the late Roman world was highly stratified. Society was broadly
divided into the honestiores and the humiliores (the honest men and the humble men)
856

856   James Gerrard

(Garnsey 1970). The honestiores were a privileged class. If guilty of crimes, they were
subject to fines, whereas the humiliores were likely to be subject to corporal punishment
and torture (Harries 2007: 36). This general distinction marked the division between the
elites and the lowly masses. Yet even within the elite groups further gradations of status
and stratification were evident and formally marked by titles of status (Näf 1995: 20–​21;
Mathisen 2001).
This social stratification and the inequalities that it embodied produced a situation
in which the humiliores came to depend on their powerful patrons (Brown 1971: 37). It
was the local members of the elite who had the connections and money to grease the
wheels of the legal system, or to negotiate with a powerful neighbouring landowner over
property rights and long-​standing custom (Harries 1999: 79; Kelly 2004). This depend-
ent relationship between the client and the patron will have been marked in other ways.
Many clients will have been tenants, debtors, or otherwise obligated to their patron.
Such power relationships and their inequalities offered significant scope for exploitation
by elites, and the tax system was particularly vulnerable to such interventions.
It is sometimes convenient to assume that late Roman taxation was a perfect system
of annual assessments and payments. This view of a cumbersome and no doubt bureau-
cratic system is almost certainly erroneous (Brown 1971: 36). Taxation was likely to have
been infrequent. Annual assessments may have stacked up and been demanded for
payment in a single season, and bad harvests were likely to have been invisible to those
charged with finding sufficient revenue to fill the state’s coffers. Periodic remissions and
extra taxes, coupled with imperfect record-​keeping, probably further complicated the
situation. For the agricultural producers coping with such a system, life would have
been far from easy (Salvian De Gub. Dei IV.3). Here the local patron offered a means to
intercede between those at the bottom of the social ladder and the inexorable demands
of late Roman government. In so doing, the lines between what was due from the cli-
ent or tenant to the patron or landlord and what was due to the state will have become
blurred. Rent and tax may have been rolled together as one burden due to the local lord
(Wickham 1984). That patron, in turn, could have used his position and influence to
gain exemptions from paying tax, or could levy increased charges on his clients and ten-
ants by claiming that state had demanded more. In short, the position of the local elite,
sandwiched between those producing an agricultural surplus and the demands of the
state, was an invitation for them to line their own pockets.
Archaeologically these elite groups are visible through the architecturally elaborate,
richly decorated and opulently furnished villas and townhouses that characterize much
of lowland Britain in the fourth century (e.g. Smith 1997; Scott 2000). These residences,
all too often seen as the dwelling places of supremely successful farmers or economically
successful businessmen, were in fact the residences of the late antique aristocracy, and
their wealth was staggering. The Hoxne hoard (Guest 2005; Johns 2010), with its plate
and jewellery bearing the names of Aurelius Ursicinus and the Domina Juliane, is one
manifestation of this wealth. It is equivalent to 5234.8 g of gold (Hobbs 2006: table 73),
and this is a sum that would have sufficed to pay the taxes of Timgad in the fourth cen-
tury (Kelly 2004: 145–​150). Yet it must represent only a fraction of the owner’s wealth.
 Economy and Power in Late Roman Britain    857

The silver plate that might be expected is largely absent, and an antiquarian find of
several hundred gold solidi nearby may indicate that the owners did not bury all their
golden eggs in one basket (Guest 2005: 16–​33).
On an imperial scale the owner(s) of the Hoxne hoard or a villa like Turkdean in
Gloucestershire (Holbrook 2004) were small fry (Figure 41.2). A middling fourth-​cen-
tury senator would have an annual income many times the value of the Hoxne hoard.
Nevertheless, in lowland Britain this provincial elite exercised local power from a posi-
tion far exalted above that of the average tenant cultivator. Yet it would be a mistake
to characterize the fourth-​century Romano-​British elites in the singular. Social strati-
fications will have divided the honestiores—​both horizontally and vertically—​produc-
ing competing interest groups and nested patron–​client relationships (Wallace-​Hadrill
1989; Brown 1992).
The analysis of late Roman settlement hierarchies has had a tendency to distinguish
between villa sites and other forms of rural settlement. Implicit within this is the sug-
gestion that such a division marks that between high-​and low-​status communities. This
may well be true, although in some parts of Britain, like the Midlands, other form of
architectural display, like aisled buildings, were prevalent (Taylor 2001: 52). However,
the cartographic interpretation of settlement distributions flattens elite social strati-
fication and provides a misleading view of power relationships in the fourth-​century
landscape.
The hinterland of the small town of Lindinis (Ilchester) in southern Somerset is one
of the most densely occupied late Roman landscapes in lowland Britain. Sites such as
Castgore (Leech 1982), Bradley Hill (Leech 1981), and Yeovilton (Lovell 2006) seem to be
those of the rural population linked by roads and trackways that ran between complex

0 100 m

Figure 41.2  Plan of the Roman villa at Turkdean.

Source: after Holbrook (2004: figure 4). © James Gerrard/​Andrew Agate.
858

858   James Gerrard

field systems identified by aerial and geophysical survey. Within this densely populated
landscape were a large number of villas (Figure 41.3). South of the town were villas at
Seavington St Mary (Graham and Mills 1995), Dinnington (Adcock and Wood 2005;
Prof. A. King, pers. comm.), and Lopen (Brunning 2001) in the south-​west; Lufton
(Hayward 1952, 1972), Westlands (Radford 1928), and Bedmore Barn (Ham Hill) (Beattie
and Pythian-​Adams 1913; Sharples et al. 2012: 39) in the centre, and East and West Coker
(Moore 1862; Haverfield 1906: 330). All of these sites have yielded stone building, mosa-
ics, wall plaster, and artefactual evidence of late Roman occupation. This density of set-
tlement in a small area is striking, and, given that both Dinnington and Lopen are recent
discoveries, it might be expected that further villa sites remain to be discovered within
this region.

n
col
Lin
To
River
Yeo

Ilchester
4

ter
Exe
To
To

1
Do

6
rch

3
est

2
er

8
0 4 km 7

Figure 41.3  The distribution of villas near Ilchester: 1) Dinnington, 2) Lopen, 3) Seavington St

Mary, 4) Ilchester Mead, 5) Batemoor Barn, 6) Lufton, 7) West Coker, 8) East Coker, 9) Westlands.
Source: © James Gerrard/​Andrew Agate.
 Economy and Power in Late Roman Britain    859

It has generally been suggested that these sites were the pinnacle of the late Roman set-
tlement hierarchy during the fourth century. This is surely true. Buildings like that exca-
vated at Dinnington represent an architectural form that embodied high-​status civilian
elite display and the aristocratic values of paideia that were shared across the Roman
Empire (Ellis 1991; Brown 1992; Bowes 2010). A hypothetical fourth-​century Sicilian,
familiar with the Piazza Armerina (Gentili 1999), might have sneered at the Romano-​
British interpretation of those values, but he or she would also have understood that
the locals were attempting to share the same values and traditions that the visitor held
dear (Bowes 2010). Unfortunately, the archaeological interpretation of these sites has
reduced them to being mechanistic points in a cartographic landscape. Each site is often
viewed as the dwelling of a single family, and interaction between sites reduced to hypo-
thetical estates derived from nearest neighbour analysis.
The reality of living in this landscape and these villas must have been very different
from the flat abstractions of archaeological analysis. The close proximity of many of
these sites is striking. The villas at East and West Coker are a few minutes’ walk apart,
and many of the other sites are within easy reach of one another, either by foot or by
horse. It is inconceivable that the inhabitants of these sites lived in isolation. A complex
social network is likely to have linked these communities together.
Many of these villas were occupied from at least the middle of the third century
onwards. Thus, by the middle of the fourth century, perhaps five generations had been
raised at some of these sites. Their physical proximity to one another and the duties
of the curial class—​serving on the local town council or ordo, for instance—​will have
thrown their inhabitants together. Friendship, marriage, business and political interests,
feuds, and disagreements will all have characterized this social network (Brown 1992;
Sivonen 2006). The Gallic world of Ausonius and Sidonius is brought to life by their
surviving letters, which give glimpses of such a society (Sivan 1993). There are no surviv-
ing analogues from fourth-​century Britain, although Patrick’s early fifth-​century letters
suggest that they may once have existed (Dumville 1993). The Gallic sources and settle-
ment patterns, such as that identified south of Ilchester, suggest that these landscapes
were characterized not by isolated estates, but by high-​status communities joined by ties
of kith, kin, and patronage.
Those at the pinnacle of this network are likely to have controlled directly and indi-
rectly large swathes of agricultural land. There is little direct evidence of the size of late
antique estates in the north-​western provinces. Ausonius’ ‘little inheritance’, one of four
estates the orator and imperial tutor owned, is thought to have equated to some 238
hectares (Sivan 1993: 66–​69). How typical this is remains debatable. Estate sizes in the
German provinces have been suggested as approximating to 200 hectares, but this is
based on hypothesis rather than empirical data (King 1990). Nevertheless, a number
of estates owned directly by an individual and controlled perhaps indirectly through
their family friends and clients could easily total many thousands of hectares and be
comparable in size to some of the smaller early medieval kingdoms known from docu-
ments like the Tribal Hidage (Dumville 1989; Yorke 1990: 9–​10). Of course, the Roman
estates are unlikely to have been neatly distributed or geographically contiguous, but
860

860   James Gerrard

the control of this land and its inhabitants and produce was the foundation of the late
Roman elite’s power.
The relationship between the state and the Romano-​British elites was probably char-
acterized by symbiosis and paranoia. The state required the elite to serve as its agents
and to enforce its demands, but to eschew the overt trappings of power that were an
imperial monopoly. Thus the villa-​owner may have liked to be portrayed to his clients
and tenants as an ‘emperor’ or ‘hero’ (Ellis 1991), but he had to stay firmly in the civil-
ian world, wedded to the values of paideia (Sivonen 2006:  100–​104, 115–​117). Those
who crossed this line, or were perceived to have done so, were branded by the state as
latrones (bandits) (Grünewald 2004) and bagaudae (perhaps best thought of as insur-
gents) (Drinkwater 1992; Sánchez León 1996). In part, the elite were compensated for
their lack of martial power by a legal system that was heavily biased towards those of
elite social standing and wealth. There were few tenants or free peasants who could hope
to triumph in the late Roman courts against the local landowner, his purse, and connec-
tions (Harries 1999; Kelly 2004). Ultimately the state provided a backstop to this system.
Rebellion by the peasantry would also be characterized as banditry or bagaudic activity
and met, if necessary, with the force of imperial arms.

The End of Roman Britain

The interconnection between economics and power was thrown into sharp relief dur-
ing the early fifth century. The usurpation of Constantine III in ad 407 (Drinkwater
1998) and, more significantly, the failure of the western court to reimpose imperial con-
trol over Britain during the fifth century, mark a significant historical break that is also
reflected archaeologically (Gerrard 2013). The early fifth century saw the disappearance
of the defining characteristics of Romano-​British society and the appearance of new
forms of material culture, architecture and—​particularly in eastern areas—​burial rites.
This transformation and the speed with which these changes occurred are complex
matters and ones that deserve more space than this chapter can devote to them (see also
Esmonde Cleary, this volume). The traditional view has been to see this change as the
product of a catastrophic economic collapse. The disappearance of the state and the
money-​using economy led to the dislocation of market-​based exchange and the virtual
disappearance of trade. Britain was plunged from a familiar economic system of mar-
kets into the alien, early medieval world of gift-​giving and socially embedded exchange
(Wickham 2005).
This chapter has argued that our fixation with archaeologically visible economic
indicators, such as pottery or roof tiles, has skewed interpretations of the late Roman
economy. The economic powerhouse of fourth-​century Britain was not the kilns man-
ufacturing red-​slipped tablewares in Oxfordshire, but grain, crops, meat, and animal
products such as cheese and leather. A significant (but unknown) proportion of the
agricultural surplus had been rendered to the state as tax. In the early fifth century the
 Economy and Power in Late Roman Britain    861

state disappeared. This presented the Romano-​British elites with both an opportunity
and a dilemma.
For the elite the late Roman tax burden offered an opportunity. After subsistence
requirements for their tenants and clients were fulfilled, the slice of surplus that went
to the landlord as rent and the slice that paid the tax—​which were perhaps already
blurred—​became indivisible. Both the rent and the tax went to the landlord (Wickham
1984). However, this opportunity was not without its difficulties. A tax that must have
been raised in kind was perishable. Without a means of transferring that surplus into
specie, the grain would rot, the cheese would go off, and the meat would spoil. The
clients and tenants, recognizing the weakening state and the traditional structures of
elite power, may also have tried to renegotiate their position—​a development possibly
spurred by the appearance of new and alternative ‘Germanic’ power structures in the
east of Britain (Böhme 1986, 1997).
Faced with these difficulties, the late Roman elites undoubtedly responded in differ-
ent ways. For some the fifth century probably offered oblivion, but for those who could
adapt to the changed economic situation and power structures, a new and post-​Roman
path could be discerned. Some of the surplus from their large estates could be preserved
and used to cement the loyalties of clients and retainers. These individuals could be
converted into nascent war bands now that the prohibition on the elite adopting the
emperor’s martial values had ended. The burdens on the agricultural producers could
also be lessened without impacting on the elite’s position (Esmonde Cleary 1989: 145). A
notional fourth-​century tax rate of 10 per cent could be lowered by, say, 8 per cent. This
would reduce the burden on recalcitrant agricultural producers but still leave the elite
taking a larger slice of the cake than they had been under the Roman state.
The argument advanced here is an interesting hypothesis. However, can it be sup-
ported by an interpretation of the historical and archaeological evidence? The historical
sources are limited and equivocal. Zosimus (New History 6.5.3), drawing on the contem-
porary and largely lost Olympiodorus (Blockley 1981: 28), tells us that the Britons took
up arms and expelled the Roman state, and this may be a reflection of a society choosing
to go its own way. The Life of St Germanus and Patrick’s writings (Dumville 1989) portray
an elite in the late Roman mould. Finally, Gildas—​in the sixth century—​recounts how
the superbus tyrannus supplied his Saxon mercenaries with foodstuffs, which suggests
that some individuals and communities could accumulate and distribute surpluses.
The archaeological evidence is just as problematic and capable of being interpreted
in contradictory ways. The final phases of many fourth-​century villas display a shift to
‘messy’ deposits, often associated with grain-​driers and metal-​working hearths. This so-​
called squatter occupation is often taken as evidence of economic collapse. However, it
could equally represent a realignment of priorities (Lewit 2003: 268). The old engines of
elite display—​villas—​were now serving a less hierarchical form of social organization.
Buildings that had previously been used to emphasize the separation of the elite from
their clients were now being used for economically productive activities (Ellis 1988;
Petts 1997; Ripoll and Arce 2000). Here it is interesting to note that grain-​driers may
have been used in brewing (Van der Veen 1989). Beer would be one way of preserving a
862

862   James Gerrard

grain surplus and turning it into a beverage that was suitable for use in the feasting that
sought to bind the lord and his clients with ties of loyalty.
The agricultural landscape also underwent significant changes during the fifth cen-
tury. Palynological evidence presents a complex picture, but there is little clear sup-
port for reforestation or a significant national decline in the area of land being worked
(Dark 1996, 2000; Dark and Dark 1997: 143–​144; Fyfe and Rippon 2004). Late Roman
field systems were sometimes abandoned or partially maintained (Rippon 1991, 2006;
Oosthuizen 1998, 2011; Baker 2002; Upex 2002). Again, this would seem at first sight to be
evidence of a collapse in the rural economy. Alternatively, we might speculate that a shift
occurred away from a late Roman rural arable economy driven by the demands of taxa-
tion. In the fifth century the reduced demands from the elite may have encouraged a less
intensive agricultural regime that emphasized cattle and animals. We lack suitable animal
bone assemblages to test this hypothesis (Crabtree 2010: 123), but the emergence of cattle
as a measure of wealth would seem to be a strong possibility (Hamerow 2002: 129).
Finally, recent analysis of osteological evidence has drawn attention to the fact that late
Roman populations were shorter than early medieval individuals. This is a phenomenon
that is visible across Europe (Köpke and Baten 2005) and also discernible in local and
regional studies (Roberts and Cox 2003: 389–​390; Klingle 2011). Traditionally this height
difference has been viewed as indicative of migration, with incoming Germanic groups
being taller than the indigenes. Interestingly, recent work has suggested that height might
be a proxy for health (Köpke and Baten 2005). The arguments are not straightforward, but
perhaps fifth-​and sixth-​century populations ate better than their Roman forebears.

Conclusions

Power and economics are inextricably intertwined. The study of the latter has been
largely and unfortunately neglected, leaving much of the data to rest on implicit and
poorly founded assumptions and models. A more critical interpretation of the fourth-​
century Romano-​British economy, using models developed for other regions of the
Roman Empire, perhaps suggests that archaeologists, by focusing on what survives,
have neglected the less visible agricultural economy that served as the foundation for
late Roman society.
A consideration of late Roman social stratification and the state suggests that the abil-
ity to control and exploit agricultural surpluses was critical to the maintenance and rep-
lication of power. The fifth-​century crisis, brought about by the failure of the western
Empire to bring Britain back into the imperial fold after the defeat of Constantine III’s
usurpation, left the insular power structures in a state of flux. Those capable of exploit-
ing and maintaining their hold over the agricultural economy used their surplus to buy
the loyalty of retainers and clients. By adopting the trappings of military power, these
elites could reinvent themselves and their children as post-​Roman warlords and—​ulti-
mately—​petty kings.
 Economy and Power in Late Roman Britain    863

Acknowledgements

I am grateful to Mr Andrew Agate for preparing the illustrations.

References
Primary Sources
Constantius, Life of St Germanus. ‘Constantius of Lyon: The Life of St Germanus of Auxerre’, ed.
and trans. T. Noble and T. Head, in Soldiers of Christ: Saints Lives from Late Antiquity and the
Early Middle Ages. Philadelphia: Pennsylvania State University Press, 1994, 75–​106.
Patrick, Letter to Coroticus: S. Patrick: His Writings and Muirchu’s Life, ed. and trans. A. B. E.
Hood. Arthurian Period Sources 9. Chichester: Phillimore & Co., 1978.
Salvian, De Gubernatione Dei. The Writings of Salvian the Presbyter, trans. J. O’Sullivan.
Washington: Catholic University of America Press, 1947.
Tab. Vindol. = The Vindolanda Tablets Online <http://​vindolanda.csad.ox.ac.uk/​TVII-​343>
(accessed 4 December ​2014).
Zosimus. Zosimus: A New History, trans. and commentary by Ronald T. Ridley. Sydney:
University of Sydney, 1982.

Secondary Sources
Adcock, J., and Wood, E. (2005). ‘Geophysical Survey Report: Dinnington, Somerset’.
Unpublished GSB Prospection Report 2005/​33.
Allen, M., and Fulford, M. (1996). ‘The Distribution of South-​East Dorset Black Burnished
Category 1 Pottery in South-​West Britain’, Britannia, 27: 223–​281.
Baker, S. (2002). ‘Prehistoric and Romano-​British Landscapes at Little Wittenham and Long
Wittenham, Oxfordshire’, Oxoniensia, 67: 1–​28.
Bang, P. (2007). The Roman Bazaar: A Comparative Study of Trade and Markets in a Tributary
Empire. Cambridge: Cambridge University Press.
Beattie, I., and Pythian-​Adams, W. (1913). ‘A Romano-​British House near Bedmore Barn, Ham
Hill, Somerset’, Journal of Roman Studies, 3/​1: 127–​133.
Blockley, R. (1981). The Fragmentary Classicizing Historians of the Later Roman
Empire: Eunapius, Olympiodorus, Priscus and Malchus. Liverpool: Francis Cairns.
Böhme, H. (1986). ‘Das Ende der Römerherrschaft in Britannien und die Angelsächsich
Besiedlung Englands im 5. Jahrhundert’, Jahrbuch des Römisches-​Germaniches Zentralmuseum,
33: 469–​574.
Böhme, H. (1997). ‘Söldner und Siedler im spätantiken Nordgallien’, Die Franken, 1: 91–​101.
Bowes, K. (2010). Houses and Society in the Later Roman Empire. London: Duckworth.
Broadberry, S., Campbell, B., and van Leuwen, B. (2010). ‘English Medieval Population: Reconciling
Time Series and Cross-​Sectional Evidence’. University of Warwick: Unpublished working paper.
Brown, P. (1971). The World of Late Antiquity. London: Thames and Hudson.
Brown, P. (1992). Power and Persuasion in Late Antiquity:  Towards a Christian Empire.
Madison: University of Wisconsin Press.
864

864   James Gerrard

Brunning, R. (2001). ‘Lopen Bridge Farm’, Somerset Archaeology and Natural History,
145: 142–​143.
Clark, C., and Haswell, M. (1970). The Economics of Subsistence Agriculture.
London: MacMillan.
Crabtree, P. (2010). ‘Agricultural Innovation and Socio-​Economic Change in Early Medieval
Europe: Evidence from Britain and France’, World Archaeology, 42/​1: 122–​136.
Cuttler, R., Davidson, A., and Hughes, G. (2012). A Corridor through Time: The Archaeology of
the A55 Anglesey Road Scheme. Oxford: Oxbow Books.
Dark, K., and Dark, P. (1997). The Landscape of Roman Britain. Stroud: Tempus.
Dark, P. (1996). ‘Palaeoecological Evidence for Landscape for Landscape Continuity and
Change in Britain ad 400–​800’, in K. Dark (ed.), External Contacts and the Economy of Late
and Post-​Roman Britain. Woodbridge: Boydell Press, 23–​51.
Dark, P. (2000). The Environment of Britain in the First Millennium ad. London: Duckworth.
de la Bédoyère, G. (2006). Roman Britain: A New History. London: Thames and Hudson.
Drinkwater, J. (1992). ‘The bacaudae of Fifth-​Century Gaul’, in J. Drinkwater and H. Elton
(eds), Fifth-​Century Gaul: A Crisis of Identity? Cambridge: Cambridge University Press,
208–​217.
Drinkwater, J. (1998). ‘The Usurpers Constantine III (407–​411) and Jovinus 411–​413’, Britannia,
29: 269–​298.
Dumville D. (1989). ‘The Tribal Hidage: An Introduction to its Texts and History’, in S. Bassett
(ed.), The Origins of Anglo-​Saxon Kingdoms. Leicester: Leicester University Press, 225–​230.
Dumville, D. (1993). St Patrick, ad 493–​1993. Woodbridge: Boydell Press.
Dyson, S. (2006). In Pursuit of Ancient Pasts:  A  History of Classical Archaeology in the
Nineteenth and Twentieth Centuries. New Haven: Yale University Press.
Ellis, S. (1988). ‘The End of the Roman House’, American Journal of Archaeology, 92/​
4: 565–​576.
Ellis, S. (1991). ‘Power, Architecture and Decor: How the Late Roman Aristocrat Appeared to
his Guests’, in E. Gazda (ed.), Roman Art in the Private Sphere. Ann Arbor, University of
Michigan Press, 117–​134.
Esmonde Cleary, S. (1989). The Ending of Roman Britain. London: Batsford.
Evans, J. (1981). ‘Wheat Production and its Social Consequences in the Roman World’, Classical
Quarterly, 31/​2: 428–​442.
Frere, S. (1967). Britannia: A History of Roman Britain. London: Routledge and Kegan Paul.
Fulford, M. (1979). ‘Pottery Production and Trade at the End of Roman Britain:  The Case
against Continuity’, in P. Casey (ed.), The End of Roman Britain. BAR British Series 71.
Oxford: Archaeopress, 120–​132.
Fulford, M., and Hodder, I. (1975). ‘A Regression Analysis of Some Late Romano-​British
Pottery: A Case Study’, Oxoniensia, 39: 26–​33.
Fyfe, F., and Rippon, S. (2004). ‘A Landscape in Transition? Palaeoenvironmental Evidence
for the “End” of the Romano-​British Period in South-​West England’, in R. Collins and J.
Gerrard (eds) Debating Late Antiquity in Britain ad 300–​700. BAR British Series 365.
Oxford: Archaeopress, 33–​42.
Garnsey, P. (1970). Social Status and Legal Privilege in the Roman Empire. Oxford, Oxford
University Press.
Gentili, G. (1999). La Villa Romana di Piazza Armerina Palazzo Erculio. Osimo: Fondazione
Don Carlo.
Gerrard, J. (2013). The Ruin of Roman Britain. Cambridge: Cambridge University Press.
 Economy and Power in Late Roman Britain    865

Graham, A., and Mills, J. (1995). ‘A Romano-​British Building at Crimbleford Knap, Seavington
St Mary’, Somerset Archaeology and Natural History, 139: 119–​134.
Greene, K. (1998). ‘Review Articles:  The Study of Roman Pottery in Britain in the Late
1990s: Progress and Prospects’, Britannia, 29: 447–​450.
Grünewald, T. (2004). Bandits in the Roman Empire: Myth and Reality, trans. J. Drinkwater.
London: Routledge.
Guest, P. (2005). The Late Roman Gold and Silver Coins from the Hoxne Hoard. London: British
Museum Press.
Hamerow, H. (2002). Early Medieval Settlements: The Archaeology of Rural Communities in
Northwest Europe 400–​900. Oxford: Oxford University Press.
Harries, J. (1999). Law and Empire in Late Antiquity. Cambridge: Cambridge University Press.
Harries, J. (2007). Law and Crime in the Roman World. Cambridge:  Cambridge
University Press.
Haverfield, F. (1906). ‘Romano-​British Somerset’, Victoria County History of Somerset, I.
207–​371.
Hayward, L. (1952). ‘The Roman Villa at Lufton, near Yeovil’, Proceedings of the Somerset
Archaeological and Natural History Society, 97: 91–​112.
Hayward, L. (1972). ‘The Roman Villa at Lufton, near Yeovil’, Proceedings of the Somerset
Archaeological and Natural History Society, 116: 59–​77.
Heather, P. (2005). The Fall of the Roman Empire: A New History. London: Macmillan.
Hobbs, R. (2006). Late Roman Precious Metal Deposits c. ad 200–​700: Changes over Time and
Space. BAR International Series 1540. Oxford: Archaeopress.
Hodder, I. (1974). ‘Some Marketing Models for Romano-​British Coarse Pottery’, Britannia,
5: 340–​359.
Holbrook, N. (2004). ‘Turkdean Roman Villa, Gloucestershire: Archaeological Investigations
1997–​1998’, Britannia, 35: 39–​76.
Johns, C. (2010). The Hoxne Late Roman Treasure:  Gold Jewellery and Silver Plate.
London: British Museum Press.
Jones, A. (1964). The Later Roman Empire 284–​602:  A  Social, Economic and Administrative
Survey. Baltimore: Johns Hopkins University Press.
Kelly, C. (2004). Ruling the Later Roman Empire. Harvard: Harvard University Press.
King, A. (1990). Roman Gaul and Germany. Los Angeles: University of California Press.
Klingle, D. (2011). ‘The Use of Skeletal and Mortuary Evidence to Understand the Transition
from Roman to Anglo-​Saxon England in the Cambridgeshire Region’. Unpublished Ph.D.
thesis, Department of Archaeology, Cambridge University.
Kolb, M., and Snead, J. (1997). ‘It’s a Small World After All:  Comparative Analyses of
Community Organization in Archaeology’, American Antiquity, 62/​4: 609–​628.
Köpke, N., and Baten, J. (2005). ‘The Biological Standard of Living in Europe during the Last
Two Millennia’, European Review of Economic History, 9/​1: 61–​95.
Lawrence, S., and Smith, A. (2009). Between Villa and Town: Excavations of a Roman Roadside
Settlement and Shrine at Higham Ferrers, Northamptonshire. Oxford: Oxbow Books.
Leech, R. (1981). ‘The Excavation of a Romano-​British Farmstead and Cemetery on Bradley
Hill, Somerset’, Britannia, 12: 177–​252.
Leech, R. (1982). Excavations at Catsgore 1970–​73: A Romano-​British Village. Bristol: Western
Archaeological Trust.
Lewit, T. (2003). ‘Vanishing Villas: What Happened to Elite Rural Habitation in the West in the
5th–​6th Centuries?’, Journal of Roman Archaeology, 16: 260–​274.
866

866   James Gerrard

Lovell, J. (2006). ‘Excavation of a Romano-​British Farmstead at RNAS Yeovilton’, Proceedings

of the Somerset Archaeological and Natural History Society, 149: 7–​70.
Maltby, M. (2010). Feeding a Roman Town:  Environmental Evidence from Excavations in
Winchester 1972-​1985. Winchester: Winchester Museums.
Mathisen, R. (2001). ‘Imperial Honorifics and Senatorial Status in Late Roman Legal
Documents’, in R. Mathisen (ed.), Law, Society and Authority in Late Antiquity.
Oxford: Oxford University Press, 179–​207.
Mattingly, D. (2007). An Imperial Possession: Britain in the Roman Empire. London: Penguin.
Mayerson, P. (1984). ‘Roman Wheat Production:  An Addendum’, Classical Quarterly, 34/​
1: 243–​245.
Millett, M. (1990). The Romanization of Britain:  An Essay in Archaeological Interpretation.
Cambridge: Cambridge University Press.
Moore, J. (1862). ‘Roman Villa at West Coker, Somersetshire’, Journal of the British
Archaeological Association, 18: 392–​395.
Moreland, J. (2001). Archaeology and Text. London: Duckworth.
Moreland, J. (2006). ‘Archaeology and Texts: Subservience or Enlightenment?’, Annual Review
of Archaeology, 35: 135–​151.
Näf, B. (1995). Senatorisches Standesbewusststein in spätrömischer Zeit. Beiträge zur Geschichte
der altchristlichenen Literatur und Theologie 40. Freiburg, Paradosis.
Oosthuizen, S. (1998). ‘Prehistoric Fields into Medieval Furlongs? Evidence from Caxton,
South Cambridgeshire’, Proceedings of the Cambridge Antiquarian Society, 86: 145–​152.
Oosthuizen, S. (2011). ‘Anglo-​Saxon Fields’, in D. Hinton, S. Hamerow, and S. Chapman (eds),
Handbook of Anglo-​Saxon Archaeology. Oxford: Oxford University Press, 377–​401.
Parker Pearson, M., and Ramilisonina, R. (1998). ‘Stonehenge for the Ancestors: The Stones
Pass on the Message’, Antiquity, 72/​276: 308–​326.
Petts, D. (1997). ‘Elite Settlements in the Roman and Sub-​Roman period’, in K. Meadows, C.
Lemke, and J. Heron (eds), TRAC 96: Proceedings of the Sixth Theoretical Roman Archaeology
Conference, Sheffield 1996. Oxford: Oxbow Books, 101–​112.
Pharr, C. (1952) (trans.). The Theodosian Code and Novels and the Sirmondian
Constitutions:  A  Translation with Commentary, Glossary and Bibliography.
Princeton: Princeton University Press.
Radford, C. (1928). ‘The Roman Site at Westland’, Proceedings of the Somerset Archaeological
and Natural History Society, 74: 122–​143.
Ripoll, G., and Arce, J. (2000). ‘The Transformation and End of Roman villae in the West
(Fourth to Sixth Centuries): Problems and Perspectives’, in G. Brogilo, N. Gauther, and N.
Christie (eds), Towns and their Territories between Late Antiquity and the Early Middle Ages.
Transformation of the Roman World 9. Leiden: Brill, 63–​114.
Rippon, S. (1991). ‘Early Planned Landscapes in South-​East Essex’, Essex Archaeology and
History, 22: 46–​60.
Rippon, S. (2006). Landscape, Community and Colonisation: The North Somerset Levels during
the 1st to 2nd Millennia ad. London: Oxford University Press and the British Academy.
Roberts, C., and Cox, M. (2003). Health and Disease in Britain: From Prehistory until the Present
Day. Stroud: Sutton.
Salway, P. (1981). Roman Britain. Oxford: Clarendon Press.
Sánchez León, J. (1996). Les Sources de L’Histoire des Bagaudes: Traduction et Commentaire.
Besancon: Annales Littéraires de l’Université de Franche-​Comté.
 Economy and Power in Late Roman Britain    867

Scott, S. (2000). Art and Society in Fourth-​Century Britain: Villa Mosaics in Context. Oxford
University School of Archaeology Monograph 53. Oxford:  Oxford University School of
Archaeology.
Sharples, N., Evans, C., Slater, A., Payne, A., Linford, P., and Linford, N. (2012). ‘Ham Hill’,
British Archaeology, 123: 34–​39.
Sivan, H. (1993). Ausonius of Bordeaux: Genesis of a Gallic Aristocracy. London: Routledge.
Sivonen, P. (2006). Being a Roman Magistrate:  Office Holding and Roman Identity in Late
Antique Gaul. Bibliotheca Historica 101. Helsinki: Bibliotheca Historica.
Smith, J. (1997). Roman Villas: A Study in Social Structure. London: Routledge.
Sorkin, A. (2009). Too Big to Fail. London: Viking.
Taylor, J. (2001). ‘Rural Society in Roman Britain’, in S. James and M. Millett (eds), Britons and
Romans: Advancing an Archaeological Agenda. CBA Research Report 125. London: Council
for British Archaeology, 46–​59.
Taylor, J. (2004). ‘The Distribution and Exchange of Pink, Grog-​Tempered Pottery in the East
Midlands’, Journal of Roman Pottery Studies, 11: 60–​66.
Taylor, J. (2007). An Atlas of Roman Rural Settlement in England. CBA Research Report 151.
London: Council for British Archaeology.
Timby, J. (1998). Excavations at Kingscote and Wycomb, Gloucestershire: A Roman Estate Centre
and Small Town in the Cotswolds. Cirencester: Cotswold Archaeological Trust.
Upex, S. (2002). ‘Landscape Continuity and Fossilization of Roman Fields’, Archaeological
Journal, 159: 77–​108.
Van der Veen, M. (1989). ‘Charred Grain Assemblages from Roman-​Period Corn Driers in
Britain’, Archaeological Journal, 146: 302–​319.
Van der Veen, M., Lavarda, A., and Hill, A. (2007). ‘The Archaeobotany of Roman Britain:
Current State and Identification of Research Priorities’, Britannia, 38: 181–​210.
Wallace-​Hadrill, A. (1989) (ed.). Patronage in Ancient Society. London: Routledge.
Ward-​Perkins, B. (2005). The Fall of Rome and the End of Civilization. Oxford:  Oxford
University Press.
Wickham, C. (1984). ‘The Other Transition:  From Ancient World to Feudalism’, Past and
Present, 103: 3–​36.
Wickham, C. (2005). Framing the Early Middle Ages: Europe and the Mediterranean, 400–​800.
Oxford: Oxford University Press.
Yorke, B. (1990). Kings and Kingdoms of Early Anglo-​Saxon England. London: Routledge.
868
 Index

Figures and tables are indicated by “f ” and “t” following page numbers.

Acculturation, 276, 288, 386, 836 Anglesey, Roman conquest of, xxx, xxxi

Actor–​Network Theory, 499 Animals, 791–​806
Adams, J. N., 578, 588–​589, 592n8 birds, 654, 800–​801
Adler, P. S., 225 cattle, 651, 792–​795, 814
Adolescence. See Childhood and adolescence dogs and cats, 653–​654, 799
ADS (Archaeology Data Service), 50 donkeys and mules, 799
Aerial photography, 54–​55, 30, 700, 703, fish, 801
715, 716f horses, 798–​799
Age. See also Life course hunted animals, 799–​800
biological, 322, 323 overview, 791
burial patterns and, 313–​314, 322–​323, pigs, 797–​798
325–331, 327f, 328t, 330t ritual deposition of, 646, 650, 651, 653–​654
chronological, 322, 331 sheep and goats, 650, 795–​797
of decapitated burials, 391f, 392–​393 zooarchaeological evidence for,
epitaphs and, 327–​328, 329 791–​792, 802
gender and, 322 Antiquarians, artefact studies by, 63–​64
identity and, 304, 314, 322 Antonine Wall
at marriage, 305, 329 abandonment of, xxxii
of skeletal remains, 307, 323 building process, 185
social, 322, 324 forts along, 14
Agricola (Tacitus), 4, 8, 110, 226 as frontier, 224
Agricola, military campaigns under, inscriptions from, 8, 293
xxx–​xxxi, 184 votive altars at, 622
Agriculture. See Farming Antoninus Pius, xxxi–​xxxii, 183f, 185
Akerman, J., 65 Appearance. See Personal appearance
Albarella, U., 798, 800 Arable farming, 808–​810, 812–​814
Allason-​Jones, Lindsay, 365, 366, 464 Archaeologia Aeliana (journal), 23, 28, 64
Allectus, xxxiii–​xxxiv Archaeology. See also Bioarchaeology;
Allen, Denise, 470 Romano-​British archaeology
Allen, J., 517–​518 behavioural archaeology, 486
Allison, Pim E., 362, 366, 373, 450 cognitive archaeology, 487, 488–​489
Amalgam-​gilding technique, 534 funding for projects, 23, 24, 48–​50, 179–​180
Ammianus Marcellinus, 110 in higher education, 32–​33, 35
Amphitheatres, 759, 779, 780, 782 memory and, 682–​684
Anderson, A. W., 227 multiculturalism in interpretations,
Androcentricism, 366, 367 216–​217
870
870   Index
Archaeology (Cont.)
New Archaeology movement, 32, 382,
383, 485
professionalization of, 44, 48
rescue archaeology, 44, 47, 50, 26, 27, 32
symmetrical archaeology, 487, 488
in United Kingdom, 43–​44
zooarchaeology, 791–​802. See also Animals
Archaeology Data Service (ADS), 50
The Archaeology of 'Race': Exploring the
Northern Frontier in Roman Britain
exhibition (2009), 216
Archaeology of Ritual and Magic
(Merrifield), 82
The Archaeology of Roman Britain
(Collingwood & Richmond), 77, 483
Arch of Constantine (Rome), 685
Aristotelian approach to identity, 246
Armour, 452, 465–​468
Army. See Military community
Art, 599–​618
Celtic, 181–​182
coinage as, 600
dating challenges for, 602
distribution of, 600–​602, 601f
interpretations of, 612–​614, 615
mosaics. See Mosaics
overview, 599–​600
paintings. See Wall-​paintings
patron–​client relationships in, 607–​608
pottery as, 600
producers of, 606–​609
quality of, 602–​604, 606, 614–​615
religious, 610, 613
sculpture. See Sculpture
users and uses of, 609–​611, 615
Artefact studies, 63–​96. See also Material
culture; Materiality; specific types of
artefacts
in 1930s–​1970s, 76–​79, 79
in 1970s and beyond, 79–​83
antiquarian interest in, 63–​64
archaeological approaches to,
64–​73, 66–​7 1t
classification and interpretation of,
65, 69–​76
overview, 63
profile of artefact researchers, 82–​83
publication of, 73–​74
social use of space derived from, 188
technological innovations in, 83–​84
text-​based scholarship vs., 83, 84
by women, 72, 77, 78f, 81–​82, 82–​83
Arthropathy, 397
Arthur, P., 519
Ashanti tribe, 388
Aspöck, E., 389, 400
Assman, Aleida and Jan, 682
Atlas of Roman Rural Settlement (Taylor),
687, 708
Augustus, xxix, 152, 163, 685
Aurelian, xxxiii
Aylesford–​Swarling tradition, 159–​160,
511, 689
Babenberg dynasty, burial practices of, 388
Baker, Patricia, 367, 555
Balty, J. C., 769
Baptism, 670–​672
Barates of Palmyra, 214, 238, 240, 298
Barbarino, J. L., 587
Barbarous radiate coinage, 842–​843
Barber, B., 387
Barber, E. W., 419
Barker, Phil, 27, 31
Barley, 808, 814
Barrett, A. A., 105
Basilicas, 122, 492, 495, 666, 769, 772
Bathing, 349, 351, 373, 495, 714, 780–​782
Bath tablet collections, 100, 108, 109, 580–​583,
581f, 583f
Bayesian statistical analysis, 179
Bayford grave assemblages, 349, 351
Bayley, J., 420
Beaumont, J., 315
Beer production, 810, 812–​813, 861–​862
Behavioural archaeology, 486
Biddle, Martin, 27
Bignor villa, excavation of, 9, 10f, 26
Bilingualism, 577–​578, 579, 584, 588, 591
Bioarchaeology
of childhood, 306–​308, 315, 316
in gender studies, 372, 374
medicine and, 567
 Index   871

Biological age, 322, 323 Brooches

Birdoswald site artefact studies of, 71, 72, 73, 78
cemetery excavations at, 236, 457–​458 British adoption of, 159
geophysical surveys of, 55 at burial sites, 257–​257
purse hoard from, 840 Celtic nature of, 76, 82, 188
Birds, 654, 800–​801 chronological and regional distribution of,
Birley, Anthony R., 31, 37, 455 411–​413, 412t, 414f, 415–​416
Birley, Eric, 43, 25–​26, 27–​28, 455 coloured, 415–​416
Birley, Robin, 26 cross-​bow, 275–​276, 370, 13, 417
Biró, M., 622 evidentiary sources for, 420
Bishop, Mike, 468 gender and, 135, 370, 413, 415
Black, E., 654 metal content of, 540, 540f
Blacksmithing, 548, 549 military, 469
Blagg, Tom, 36, 774 in mobility studies, 207, 249, 250, 250f,
Boar, 800 255–​258
Bonsall, L., 374 ritual deposition of, 655
Boon, G. C., 76, 78 Brougham burial site
Booth, Paul, 434, 522 age-​related patterns at, 343, 368
Boudican Revolt (AD 60), xxx, 117, 119, 121, child burials at, 324
459–​460, 514 cremation at, 437
Bowman, A. K., 96, 104 diversity of remains at, 214
Boylston, A., 384 gender patterns at, 343, 368, 372
Bradley, Richard, 50, 683 inscriptions at, 439, 458
Brailsford, J. W., 517 material culture at, 236–​237, 275, 346
Brass, 532, 538, 541 Bruce, John Collingwood, 14
Braund, D., 163 Bruhn, J., 275
Bread wheat, 814 BSR (British School at Rome), 28, 29
Breastfeeding, 308–​309 Building dedications, 287, 292
Breeze, David, 36, 227 Burials, 341–​362. See also Cremation;
Brigantes tribe, xxx Funerary process; Tombstones; specific
Britain. See Roman Britain; United burial sites
Kingdom (UK) age-​related patterns in, 313–​314, 322–​323,
Britannia (Camden), 3–​4, 5, 7 325–​331, 327f, 328t, 330t
Britannia: A History of Roman Britain (Frere), Bronze Age, 314, 345
43, 30, 483 brooches at, 256–​257
Britannia: The Creation of a Roman Province of children, 310–​315
(Creighton), 125 Christian, 672–​674
British School at Rome (BSR), 28, 29 coffins for, 348–​348, 394f, 396, 428, 430–​431
Brochs, 179, 180, 190, 194, 606 decontextualization of, 145
Bronze, 493, 532, 538, 540, 839, 844 demarcation of, 353–​351, 439
Bronze Age destruction and degradation of, 343
burials, 314, 345 deviant. See Deviant burials
continental contacts in, 155 family circle graves, 16, 17f
feasting in, 161 in frontier zones, 236–​237
gold in, 157 gender patterns in, 313, 326–​329, 328t, 330,
hillfort construction in, 156 330t, 368–​369
monuments of, 686–​687 of infants, 311–​312, 325, 369, 428, 715
872

872   Index

Burials (Cont.) Camulodunum site

Iron Age, 151, 159–​161, 344, 345
material culture in, 145, 236–​237, 349–​352,
350f, 352f, 437–​438
in military community, 457
overview, 341–​342
plaster, 347, 430
pre-​interment rituals for, 345–​347
regional variations in, 331–​333, 332–​333f
spatial organization of, 352–​353, 368–​369,
433–​434
status and, 309–​310, 314, 341–​345, 354–​356
urban vs. rural, 343, 353–​354, 431–​432,
735–​736, 736f
Bushe-​Fox, J. P., 26, 28, 73, 74, 75
Butcher, S., 420
Butt Road cemetery
age-​related patterns at, 322
baptismal font at, 671
child burials at, 309–​310, 313–​314, 325
church structures at, 667
clothing at, 407
gender patterns at, 313
spatial organization of, 433, 673
Caerwent site
amphitheatre at, 779
bathing complex at, 781
cattle at, 794
coin assemblages from, 844
development of, 745, 755
forum at, 492, 768, 769, 770f, 772
pigs at, 797, 798
transformation of, 495
twentieth-​century excavations of, 23
Caesar. See Julius Caesar
Caistor-​by-​Norwich
bathing complex at, 781
development of, 755
forum at, 769, 772
public buildings at, 785
temples at, 776–​777, 778
theatre at, 779
Caligula, xxix
Cambridgeshire, decapitated burials in, 389–​
399, 390t, 391–​395f, 398f
Camden, William, 3, 5–​6, 7, 18, 19
construction of, 756
continuity at, 128–​130
Iron Age settlement at, 26
pottery from, 28
Canterbury
bathing complex at, 495, 781
burial sites at, 167, 356, 467
coin assemblages from, 844
establishment of, 165, 755
grave goods at, 438
silver hoard from, 664
theatre at, 789, 779
urban structure of, 774, 775f
Cantonal system, 748
Caracalla, xxxiii, 185, 255, 344
Carausius, xxxiii, 452
Carbon isotope analysis, 212–​213
Care of children, 307–​310
Carlisle site, 98, 99, 184, 746
Carroll, M., 204
Casey, John, 32, 80, 85
Cassius Dio, 110, 119, 152, 248, 746–​747
Castgore site, 853, 854–​855, 857–​858
Cats, 799
Catterick burials, 370, 371
Cattle, 651, 792–​795, 814
CBA (Council for British Archaeology), 29, 52
The Celt, the Roman and the Saxon
(Wright), 11
Celtic–​Latin bilingualism, 577–​578, 579, 584,
588, 591
Celtic peoples
art of, 181–​182
languages of, 575–​578, 580–​584, 581f
as subalterns, 4–​5, 8–​10, 13, 18
Cementation technique, 541
Cemeteries. See Burials; Tombstones
Centuries of Childhood (Ariès), 303
Ceramics. See Pottery
Cereal crops, 808, 809, 813
Chamberlain, A. T., 311
Champion, Timothy, 150
Charles-​Edwards, T., 588
Charlesworth, D., 76
Cheesman, G. L., 24
Chesters, excavated forum at, 14, 16f

Chickens, 800
Childhood and adolescence, 303–​321. See also
Infancy; Life course
age and sex determination, 306–​307
bioarchaeology of, 306–​310, 315, 316
classical perspectives on, 304–​306
culturally specific constructions of, 303,
304, 306
diet in, 308–​309
diseases of, 308, 309
funerary evidence of, 310–​315
future directions for research, 315–​316
health and care considerations in, 307–​310
overview, 303–​304
rites of passage in, 305
Childhood in the Past (journal), 304
Chi-​rho symbol, 663, 664, 668–​670, 669f, 671
Chorography, 5–​6
Christianity, 660–​680. See also Religion
baptism in, 670–​671
in Britain, 11, 143, 387
churches and, 143–​144, 409, 454, 665–​670,
667f, 669f
death and burial in, 672–​674
documentary evidence for, 661–​663, 674
early Christianity, 660–​661
in Gaul, 143
hairstyles influenced by, 419
imagery and symbolism in, 663–​664,
668–​670
martyrdom and, 662
material culture and, 663–​665, 665f
in military community, 454
in mosaics, 669–​670, 671
overview, 660
paganism, relationship with, 675
regional variations in, 675–​676
in wall-​paintings, 143, 661, 668–​669, 669f
Chronological age, 322, 331
Churches, 143–​144, 409, 454, 665–​670,
667f, 669f
Cicero, 681, 682
Cirencester site
amphitheatre at, 759, 780
cattle at, 793, 794
fort at, 754
forum at, 769, 771, 772
Index   873
graves goods at, 329, 330t
organization of excavations at, 27, 45, 45f
pottery at, 523
public buildings at, 785
sheep at, 795
sustained occupation of, 14
temple at, 776
theatre at, 782
Cities. See Urban areas
Citizenship, Roman, 232–​233, 452
Civil districts, 14, 15f, 18
Clarke, G., 208
Clarke, S., 188, 642, 652
Classicianus, 103, 103f, 110, 293, 353, 434
Claudia Rufina, 7, 254, 410
Claudius, xxix, 6, 119, 127, 341
Clay, C., 204
Cleere, H. F., 546, 549
Client-​kingdoms, 510, 511–​513, 514
Clodius Albinus, xxxii
Clothing, 406–​424. See also Brooches;
Personal appearance
accessories, 135–​136, 159, 249
evidentiary sources for, 406–​410, 419–​420
footwear, 410, 420, 438, 450
gender and, 369–​371, 409
manufacturing process for, 407–​408
military, 416–​417, 419–​420, 452
overview, 406
religion and, 418
styles of, 409–​409
togas, 409, 418
tube dresses, 409, 413
tunics, 409, 416
Coffins, 348–​349, 394f, 396, 428, 429–​430
Cogidubnus inscription, 6–​7, 7f, 8
Cognitive archaeology, 487, 488–​489
Cohen, R., 214
Coinage, 834–​849
artefact studies of, 64, 65, 75, 76, 79, 85
as art object, 600
bronze, 493, 839, 844
cessation of coin supply, 844–​845
distribution of, 600
as economic evidence, 77
in first century AD, 838–​840
in fourth century AD, 843–​844
874
874   Index
Coinage (Cont.)
in Gallo-​Roman horizon, 162, 266, 512,
512f, 513
gold, 137, 157–​158, 159, 163–​164, 532
inscriptions on, 9, 101, 164, 575–​576
Iron Age, 107, 152, 157–​159, 161–​162,
836–​838, 837f
Latin on, 575–​576
overview, 834, 836
pioneering studies on, 32
radiate coinage, 841–​843
regional distribution of, 512, 512f, 513, 516f
ritual deposition of, 161–​162, 644–​645, 647,
648, 841
Romanization of, 16, 163–​164, 510
in second and third century AD, 840–​843
silver, 137, 164, 195, 493, 537, 541, 541f
striking and distribution of, 136–​137
textual evidence from, 107
transformations in, 491–​492, 493–​494
Coins and the Archaeologist (Casey &
Reece), 80
Coins of the Ancient Britons (Evans), 151
Colchester site
age-​related burial patterns at, 326, 327f, 328
cattle at, 793, 795
church structures at, 666–​667
construction of, 752–​754, 754f
continuity at, 128–​130
dating evidence at, 128
in early Roman horizon, 121, 122
forum at, 772
industrial activity at, 783
industrial production at, 164–​165
infant burials at, 325
location of, 750
military settlements at, 120
pigs at, 797
plaster burials at, 347
pottery from, 513, 519, 730
sheep at, 796
site reports from, 80
temple complex at, 774, 776, 785
theatre at, 779, 781
veterans at, 459–​460
Coleraine hoard, 845, 845f
Collective memory, 681, 682, 691
Collingwood, R. G., xxvii, 24–​25, 76, 100, 105,
482–​483, 602, 603, 720–​721, 722, 731, 737
Collins, R., 417
Collyrium stamps, 560–​562
Colonial diasporas, 214, 214
Colostrum, 309
Coloured brooches, 415–​416
Commemoration stage of funerary process,
438–​439
Commius, 163, 264
Commodus, xxxii
Community archaeology groups, 52
Condron, F., 549
Conkey, M. W., 363
Conolly, R., 579
Constans, xxxiv
Constantine, xxxiv, 454, 662, 685
Constantine III, xxxv, 860, 862
Constantius, xxxiii–​xxxiv
Construction process for urban areas, 752–​
760, 753–​754f, 757f
Cook, James, 150
Cool, Hilary, 214, 236–​237, 240, 327, 328, 406,
420, 451, 458, 520, 522, 526
Cooper, N., 731
Coote, Henry, 11
Copper and copper alloys, 532, 538–​542, 540–​
541f, 548–​549
Corbridge hoard, 467, 472
Core-​periphery models, 689, 721
Corn-​driers, 808, 813
Cosh, Stephen, 37
Cosmetics, 411, 416
Cotswold Archaeology, 28, 50, 699
Council for British Archaeology (CBA), 29, 52,
Council of Arles, 662, 663
Cox, John, 72
Craniomorphometric analysis, 215–​216, 215f
Creighton, J., 125, 163, 164, 267, 277, 510, 511–​
513, 516, 749, 755, 758
Cremation. See also Funerary process
ageing and sexing remains from, 323, 368
Aylesford–​Swarling tradition, 159–​160,
511, 689
British adoption of, 168, 434
containment methods, 159–​160,
347–​348, 437
 Index   875

of horses, 236 spatial organization of, 397–​399, 398f

Late Iron Age, 151, 263–​264, 263f, 265f, 428, Deer, 799–​800
436–​437 Defixiones. See Curse tablets
locations for, 431–​433 Deities, 619–​640
pre-​interment rituals for, 345–​346 curse tablets and, 100, 108, 109, 622, 627
rituals in, 435 distribution and dating of, 622–​626,
as technology of remembrance, 686 623f, 625f
Creolization, 214, 437 double-​named, 633–​635
Crerar, Belinda, 381 evidentiary sources on, 620–​622, 635–​636
Cribra orbitalia, 309 geographic nature of, 630–​632, 631f
Croom, A., 419 healing powers of, 564
Crops. See Farming imperial vs. local, 629–​630
Cross-​bow brooches, 275–​276, 370, 413, 417 inscriptions dedicated to, 108, 292,
Croxford, Ben, 599 621–​622
Crummy, N., 75, 80, 82, 154, 560 of military community, 456, 619
Cult practices, 453–​458, 778, 779 offerings to. See Ritual deposition;
Cunetio hoard, 834, 841, 843 Votive altars
Cunliffe, Barry, 27, 28, 32, 127, 153–​154, 156, 517 overview, 619–​620
Curle, Alexander, 73 sculptures involving, 610
Curle, James, 73, 74, 80, 186 vows pledged to, 626, 627
Currency bars, 155 DeLaine, J., 781
Curse tablets (defixiones) Demographic determinism, 311
content and subject matter of, 108–​110 Demography, statistical analysis on, 292–​293
deities and, 100, 108, 109, 622, 627 Dental diseases, 733–​734
in Gaulish, 582 Deposition stages of funerary process,
in Latin, 110, 585–​587, 586f 436–​438
sociolinguistic evidence from, 574 Derks, T., 237, 272, 633, 634, 635
textual evidence from, 99–​101, 99f Deviancy, defined, 381, 382
Uley curse tablets, 99f, 100, 108–​109, 574, Deviant burials, 381–​405
582, 628 characteristics of, 385
Cybele, cult of, 778 decapitated. See Decapitated burials
explanations for, 384–​385
Dalton, O. M., 73, 76 funerary process for, 428
Danebury Environs project, 706, 708, 809 historical approaches to, 381–​386
Daniel, Glyn, 52 individuals associated with, 385–​386
Davidson, C., 368 overview, 381–​382
Decapitated burials, 389–​399 in Romano-​British archaeology,
age of, 391f, 392–​393 388–​389, 400
coffins for, 394f, 396, 428 vicarious human sacrifice and, 384
explanations for, 384–​385, 387–​388 Diasporas in Roman Britain project (2007–​9),
funerary process for, 428 207, 211–​213, 215–​216, 216
gender of, 392f, 393–​394 Diaspora theory, 213–​214, 216–​217
health, disease, and disability in, 397 Díaz-​Andreu, M., 363
material culture in, 393f, 394–​396 Dickinson, Brenda, 470
methodology for analysis, 389–​392, 390t Diet and nutrition. See also Farming
as pagan rites, 387 in childhood, 308–​309
posture of, 395f, 396–​397 cultural influences on, 272–​274
876
876   Index
Diet and nutrition (Cont.)
dental diseases and, 733–​734
gender and, 374
lead in, 308, 427
meat in, 792–​793, 794, 802
North African, 821
for ritual and ceremonial activities, 822
in urban vs. rural areas, 734
variety and diversity in, 81, 819–​823
Dietary deficiency diseases, 308, 309, 315
Dietler, M., 161
Diocletian, xxxiii, 454
Diodorus Siculus, 152, 542
Disability among decapitated burials, 397
Diseases. See also Medicine
agricultural, 810, 811–​812f
of childhood, 308, 309
decapitation burials and, 397
dental, 733–​734
dietary deficiency, 308, 309, 315
gender and, 567
Diversity. See Ethnicity;
Multiculturalism; Race
Division of labour, 371–​372, 374
Dobney, K., 524, 794
Doctors, 563, 564
Dogs, 653–​654, 799
Domestic structures, 373–​374
Domitian, 184
Donkeys, 799
Dorchester settlement
amphitheatre at, 780
bathing complex at, 781
cattle at, 794, 795
chronology of occupation, 157
construction of, 752, 753f, 755, 783
lead coffin liners at, 347, 348
pigs at, 797, 798
ritual deposition in, 653–​654
sheep at, 796
Double-​named deities, 633–​635
Dress. See Clothing
Druids
abolishment from Gaul, 560
military campaigns against, xxx
pictorial representations of, 153
training of, 270
Duality of structure, 246
Ducks, 801
Dungworth, David, 532
Dunning, G., 76, 151
Durham University, 25, 27, 45, 51, 33, 216
Early Roman horizon, 117–​133
Colchester in, 121, 122
Colonia Claudia in, 121, 122
continuity in, 126–​130
evidentiary challenges and biases in study
of, 117–​118
Londinium in, 118–​119, 121, 122–​123
material culture of, 124–​125, 130–​131
overview, 117
public buildings in, 123–​124
regional diversity in, 118
towns in, 123–​126
urbanization in, 125–​126
Verulamium in, 121–​122
East Anglia
building types in, 142
Christianity in, 675–​676
folk customs in, 65
hinterlands of, 725–​726, 728–​730
pottery in, 511
revolts in, xxx
Eckardt, Hella, 203, 683
Economics
gender and, 371–​372
gross domestic product and, 853
laissez-​faire, 722
power and, 850, 856, 860, 862
systems of, 138–​140, 608
taxation and, 136, 227, 526, 721–​722,
855, 856
theoretical approaches to, 851, 852
urban and rural relationships, 720–​722,
723–​724, 729–​731
Edict of Milan (AD 313), 660, 661, 663
Edict of Thessalonica (AD 380), 661
Educational networks, 33, 34–​35t
Educational programs for archaeology,
32–​33, 36
Elite auto-​representation, 142, 143, 144
England
historic preservation responsibilities in, 44
 Index   877

research framework in, 47 Family circle graves, 16, 17f

English Heritage Family structures, 372–​374
historic preservation responsibilities of, 44 Farley, J. M.-​A., 168
National Mapping Programme, 54–​55 Farming, 807–​833
research framework developed by, 47 arable farming, 808–​810, 812–​814
research funding from, 50 decline in, 862
Entertainment in urban areas, 779–​780, 781 diet impacted by, 817, 819–​823
Epigrams (Martial), 7 future research needed, 827
Epigraphic habit germination of grains in, 809
dynamic nature of, 293 horticulture and, 814–​817, 815f, 818f
inscriptions and, 288–​290, 456, 622 modes of preservation for, 815–​816, 815f
in military community, 95, 237 overview, 807–​808
mortuary, 439 pest and diseases in, 810, 811–​812f
sculpture and, 609 regional variations in, 823–​827, 828
Epigraphy scale of production in, 809, 813–​814
Christian, 673–​674 vineyards, 817
inscriptions and, 204–​205, 247, 248–​249, Farmsteads, 127, 128–​129, 131, 706, 708
253–​254, 253f Faulkner, N., 783
military community and, 471 Faussett, B., 64
in mobility studies, 204–​206, 205f, 248–​249, Feasting deposits, 161, 690, 715
251–​255, 252–​253f Females. See Gender; Women
Epitaphs Field, D., 687
age-​related patterns in, 327–​328, 329 Figs, 819, 820
on coffins, 348 Fish, 801
content of, 291–​293 Fishbourne site
decline in, 300 dating evidence at, 127
function of, 107, 287 pigs at, 797
gender and, 326, 328, 329 research framework for, 47, 27
prevalence of, 287, 289 villa at, 270, 704–​706, 705f
Erdrich, M., 193 Fitzpatrick, A. P., 161
Esmonde-​Cleary, Simon, 129, 427, 767–​768 Flavian invasion of Scotland, 183f, 184–​185
Ethnicity. See also Multiculturalism; Race Flavius Cerialis, 104
definitions of, 223–​224, 246 Folkestone villa, 26, 74
identity and, 246, 257, 278, 458–​459 Folly Lane burial site
material culture and, 136, 207 age-​related patterns at, 323
in military community, 458–​459 creation of, 344
Evans, Arthur John, 16–​17, 73, 151, 153 cremation at, 264
Evans, C., 75 laying-​out stage at, 428
Evans, Jeremy, 108, 208–​209, 211–​212, 510, 521, military items at, 167
523–​524, 592n6 pottery from, 730
Evans, John, 151 wealth at, 160, 165
Everill, P., 45 Food. See Diet and nutrition
Excarnation, 434 Footwear, 410, 420, 438, 450
Excavations in Cranborne Chase Forensic ancestry assessment, 215–​216
(Pitt-​Rivers), 71 Forum, as central feature of urban areas,
Eye remedies, 561, 562 768–773, 770–​771f
Eyers, E., 311–​312 Fowl, 800
878

878   Index

Fox, Aileen, 27, 65 interactions with Britain, 262, 263–​272, 265f,

Franks, A. W., 71–​73, 306
Freestone, I. C., 511
Frere, Sheppard, 43, 27, 30, 31, 32, 79, 121, 483,
748, 755, 769, 783
Frisian peoples, 235–​236, 459, 470
Frocester villa, 515, 706, 707f
Frome hoard, 54, 841, 842f, 843
Frontier zones
burial sites in, 236–​237
identities in, 225–​227
indigenous populations of, 226
lifestyle in, 182, 185–​189, 187f, 194
material culture in, 188–​194, 191f, 226
Fulford, Mike, 26, 36, 36, 234, 517–​518,
644, 652
Fuller, B. T., 308–​309
Funding for archaeological projects, 44, 45,
48–​50, 179–​180
Funerary process, 425–​447. See also Burials
commemoration stage, 438–​439
deposition stage, 436–​438
evidentiary sources for, 425, 426t
interpretation and translation of, 440–​441
laying-​out ceremonies in, 429–​431
locations for, 431–​434
modification stage, 434–​435
preparation stage, 427–​431
selection stage, 426–​427
Gaffney, V., 725
Gager, J. G., 109
Gale, Robert, 6
Galen, 307, 735
Galinsky, Karl, 684
Gallic empire, xxxiii
Gallicization, 262, 273, 274, 276–​278
Gardening, 814, 816, 817, 823, 826
Gardner, Andrew, 228, 406, 481
Garments. See Clothing
Gaul
abolishment of druids from, 560
building types in, 141, 142, 143–​145
cemeteries in, 145
Germanic peoples in, 144, 145
immigrants from, 274
inscriptions in, 288
269f, 273–​275, 276–​278
militarization of, 136, 141, 143
patron–​client relationships in, 158–​159
pottery from, 73, 139, 140, 273, 470, 511
religion in, 143
revolts in, xxxiv
Roman conquest of, xxviii, xxxiii, 491
withdrawal of imperial presence from,
135, 137
GDP (gross domestic product), 853
Geese, 801
Gell, Alfred, 485, 488, 499
Gender, 363–​381. See also Women
age and, 322
artefact studies and, 71, 76, 78f, 80–​81, 82–​83
being vs. doing, 364
bioarchaeology of, 372, 374
brooches and, 135, 370, 413, 415
burial patterns and, 313, 326–​329, 328t, 330,
330t, 368–​369
cultivation techniques and, 826–​827
of decapitated burials, 392f, 393–​394
defining, 363–​365
diet and, 374
diseases and, 567
division of labour and, 371–​372, 374
in dress, adornment, and appearance,
369–​371, 409
economic activity and, 371–​372
epitaphs and, 326, 328, 329
in family and domestic structures, 372–​374
future directions for research, 374
identification of, 367–​374
in life course, 305
material culture and, 367–​371
mobility and, 205, 252, 254, 256
overview, 363
Romano-​British gender studies, 365–​367
as social construct, 363–​364
Geographia (Ptolemy), 110, 181, 226, 747
Geophysical surveys, 55, 55f, 38, 700,
701–​703, 702f
Germanic Mode of Production, 154
Germanic peoples
in Gaul, 144, 145
on Hadrian's Wall, 204
 Index   879

languages of, 560, 561 Hadrian's Wall

in Roman Britain, 144 armour and weaponry from, 468
Germanization, 262, 273, 275, 276 coinage from, 844
Germany conflicts regarding, 185
diplomacy and interference in, 194–​195 as defensive system, 137
immigrants from, 274 forts along, 14, 138
inscriptions in, 288–​289 frontier life beyond, 182, 185–​189, 187f, 194
institutional support for archaeology in, 73 geophysical surveys of, 55
interactions with Britain, 262, 268, 272, Germanic peoples on, 204
273–​278 inscriptions from, 8, 291
Germination of grains, 809, 810 invasions across, xxxii
Gerrard, James, 850 keyhole excavation on, 24
Gildas (monk), 8, 454, 662, 861 materiality of, 234–​235
Gillam, John, 28, 79 military presence along, 227–​231, 229f, 230–​231t
Gini coefficient, 735 multiculturalism at, 217, 224–​225, 228–​2331,
Globalization, 484, 487, 499, 500, 737 229f, 230–​231t, 240–​241
Goats, 795–​796 Newcastle excavations on, 45, 51
Gods. See Deities purpose of, xxxi
Gold research framework for, 47
coinage produced from, 137, 157–​158, 159, as tourist attraction, 52
1663–​164, 532 votive altars at, 622
sites for extraction of, 533, 533f writing tablets from, 26
techniques for working with, 533–​534, Hairstyles, 410, 413, 419, 431
535, 549 Halbwachs, Maurice, 682
uses for, 532, 534–​535 Halkon, Peter, 46
Golden, M., 303 Hall, J., 387, 782
Gosden, C., 167, 168 Hamp, E. P., 592n8
Gowland, Rebecca, 303, 307, 308, 311, 313, 322, Handley, M., 205, 206
324, 325, 328, 330, 331 Hanson, W. S., 579
Graffiti scratches, 100, 101f, 105, 108, 592n6 Hapsburg–​Lorraine dynasty, burial practices
Grain weevils, 810, 811f, 819–​821 of, 388
Gratian, xxxv Hares, 800
Gratwick, A. S., 577, 585 Harris, W. V., 102, 622
Graves. See Burials; Tombstones Haselgrove, C. C., 721
Greenfield, Ernest, 44 Hassall, Mark, 33, 582
Griffin, R., 310, 735 Haverfield, Francis, 3, 4–​5, 13–​14, 16–​17, 18–​19,
Grimes, W. F., 27 22, 23–​24, 73, 74, 75, 482, 576, 747–​748
Grog-​tempered pottery, 513, 514, 515, 516 Hawaii, social transformation of, 150
Grooming tools, 411 Hawkes, Christopher, 26, 28
Gros, P., 767 Hawkes, S., 76, 151
Gross domestic product (GDP), 853 Hayling Island temple, 270, 272, 689, 690, 837
Grupe, G., 209 Haynes, Ian, 228, 448
Guildhall Museum exhibition (1962), 29 Health. See also Diseases; Medicine
Gutteridge, A., 685 in childhood, 307–​310
decapitated burials and, 397
Habitus, 246 inequality and, 731, 735
Hadrian, xxxi, 185 in urban vs. rural areas, 373–​734, 733f
880

880   Index

Health stress, 307 Identity

Heidegger, Martin, 485 age, 304, 314, 322
Henig, M., 105, 603, 621, 633, 635 appearance and, 327, 406, 417–​421
Herbs, 815–​817, 820–​821, 822, 824–​825 categories of, 248
Heritage Lottery Fund, 44, 52 cult practices and, 453–​458
Herodian, 344, 434 defined, 225
Herodotus, 542 diasporic, 214
HERs. See Historic Environment Records ethnic, 248, 257, 278, 458–​459
Heterarchy, 154 fluidity of, 233, 248, 277, 304
Heterosexual bias, 364, 365 in frontier zones, 225–​227
Higham, N. J., 590 individual expressions of, 237–​240, 239f
Higham Ferrers settlement, 644, 646–​648, inscriptions and, 295–​298
708, 709–​7 10f, 853, 854f material culture as indicator of,
Hill, J. D., 82, 154, 365, 373, 511, 512, 525, 689 206–​207, 226
Hillforts, 154, 156, 491, 690, 741–​742 of military community, 417, 448–​449,
Hingley, Richard, 3, 216, 224–​225, 240, 373, 451–​459
484, 644, 652, 721–​722, 729, 730, 745 mobility and, 245–​246
Historic Environment Records (HERs), 44, 50, Identity and Violence (Sen), 448
54, 579 Immigrants. See also Mobility
Hoare, Richard Colt, 11, 17 foods brought by, 821
Hodder, Ian, 32, 82, 487, 489, 851, 852 from Gaul and Germany, 274
Hodgson, N., 226, 227 identification of, 214–​217, 248–​250, 250f
Hod Hill site, 28 Imperial diasporas, 214–​215, 217
Holland, Sally, 310 Imperial economy, 138, 139
Holocaust memorials, 684 Imperialism
Holtorf, C., 683 materiality and, 481, 482, 489, 492–​493
Homogenization, superficial, 248 rural areas impacted by, 721, 732, 736
Homosexuality, 364–​365 An Imperial Possession (Mattingley), xxvii, 449
Honorius, xxxvi Inchtuthil site, 28, 184–​185, 566
Hope, Valerie M., 237, 285 Indigenous populations. See also
Horses, 798–​799 specific groups
Horticulture, 814–​817, 815f, 818f of frontier zones, 226
Hospitals, 563, 565–​566 land incorporated into Roman Empire, 117
House churches, 409, 668–​669 Romanization of, 4, 13–​14, 17
Housesteads fort Inequality
architecture of, 470–​471, 495 in rural areas, 732–​736, 736f, 737
church structures at, 669, 670 social stratification and, 856
medical evidence at, 564, 566 Infancy. See also Childhood and adolescence;
textual evidence at, 472 Life course
Hoxne hoard, 856–​857 burial practices in, 311–​312, 325, 369,
Hull, M. R., 28, 77 428, 715
Hunter, Fraser, 179, 226 dependency in, 306
Hunting, animals for, 799–​800 dietary deficiency diseases in, 308, 309
Hurst, Henry, 95, 784, 786 feeding practices in, 308–​309
Hutton, R., 687 personhood in, 312, 325–​326
Hybridization, 214, 276 Infanticide, 311–​312, 325
Hydrocephalus, 397 Ingold, Tim, 488

Inscriptions, 285–​302. See also Epitaphs;
Textual evidence
abbreviations in, 291
building dedications, 287, 292
categories of, 287
Cogidubnus inscription, 6–​7, 7f, 8
on coinage, 9, 101, 164, 575–​576
communication to audiences through,
290–​295
dating, 289
for deities, 108, 292, 621–​622
epigraphic analysis of, 204–​205, 247, 248–​
249, 253–​254, 253f
epigraphic habit and, 288–​290, 456, 622
from Hadrian's Wall, 8, 291
identity and, 295–​298
in Latin, 5, 8, 18, 106, 240, 290–​291
literacy and, 291
medical-​related, 562–​564
military, 14, 287, 288–​291, 295–​296, 628
Ogam inscriptions, 588–​590, 589f
on pottery, 101
prevalence and distribution of, 286–​287
statistical analysis on demography using,
292–​293
on stone, 286
on tombstones, 237–​240, 239f, 289,
296–​298, 439
on votive altars, 292, 294, 297, 298, 622
Institute of Archaeology (London), 25, 33, 483
Instrumentation, medical, 558–​560
Inventory of Romano-​British Coin Hoards
(Robertson), 834
Inveresk military complex, 186, 187f
Iron, 532, 542–​548, 544–​545f, 549
Iron Age
archaeological evidence for, 152–​153
burials, 151, 159–​161, 344, 345
Celtic languages in, 575
changing perceptions of, 153–​154
coinage from, 107, 152, 157–​159, 161–​162,
836–​838, 837f
conceptualizing societies in, 180–​182
hillforts in, 742
late. See Late Iron Age (LIA)
material culture in, 10, 26, 155, 159, 167, 689
metalworking in, 490
Index   881
political authority in, 153, 154, 162
pottery from, 128, 155, 159–​160
ritual practices in, 159–​161, 272
Roman, 180
rural landscape during, 49
in Scotland, 179–​182
settlement and economy of, 10, 26,
155–​157
social organization in, 153–​154
wealth and power in, 157–​162
Iron Age and Roman Coin Finds from Wales
Project, 834
Isotope analysis
archaeological impact of, 38
carbon, 212–​213
lead, 212–​213
in mobility studies, 204, 207–​214, 212f,
217–​218
nitrogen, 212–​213, 308
oxygen, 208, 209, 210–​212, 212f
strontium, 208, 209, 210–​211, 212f
of tooth enamel, 308, 309, 315
Itinerarium Curiosum (Stukeley), 6
Ivleva, Tatiana, 207, 245
Ivory Bangle Lady, 214, 216, 216f
Jackson, K. H., 576–​577, 578–​579, 584–​585, 586,
587–​588, 592n7
Jackson, R., 560
James, A., 306
James, Simon, 228, 231, 277, 451
Jenny, L. L., 309–​310, 313
Jobey, George, 30, 235
Johns, C., 621
Jones, Barri, 33
Jones, M., 813
Jones, R. F. J., 80, 342–​343
Jongman, William, 732, 735
Journal of Roman Military Equipment
Studies, 466
Journal of Roman Pottery Studies, 80
Joy, J., 161
Joyce, James Gerald, 14
Julius Caesar
on appearance of British, 410
expeditions to Britain, xxviii–​xxix, 162–​163,
510–​511, 575
882

882   Index

Julius Caesar (Cont.) gender patterns at, 313, 326–​327, 328, 330

Gallic War account by, 264, 268 infant burials at, 325
writings of, 4, 9, 128, 152, 154, 158 isotope analysis of, 211, 212f
Jupiter columns, 272 material culture at, 275, 276, 329, 342,
350, 417
Kakoschke, A., 204 spatial organization of, 433
Kamash, Zena, 681 Late Antique period, mobility in,
Kampen, N. B., 371 205–​206, 205f
Kennedy, D., 251 Late Iron Age (LIA)
Kenyon, Kathleen M., 27, 28, 75–​76 coinage in, 836–​838, 837f
Kiernan, P., 646 community identities in, 226–​227
Kincheloe, J. L., 225 continuity in, 126, 128–​129
King, Anthony, 273, 274, 646, 647, 690, 723, 792 cremation in, 151, 263–​264, 263f, 265f, 428,
King, Charles, 72 436–​437
King Harry Lane burial site, 165, 343, 411, 413, feasting in, 161
429, 436 funerary process in, 428, 430, 432–​437
Kingscote settlement, 602, 709–​7 10 models of interaction in, 166–​169
Kipling, Rudyard, 13 modes of interpretation for, 151–​152
Kirk, J., 77 political authority in, 153, 154
Knight, J. K., 673 pottery from, 128
Knook Castle, 11, 12f relationship with early Roman horizon, 117,
Kolb, M., 854 118, 126, 128
Küchler, Susanne, 683–​684 shrine structures in, 270–​271
Kyphosis, 381 transitions in, 125, 126, 127
Latinitas Britannica (Jackson), 584–​585, 588
Labour, division of, 371–​372, 374 Latinization, 574, 579, 591
Laissez-​faire economics, 722 Latin language
Landownership, 701, 745 adoption of nomenclature from, 583–​584
Landscapes. See also Rural areas; Urban areas Celtic–​Latin bilingualism, 577–​578, 579,
aerial photography of, 30, 54–​55, 700, 703, 584, 588, 591
715, 716f on coinage, 574–​576
chorography and, 5–​6 on curse tablets, 110, 585–​587, 586f
geophysical surveys of, 55, 55f, 37–​38, 700, in elite contexts, 584, 588, 592n11
701–​703, 702f emergence in Britain, 575
memory and, 683 filiation markers in, 582–​583, 583f
reorganization of, 687, 688 inscriptions in, 5, 8, 18, 106, 240, 290–​291
research philosophy on, 32 levels of, 577
Language. See Sociolinguistics literacy in, 168, 576, 579
Language and History in Early Britain loanwords from, 577–​578, 584–​585,
(Jackson), 576–​576 587, 593n12
Lankhills cemetery nature of British Latin, 584–​588, 592n8
age-​related patterns at, 313, 322, 324, on Ogam stones, 588–​590
326–327, 327f, 328 in post-​Roman Britain, 591
developer-​funded excavations of, 49 on tombstones, 439
deviant burials in, 384, 428 Latour, Bruno, 485, 488, 499
diversity of remains at, 209 Laurence, R., 307, 767–​768
 Index   883

Laying-​out ceremonies, 429–​431 material culture in, 124–​125, 130

Lazzari, M., 233 public buildings at, 123, 124
Lead Lost-​wax casting technique, 539
in coffins, 309, 348, 430, 431 Love, J. R., 534
in diet, 308, 427 Lucas, G., 65, 71, 73–​74, 75
isotope analysis using, 213–​214 Lucy, S., 225
sites for extraction of, 535, 533f Lullingstone site
techniques for working with, 535, 537 burials at, 348
uses of, 532, 538, 540, 540f, 548 church structures at, 409, 668–​669
Leaf-​gilding technique, 534 sculptures from, 608
Leather products, 410 textual evidence from, 105, 106f
Lewis, M. E., 307–​308, 309 wall paintings at, 151, 409, 668–​669, 669f
Lewis, M. J. T., 29 Lysons, Samuel, 9
Lexden burial site, 160, 165, 352, 429
LIA. See Late Iron Age Macaulay, Thomas, 150, 151
Life course, 321–​340. See also Age MacDonald, George, 24
age identity and, 304 MacDonald, J. L., 384
childhood in. See Childhood and MacKinnon, M., 798
adolescence Mackreth, D. F., 412, 415, 420
future research needs, 334 MacMullen, R., 287, 622
gender in, 305 Magnentius, xxxiv, 663
infancy in. See Infancy Magnesian Limestone ridge, 49, 703
local and regional identities in, 331–​333 Magnetometry, 701, 702f
older adulthood in, 329–​331, 330t Magnus Maximus, xxxv, 135
overview, 321–​322 Make-​up, 409, 416
as social construct, 305, 306 Males. See Gender
in theory and practice, 322–​324 Maltby, Mark, 791
young adulthood in, 328–​329, 328t Mann, J., 592n5, 622
Lilley, I., 214 Mann, M., 233
Limes Congress, 56, 27 Manning, W., 79
Ling, R., 105 Manufacturing process for textiles, 407–​408
Linguistics. See Sociolinguistics Marchitello, Howard, 5–​6
Literacy. See also Inscriptions; Textual evidence Market economy, 138–​139, 608
defined, 102 Market gardening, 817, 823, 826
inscriptions and, 291 Marriage, age at, 305, 329
in Latin, 168, 576, 579 Martial, 7, 254, 410
in military community, 104–​105, 471–​472 Martin, M., 109
in Roman Britain, 102–​102, 104–​106, 166 Martin of Tours, 143
Liversidge, J., 76, 77, 78, 384 Martyrdom, 662
Livia, 408 Maryport site, 55, 294, 298, 454
Loanwords, 577–​578, 584–​585, 587, 593n12 Material culture. See also Artefact studies;
Locker, A., 801 Materiality
Lockyear, K., 84 at burial sites, 145, 236–​237, 349–​352, 350f,
Londinium 352f, 437–​438
in early Roman horizon, 119–​120, 121, Christianity and, 663–​665, 665f
122–​123 in decapitated burials, 393f, 394–​396
884

884   Index

Material culture (Cont.) bioarchaeology and, 567

of early Roman horizon, 124–​125, 130–​131 collyrium stamps and, 560–​562
ethnicity and, 136, 206 doctors and, 563, 564
food as, 822, 826 herbs in, 822
in frontier zones, 188–​194, 191f, 226 historical overview, 556–​557
gender identification through, 367–​371 hospitals and, 563, 565–​566
of Hadrian's Wall, 234–​236 instrumentation used in, 558–​560
homosexuality represented in, 364–​364 in military community, 562–​567
hybrid forms of, 188 Romanization of, 555–​556, 557
identity and, 206–​207, 226 Vindolanda tablets and inscriptions
in Iron Age, 10, 26, 155, 159, 167, 689 regarding, 562–​564
memory and, 682–​683 vision remedies, 561, 562
of military community, 451–​452 Memoria Romana project (Galinsky), 684
mobility and, 206–​207, 255–​258 Memory, 681–​696
multiculturalism and, 233–​234 archaeology and, 682–​684
in rural areas, 712, 714 collective, 681, 682, 691
selection processes for, 190 history of, 681–​682
transformations in, 491–​492, 493–​494, 496 landscapes and, 683
Materiality, 481–​519. See also Artefact studies; limitations of, 688, 691
Material culture material culture and, 682–​683
agency and, 484, 486–​487, 489, 497 overview, 681
in contemporary archaeological thought, prospective, 683, 688, 691
487–​489 retrospective, 683, 687, 688, 691
continuity and, 495 in Roman Britain, 685–​691
globalization and, 484, 489 in Roman society, 684–​685
historical approaches to, 482–​484 visibility of, 691
imperialism and, 481, 482, 489, 492–​493 Men. See Gender
interdisciplinary studies of, 485, 487–​488 Menarche, 305, 307
material horizons in Roman Britain, Merrifield, R., 82, 440
489–​496 Meskell, L., 233
theoretical approaches to, 485–​487, 488–​ Metabolic disease, 308
489, 499–​500 Metal detectors, 53–​54
Matthews, C. L., 384 Metals and metalworking, 532–​554
Matthews, K., 364–​365 blacksmithing, 548, 549
Mattingley, David, xxvii, 56, 102, 109, 228, 238, brass, 532, 538, 541
449, 579, 624, 628, 689, 722, 745, 749, 851 bronze, 493, 532, 538, 540, 839, 844
Mauss, Marcel, 485 copper and copper alloys, 532, 538–​542,
Maximian, xxxiii, xxxiv 540–​541f, 548–​549
Mays, S., 311–​312 iron, 532, 543–​548, 544–​545f, 549
McKinley, Jacqueline, 435 in Iron Age, 490
McManus, D., 585 lead. See Lead
McWhirr, Alan, 27 as motivation for Roman conquest of
Media coverage of Romano-​British Britain, 548
archaeology, 51–​52 overview, 532
Medicinal herbs, 822 pewter, 532, 542
Medicine, 555–​572 precious metals. See Gold; Silver
anthropological studies of, 555, 556 in Scotland, 181
 Index   885

tin, 532, 542, 548 epigraphy in study of, 204–​206, 205f, 248–​

in urban areas, 782, 783 249, 251–​255, 252–​253f
zinc, 532, 538, 540, 541 gender and, 205, 252, 254, 256
Middle ground theory, 168 identity and, 245–​247
Migration. See Mobility immigrants, identification of, 215–​217
Miles, David, 32 isotope analysis and, 204, 207–​214, 212f,
Military community, 448–​477 215–​217
armour and weaponry of, 452, 464–​468 in Late Antique period, 205–​206, 205f
burial practices within, 457 material culture and, 206–​207, 255–​258
Christianity in, 454 models for interaction, 213–​214
clothing of, 416–​417, 419–​420, 452 overview, 203
cult practices within, 453–​458 scientific techniques in study of, 207–​213
daily life in, 470–​471 Modification stage of funerary process,
epigraphic evidence for, 471 434–​435
epigraphic habit in, 95, 237 Modood, T., 225
extramural settlements associated with, Molleson, T. I., 308
450–​451, 470, 471, 472–​473 Mommsen, Theodor, 13, 482
on Hadrian's Wall, 227–​231, 229f, 230–​231t Money. See Coinage
identity of, 417, 448–​449, 451–​459 Mons Graupius, battle of (AD 83), xxxi, 184
inscriptions by, 14, 287, 288–​291, Montgomery, J., 212
295–​296, 628 Moore, Alison, xxvii, 312, 321, 322, 363, 372
literacy in, 104–​105, 471–​472 Moore, Tom, 28, 50, 154, 226–​227, 262
material culture of, 451–​452 Moorhead, Sam, 834, 836
medicine and hospitals within, 562–​567 Moreland, J., 104, 112
multiculturalism in, 227–​231, 229f, 230–​231t, Morris, J., 654
458–​459 Mortaria, 107, 470, 517, 520
overview, 448–​449, 464–​465 Mosaics
personal belongings of, 468–​470 Christian, 669–​670, 671
population estimates for, 464 distribution of, 600
in rural areas, 712 early interpretations of, 9
servants of, 450 figurative imagery on, 105
sociolinguistics within, 578–​579 function of, 610
veterans in, 459–​460 subject matter for, 141, 142, 351
women in, 473 textual evidence from, 105, 106f
Military districts, 14, 15f, 16, 17–​18 on villa flooring, 29, 141
Miller, D., 233, 488 Moulding techniques, 539
Millett, Martin, xxvii, 46, 22, 231, 343, 483–​484, Müldner, Gundula, 203
519, 699, 721, 722, 723, 724, 729, 730, 748, Mules, 799
755, 774 Mullen, Alex, 251, 573
Milne, J., 557 Multiculturalism. See also Ethnicity; Race
Mionnet, Théodore Edme, 64 in archaeological interpretations, 215–​216
Mithras, cult of, 456 definitions of, 225
Mnemonics, 682 Hadrian's Wall and, 216, 224–​225, 228–​231,
Mobility, 203–​223, 245–​258 229f, 230–​231t, 240–​241
archaeological evidence for, 204 materiality of, 233–​234
cultural change and, 274–​275 in military community, 227–​231, 229f, 230–​
diaspora theory and, 214–​215, 217 231t, 458–​459
886

886   Index

Nash-​Williams, V. E., 30, 77 Oriental cults, 779

National Planning Policy Framework (2012), Orthopraxy, 635
44, 51, 58 O'Shea, J. M., 314
National Roman Pottery Fabric Reference Osteoarthritis, 397
Collection, 82 Osteological analysis, 306–​307, 323, 324,
Neal, David, 37 345–​346, 427
Neolithic monuments, 686–​687 Osteological paradox, 732
Nero, xxx, 610, 758 Oswald, F., 74
Nesbitt, Claire, 224 Owles, E., 549
Networks, educational, 33, 34–​35t Oxford Archaeology, 45, 45f, 49, 33
New Archaeology movement, 32, 382, 383, 485 Oxford History of England, xxvii
Newborns. See Infancy Oxygen isotope analysis, 207, 208,
Newstead site 210–​212, 212f
artefact studies from, 69, 73, 80, 457
geophysical surveys of, 186 Pader, E. J., 385
ritual deposition at, 654, 655 Paganism, relationship with Christianity, 675
twentieth-​century excavations of, 23 Paintings. See Wall-​paintings
Niblett, R., 127, 429 Parker, A., 800
Nitrogen isotope analysis, 212–​213, 308 Parks, K., 809
Nora, Pierre, 682 Parsons, D. N., 592n4
Normanby hoard, 834, 841 Parting technique for goldworking, 535
North African populations PAS. See Portable Antiquities Scheme
diet, 821 Patron–​client relationships, 158–​159, 161,
at Hadrian's Wall, 234–​235, 459 607–​608
pottery from, 234, 459 Paul (apostle), 232
Nucleated settlements, 156, 266, 526, 708–​7 10, Peachin, M., 355
709–​7 10f Peacock, David, 32, 80, 511, 525
Numismatics. See Coinage Pearce, John, 325, 341, 368, 426–​427, 440
Nutrition. See Diet and nutrition Pearson, Joseph, 684
Perring, D., 514
Occupational communities, 451 Personal appearance. See also Clothing
Offerings. See Ritual deposition; Votive altars burial items related to, 328, 328t, 351
Ogam inscriptions, 588–​590, 589f cosmetics and, 411, 416
Older adulthood, 329–​331, 328t. See also Life course gender and, 369–​371
O'Neil, J., 76 hairstyles and, 410, 413, 419, 431
Oppida (settlements) identity and, 327, 406, 417–​19
in Britain vs. Gaul, 266 toilet instruments and, 166, 264, 265,
construction of, 759 369–370, 411, 512
early Roman horizon in, 118, 121, 122, Personhood, 312, 325–​326, 488
128, 129 Pertinax, xxxii
emergence of, 742 Petts, David, 660
in Iron Age, 154, 156, 490 Pewter, 532, 542
ownership of, 746 Pheasants, 801
pottery assemblages in, 514, 515 Philpott, Robert, 36, 80, 425, 434, 438
Romanization of, 512 Phosphoric iron, 543, 545
typology of, 156 Photography, aerial, 54–​55, 30, 700, 703,
villas as, 704 715, 716f

Picts tribe, xxxiv, 186, 195, 196
Piddington excavations, 46, 46f
Piece-​moulding techniques, 539
Pigs, 797–​798
Pine cones, 824, 824f
Pitt Rivers, A. H. L. F., 17, 65, 71, 73
Pitts, Martin, 161, 514, 720, 735
Planning Policy Guidance 16: Archaeology and
Planning (PPG 16), 43–​44, 45, 49, 50, 484
Planning Policy Statement 5: Planning and
the Historic Environment (PPS 5), 44,
50–​51, 52
Plaster burials, 347, 430
Plautius, 119
Pliny, 65, 74, 561
Plouviez, J., 415
Politics and religion, 627
Polluter pays principle, 44
Polybius, 158
Portable Antiquities Scheme (PAS)
coinage recorded by, 834, 835f, 836, 837
databases of, 54, 56, 82, 188, 451
metalwork findings recorded by, 54, 38,
82, 152
styli distributions recorded by, 579, 580f
Postcolonial theory, 214, 216, 484, 485, 487, 500
Post-​processualism, 84, 383, 384, 385, 483
Posture of decapitated burials, 395f, 396–​397
Pottery, 510–​531
artefact studies of, 64, 71, 73–​74, 75,
77, 78–​79
as art object, 600
in burials, 349, 350f
characterization and sourcing of, 32
chronological studies of, 27–​28
civilian traditions, 518–​521
in Claudio-​Neronian periods, 510–​513
for cremation containment, 159–​160
development of Romano-​British traditions,
513–​521, 525
as economic evidence, 77, 139–​140
Frisian, 235–​236, 459, 470
Gaulish, 73, 139, 140, 272, 470, 511
graffiti scratches on, 108, 592n6
grog-​tempered, 513, 514, 515, 516
inscriptions on, 101
Iron Age, 128, 155, 159–​160
Index   887
late Roman period, 522–​525
marketing patterns and, 851
North African, 234, 459
overview, 510
regional types, 521–​522, 525–​527
Romanization of, 16, 17
stamps on, 101, 105–​106, 107
transformations in, 492, 494
in urban vs. rural areas, 514–​515, 518, 723–​
724, 725–​731, 727f, 729f
wheel-​thrown, 159–​160, 511, 515
The Pottery Kilns of Roman Britain (Swan), 47
Poundbury cemetery
age-​related patterns at, 326, 327f
child burials at, 308, 309, 323, 325
coffins in, 348
demarcation of burials at, 354
gender patterns at, 326, 330, 372
material culture at, 329, 370
spatial organization of, 433, 672
wall paintings at, 409
Powell, L. A., 308–​309
Power
economics and, 850, 856, 860, 862
wealth and, 157–​162, 856–​857
Powleslend, Dominic, 701
PPG 16. See Planning Policy Guidance
16: Archaeology and Planning
PPS 5. See Planning Policy Statement
5: Planning and the Historic Environment
Precious metals. See Gold; Silver
Pre-​interment rituals, 345–​347
Preparation stage of funerary process,
428–​431
Preservation by record, 44, 31
Price, J., 80
Priorities for the Preservation and Excavation
of Romano-​British Sites (Roman
Society), 47
Processualism, 84, 383, 385, 483, 486, 488–​489
Professionalization of archaeology, 44, 48
Prone burials, 384, 387, 388, 395f, 396–​397, 428
Propertius, 410
Prospective memories, 683, 688, 691
Prowse, T. L., 209
Ptolemy, 110, 180–​181, 226, 747
Puberty, 305
888
888   Index
Public engagement with archaeology,
51–​33, 53f
Puck of Pook's Hill (Kipling), 13
Purbeck marble plaque, 626
Puttock, S., 645
Querns, 155, 470, 712, 715
Rabbits, 800
RAC (Roman Archaeology Conference), 47, 57
Race. See also Ethnicity; Multiculturalism
controversy in identification of, 215–​216
mixed-​race individuals, 215, 240
Radiate coinage, 841–​843
Rahtz, Philip, 44
Ralph, S., 161
Reciprocity, in economic systems, 139
Redfern, R. C., 307, 308, 567
Reece, Richard, xxvii, 32, 33, 80, 85, 111, 276,
483, 836, 840–​841
Regina (Catuvellaunian freedwoman), 238–​
240, 239f, 298, 369
Religion. See also Christianity; Deities; Ritual
deposition; Temples
in art, 610, 613
in Britain, 11, 143, 387
clothing and, 418
cult practices, 453–​458, 778, 779
in forum, 772
in Gaul, 143
politics and, 627
social renegotiation through, 688
theories of transmission and adaptation, 619
worship spaces in urban areas, 772–​779,
775f, 777f
Remote-​sensing techniques, 55
Renfrew, Colin, 487
Rescue archaeology, 44, 47, 50, 26, 27, 31
Res Gestae (Augustus), 152
Retrospective memories, 683, 687, 688, 691
Revell, Louise, xxvii, 321–​322, 326, 328, 330,
332, 367, 371, 372, 373, 635, 767
Rice, 822
Richborough site
artefact studies from, 65, 71
baptismal font at, 670
church structures at, 668
coin assemblages from, 844
dating evidence at, 127
textual evidence from, 106
Richmond, Ian, 28, 79, 483
Rickets, 308, 309, 397
Rigby, V., 511
Rigor mortis, 429
Rites of passage, 305, 322, 326
Ritual deposition, 641–​659
of animals, 646, 650, 651, 653–​654
of coinage, 161–​162, 644–​645, 647, 648, 839
offering zones for, 646–​649
overview, 641–​643
in settlement contexts, 651–​654, 653
special deposits, 642–​643, 650, 653–​654, 653
structured, 649–​651
in temple settings, 642, 643–​644, 649–​651
votive offerings within cult sites,
644–​646, 645t
in water, 160, 642, 671–​672
Rivet, A. L. F., 49, 31, 723
Roach-​Smith, Charles, 64–​65, 73
Robinson, H. Russell, 467
Robinson, M., 813
Rogers, Adam, 741
Roman Archaeology Conference (RAC), 47, 57
Roman Britain
animals in. See Animals
archaeological study of. See Romano-​
British archaeology
artefact studies of. See Artefact studies
art in. See Art
building types in, 142, 144
burials in. See Burials
Celtic subaltern image of, 4–​5, 8–​10, 13, 18
childhood in. See Childhood and
adolescence
civil and military districts of, 14, 15f,
16, 17–​18
civilizing influence of Romans in, 4, 5–​7,
8, 10–​11
clothing in. See Clothing
coinage in. See Coinage
continental influences in, 268–​274,
269f, 271f
contrasting views of, 10–​11, 13
defensive systems in, 135–​136

diet and foodways in, 272–​274
division into provinces, xxxii
early studies of, 3–​10
economic systems of. See Economics
elite auto-​representation in, 142, 143, 144
emergence of, 489–​492. See also Early
Roman horizon
end of, 858–​860
farming in. See Farming
fragmentation of, 494–​496, 498
framework for study of, xxvii, xxviii, 25
gender in. See Gender
Germanic peoples in, 144
independence of, xxxv
interactions with Gaul and Germany, 262–​
272, 265f, 269f, 273–​278
language in. See Sociolinguistics
life course in. See Life course
material culture of. See Material culture;
Materiality
medicine in. See Medicine
memory in, 685–​691
metalworking in. See Metals and
metalworking
military in. See Military community
mobility in. See Mobility
political evolution of, xxix
population estimates, 700, 852–​853
religion in. See Religion
rural areas of. See Rural areas
textual evidence from. See Textual evidence
timeline of, xxviii–​xxxvi
urban areas of. See Urban areas
withdrawal of imperial presence from,
135–​138, 144
Roman Britain and the English Settlements
(Collingwood), 24, 25
Roman Crafts (Strong & Brown), 80
Roman Empire
expansion of, 151, 152, 153, 232
indigenous lands incorporated into, 117
literacy in, 102
multiculturalism of, 225, 247–​248
unification of populations across, 5, 213
The Roman Era in Britain (Ward), 73
Roman Finds Group, 81
Roman Grey Literature Project, 50
Index   889
The Roman Inscriptions of Britain
(Collingwood & Wright), 37, 95, 100–​101,
105–​107, 561–​562, 609
Roman Iron Age, 180
Romanization
age construction and, 324
citizenship and, 232–​233, 452
of coinage, 16, 163–​164, 510
conceptualizations of, 231–​232
emergence of, 482
inadequacy in explanation of cultural
transformations, 262
of indigenous populations, 4, 13–​14, 17
of medicine, 555–​556, 557
of pottery, 16, 17
of rural areas, 723
in urban areas, 14, 16, 492–​493, 747–​749
The Romanization of Britain (Millett), 483, 748
The Romanization of Roman Britain
(Haverfield), 4, 23, 482, 747–​748
Romano-​British archaeology, 43–​64
in 1940s–​1960s, 26–​30
in 1970s and 1980s, 30–​33, 36
in 1990s and beyond, 36–​38
approaches and techniques in, 54–​55, 55f, 38
developer-​funded, 48–​50
deviant burials in, 388–​389, 400
future challenges for, 57–​58
history and evolution of, 44–​45, 45f, 56–​57
in interwar years, 24–​26
metal detectors in, 53–​54
non-​publication of, 50–​51, 37
overview, 43
public engagement with, 51–​53, 53f
research frameworks, 46–​48
specialist groups in, 48
twentieth-​century foundations of, 22–​24
Romano-​Celtic temples
baptismal fonts at, 670
design of, 270–​272
distribution of, 271, 271f, 277
reuse of ancient monuments by, 689
ritual deposition at, 690
spatial configuration and patterning of, 691
in urban areas, 773–​774, 776–​778
The Roman Wall in Scotland (MacDonald), 23
Rooke, Hayman, 9
890

890   Index

Ross, Ann, 29 Sandwich Islands, 150

Roundhouses, 18, 180, 190, 267, 496–​497,
706, 759
Rudling, D., 652
Rummymede Trust, 216
Rural areas, 699–​740
burials in, 343, 353–​354, 431–​432,
735–​736, 736f
characteristics of, 700–​703, 701f
diet in, 734
dynamic understanding of, 710–​7 17
economic relationship with urban areas,
720–​722, 723–​725, 729–​731
farmstead diversity in, 706, 708
health conditions in, 733–​734, 733f
imperialism impacting, 719, 732, 736
inequality in, 732–​736, 736f, 737
in Iron Age, 49
material culture in, 712, 714
military community in, 712
nucleated settlements in, 708–​7 10, 709–​7 10f
overview, 699–​700
population estimates for, 700
pottery in, 514–​515, 518, 723–​724, 726–​731,
727f, 729f
regional variations in, 699–​700
Romanization of, 723
small worlds concept applied to, 854–​855
transformation of, 721, 722–​724, 737
of urban hinterlands, 725–​731, 727f,
728t, 729f
villas in, 703–​706, 705f, 707f, 723
Rural Settlement in Roman Britain (Hingley),
30, 484
Russell, P., 251, 592n3
St Albans site
cemetery at, 343
coinage from, 844
inscriptions at, 287
organization of excavations at, 27
pottery from, 730
wealth at, 160
St Germanus, xxxvi
St Joseph, J. K., 54, 30
Saller, R., 292
Salway, Peter, xxvii, 30, 387–​388, 465
Saxe, A., 382, 383–​384, 388
Saxon Shore site, 137–​138
Scarth, Henry Mengden, 11
Scheidel, W., 209
Schiffer, Michael, 486
Schrijver, P., 591
Schwartz, B., 683
Schweissing, M. M., 209
Scotland
coinage in, 839
diplomacy and interference in, 194–​195
frontier life in, 182, 185–​189, 187f, 194
historic preservation responsibilities in, 44
Iron Age in, 179–​182
long-​term impact of Roman influence on,
195–​196
material culture in, 188–​194, 191f, 226
metalworking in, 181
military community in, 469
military history of, 182–​186, 183f
research framework in, 47
Roman conquest of, xxxi, xxxiii, xxxiv,
181–​185, 183f
settlement patterns in, 180–​181
twentieth-​century excavations in, 23
Scott, Eleanor, 365, 366, 369, 483
Sculpture
destruction of, 683
distribution of, 600–​602, 601f, 614
funerary, 353
military-​related, 416, 606
producers of, 607–​609
quality classifications for, 604–​606, 605f
religious use of, 610
as status display medium, 141
subject matter for, 141, 606
textual evidence from, 292
uses and users of, 610–​611
in villas, 139, 142, 610–​611
Scurvy, 308
Sears, G., 767–​768
Selection stage of funerary process, 426–​428
Sen, Amartya, 448
Septic arthropathy, 397
Septimius Severus, xxxii–​xxxiii, 185, 216, 234,
344, 434, 743
 Index   891

Sewing kits, 468 Sims-​Williams, P., 592n10

Sex determination of skeletal remains, 306, Site reports, 65–​69, 66–​68t, 73–​74
327, 368 Skeleton Green site, 126
SFBs (sunken-​featured buildings), 144, 145 Small finds, origins of term, 74–​75
SGRP (Study Group for Roman Small worlds concept, 854–​855
Pottery), 48, 79 Smiles, S., 152
Sharples, Niall, 154, 689, 690, 759 Smith, Alex, 641
Shaw, B. D., 292, 305 Smith, C., 585, 592n7
Shay, T., 383, 385 Smith, David, 29, 31
Sheep, 650, 795–​796 Smith, K., 654
Sheepen settlement, 128 Snead, J., 854
Sherratt, Melanie, 363 Social age, 320, 324
Shiel, N., 592n5 Social mapping, 453
Shiptonthorpe settlement, 369, 712, 713, Social stratification, 128, 855–​856, 857, 862
713f, 715 Society of Antiquaries
Shoes. See Footwear founding of, 63
Silbury Hill site, 687 meetings of, 8
Silchester site research reports of, 74, 75
baptismal font at, 671 Silchester excavations by, 14, 23
bathing complex at, 781 subject catalogue, 64, 72
cattle at, 794 women in, 77
Christian symbols at, 663 Sociolinguistics, 573–​598. See also Latin
church structures at, 665, 667f language
construction of, 755–​756, 757f of Celtic languages, 575–​578, 580–​584, 581f
cupellation at, 537 evidentiary sources for, 574
foods at, 819, 824 Latinization and extent of bilingualism,
forum at, 768, 769, 770–​771, 771f, 772 576–​584, 588, 591
housing at, 784 loanwords and, 577–​578, 584–​585,
inscriptions at, 287 587, 593n12
pigs at, 797, 798 in military community, 578–​579
public buildings at, 785 nature of British Latin, 584–​588, 592n8
report from, 65 of Ogam stones, 588–​590, 589f
ritual deposition at, 652–​653, 655, 687 overview, 573–​574
sustained occupation of, 14 post-​Roman inscriptions and Germanic
temples at, 776, 777–​778 languages, 588–​591, 591f
theatre at, 780 in pre-​Roman Britain, 575–​576
twentieth-​century excavations of, 23 Soden Smith, Robert, 72
university excavations at, 46 Soldiers. See Military community
urban features of, 7, 266, 492 Spector, J., 363
Silver Spelt grains, 808, 809–​810, 811–​812f, 812–​813
coinage produced from, 137, 164, 193, 493, Springhead temple complex, 162, 325, 348,
537, 541, 541f 647, 651
hoards of, 664 Stair, John, 7
sites for extraction of, 535, 536f Stamps
techniques for working with, 535, 537, 538 collyrium stamps, 560–​562
uses of, 535, 537 on pottery, 101, 105–​106, 107
Simonides of Ceos, 682 Stanford University, 51
892

892   Index

Stanway burial site ritual deposition in, 642, 643–​644, 649–​651

continuity in, 128, 129 in urban areas, 773–​779, 775f, 777f
creation of, 344 Temples, Churches and Religion in Roman
evidence of literacy at, 166 Britain (Rodwell), 644
individualized nature of, 267, 268 Tertullian, 454, 661
laying-​out stage at, 429 Textile production, 407–​408. See also Clothing
material culture at, 350 Textual evidence, 95–​116. See also Inscriptions;
medical instrumentation discovered at, Literacy
559–​560 ancient uses and responses to writing,
pottery from, 730 102–108, 103f, 106f
warrior burial, 264 of Christianity in Britain, 661–​663, 674
wealth at, 160, 165 classification as artefacts, 104
Statistical analysis on demography, 292–​293 curse tablets. See Curse tablets
Status and burial, 309–​310, 314, 341–​345, evolution of study of, 112
354–​355 overview, 95–​96
Steinberg, S. R., 225 relationship with archaeological evidence,
Stewart, P., 621, 622 110–​112, 118
Stilicho, xxxv stylus tablets, 95, 98–​99, 98f
Stinking mayweed, 813 types of, 96–​101, 97–​99f, 101f
Stolpersteine plaques, 684, 686 of urban development, 746–​747
Stonehenge, 6, 8, 47 Thalassemia, 309
Strabo, 152, 163, 267, 542, 837–​838 Theatres, 773–​774, 776, 778, 779–​780, 782
Strathern, Marilyn, 488 Theodosian Code, 855
Strontium isotope analysis, 207, 208, Theodosius, xxxv, 660–​661
209–​210, 212f Theoretical Roman Archaeology Conference
Structural Marxism, 154 (TRAC)
Study Group for Roman Pottery artefact studies and, 81
(SGRP), 48, 79 diversity of approaches represented
Stukeley, William, 6–​7, 9, 11, 18 by, 852
Stylus tablets, 96, 98–​99, 98f establishment of, xxvii, xxviii, 483
Sumner, G., 419–​420 on gender, 363, 366
Sunken-​featured buildings (SFBs), 144, 145 on identity, 406
Superficial homogenization, 248 influence of, 35, 36–​37
Supine burials, 395f, 396, 437 internationalization of, 57
Surgical instrumentation, 558–​559 on materiality, 484, 488
Surgical Instruments in Greek and Roman Thomas, J. D., 96
Times (Milne), 557 Tilley, C., 488
Swan, Vivien, 47, 234, 235, 274, 459, 470, 724 Timby, J. R., 515
Swift, Ellen, 63, 207, 417, 420 Tin, 532, 542, 548
Symbols in Action (Hodder), 82 Titus, 184
Symmetrical archaeology, 487, 488 Togas, 409, 418
Togidubnus, 758
Tacitus, 4, 8, 9, 110, 111, 185, 224, 268, 459–​460, Toilet instruments, 166, 264, 267, 369–​370, 411, 512
548, 575, 747, 758, 839 Tollgate grave assemblages, 347, 349, 350f, 351
Taxation, 138, 227, 524, 721–​722, 855, 856 Tombstones. See also Burials
Taylor, Jeremy, 35, 37, 687, 700, 708, 717 expressions of identity through,
Technological determinism, 525 237–​240, 239f
Temples. See also Romano-​Celtic temples inscriptions on, 237–​240, 239f, 289, 296–​298, 439
 Index   893

Latin on, 439 diet in, 734

stock images for, 608 in early Roman horizon, 125–​126
on Via Appia route, 685 economic relationship with rural areas,
Tomlin, R. S. O., 99, 108, 580, 582, 635 720–​722, 723–​724, 729–​731
Tonnochy, A. B., 77 entertainment and spectacles in,
Tooth enamel 779–​780, 782
health stress and, 307 forum as central feature of, 768–​773,
isotope analysis from, 308, 309, 315 770–​771f
Toponyms, 592n4 health conditions in, 733–​734, 733f
Town and Country in Roman Britain hinterlands of, 725–​731, 727f, 728t, 729f
(Rivet), 29 housing in, 782, 783–​784
Towns. See Urban areas industrial activity in, 782–​783
The Towns of Roman Britain (Wacher), 748 legal hierarchy of, 742–​744
Toynbee, Jocelyn, 29, 76, 348, 602–​603 location of, 749–​752
TRAC. See Theoretical Roman Archaeology major settlements in, 744, 744f
Conference overview, 741–​742
Trade diasporas, 213, 216 political and administrative spaces in, 768–​
Trajan, 292, 293, 467 773, 770–​771f
Traprain Law site, 179, 192, 193, 664, 845 pottery in, 514–​515, 518, 723–​724, 726–​731,
Treggiari, S., 371 727f, 729f
Tube dresses, 409, 419 regional variations in, 784–​786
Tunics, 409, 416 Romanization in, 14, 16, 492–​493, 747–​749
Turner, E. G., 96 size of, 745, 784
Tyers, Paul, 37, 64, 85, 519 textual evidence of urban development,
Tylecote, R. F., 549 746–​747
worship spaces in, 773–​779, 775f, 777f
Uley temple
curse tablets at, 99f, 100, 108–​109, 572, Vale of Pickering, geophysical survey of, 55,
582, 628 55f, 701–​703, 702f
goats at, 795 Vale of York, 515, 711
images at, 607, 613 Vance, Norman, 4
infant burials at, 326 Van der Veen, Marijke, 807
ritual deposition at, 644, 649 Van Driel-​Murray, Carol, 235, 365–​366, 450
United Kingdom (UK). See also England; Verulamium site
Scotland; Wales bathing complex at, 781
archaeology in, 43–​44 continuity at, 127
higher education in, 32–​33 cult practice at, 778
University of Birmingham, 27, 46 in early Roman horizon, 121–​122
University of Cambridge, 51 forum at, 769, 772
University of Newcastle, 45, 51 industrial activity at, 165, 783
University of Reading, 46, 50, 203, 699 material culture in, 130–​131
University study of archaeology, 32–​33, 36 mortuary complex at, 341, 758
Urban areas, 741–​790 organization of excavations at, 25, 27
bathing in, 780–​782 pottery from, 513–​514, 519
burials in, 343, 353–​354, 431–​432, public buildings at, 123–​124, 755, 784–​785
735–​736, 736f sustained occupation of, 14
categorization of, 742–​746 temples at, 773–​774, 776, 778
construction of, 752–​760, 753–​754f, 757f, 767 theatre at, 780
894

894   Index

Vespasian, xxix, 184 research framework in, 47

Veterans, 459–​460 Roman conquest of, xxx, xxxi
Via Appia, tombstones lining route of, 685 Wallace, Lacey, 117, 784
Vicarious human sacrifice, 384 Wallerstein, Immanuel, 721
Victim diasporas, 213 Wall-​paintings
Victor (North African freedman), 235, 237–​238 Christian, 143, 661, 668–​669, 669f
Victoria County Histories, 23 clothing styles in, 409
Villas distribution of, 600
architecture of, 10, 10f, 141–​142, 268–​269, function of, 610
269f, 493, 857, 857f textual evidence from, 105
baptismal fonts at, 671 Walton, Philippa, 834, 842
churches in, 668–​669 Ward, John, 73
mosaic flooring in, 29, 141 Warner, Richard, 64
regional distribution of, 704, 858–​859, 859f Warton, Thomas, 3, 10–​11
in rural areas, 703–​705, 705f, 707f, 723 Water Newton hoard, 143, 663, 664, 675
sculpture in, 141, 142, 610–​611 Way, Albert, 72
Vindolanda site Wealth and power, 157–​162, 856–​857
churches at, 454 Weaponry, 465–​468
church structures at, 668 Webster, Graham, 27, 30, 77, 79
clothing at, 406–​407, 408 Webster, J., 558, 621, 633, 634
cult practices at, 455 Weekes, Jake, 423, 440
extramural settlements at, 450 Weiss-​Krejci, E., 388
medical-​related tablets and inscriptions at, Welch, K., 780
562–​564 Wells, P., 272
sociolinguistic evidence from, 574, 578 Welwyn burial site, 151, 164, 165, 264, 265f, 349
textual evidence from, 37, 37, 96–​98, 97f, Wheeler, Mortimer, xxvii, 52, 57, 25, 75
104–​105, 471–​472, 839 Wheeler, Tessa Verney, 25, 75
volunteers at, 51, 53, 53f Wheel-​thrown pottery, 159–​160, 511, 515
Vineyards, 817 White, Richard, 166, 725
Vision remedies, 561, 562 Whittaker, C. R., 627, 629
Vitamin C deficiency, 308 Wild, J. P., 409, 419
Vitamin D deficiency, 308, 309 Wild boar, 800
Viticulture, 817 Wilkes, John, 33
Volunteer archaeology groups, 51, 51, 27 William, John, 11
Votive altars Williams, Howard, 351, 683, 686, 689
distribution of, 622–​624, 623–​624f, 630, 631f Williams, J. H. C., 107, 166, 168, 574
female construction of, 632 Wilmott, T., 779
inscriptions on, 292, 294, 297, 298, 622, Wilson, David, 54
628–​629 Wilson, Peter, 43
offerings within cult sites, 644–​645, 645t Winbolt, S. E., 74
Vows pledged to deities, 626, 627 Winchester site
cattle at, 794, 795
Wacher, John, 44, 27, 31, 721, 741, 746, 748, 768 cupellation at, 537
Wales gold artefacts at, 161, 413
historic preservation responsibilities in, 44 isotope analysis of, 211, 212f
pottery in, 513, 514–​515, 522–​523, 525–​526 occupation of, 157
pre-​Norman inscriptions in, 592n10 pigs at, 797, 798
 Index   895

research framework for, 27 Wroxeter site

sheep at, 796, 797 bathing complex at, 781, 782
Windle, Bertram, 11, 13 cattle at, 795
Witcher, R., 225 coinage from, 844
Women. See also Gender construction of, 754
artefact studies by, 71, 77, 78f, 80–​81, 82–​83 cupellation at, 537
breastfeeding by, 308–​309 forum at, 771
gardening by, 826 hinterlands project, 725
hairstyles of, 410, 413, 419, 431 inscriptions at, 287
in military community, 473 pigs at, 797
mobility of, 205, 252, 254, 256 public buildings at, 785
social roles of, 329, 331 remote-​sensing techniques for, 55
in Society of Antiquaries, 77 temples at, 775, 777f
in textile production, 408 twentieth-​century excavations of, 23
visibility of, 413, 415 university excavations at, 46, 27
votive altars constructed by, 632 urban features of, 492
Woodfield, C., 524
Woolf, G. D., 107, 152, 155, 166, 231–​232, Yalden, D., 800
635, 785 Young adulthood, 328–​329, 328t. See also
World-​systems theory, 689, 721 Life course
Wright, R. P., 100
Wright, Thomas, 11 Zinc, 532, 538, 540, 540–​541f, 541
Writing on the Wall project (2006), 216 Zoll, Amy, 619
Written evidence. See Inscriptions; Textual Zooarchaeology, 791–​802. See also Animals
evidence Zosimus, 861
896
898
900
902
904
906
908