Romano-British Settlement and Cemeteries at Mucking: Excavations by Margaret and Tom Jones, 1965–1978

Acknowledgements

There are many people to whom we owe a debt of thanks in the preparation of this volume. First must, of course, come our funders. English Heritage (now Historic England), under the Aggregates Levy Sustainability Fund, supported both the first stage of the project and the costs of this publication, and we remain grateful to Kim Stabler, Kath Buxton and Barney Sloane for their interest and enthusiasm throughout. The second stage of the project was jointly funded by the British Museum, the Roman Research Trust and the Society of Antiquaries of London. At the British Museum, we would like to thank J.D. Hill, firstly for his initial approach to the CAU for involvement in the Mucking publication project, and secondly for his unceasing support and encouragement, alongside that of Jonathan Williams. John Pearce, as well as being an invaluable source of information on Roman cemeteries, assisted with our applications to the Roman Research Trust, and we would like to record our thanks to both. The Society of Antiquaries of London, who were serendipitously able to fund much of the analytical work of the project through the Jones Bequest, have also provided valuable support, and we are particularly indebted to David Gaimster, John Lewis and Martin Millett.

On starting this project, we were faced with a mass of archival data and artefactual material to make sense of. There are many people to acknowledge for their help with that. Firstly we must record our thanks to the British Museum/English Heritage post-excavation team: Ann Clark, John Etté and Chris Going, who made their personal archives available to us, and did much to help us understand the various phases of post-excavation research (i.e. who had done what, when and why). Members of the original MPX (Mucking Post-Excavation) team established by Margaret and Tom Jones also provided invaluable assistance. We would particularly like to thank Jonathon Catton, Ruth Leary (née Birss), Paul Barford and Helena Hamerow for illuminating discussions and helpful information. While not directly part of MPX, Mark Hassall, Warwick Rodwell, Isobel Thompson, Nigel Brown, Val Rigby, Leslie Webster, Ian Bailiff and Sarnia Butcher all helped with information relating to the earlier phases of work, as well as, of course, Sue Hirst, who helped with information on the Anglo-Saxon cemeteries at Mucking. Our utmost thanks here, however, must go to Rosemary Jefferies (now Arscott) for her assistance both with the preparation of this volume, but also in helping us to make sense of the archives. She has been unfailingly generous with her time and her detailed knowledge of Roman pottery, kiln studies and Roman archaeology in general.

Others have also assisted us in the collation of data. Within English Heritage, we would like to thank Claire Jones and Kirsty Stonell-Walker for providing copies of old AML reports relating to Mucking; Justine Bayley for tracking down old reports (and samples!) and Simon Crutchley for answering queries on aerial photographic coverage. Within the British Museum, our thanks go to Tony Spence and Cyril Sylvestre, for helping us track down data tapes, to Matthew Harvey for facilitating access to British Museum archives for the project, and particularly to Marta Flanelly and Elena Jones, who showed unceasing patience and humour during our periodic archival rummages.

Being able to retrieve the data from the MPX computerisation project was the real key to being able to produce the site interpretations presented here, and there are several people to thank. Primarily, our gratitude goes to Graeme Mowbray, Richard Richner and Debbie Williams of eMag Solutions Ltd, who managed to turn archive boxes of eight-inch floppy disks last saved in 1983 into readable data (thanks must also go to Mike Lucy for identifying them as data retrieval experts). Mark Tomsett wrote the VBA code that converted this data into usable Excel spreadsheets. Catherine Hardman and Tim Evans of the Archaeology Data Service were extremely helpful in offering advice on data format and storage.

We also had enormous support from our specialists. Wherever possible we tried to locate the authors of reports produced during MPX so that they could be asked to update them; even after decades of silence, and without exception, their unfailing enthusiasm for the project was extremely gratifying. Our thanks here go to Quita Mould, Donald Bailey, David Buckley, Hilary Major, Joanna Bird, Brenda Dickinson, Kay Hartley, David Williams, Martin Henig, Catherine Johns, Paul Barford, Marijke van der Veen and Graham Morgan. Where the original specialists were no longer available, other colleagues stepped in to help, and we are grateful to Colin Haselgrove, Richard Reece, Jennifer Price, Penelope Rogers and Krish Seetah.

We would like to particularly thank Richard Reece and Mike Fulford, who acted as our academic readers. Both they and other academic colleagues, including Mark Atkinson, Edward Biddulph, John Collis, James Gerrard, Lise Bender Jørgensen, Ruth Leary, Bill Manning, Maria Medlycott, Rob Perrin, Dom Perring, Bill Sillar, Alex Smith, Ellen Swift, Jeremy Taylor, Jess Tipper and Todd Whitelaw, variously provided information, due critique and suggestions. We are particularly grateful to Edward Biddulph and Maria Medlycott for providing information on their sites in advance of publication , and to Paul Bidwell for permission to use unpublished and published data relating to the incidence of SE grey wares on the northern frontier. A youthful digger at Mucking, Mike Pitts kindly allowed us to use his site photographs that feature in Chapter One.

Within the CAU, we remain indebted to our excellent graphics team. The conversion, collation and subsequent visualisation of such a large set of records demanded a robust Geographic Information System and we would like to thank Iain Forbes for the creation, implementation and operation of the Mucking GIS, without which the distribution plots would not have been possible. Andrew Hall managed the graphical programme, ably supported by Iain Forbes, Jane Matthews, Vicki Herring and Bryan Crossan. Nicole Taylor also proved invaluable for her work on the cemetery graphics, cemetery gazetteers and archival research. In addition, we would like to thank other CAU colleagues for their help and advice: Katie Anderson (Roman pottery) and Jo Appleby (for help with archival sources and input into the cemetery analysis).

Finally, these acknowledgments should really pay tribute to Margaret and Tom Jones. Without their sheer determination in excavating such a huge site and recording it to what really were high standards, and particularly Margaret’s astonishing foresight shown in the computerisation programme, this volume would not have seen the light of day, at least not in this much detail. It turns out that she was right: with sites of this scale, the computer really can help provide the answers, and all her data was structured in such a way to do just that.

Summary

Excavations at Mucking, Essex, between 1965 and 1978 revealed extensive evidence for Romano-British burial and settlement (as well as an extraordinary prehistoric sequence, and what remains the largest Anglo-Saxon settlement yet discovered in Britain). Retrieval of computer data recorded as part of the Mucking Post-Excavation project has enabled the site to be phased and interpreted as a multi-phase rural settlement, perhaps an estate centre, with five associated cemetery areas, some of which saw contemporaneous use, with different burial areas reserved for different groups within the settlement.

The settlement demonstrated clear continuity from the preceding Iron Age occupation (evidenced by pottery production, retention and re-use of enclosures and similarities in building styles, among other things). Dateable artefact types, such as coins and brooches, saw unbroken sequences through the first century AD, while the mass of pottery data strongly suggests an indigenous industry that developed new forms and fabrics, but operated very much within an existing technological and social framework. Burial in the first century AD also reflected later Iron Age practices.

The site saw a dramatic transformation in the later decades of the first century AD, primarily marked by the laying out of the Central Enclosure, with its internal structures and divisions. While its position and orientation clearly paid some respect to existing features, together with the establishment of the accompanying Southern Enclosure system, it represents a dramatic remodelling of the settlement, complete with organised water supply in the form of two large wells within it. This area may still, however, have retained a ritual function.

This phase also marked the start of large-scale pottery production at Mucking. Phase 1 saw the production of ledge-rimmed jars in shell-tempered fabrics (part of the Thames-side pottery industry of the later first to early second century AD); in Phase 2 (early to mid second to mid third century AD) production shifted to grey and black burnished wares produced in large kilns; this pottery evidently formed part of the exports to Hadrian’s Wall (perhaps along with agricultural produce from the site; brewing activity is also evident).

Internally, the site saw other evidently dramatic events: after the mid second century AD, the major wells of the Central Enclosure were abandoned (its main well being infilled with the remains of an apparently high status major household assemblage and associated burnt down building). This may coincide with the abandonment of the Central Enclosure for everything other than continued burial in its associated cemetery, and attests to a major act of enclosure-ground clearance. The pottery production on the site seems to continue, however, with four kilns (Nos. 2–5) still in use at the end of the second century, and probably into the mid third century. By this point, settlement activity appears to have shifted its focus more to the Southern Enclosure system.

Settlement activity at Mucking after the mid third century AD is more opaque, and very few cut features could be assigned to this phase (Phase 3), although continued burial, pottery and artefactual deposition indicate that a form of settlement continued, possibly with some low-level pottery production in one of the kilns. Towards the end of the fourth century, or perhaps in the very early decades of the fifth, some of the latest Roman pottery was strongly associated with the earliest Anglo-Saxon style pottery and other material at the site, suggesting contemporary and equivalent use. It implies the existence of a terminal Roman settlement phase (Phase 4) that essentially involved an ‘Anglo-Saxon’ community. Given recent revisions of the chronology for the early Anglo-Saxon period, this casts an intriguing light on the transition, with radical implications for understandings of this period.

Each of the cemetery areas was in use for a considerable length of time: Cemetery I from the mid first until the fourth century; Cemetery II from the later first to the mid third century (with even earlier Late Iron Age burials perhaps acting as a focus). Cemetery III was possibly more short-lived, with cremation burial taking place from the later first to the late second century, but its inhumations may extend this span. Cemetery IV saw cremation burial take place from the early second to the mid third century, but also saw the site’s last Roman burial, in a stone coffin and dating to the later fourth century (and possibly relating to the site’s ‘transitional’ Romano-Saxon usage). Cemetery V, in contrast, seems to have been fairly short-lived, bridging the Conquest Period.

Taken as a whole, Mucking was very much a componented place/complex; it was its respective parts that fostered its many cemeteries, whose diverse rites reflect the variability and roles of the settlement’s evidently varied inhabitants.

Résumé

Des fouilles à Mucking, Essex, entre 1965 et 1978 ont révélé de considérables témoignages d’inhumation et d’occupation romano-britannique (ainsi qu’une extraordinaire séquence préhistorique, et ce qu’il reste de la plus grande occupation anglo-saxonne jamais découverte en Grande-Bretagne). Le recouvrement de données numériques enregistrées dans le cadre du projet de Suivi de Fouilles de Mucking a permis de mettre en phases le site et l’a interprété comme une occupation rurale à plusieurs phases , peut-être le centre d’un domaine, avec cinq zones d’inhumation associées dont certaines furent contemporaines, avec différentes zones d’inhumation réservées à différents groupes à l’intérieur de l’occupation.

Le campement attestait d’une évidente continuité avec l’occupation de l’âge du fer qui l’a précédé (ce qui est démontré par la production de poterie, la rétention et la réutilisation des enclos, et des similarités dans le style des constructions, entre autres). Certains types d’objets façonnés datables, tels que des monnaies et des broches, ont connu des séquences sans interruption pendant le premier siècle ap. J.-C., tandis que l’abondance de données relatives à la poterie indique fermement une industrie indigène qui a développé de nouvelles formes et matières , mais opérait bien à l’intérieur d’un cadre technologique et social existant. Les inhumations du premier siècle ap. J.-C. reflètaient aussi des pratiques de la fin de l’âge du fer.

Le site a été témoin d’une transformation dramatique dans les dernières décennies du premier siècle ap.J.-C., essentiellement marquée par l’aménagement de l’Enclos Central, avec ses structures et divisions internes. Tandis que sa position et son orientation rendaient de toute évidence un certain hommage aux éléments existants, associé à l’établissement du système d’Enclos Sud qui l’accompagnait, il représente un remodelage dramatique de l‘occupation, le tout avec une alimentation en eau organisée sous la forme de deux grands puits à l’intérieur. Il se peut que cette zone ait, toutefois, conservé encore une fonction rituelle.

Cette phase marqua aussi le début d’une production de poterie sur une grande échelle à Mucking. La Phase 1 vit la production de cruches à lèvres en bourrelet en matériau dégraissé à la coquille (partie de l‘industrie céramique du bord de la Tamise de la fin du premier au début du deuxième siècle ap.J.-C.), dans la Phase 2 (du début-milieu du deuxième au milieu du troisième siècle ap. J.-C.) la production a évolué vers de la vaisselle brunie grise ou noire produite dans de grands fours ; cette poterie faisait de toute évidence partie des exportations vers le Mur d’Hadrien (peutêtre à côté de produits agricoles du site, une activité de brassage est également en évidence).

Intérieurement, le site a vu d’autres événements tout aussi dramatiques après le milieu du deuxième siècle ap. J.-C., les importants puits de l’Enclos Central furent abandonnés (son puits principal fut rempli des débris d’un majeur assemblage domestique apparemment de haut rang et d’un bâtiment associé incendié) Cela pourrait coïncider avec le total abandon de l’Enclos Central hormis pour la continuation des inhumations dans son cimetière associé et atteste d’une opération majeure de défrichage d’un terrain d’enclos.La production de poterie sur le site semble cependant avoir continué, quatre des fours (Nos 2–5) étant encore en usage à la fin du deuxième, et probablement jusqu’au milieu du troisième siècle. A ce moment-là le centre de l’activité du site semble s’être déplacé plus vers le système d’Enclos Sud.

Après le milieu du troisième siècle ap.J.-C. l’activité de l’occupation de Mucking devient plus opaque , et très peu de traits de coupure ont pu être attribués à cette phase (Phase 3), bien que la poursuite des inhumations, de la céramique et de la déposition d’objets façonnés indique qu’une forme d’occupation avait perduré, peut-être avec une production de poterie mineure dans un des fours. Vers la fin du quatrième siècle, ou peut-être dans les toutes premières décennies du cinquième, certaine de la poterie romaine la plus récente était étroitement associée au plus ancien style de poterie anglo-saxonnne et à d’autre matériel du site, ce qui indique une utilisation contemporaine et équivalente. Cela implique l’existence d’une phase d’occupation romaine finale (Phase 4) qui comportait essentiellement une communauté anglo-saxonne. Compte tenu des récentes révisions de la chronologie du début de la période anglo-saxonne, ceci projette une lumière intriguante sur la transition, avec des implications radicales pour la compréhension de cette période.

Chacune des aires de cimetière fut utilisée pendant une considérable période de temps, Cimetière 1 du milieu du premier au quatrième siècle; Cimetière II de la fin du premier jusqu’au milieu du troisième siècle (avec même de plus anciennes inhumations de l’âge du fer final agissant peut-être comme point focal). Cimetière III fut peut-être de plus courte durée, des inhumations à incinération y ayant lieu de la fin du premier à la fin du deuxième siècle, mais il se peut que ses inhumations aient dépassé ce laps de temps. Cimetière IV fut témoin du déroulement d’inhumations à incinération à partir du début du deuxième jusqu’au milieu du troisième siècle , mais il vit aussi la dernière inhumation romaine du site, dans un cercueil de pierre et datant de la seconde partie du quatrième siècle (et ayant peut-être un lien avec l’utilisation romano-saxonne `transitionnelle’du site). Cimetière V, au contraire, semble avoir été d’assez courte durée, faisant le pont avec la Période de la Conquête.

Pris dans son ensemble, Mucking fut réellement un lieu/un complexe à multiples composants; ce furent ses parties respectives qui entretinrent ses nombreux cimetières dont les divers rites reflètent la variété et les rôles évidemment divers des habitants de l’occupation.

Zusammenfassung

Während der zwischen 1965 und 1978 durchgeführten Ausgrabungen bei Mucking in der Grafschaft Essex wurden umfangreiche Hinweise auf romano-britische Bestattungen und Besiedlung freigelegt (abgesehen von einer außergewöhnlichen Abfolge prähistorischer Befunde und der bislang größten in Großbritannien entdeckten angelsächsischen Siedlung). Dank der Wiederherstellung von Computerdaten, die im Rahmen der Auswertung der Grabung Mucking aufgenommen wurden, konnte die zeitliche Abfolge der verschiedenen Phasen ermittelt und der Fundplatz als mehrphasige ländliche Siedlung – möglicherweise der Zentralort eines Landguts – interpretiert werden. Dazu gehörten fünf Gräberfeldbereiche, einige davon gleichzeitig benutzt, mit verschiedenen Bestattungsbereichen, die unterschiedlichen Gruppen innerhalb der Siedlung vorbehalten waren.

Die Siedlung wies eine klare Kontinuität zur vorangehenden Besiedlung der vorrömischen Eisenzeit auf (was sich u.a. anhand der Keramikproduktion, der Beibehaltung und erneuten Nutzung von Einfriedungen und Übereinstimmungen von baulichen Konstruktionsmerkmalen zeigen lässt). Datierbare Fundgruppen, wie Münzen oder Fibeln, wiesen ununterbrochene Entwicklungsabfolgen über den gesamten Zeitraum des 1. Jahrhunderts n. Chr. auf, während die Unmenge an Daten zur Gefäßkeramik deutliche Hinweise auf eine lokale Töpfereiindustrie lieferten, die zwar neue Formen und Warenarten entwickelte aber eindeutig innerhalb eines bereits existierenden industriellen wie sozialen Netzwerkes arbeitete. Ebenso reflektierten Bestattungen des 1. Jahrhunderts n. Chr. Gebräuche der späten vorrömischen Eisenzeit.

In den späteren Jahrzehnten des 1. Jahrhunderts n. Chr. erfuhr der Fundplatz eine dramatische Transformation, die sich vor allem in der Anlage der zentralen Einfriedung mit ihren internen Strukturen und Unterteilungen niederschlägt. Während ihre Lage und Orientierung offensichtlich Bezug auf bereits existierende Strukturen nahm, ist sie dennoch, zusammen mit der Anlage des südlichen Einfriedungskomplexes, Ausdruck einer dramatischen Neugestaltung der Siedlung, die auch eine organisierte Wasserversorgung in Form von zwei großen Brunnen in ihrem Innenbereich umfasste. Dennoch scheint dieser Bereich weiterhin eine rituelle Funktion beibehalten zu haben.

Diese Phase markiert auch den Beginn großmaßstäbiger Keramikproduktion in Mucking. In Phase 1 fällt die Produktion von Krügen mit Absatzrändern (ledge-rimmed jars) in muschelgrusgemagerten Warenarten (Teil der Themse-Ufer Keramikindustrie des späten 1. und frühen 2. Jahrhunderts n. Chr.); in Phase 2 (frühes bis mittleres 2. bis mittleres 3. Jahrhundert n. Chr.) verlagerte sich die Produktion auf grau und schwarz geglättete Waren, die in großen Öfen produziert wurden. Die Töpferwaren bildeten offensichtlich Teil der Exporte an den Hadrianswall (vielleicht zusammen mit landwirtschaftlichen Erzeugnissen des Fundplatzes; auch Brauerei lässt sich nachweisen).

Im Inneren der Siedlung ereigneten sich andere, offenbar dramatische Ereignisse: nach der Mitte des 2. Jahrhunderts n. Chr. wurden die beiden größten Brunnen der der zentralen Einfriedung aufgegeben (der Hauptbrunnen wurde mit den Überresten eines bedeutenden, offensichtlich sozial hoch stehenden Haushalts samt zugehörigem abgebrannten Gebäude verfüllt). Dies erfolgte möglicherweise zeitgleich mit dem Ende jeglicher Nutzung der zentralen Einfriedung, abgesehen von dem weiterhin für Bestattungen genutzten zugehörigen Gräberfeld, und bezeugt eine ausgedehnte Räumung ihrer Innenfläche. Die Keramikproduktion scheint jedoch weiterzulaufen, denn vier Töpferöfen (Nr. 2–5) sind Ende des 2. Jahrhunderts, und wahrscheinlich bis ins mittlere 3. Jahrhundert, weiterhin in Benutzung. Bis zu diesem Zeitpunkt scheint sich der Siedlungsschwerpunkt mehr in Richtung des südlichen Einfriedungssystems verschoben zu haben.

Nach der Mitte des 3. Jahrhunderts n. Chr. ist die Besiedlung in Mucking schwerer zu fassen, so lassen sich nur wenige eingetiefte Befunde dieser Phase (Phase 3) zuweisen; dennoch deuten Bestattungen, Töpferei und Niederlegungen von Gegenständen ein wie auch immer geartetes Fortlaufen der Siedlungsaktivität an, möglicherweise unter Einschluss kleinmaßstäbiger Keramikproduktion in einem der Töpferöfen. Gegen Ende des 4. Jahrhunderts, oder vielleicht in den allerersten Jahrzehnten des 5., ist ein Teil der spätesten römischen Keramik deutlich mit der frühsten angelsächsischen Keramik und anderem Material in der Siedlung vergesellschaftet, was gleichzeitige und vergleichbare Nutzung andeutet. Dies impliziert die Anwesenheit einer angelsächsische Gemeinschaft vor Ort während der letzten römischen Siedlungsphase (Phase 4). Angesichts der jüngsten Revisionen der Chronologie der frühen angelsächsischen Periode beleuchtet diese Tatsache einen faszinierenden Aspekt der Übergangsphase, mit radikalen Implikationen zum Verständnis dieser Periode.

Alle Gräberfeldbereiche wurden über verhältnismäßig lange Zeiträume genutzt: Gräberfeld I vom mittleren 1. bis ins 4. Jahrhundert; Gräberfeld II vom späteren 1. bis ins mittlere 3. Jahrhundert (wobei hier sogar wesentlich ältere Bestattungen der späten vorrömischen Eisenzeit als Fokus gedient haben können). Mit Brandbestattungen vom späteren 1. bis ins späte 2. Jahrhundert hatte Gräberfeld III möglicherweise eine kürzere Belegungsdauer, aber die hier ebenfalls niedergelegten Körperbestattungen gehen vielleicht über diesen Zeitraum hinaus. Gräberfeld IV wurde vom frühen 2. bis ins mittlere 3. Jahrhundert für Brandbestattungen genutzt, aber dort fand auch die späteste römische Bestattung des Fundplatzes statt, in einem Steinsarg und in das spätere 4. Jahrhundert datierend (und möglicherweise in Verbindung mit der Nutzung während der römischangelsächsischen Übergangsphase). Gräberfeld V scheint hingegen relativ kurzlebig gewesen und nur im Zeitraum um die römische Eroberung belegt worden zu sein.

Zusammengenommen stellt sich Mucking deutlich als ein aus seinen Komponenten zusammengesetzter Platz/Komplex dar; es waren diese verschiedenen Teile, die die zahlreichen Gräberfelder unterhielten, deren unterschiedliche Riten als Niederschlag der Variabilität und der Rollen der offensichtlich verschiedenartigen Bewohner der Siedlung zu werten sind.

Übersetzung: Jörn Schuster

(ARCHÆOLOGICALsmallFINDS)

Chapter 1

Roman Mucking

Many Things

This volume presents the results of excavations at Mucking, Essex (Figs. 1.1–2) relating to the site’s Romano-British occupation, which took place between 1965 and 1978. While a ‘rescue’ excavation occurring in advance of (and often during) quarrying operations, it also aimed at near-total excavation of the archaeological features. The site was directed by Margaret Jones, but her Assistant Director (and site photographer, and husband) – W.T. (Tom) Jones – took the main responsibility for directing the cemetery excavations, co-ordinating much of the initial post-excavation research on them.

This volume is a companion to that which deals with the site’s prehistoric sequence (Evans et al. 2016). That also deals with both the logistical/operational background of the excavations and their conceptual context. This ranges from the very notion of ‘total archaeology’ and the predominant ‘dynamic’ culturehistorical framework of the site’s interpretation, to the problems posed by its notebook-based record and the ‘type’-based approaches applied to its postexcavation. It is, however, the intention that this book can also serve as a stand-alone account for Romanspecialist readers. Whilst therefore risking a degree of repetition, some of these issues must also be touched upon here, with all being more fully explored within its companion volume.

This aspired to be a much more straightforward and far less ‘problematised’ book than the Prehistory Volume, as it was thought that there was less need for contextualisation. This is largely due to the fact that there was less ambiguity regarding feature attributions and, as outlined below, there was an existing draft manuscript covering elements of the period’s archaeology. Accordingly, this book’s draft text was completed first, which was entirely in keeping with the site’s ‘archaeology backwards’ programme – proceeding from most recent to earlier (and most readily to least ‘knowable’). That said, such ‘time-slicing’ approaches invariably create problems at their temporal edges and sequences are rarely neat. This was certainly the case here, for having worked through the site’s prehistory – particularly the extraordinary character of its later/Late Iron Age – considerable revision of Mucking’s Conquest Period/ Early Roman landscape was then needed. Coming to terms with the site’s Late Roman usage also proved difficult. This in turn now demands a significant reappraisal of its earliest Anglo-Saxon occupation, and in all this there seems an underlying circularity (see Chap. 5 on spiralling towards knowledge).

While Mucking’s Roman sequence is relatively uncomplicated, its publication poses major challenges. Further detailed below, there is, on the one hand, the scale of its assemblages and vast finds numbers. On the other hand, there are its many parts. Aside from its three main settlement foci – each seemingly having quite different functions – there are its five cemeteries. The challenge here is how to allow for their meaningful articulation.

Backgrounding and Prehistoric Sequence

The decision to start the excavation came as a result of the aerial photograph published by J.K. St Joseph (1964; Fig. 1.3); this clearly depicted the South Rings (then considered to probably be a Neolithic henge monument). While the Ministry of Public Buildings and Works had intended to schedule the cropmark area, this was not done (Jones 1993, 6). The trial excavations in 1965 were, therefore, conducted in order to gain familiarity with the terrain before tackling the monument itself (the Rings being excavated in 1966); what, of course, was revealed was an incredibly dense settlement palimpsest, of which the unexpected Romano-British Cemetery I formed a part. The aim of the excavations then became the ‘rescue of a multi-period landscape’ (ibid.).

Figure 1.1. General Roman landscape location map (top) and, below, excavation area (red) shown in relationship to its environs cropmarks.

Figure 1.2. MPX-phase rendering of all features (top) and, below, site contours.

The Romano-British settlement at Mucking exhibits direct continuity from the later Iron Age, particularly in terms of continuing use and development of the Western Enclosure-area for industrial and craft-related activity, especially pottery production (this area was termed the ‘Banjo’ enclosures during MPX; see below). While a new kiln technology was introduced around the time of the Conquest, the pottery forms and fabrics display a continuous sequence throughout this transition. What is also new, probably in the immediate post-Conquest Period, is the laying out of a more formal rectilinear enclosure system to the east of the Western Enclosures, comprising a ditched central settlement area (the Central Enclosure),¹ with a partitioned enclosure system to its south (the Southern Enclosures). In the second century AD, these more ‘formal’ parts of the settlement continued to be occupied and developed, but the western enclosure area went out of use. Both of these phases saw contemporary burial, within distinct cemetery areas. In the third and fourth centuries, the picture is more obscure, with much of the distinctly later fourth-century material culture found in features and deposits strongly associated with the Anglo-Saxon phase of occupation. Very few cut features can be assigned to the later Roman period; some burials can be seen to date to the third century, particularly the earlier part, with only a handful, at most, dating to the fourth century, but all indications are that the excavated area saw a decline in settlement intensity after c. AD 250/60. There are obviously significant dating issues that need to be addressed, and these will be fully dealt with below.

Figure 1.3. Aerial photographs; top, looking east with 1100 Enclosure and Belgic Banjo in middle ground and South Rings, ABC and RBI Enclosures upper right; bottom left, vertical shot, with South Rings, RBI and ABC Enclosures (note that comparing the structural/enclosure detail visible in this image with Fig. 1.13 attests to just how much was lost through ‘early days’ quarry machining); bottom right, ABC Enclosures as stripped and immediately prior to excavation.

Romano-British cemeteries were located during the first season of excavation, and in the very last formal excavation (Cemetery IV, excavated in 1978 in advance of reinstatement of farmland at the quarry edges). The final burial to be excavated was the site’s one stone coffin burial (Grave 1086), discovered late in 1978 during quarry roadworks and after the formal end of excavations. (An emergency excavation was hastily conducted after the find was reported by the drag-line operator.) Excavation of the Romano-British burials at Mucking thereby frames the entire history of the excavations, with Grave 1 and Grave 1086 both published in this volume. Burials were also found within the settlement areas themselves, and were recorded using the same methods (see below and Prehistory Vol.).

The Joneses essentially viewed Roman Mucking as the infield and outfield of a Roman villa settlement, and they thus interpreted the cemeteries in this light. No villa building was ever found and, therefore, it was assumed that the main area of … Romano-British and Anglo-Saxon settlement must have lain on more fertile land further down the terrace slopes (Jones 1993, 8). They eventually thought that the villa proper lay in the area of Walton Hall some 400m southeast of the site.² It would, though, have to be said that the interpretation of Mucking’s Roman landscape was relatively little developed, at least in comparison to its Anglo-Saxon settlement (the same arguably being true of much of its prehistory; see below and Prehistory Vol.). Apart from highlighting its corn-drier, and the Central Enclosure’s granary and main well in various Panorama notices (aka Journal of Thurrock Local History Society: e.g. 1974, 1979 & 1985), it was the associated pottery production that featured within interim reports (Jones 1972; Jones & Rodwell 1973; see also Jones 1979).

Figure 1.4 is intended to convey something of the flavour of the Roman settlement’s interpretation. Roger Massey-Ryan’s 2004 reconstruction painting, ‘Roman Villa at Mucking’ is telling. Commissioned by Essex County Council, liberties have certainly been taken with the excavators’ interpretations, so that the postulated villa has now migrated up from the terrace-slope to its crown. This was achieved by merging the two phases of the southern aisled barn building, thereby allowing its separate post-lines to become a colonnade. With no industrial activity depicted and by having its field/horticultural plots emphasised, the whole thing has been turned into a well-manicured farm-estate. Through the careful deployment of wood-scrub behind, the ambiguous status of the Western Enclosures/Belgic Banjo-area compounds at that time has also been masked.

A gem of Mucking’s archives is Figure 1.5’s uppermost illustration, as it effectively amounts to a ‘cognitive map’ of the site’s main enclosure components (and must relate to ideas contemporary with the fieldwork). Indicating that the otherwise omitted prehistoric roundhouses and Anglo-Saxon sunken huts dotted the terrace, the North and South Rings are clearly entitled ‘Late Bronze Age Hillforts’. The Iron Age North Enclosure is labelled as a ‘Native Earthwork’ in direct contrast to the Banjo complex’s ‘Belgic Settlement’ (i.e. introduced populace), and also relevant is RBI’s appellation as a ‘?Roman Military Earthwork’. In short, it tells of the project’s dynamic invasion-/migration-dominated framework. Add to this its distinctly militaristic emphasis, and the sketch plan expresses well the project’s (increasingly outdated) culture-historical paradigm (see, though, Chap. 5). (While the Double Ditch/Central Enclosure is shown as ‘Early RB farmstead’, the eastern terraceside is labelled ’Ditch enclosed fields of Romano-British villa’.)

The same figure also includes the MPX-issued plans of the site’s Iron Age and Roman phases (Fig. 1.5). Though the character of its Iron Age layout is discussed below, these early 1980s interpretations emphasise just how much of the site’s sequence was then understood. Admittedly, this needs to be balanced by the ‘Uncertain’ phase plan in the same series (see Prehistory Vol., fig. 1.19) and its many thousands of small features (most of which appear to have been prehistoric). Nevertheless, we estimate that the Joneses had got 80–90% of Mucking’s sequence ‘right’, with the only serious omission – and which amounts to the main ‘discovery’ of our programme – being the Late Iron Age’s The Plaza ceremonial ground (see Prehistory Vol. chap. 6 concerning issues of the site’s understanding vs. proof). Given this, it is only appropriate that at this point the site’s prehistory is reviewed.

Mucking’s earlier prehistory is relatively nuanced, when compared to many of the region’s recent huge-scale massstripping excavations, as its intense overall excavationsample generated substantive quantities of residual-status pottery and flintwork (there being, for example, 26,740 entries recorded for the latter). As shown in Figure 1.6 (and detailed in the Prehistory Vol.), this has allowed for the mapping of its successive earlier Neolithic, Grooved Ware and Beaker occupation clusters. Almost no distinctly Early Bronze Age pottery (i.e. Collared Urn) was recovered, but by this time the landscape must have been largely cleared as there followed the construction of its eight barrows. Dating to c. 1700–1400 cal. BC, and typical of the region, these were relatively small (c. 4.25–14.0m diam.) and predominantly inhumation-associated.

Figure 1.4. Roman Mucking ‘issues’: top, Roger Massey-Ryan’s 2004 reconstruction painting, ‘Roman Villa at Mucking’ (Essex County Council); centre, fieldwork issued postcard of kilns; below, interim publications.

The middle centuries of the second millennium BC were marked by a fieldsystem, with evidence of a broadly contemporary Deverel Rimbury Ware-associated settlement within the site’s northern quarter.

Up to this point the terrace’s remains could be considered as fairly typical of South Essex sequences generally, but it was during the Late Bronze Age that Mucking became ‘special’. This is demonstrated by its two ringwork monuments: the single-circuit North Ring – dug by the Central Unit some 230m beyond the main site (Bond 1988) – and the South Rings at the site’s southern end. The latter was double-circuited, with another much smaller concentric ring within its interior; both ringworks were settlement-related, and had evidence of metalworking and salt production/trade-control. Based on its ceramics, the South Rings appears to have hosted mass-gatherings in its final usage and – almost folly-like – it may well have been to this activity that its innermost ring related.

Figure 1.5. Entitled enclosure sketch plan, presumably intended for public display purposes (note labels, ‘Ditch enclosed fields of Romano-British Villa’ and ‘Native Earthwork vs. ‘Belgic Settlement’); below, MPX Iron Age and Roman phase-plans.

Figure 1.6. Prehistoric phase-plan ‘cartoons’.

Figure 1.7. Iron Age and Romano-British phase plans.

Evidence of broadly contemporary salt-related production, as well as a scattering of roundhouses and other settlement features (e.g. pits and four-posters) – plus a small cremation cemetery – had been distinguished within the site’s northern quarter. However, when faced with only the partial study of the site’s prehistoric pottery, both for that period and the subsequent Early Iron Age, the extensive scale of its settlement could only really be plotted by analysis of the MPX pottery drawings and the distribution of other diagnostic finds-types (saddle querns and pyramidal loomweights, etc.). While based on what pottery studies have been conducted a few of the designated roundhouses could be assigned to the latter period (including the very large Roundhouse 39), through close scrutiny of the site’s base-plans further post-built Late Bronze/Early Iron Age candidates were identified. This has taken the possible number of buildings proper (vs. ‘posters’) attributable to that time up to nearly 50 (see Prehistory Vol., fig. 4.24).

There can be more assurance of the total number of roundhouses assigned to the Middle/Late Iron Age: these gullied buildings number 86 (Fig. 1.7). While many were directly associated with the site’s main settlement compounds of the period – the so-named North and ABC Enclosures, RBI and the Belgic Banjo – more than half of the buildings were unenclosed and lay exterior to settlement compounds, with these tending to be the more robust gullies.

Figure 1.8. Quarry-face Excavation (I; Iron Age): left, on-going excavation of the Belgic Banjo complex (around whose circuit the intervening segment-baulks would later be removed); right, the northern square barrow range being dug hard by the conveyor belt.

Only the pottery from the North Enclosure compound was studied in detail, which indicated its probable establishment in the late fourth century BC and that it continued in use through to the mid/later first century BC. The site’s three other compounds were also of Middle Iron Age origin, but of these only the ABC Enclosures in the south also ceased to function prior to the Late Iron Age proper (although it survived as an upstanding earthwork and, as such, went on to frame one of the Romano-British cemeteries).

The sequences of both the RBI and Belgic Banjo are fraught with ambiguity. Given the state of the site’s pottery analyses, the distinction of its Late Iron Age wares proved difficult, and their eventual identification is charted through a series of case-studies within the Prehistory Volume. What proved most troublesome was the Belgic Banjo’s sequence. Having a most extraordinary accumulative plan (Fig. 1.8), the morphology of its parts would certainly not suit a Roman date, although the distribution of the site’s Early Roman pottery suggested this. Eventually it was realised that most of the Late Iron Age wheel-thrown wares had been subsumed within the Roman pottery series and that a number of the Early Roman kilns had been inserted into the upper profile of the earlier, Iron Age ditches. It was only by analysing the pottery series by feature-depth that a resolution was finally achieved, with the result that only a series of later inserted small paddock-enclosures there – plus two granaries and a possible shrine structure – seem to be of post-Iron Age attribution.

The reason why so much effort was expended upon the untangling of the site’s Late Iron Age pottery relates to the dating evidence of its metalwork and cremations (attesting to its ‘Late’ usage), and also the distinction of The Plaza’s ceremonial ground. Throughout much of the site’s fieldwork, the two arrays of conjoining small square barrows were evidently thought to have been ‘Belgic houses’ (Figs. 1.8–9 & 2.2; this being an issue discussed at length in the Prehistory Vol.). Although they were eventually recognised as barrow settings, they were never put together with other ‘special’ components in the immediate area. As well as the row of raised granaries (four-, six- and, even, a nine-poster) framing the western side of the two barrow groupings, there was also a series of densely set fence-lines arranged in a complementary manner between the barrows. These latter alignments had earlier been interpreted as relating to horticultural plots within the ensuing Roman Central Enclosure (see Figs. 1.7, 1.9 & 2.2). However, with their vague affinities to the layout of the great shrine complex at Fison Way, Thetford (Gregory 1991), we instead relate them to the area’s Late Iron Age layout. As these settings truncated the barrows and seem to have been secondary within The Plaza’s sequence, a Conquest Period attribution seems most apt. They evidently continued to stand throughout the later first century AD and probably into the next; a crucial point is that this enclosed area remained as ‘reserved space’ and no unrelated Roman structures were ever sited within it.

It was the distinction of this ceremonial complex that determined much of the interpretation of the site’s later Iron Age usage. We cast around widely in an attempt to find a close British parallel for Mucking’s square barrows and, whilst some offered generic similarities, it was only in France’s Champagne District that a direct match was found for their conjoining ‘cell’ arrangements (Ville-Sur-Retourne: Stead et al. 2006). This recognition of a Gaulish connection was then explored in relationship to the site’s imports: Dressel 1 amphora, Terra Nigra and Terra Rubra (the latter both as genuine imports and local imitations). While there are certainly no grounds to argue that this was a markedly high status community given their limited numbers, as well as the modest character of Mucking’s Late Iron Age cremation burials, that it reflected a cross-channel connection nevertheless seems sound.

Figure 1.9. ‘Phasing wall’ plan of Central Enclosure and The Plaza’s components (see Fig. 2.2 and Prehistory Vol., fig. 1.14).

Recognising this, there was one final element to consider: the evidence of the west-flanking granary alignment, termed by us ‘Granary Row’. With their collective footprint amounting to 158sqm, it is estimated that they would have had sufficient storage capacity for approximately 600ha of arable produce as an absolute minimum, enough to feed at least 100–150 families. This would be as opposed to the some four to 10 households that were then, perhaps, resident on the terrace, and whose arable requirements would have only required some 18ha. This must attest to centralised grain storage. It can only be presumed that this was for export and the issue of whether – following classical sources (e.g. Strabo IV.5.2) – this was direct to Gaul or organised via (or otherwise for) a more immediate regional authority (e.g. Camulodunum) is duly rehearsed within the Prehistory Volume, but needless to say is without any definite resolution. The Joneses’ predominant ‘transhumant Belgic shepherds’ interpretation of the site’s Iron Age inhabitants, while reflective of broader trends of the day, seems rather underdeveloped in relation to this.

In addition to various cremations scattered around the site, two other Iron Age cremation cemeteries were excavated, Prehistoric Cemeteries II and IV (IV being the square barrow-area); Prehistoric Cemetery V, located in the south-centre of the site, appears to have seen interment into the Conquest Period and is further discussed in Chapter Two.

A significant difference with MPX relates to what has been assigned as Romano-British Cemetery V. With its main series of inhumations aligned on the back/northern side of the Belgic Banjo’s outer ‘squared’ concentric circuit, this was originally thought to have been of later Iron Age date (and duly assigned as Prehistoric Cemetery III). However, the inhumation and cremation ‘mix’ of its interments – very few of which have any grave-goods – seems to broadly match that of RB Cemetery III. This, the dating of its grave-goods and the fact that some of its inhumations included hobnails, suggests that it was instead of Conquest Period date and that interment there continued into Early Roman times (i.e. Phase 1). It is therefore included with this volume’s cemeteries; that it retains a prehistoric cemetery assignation reflects that fact that it probably had Iron Age origins.

Within the Prehistory Volume it is argued that Mucking’s sequence effectively took off from the Late Bronze Age, variously propelled by salt and grain production. As witnessed by its extraordinarily high levels of buildings and burials per period, it thereafter supported sustained population levels until the end of the early Anglo-Saxon period (following a marked hiatus in the third/fourth centuries AD). A range of factors contributed to this, particularly the terrace’s immediate access to both marsh/estuarine resources and water transportation (see below).

The Joneses consistently emphasised Mucking’s strategic location, especially in the light of its proximity to the Thames’ Tilbury crossing. While they also characterised the site’s sequence as a ‘Landscape Palimpsest’ – in other words, as an all-period sample of the archaeology in this region – in the companion volume what is highlighted is the absolute specificity of Mucking as a place. Many terrace-locales along this portion of the Thames could be said to have comparable strategic qualities, but there and there alone would have furnished a viewshed straight down the Thames’ lower reaches (Fig. 1.10). In this capacity, the site’s long-term Continental connections from the first millennium BC are outlined and it is argued that from the Late Iron Age these granted its sequence quasi-historical properties, whose specificities cannot by entirely comprehended. In some respects, this is a matter of interpretation now coming full-circle, as it resonates with Mucking’s earlier, dynamic culturalhistorical paradigm, following decades when British archaeology was dominated by strict perspectives of insular continuity.

Situation, Excavation Context and Methods

The Romano-British enclosures at Mucking lay on the relatively flat terrace-top at 28–32m OD (see Figs. 1.1–2), although further unexcavated areas of the site no doubt lay downslope to the east. The subsoil throughout was well-draining gravel, with areas of brickearth; the gravel produced excellent cropmarks, whereas the brickearth did not, and these differences were also evidently reflected in how easily archaeological features were discerned. Similarly, the site’s geology greatly affected bone preservation (both human and animal), with it being very poor on the gravels.

One of the key factor’s determining Mucking’s extraordinary long-term sequence was its immediate location (see Fig. 1.10). Perched high on a terrace above riverside marshes, it would have allowed ready access to both seasonal pasture and tidal creeks for the purposes of saltmaking. The potential of Essex’s coastal/estuarine archaeology has only been fully recognised during recent decades (see e.g. Wilkinson 1988, Wilkinson & Murphy 1995, Sidell et al. 2000, Wilkinson et al. 2012 and Biddulph et al. 2012, chap. 2) and, accordingly, in the Prehistory Volume Pete Murphy provides an overview of the site’s environmental situation.

With its many parts – multiple cemeteries and enclosure foci – it has clearly proven difficult to do justice to the many facets of Mucking’s Roman usage and it evades easy characterisation. In the first of the ‘Archaeology of/in Essex’ overview conference volumes (Buckley 1980), aside from mentioning the evidence of the site’s fifth- to sixth-century Germanic immigrants (Margaret and Tom Jones respectively providing papers on Mucking and Essex’s Anglo-Saxon settlement and cemeteries), Drury and Rodwell’s Late Iron Age/Roman settlement study focused on Mucking’s Roman military evidence (and the RBI enclosure), discussing this in the light of their posited ‘planned’ Thurrock-area landscape (1980, 64; see also Chap. 5, below). In the Writtle conference proceedings of 15 years later, Going (1996) in his ‘Roman Countryside’ contribution paid greater heed to Mucking, not unsurprisingly given his recent post-excavation involvement (see below). Aside from discussing its structures and cemeteries, he included a plan of the Central/Double Ditch Enclosure’s ‘farmstead’ (with no mention of the site’s military connection or of RBI; see also Going 1993a). Reflecting how few sites of the period had still then been excavated at any scale (and published) he noted: It is salutary to reflect that DD Enc [Double Ditch Enclosure], together with the Orsett ‘Cock’ site (Toller 1980; Carter forthcoming [1998]) are perhaps the sole examples of what was perhaps the most typical rural Roman settlement type in Essex to have been excavated on the scale necessary to elucidate with any confidence their layout, chronology, and morphological development (ibid., 100). By the time of the most recent Essex conference (2011), while a plan of Mucking’s main enclosure appears alongside a number of other sites in a ‘medium and smaller farmstead’ illustration within Medlycott and Atkinson’s ‘Roman‘ paper (2013, fig. 8), the site was otherwise not mentioned in their contribution. This tells both of the impact of so much recent developerfunded fieldwork and, too, the fortunes of sites whose publication is awaited for too long.

Figure 1.10. The site’s down-river viewshed (upper left) and topographic profile (with exaggerated vertical scale; below) and location map: 1) Mucking excavations; 2) North Ring; 3) Linford; 4) Rainbow Wood; 5) Orsett ‘Cock’; 6) Orsett causewayed enclosure; 7) Rectory Road, Orsett; 8) Stanford Wharf/London Gateway Compensation Site A (after Clark 1993, fig. 2).

Various other finds and settlement evidence have been reported in the vicinity. Romano-British fieldsystem/paddock boundaries extended across the area of the adjacent Linford Quarry investigations (Barton 1962), and were probably directly related to Mucking’s settlement (see Prehistory Vol., chap. 1 and figs. 1.8 and 6.15). A kiln and probable settlement site were found 700 yards southeast of Chadwell St Mary parish church and a portion of a tessellated pavement during building work at the crossroads in Chadwell (Manning 1962, 136). Finds from the gravel pit in Sandy Lane, Chadwell St Mary comprise pottery, probably of the second to fourth centuries, and a coin hoard with a TPQ of the reign of Caracalla (AD 196–217: ibid., 136–40). Excavations there in 1959 produced evidence for a settlement probably of first- to later second- or third-century date, together with one heavily truncated wooden coffin burial (Manning 1962). Further discussed below, more recent excavations 2km away from Mucking at Stanford Wharf (Biddulph et al. 2012) have produced good evidence for a salt-production site in use during the first century AD, and again from the third to late fourth century AD. Otherwise, during the time of Mucking’s fieldwork and its 1980s post-excavation aftermath, relevant comparative material would have been provided by the earlier and on-going excavations at Colchester and Chelmsford.

Within the site’s immediate environs, though, of particular note was the fieldwork at Gun Hill, Tilbury (1969–70: Drury & Rodwell 1973; see Fig. 5.2.3), Orsett ‘Cock’ (1976: Carter 1998; see Fig. 5.2.4) and the Gray’s Bypass sites (1979–80: Wilkinson 1988). Further afield in Essex, amongst others would have been the Kelvedon (1970–3: Rodwell 1988), Little Waltham (1970–1: Drury 1978), Chignall (1978–81: Clarke 1998) and Ivy Chimneys excavations (1978–83: Turner 1999), and later the 1986–91 work at Stansted (Havis & Brooks 2004) and Great Holts Farm, Boreham (1992–4: Germany 2003). With the exception of Chignall and Stansted, these were all quite small sites, usually of a hectare or less; it was really only with Elms Farm, Heybridge (1993–5; Atkinson & Preston 1998) that a Roman settlement was tackled within the region at a scale comparable to Mucking (c. 21ha exposed in total, with c. 9.35ha of investigation of dense Late Iron Age and Roman remains; ibid. and see Medlycott & Atkinson 2013).

The main point is just how little excavation context precedent there was when Mucking was dug, and this is true for all its periods. Today we can take for granted that, in most instances, some manner of site parallels and comparable building/settlement layouts can be turned to, but these were simply not available in the 1960s and 1970s. In contrast, what is now known to be the density of southeast England’s Roman settlement can be readily appreciated in Taylor’s Atlas (2007), with the results of the vast numbers of post-1990 developer-funded excavations currently being compiled through the joint University of Reading and Cotswold Archaeology’s ‘Roman Rural Settlement Project’.

Further to this theme, the staggering amount of huge-project fieldwork recently conducted in Kent also provides crucial context within the ‘Thames Gateway’ (see Millett 2007 for summary); this will be referred to on many occasions in this volume. However, the most relevant site for Mucking’s Roman usage is now Oxford Archaeology’s neighbouring 2008–9 Stanford Wharf excavations (London Gateway Compensation Site A; Biddulph et al. 2012). Extending over 44ha along the north side of Mucking Creek and the Thames mudflats (including one of Essex’s ‘Red Hills’; Figs. 1.10 & 5.17), its significance here is its complementary location down in the riverside marshes off the terrace. Although excavations per se were limited to approximately a third of the area (in addition to watching brief monitoring throughout), they were accompanied by a programme of geoarchaeological sampling. For our purposes, the main result was its evidence of Late Iron Age to Late Roman salt production (see, also, e.g. Drury 1973, 118–22 and Rodwell & Rodwell 1973 for earlier findings from the Stanford-le-Hope area).

Figure 1.11. Quarry-face Excavation (II; Roman): Mike Pitts’ site photographs of Central Enclosure area (top) and, below, excavation of a cremation.

Prior to discussing its fieldwork methods, the circumstances of Mucking’s excavation have to be understood. The digging took place in a working gravel quarry, with excavation occurring close to the quarry face and its drag-line (Fig. 1.11; Jones 1993, 6). While the trial excavation of 1965 took place in an area clear of quarrying, thereafter the race to keep ahead of the dragline was constant (and some areas just had to be abandoned). As indicated in Figure 1.12, due to the location of vehicle haul-tracks and the quarry’s working pace etc., the archaeology went uninvestigated across substantial areas of the site to varying degrees. This is particularly true of its southern portion and, given this volume’s concerns, it most severely impacted upon the recovery of features within the area of RBI and the Southern Enclosures (Fig. 1.13; see also Fig. 2.46).³ As is highlighted in Chapters Two and Three, across portions of those areas the damage was such as to make detailing of their sequence impossible. Accordingly, at least part of the later Roman sequence analysis will have to be based largely on finds distributions, rather than specific stratigraphic attribution as such.

Because of its early excavation, the area of Cemetery I saw a different stripping strategy to the other cemeteries and the main settlement areas. It lay in what was then farmland, and a front-loading Drott (rather than the requested box-scraper) was used to remove the ploughsoil after the harvest (Jones 1993, 6). Experienced workmen were then used to remove the subsoil, and a small team of archaeological students carried out planning and recording. Cemeteries and settlement enclosures excavated in subsequent years lay in areas that were stripped using tracked vehicles towing box-scrapers; the excavation budget bore some of the cost of this in order to be able to choose a smaller machine (when available) and to enable supervision of the depth and direction of scraping. There was thus a limited amount of time between an area being scraped and it being quarried, in which excavation had to occur. From 1968, work continued all through the year, and excavation of some features was therefore carried out in far from ideal conditions (e.g. frozen ground surface, high winds, driving rain). Excavators and site supervisors were provided with a very detailed series of notes, complete with sample plans and sections, to try to ensure a consistent quality of recording (see Prehistory Vol., fig. 1.10).

The aim was the total excavation of features (rather than the selective sampling that is now standard). In some areas this had to be compromised, due to the speed of digging required to keep ahead of the dragline, but it is estimated that for the site as a whole, c. 75% of the archaeological remains were fully excavated and recorded (though some whole areas had to be abandoned with only cursory examination). Most features were tackled by spit-digging, where a predetermined level (usually 3 or 6 inches) was taken down in plan, with any finds recorded as coming from this ‘level’ within this feature. Often, the plans were superimposed onto one another on the plan sheets, resulting in a cumulative section drawing alongside a ‘contoured’ plan (Fig. 1.14). Barford (1995) offers an account of the excavation strategy, and the conditions encountered, which serves as a useful corrective to those accounts that suggest methods used at Mucking were sub-standard (e.g. Dixon 1993, Clark 1993).

As the aim was complete excavation of all features on the site, in a sense the burials were not treated differently from any other feature (today, while burials would obviously be completely excavated, a pit would probably be 50% sampled and a ditch 5–10% sample-dug). In ideal circumstances, where a burial at Mucking had been recognised as such from surface soil-marks it would be taken down in plan, rather than half-sectioned first, as a pit or other discrete feature would be treated (or dug in metresections like a ditch). Usually plan-drawings were made at successive levels (Fig 1.15). Photographs would also be taken during and after excavation, and (as today) any grave-goods and skeletal remains or staining were indicated in notes and on the plans. In unhurried excavation conditions, recording can generally be considered good: the inhumations within wooden coffins in Cemetery I, for example, had every nail recorded and numbered on the plan series; moreover, their orientations within the grave were written on the individual finds bags for later integration with the records. The corollary to this, of course, is that when burials had to be dug in haste, recording could be cursory. Much depended on the individual excavator and on the supervisor maintaining the notebook records.

Tom Jones maintained the grave index in the field (as well as site notebooks specifically for burials), and a continuous numbering sequence for burials was implemented, such that Grave 24 could be followed by Cremation 25 (inhumation burials were termed Grave, or GR, while cremation burials were termed CREM). The number of a burial can, therefore, give an idea of when it was excavated. Also encompassing Anglo-Saxon and prehistoric interments, all burials were included in this sequence, and all were located by the northing and easting of the grave’s centrepoint. Although context recording was not used, the employment of recording by levels meant that most of the burials and their assemblages could generally be reconstructed with some accuracy.

The Joneses always intended to publish their excavations. The ‘draft for definitive report’ produced in November of 1976 listed, within the fascicule on ‘Belgic to Romano-British’, the three cemeteries then known (I, II & III), together with isolated cremations. The intention seems to have been to publish these as descriptions, followed by separate treatment of the finds, detailed treatment of specific grave groups (Graves 903 & 911 are described as examples), and a section on coffin nails. This publication method represents the standard treatment of Romano-British cemeteries at the time (cf. Down & Rule 1971 on the St Pancras, Chichester Roman cemetery). However, as is well documented (Clark 1993, 13–14), the demands of preparing the Level III archive for Mucking meant that any publication plans were postponed, and eventually responsibility for the site was passed over to the Mucking Management Committee, who instigated the phase of post-excavation (from 1986–92) funded by English Heritage and based in the British Museum (known as the BM/EH phase).

Figure 1.12. Site-area plan with areas of over-machining indicated by grey-tone (left) and its impact on finds distributions (centre: rotary querns; right, kilns and fire bar fragments).

Figure 1.13. ‘Raw’ base-plan rendering of the South Rings, RBI and Southern Enclosures, with white swathes indicating unrecorded portions destroyed by the quarry works (with Roman-phase structures indicated).

Figure 1.14. Worked up ‘contoured’ field drawing of Structure 1.

Figure 1.15. Field plans (by M. Jones) of Burial 899 at Mucking.

A volume on Roman Mucking, including its cemeteries, was to be an outcome of the latter postexcavation phase and publication proposals put forward in 1987 included a series of period volumes together with a general introductory volume, and the Late Iron Age and Roman volume was outlined (with a stated completion date of March 1989). This obviously never appeared, and the archive had sat, largely untouched, until the current phase of work was instigated in 2007.

Base-line Sources

As outlined more fully in the Prehistory Volume, interpretation of the settlement sequence at Mucking has been heavily dependent on the two previous phases of post-excavation: the Joneses’ Mucking Post-Excavation (MPX), which ran from 1978 to 1985, and that of the British Museum/English Heritage team (BM/EH), from 1986 to 1992. From the latter, we have drawn upon the stratigraphic sequence for the site as outlined in the Site Atlas (Clark 1993); an unpublished manuscript by Chris Going on aspects of the Roman-period occupation that drew on work by Rosemary Jefferies (now Arscott) and Ruth Birss (now Leary) has also been used as a starting point for some elements of the text that follows (albeit with significant elaboration and alteration). Within Going’s intended volume, while all the site’s finds were to be reported, its archaeology was to be addressed through a series of ‘windows’ case-studies, variously focusing upon the RBI and