Professional Documents
Culture Documents
Urbanism in the
Aegean Bronze Age
SHEFFIELD STUDIES IN AEGEAN ARCHAEOLOGY
Urbanism in
the
Aegean Bronze Age
edited by
Keith Bmnigan
Copyright © 2001 Sheffield Academic Press
A Continuum imprint
Published by
Sheffield Academic Press Ltd
The Tower Building, 11 York Road, London SE1 7NX
370 Lexington Avenue, New York, NY 10017-6550
www.SheffieldAcademicPress.com
www.continuumbooks.com
A catalogue record for this book is available from the British Library.
ISBN 1-84127-341-4
Contents
Abbreviations vi
Preface Keith Braniganv vii
List of Contributors x
Settlement Nucleation
1. CHRISTOPHER MEE
Nucleation and Dispersal in Neolithic and Early Helladic Laconia 1
Appendices
1. TODD WHITELAW
The Floor Area of 207 Minoan Houses 174
2. WILLIAM CAVANAGH
Recent References to Bronze Age Roads in the Aegean 180
Abbreviations
AAA Athens Annals of Archaeology
ADelt Arkhaiologiko Deltion
AE Arkhaiologiki Efimeris
AEMTh To Arkhaiologiko Ergo sti Makedonia kai Thraki
A]A American Journal of Archaeology
AM Mitteilungen des Deutschen Archaologischen Instituts, Athenische Abteilung
AR Archaeological Reports
ASA Annuario della Scuola Archeologica di Atene
BAM Beitrage zur Ur- und Friihgeschichtlichen Archdologie des Mittelmeer-Kulturraumes
BAR British Archaeological Reports
BCH Bulletin de Correspondance Hellenicjue
BICS Bulletin of the Institute of Classical Studies of the University of London
BSA Annual of the British School at Athens
CAJ Cambridge Archaeological Journal
CMS Corpus der minoischen und mykenischen Siegeln
CP Classical Philology
CQ Classical Quarterly
Ergon To Ergon tis Arkhaiologikis Etairias
G&R Greece and Rome
JAS Journal of Archaeological Science
JHS Journal of Hellenic Studies
JMA Journal of Mediterranean Archaeology
Kr Chron Kritika Chronika
OJA Oxford Journal of Archaeology
PAE Praktika tis en Athinais Arkhaiologikis Etairias
PBA Proceedings of the British Academy
PCPS Proceedings of the Cambridge Philological Society
SIMA Studies in Mediterranean Archaeology
SMEA Studi Micenei ed Egeo-Anatolici
Preface
Keith Branigan
The papers found in this volume were first personality. These are, after all, what makes a
presented at the fifth Round Table on Aegean town different to a village, and urban differ-
Archaeology, held at Sheffield in January ent to rural. They underpin the definition of a
2000. They were subsequently re-written, in town which I offered the Round Table, and
the light of the intensive discussion and which I unashamedly admit is taken from the
debate which they generated, for publication combined words of Louis Wirth (in 1938) and
in this volume. Two contributors to the Round Bruce Trigger (in 1972): 'a relatively large,
Table were unable, for various reasons, to con- dense and permanent settlement of socially
tribute a chapter to the book, but they con- heterogeneous individuals, which performs
tributed fully to the discussions which specialist functions, of a non-agricultural
informed the papers published here. We type, in relationship to a broader hinterland'.
would like to acknowledge the contributions Whilst that definition can be seen to
of Cyprian Broodbank and Vance Watrous, as embrace a number of both Minoan and
well as the full part played by our principal Mycenaean nucleated settlements, most
discussant, Anthony Snodgrass. Aegean prehistorians have long recognised
Our Round Table was about urbanism, and that Minoan towns were in some respects
so is this volume; it is not concerned with quite different to those of contemporary
urbanisation. That is, we focus not on the mainland Greece. This was something that
process but rather on its end-product. This is was brought out by our discussions, and
partly because we did not want the discus- indeed the differences were seen to be per-
sion to drift from urbanisation to state-forma- haps more wide-ranging than we had previ-
tion or the emergence of civilisation. ously realised.
Important and interesting as they are, these
topics have been the centre of much debate in
Aegean prehistory over the past thirty years, Minoan Urbanism
and they will be so again. The nature and
character of Bronze Age towns, however, has Todd Whitelaw rightly says that to under-
seen much less discussion, particularly at a stand Minoan urbanism we must first
generalised level. Papers on prehistoric attempt to establish its scale in human terms,
Aegean towns have largely focussed on their which means getting to grips with the diffi-
architecture, and particularly their elite or cult topic of population estimates. He pre-
public architecture, and are often restricted to sents probably the most carefully argued and
a single town or even a single building. The thoughtful paper yet published on this topic
purpose of the Round Table was to direct (and an appendix provides some of the raw
attention and thought not only to urban set- data for others to use in further research).
tlements as a whole but to their social and Keith Branigan compares urban and rural
economic roles, their demographic signifi- populations and concludes that urbanism
cance, and ultimately to their character or was a way of life for a very significant part of
viii Urbanism in the Aegean Bronze Age
the Minoan population, and that it was struc- domestic housing is a feature of this early
tured in a three-tier hierarchy. Jan Driessen, town, and that changing social structures
whilst accepting a similar hierarchical struc- may be reflected in the development of the
ture, uses the evidence provided by more town through the Middle Helladic. One of
than twenty regional surveys to argue that the most obvious points of difference
both settlement history and hierarchy varies between Minoan and Mycenaean towns, the
from region to region. Tim Cunningham size of public spaces and courts, is taken up
takes the arguement a stage further, with a and explored from the Mycenaean viewpoint
detailed examination of urban settlements by William Cavanagh. Open spaces and
and their hinterlands in east Crete. He identi- courtyards in Mycenean towns have their
fies local variability both in centre-periphery own distinctive character and functions.
relations, and in the spatial organisation of They have little to do with public meeting
towns. The same theme of temporal and places, but much to do with public ceremo-
regional variability is taken up by Use Schoep nial processions and progress. The ceremo-
in her discussion of the urban-hinterland nial roads which lead from the courts, lead
relationship as revealed in the archival evi- also beyond the urban centres to their rural
dence. She suggests that in Protopalatial hinterlands. John Cherry and Jack Davis
Crete administrative documents are restrict- explore the settlement of those hinterlands in
ed to urban centres and 'public' buildings, an attempt to understand better what sus-
whilst in the Neopalatial they are more tained the central places. Their case study of
widely distributed in town and country and the Nemea valley reveals only a handful of
appear in private as well as 'public' contexts. other potential towns in the region of
Her case studies suggest that urban-rural Mycenae, forming a second tier in the urban
relationships may have been managed in dif- hierarchy. Below this there appear to be only
ferent ways for different purposes in different villages, hamlets and farmsteads. John
times and places. Bennett and Cynthia Shelmerdine, examining
Overall, the papers on Minoan urbanism the case of Pylos and its nearest neighbours,
suggest that towns were a very significant are able to outline the growth of the nucle-
part of Minoan life, demographically and ated settlements, and by relating the archaeo-
socially as well as economically, but that the logical to the textual data, to suggest the way
ways in which the urban-rural dialogue was in which relationships between first and sec-
articulated varied considerably from region ond rank centres may have developed.
to region, as well as from Protopalatial to Stelios Andreou examines a very different
Neopalatial. region, in central Macedonia, where small-
scale societies endured for millennia. In the
Late Bronze Age a small number of signifi-
Mycenaean Urbanism cant nucleated settlements with features like
perimeter walls, spatial organization, and
The early stages of mainland urbanism have acquisition of long-distance trade objects,
received little attention and Anastasia were clearly the focus for social activity and
Dakouri-Hild's paper on Middle Helladic the exercise of power. This is the sort of com-
Thebes is therefore a particularly welcome plexity we might associate with towns but
contribution. She demonstrates that variabil- should that term be applied to these
ity in both the density and architecture of Macedonian mounds?
Preface ix
Certainly, as we noted earlier, size is not sarily prepare the way for it. But to explore
everything when it comes to defining urban- how and why towns develop is not only a
ism, the provision of social and economic long and difficult task, it is also a different
services and amenities are essential features one to that which the fifth Round Table set
of towns. It is appropriate therefore that the itself.
volume begins with Christopher Mee's dis-
cussion of nucleation and dispersal in
Neolithic and Early Bronze Age Laconia. He Acknowledgment
demonstrates that the growth of nucleated
settlements alone neither announces the We are pleased to acknowledge the financial
arrival of urbanism nor does it always neces- support of INSTAP for this Round Table.
Contributors
STELIOS ANDREOU Department of Archaeology, University of Thessaloniki,
Greece
Neil Brodie supervised a gradiometer sur- which varies in depth from 55-400 cm and
vey of the site. The variation in magnetic contains sherds, 'brick earth7 and stone.
intensity was greatest around the centre of Below this there is a layer of clay with sand
the mound, the core of the prehistoric settle- and gravel beds or lenses. It would appear
ment. Four areas of burnt earth, probably that the site was built on the floor of a
mudbrick or daub were identified but it is not drained lake, although a floodplain environ-
yet clear whether the rectilinear outlines of ment cannot as yet be ruled out. If there was
these features relate to buried structures. a lake in the Holocene, it may well have
Resistivity survey was also tried but proved stretched across the present Eurotas valley. Is
less effective. this the origin of the story in Pausanias (3.1),
Nine 7.5 cm cores were drilled in a series of that Eurotas, one of the first kings of Sparta,
transects across the site. Eight of the cores 'channelled the stagnant water from the plain
went 5 m deep and one was taken down to 10 down to the sea, and when it had drained
m. They were examined by Peter James and away he called the river which was left there
Alison Jones in the field and samples have the Eurotas'?
been taken for analysis of fossil biogenic Now that the pottery has been systemati-
material, especially pollen and diatoms. The cally studied, we know that Kouphovouno
cores revealed an anthropogenic horizon was occupied in the Middle Neolithic,
Nucleation and Dispersal in Neolithic and Early Helladic Laconia 3
(C29/216), Plitra - Goulas/Kastelli (C32/228), Table 1.1 Laconia Survey: Neolithic Sites (bold
Kotronas - Skopas (C46/203) and a number indicates that Neolithic is the sole or main period of
use). 10496 is a 'non-site'.
of the sites identified by the Laconia Survey
(Figure 1.6 and Table 1.1). In a recent paper, Site Number Size (ha) LS ii Reference
Cavanagh (1999: 34-7) lists twelve locations Bill 6.00 325-8
which have Final Neolithic finds but there is B116 0.01 328
nothing earlier. Only E48 is an obvious set- E48 0.6 339-40
E77 0.12 339
tlement site with pottery, chipped stone arte-
E81 340
facts and polished stone tools. At most of the L401 0.13 379-80
other locations there were just chipped stone R429 0.41 408-9
artefacts. The distribution of the sites is of T480 <0.01 421
interest, in that they are situated on some of T481 0.03 421
U487 0.1 424-5
the poorest soils in the region, in particular
U489 0.03 425-6
limestone outcrops which now have just a 10496 1
thin cover of terra rossa. Cavanagh (1999: 31)
notes that elsewhere in the Peloponnese
many Late and Final Neolithic sites are also on pastoralism but not extended transhu-
in agriculturally marginal locations and mance. To offset the risks inherent in this
there is an increase in the use of caves. He specialized strategy, there would have
argues that this reflects a greater emphasis been exchange with communities which
Nucleation and Dispersal in Neolithic and Early Helladic Laconia 5
Project, the Asea Valley Survey, the Berbati- time. If these communities were in fact envi-
Limnes Archaeological Survey and the ronmentally circumscribed or constrained
Southern Argolid Exploration Project, because of their reliance on restricted water
although not in the case of the Nemea Valley resources and consequently could not split
Archaeological Project where there were into smaller units, this would surely have led
fewer Late/Final Neolithic sites (Table 1.2). to the institution of some form of centralized
Nucleated Middle Neolithic settlements, like organization to regulate access to resources.
Kouphovouno, include Asea (Forsen 1996), It is not in fact necessary to invoke envi-
Asea Valley site S16 (Forsen et al 1996: 85), ronmental circumscription as an explanation
Ayioryitika (Petrakis 1992: 341), Lerna for the Early and Middle Neolithic settlement
(Johnson 1996a: 276-77), Berbati-Limnes site pattern in the Peloponnese. Perles (1999b)
FS400 (Johnson 1996b: 44-57), Tsoungiza has analyzed the distribution of EN2 sites in
(Wright et al. 1990: 624-25), and Corinth eastern Thessaly and finds that 'no positive
(Alram-Stern 1996: 222-29). Franchthi may relationship can be established between set-
have had a similar role (Alram-Stern 1996: tlement choice or settlement density and nat-
244-61) but it is difficult to estimate the orig- ural features'. She concludes that the 'main
inal size of the site because of the rise in sea- factor in settlement foundation was socio-
level (Johnson 1996a: 280). economic' (Perles 1999b: 53). However,
Johnson (1996a) has claimed that it was the Thessaly is rather different in that the settle-
need for a reliable water supply which deter- ments are typically just 1-4 km apart
mined site location in the north-east (Halstead 1995: 13-14)—Perles (1999b: 46)
Peloponnese in the Early and Middle reckons that the mean distance is 2.3 km. At
Neolithic periods. Early farmers, who were least in the Early Neolithic period, villages
dependent on the use of the hoe until the may have periodically split up when the pop-
introduction of the ard in the Final Neolithic ulation reached a critical level (Perles 1999b:
or Early Helladic period, favoured 'well- 53-54), although there were no doubt inte-
watered alluvial soils of high potential for grative mechanisms to control this tendency
arable agriculture' which 'have a strictly lim- towards fission (Halstead 1999: 89).
ited distribution in southern Greece' The size of the communities in the
(Johnson 1996a: 282-83). Van Andel and Peloponnese presumably ensured their
Runnels (1987: 70-73) have also stressed the demographic viability and generated a pool
importance of spring-fed agriculture at this of labour and surplus agricultural resources
Table 1.2 Neolithic and Early Helladic sites in the Peloponnese identified by surface survey. Some of the figures are
approximate.
four sites have pottery which may be Late Neolithic or Middle Helladic
Nudeation and Dispersal in Neolithic and Early Helladic Laconia 7
1 = Kouphovouno 10 = Palaiopyrgi
2 = Diros 11 = Skoura
3 = Alepochori 12 = A.Vasileios
4 = Goritsa 13 = Geraki
5 = Asteri 14 = A. Stephanos
6 = A. Stratigos 15 = Pavlopetri
7 = Apidia
8 - Plitra area covered by Laconia Survey
9 = Kotronas
Figure 1.5 Laconia: location of major Neolithic and Early Helladic sites and of the area covered by the Laconia Survey
8 Urbanism in the Aegean Bronze Age
Table 1.3 Laconia Survey: Early Helladic Sites (bold have operated in tandem with social com-
indicates that Early Helladic is the sole or main period
pacts which provided a further safeguard
of use). R3012, U3001, U3005 and U3006 are 'out-of-
area' sites. against the risk of crop failure and conse-
quently the threat of starvation (Talalay 1987:
Site Number Size (ha) LS ii Reference
167-69).
C126 0.01 331 Crete apparently presents us with an even
C128 0.01 331
C131 0.01 331
more extreme example of nucleation. It seems
G154 0.04 350 that Knossos was the only major Neolithic
K414 0.03 374 settlement (although doubts were expressed
L400 0.03 379
about this by Peter Tomkins at the Round
M357 0.03 383
N191 0.07 397 Table) and possibly covered more than 5 ha
N333 0.29 394 in the LN period (Evans 1994: 19 but
P262 0.05 399 Whitelaw 1992: 226-27 is not so sure). There
P263 0.01 398
P267 0.31 400 has of course been some debate about what
P269 0.10 400 this implies (Broodbank 1992; Whitelaw 1992;
P284 1.00 399 Manning 1999). Nevertheless, it is assumed
P285 0.20 397
Q360 25 403-5
that Knossos had exceeded the size threshold
R280 0.05 411 for a simple family-based community.
R287 0.18 409 A higher level of social organization,
R289 0.19 407
R428 0.47 410
marked by institutionalized inequality, will
R462 <0.01 414 have been almost inevitable once the popula-
R529 <0.01 410 tion of these settlements exceeded 500
R3012 0.71 409
(Halstead 1995:13-14; Manning 1999: 470-71).
S448 <0.01 416
S459 0.05 417 It is of course notoriously difficult to estimate
S478 0.02 418 population size but, in a typically rigorous
U490 0.20 428-9 analysis of the evidence from the prehistoric
U500 0.70 435-6
U504 0.05 437 Aegean, Whitelaw (this volume) proposes a
U520 0.11 436-7 figure of 200-225 per hectare, so we may
U3001 0.71 432-4 need to revise our perception of Middle
U3005 0.13 438
U3006 0.56 438 Neolithic society as relatively egalitarian.
the Laconia Rural Sites Project - LP7 = R3012 apex? Or was the relationship more symbi-
and LP8 = R287. On the basis of the artefact otic? If Kouphovouno did have an adminis-
distribution, geophysical data and soil analy- trative role, this should become apparent
ses, our impression is that they were small set- when the site is excavated and it may then be
tlements, a cluster of houses. Some of the possible to speak of a proto-urban centre. At
other sites may have been farmsteads occu- the moment the evidence is inconclusive.
pied by just one family. Cavanagh and Elsewhere in Laconia a number of clay seal-
Crouwel (in press) believe that there may have ings have recently been discovered in an EHII
been a two-tier settlement hierarchy in EHII. context at Geraki (C12/103) next to a pithos
The largest of the sites, P284, covers c. 1 ha. which contained carbonized grain (Weingarten
Down in the Eurotas Valley there are sev- et al. 1999). Crouwel (1999: 150) believes that
eral major EHII settlements (Figure 1.5) - the sealings must be administrative docu-
Kouphovouno, Palaiopyrgi (C4/93), Skoura - ments. In the Helos Plain there is a cluster of
Vouno Panagias (Banou 1999: 65-66) and Ayios EHII sites (CIS/145, C16/125, CIS/139, C19/
Vasileios (C7/101 - Banou 1999: 65-67) but no 136, C20/137, C22/132, C23/133, C24/142,
network of subsidiary sites, or at least not on C26/142, C27/146, C29/216) around Ayios
the east side of the river which has been inten- Stephanos (Cl7/141), which may well have
sively surveyed. So we appear to have two been the dominant settlement in southern
different settlement patterns: nucleated in the Laconia. The top of the hill covers approxi-
lowland and dispersed in the hinterland. This mately 4.5 ha and Early Helladic occupation
is not a temporal phenomenon—of dispersal was apparently widespread (Taylour 1972). It
and then nucleation—as Wiencke (1989: is a pity that Lord William Taylour did not dis-
498-99) has suggested for the Argolid cover a corridor house. Dickinson (1992:
(Cavanagh and Crouwel in press) but may be 109-10) draws attention to the size of the
a reflection of two different agricultural underwater site at Pavlopetri (C39/250) but
regimes. The situation in Late Neolithic wonders how much of this is Early Helladic.
Thessaly is quite similar. Halstead (1995: 15) The stark contrast between the Neolithic and
observes that Villages are clearly concen- Early Helladic settlement pattern in Laconia is
trated in the areas of early agricultural settle- not unexpected, particularly if we take into
ment and hamlets in the areas colonized later. account the enormous time span. However, I
Villages and hamlets thus represent alterna- had not anticipated that there would be so
tive settlement strategies in agriculturally much variation in EHII. The situation in central
core and marginal areas rather than the cen- and southern Laconia seems quite different. It
tres and satellites of local site hierarchies'. is consequently difficult to discern a regional
Although it would appear that there was less trajectory, although we can clearly see
of an emphasis on pastoralism in Early advances in organizational complexity which
Helladic Laconia, the type of economic parallel developments elsewhere.
pluralism which Cavanagh (1999: 56-58) has
proposed for the Late and Final Neolithic
periods may still have been in operation. Early Helladic Peloponnese
How did the sites in central Laconia inter-
act? Were they dependent or interdependent? Particularly in the Argolid there is a compara-
Should we envisage a settlement hierarchy ble increase in the number of sites, although
with Kouphovouno, for example, at the not in Arcadia or Messenia (Table 2). A two or
Nudeation and Dispersal in Neolithic and Early Helladic Laconia 11
first order approximation, but where addi- out and construction. For example, at the
tional information from the culture under Minoan site of Gournia, most analysts accept
study is available, it is likely to improve con- the original excavator's definition of individ-
siderably the relevance of estimates, and our ual houses, on the basis of room layout, inter-
confidence in them. connections between rooms, and regular
Attempts to side-step some of these con- differences in construction between internal
textual difficulties have involved the devel- and external walls (Boyd Hawes 1908: plan;
opment of region-specific formulae relating Fotou 1993). Such boundaries between resi-
settlement population to overall settlement dences can be complicated by additions and
area (e.g. Frankfort 1950; Cook and Heizer modifications in property boundaries during
1968; Drager 1976; Aurenche 1981; Kramer the life of a structure, since the households
1980; Sumner 1979; Adams 1981; Shiloh 1980; using them also change in size and composi-
Kolb 1985; Carothers and McDonald 1979; tion according to domestic cycles (Fortes
Haviland 1972; Sanders, Parsons and Santley 1958), and may adapt the houses physically
1979; Zubrow 1974), by using ethnographic to conform to their changing social needs
data on settlements in the same geographical (e.g. Stone 1981; 1987; 1996; Castel 1984). But
region as the archaeological culture of inter- regardless of such modifications, it is likely
est. These rely on an assumption of direct that changes in house size and layout will
historical continuity between the present and also be accompanied by architectural re-defi-
past cultures, which is rarely justified by an nition of spatial boundaries, so that the social
explicit argument of relevance. However, units are spatially recognizable, at least to the
even with the latter, we still have little basis inhabitants of the community.
for assessing the additional assumption of
continuity in spatial behaviour which is
necessary for them to be of any interpretive Identifying and Interpreting the
value. Neopalatial Household
What about the alternative, more variation-
sensitive approach: working from identifi- In many cases, we could go on to debate
able houses, multiplying-up by estimated whether it is appropriate to interpret such
average residential group - or household - architecturally-defined units as independent
size, and then multiplying this by an estimate residential households, but for most of our
of the density of houses within the commu- Neopalatial examples, this does not appear to
nity? In some archaeological cases, particu- be seriously in doubt (McEnroe 1982). What
larly where there is agglomerative architecture, we do need to establish is what sort of house-
it may be difficult to disentangle individual holds we are dealing with, since residential
residence units, though this is often possible units can represent various different types of
on the basis of detailed architectural study of social group (Bender 1967; Yanagisako 1979;
the sequence and pattern of construction (e.g. Wilk and Netting 1984; Hendon 1996; Price
Rohn 1965; Hill 1970; Dean 1970; Wilcox 1975; 1999). Neopalatial Gournia provides a good
Adams 1983; Crown 1991; Lowell 1991; methodological example for exploring this
Whitelaw 1983; Ling 1997). On the other question. One particularly marked feature of
hand, for many sites, individual residential the houses which can be readily defined at
units are fairly readily identifiable architec- Gournia is their standardization in size
turally, through cultural conventions in lay- (Figure 2.2).
18 Urbanism in the Aegean Bronze Age
Figure 2.2 House size variability with family form: left: 'Aliabad', Iran, 1975; right: Gournia, Crete, Neopalatial period.
Having criticized the basis for Naroll's The reason for proposing this is precisely
cross-cultural equation of 10 m2 per person the standardization in scale. In any society
of roofed floor area, we can't really turn with regular polygamous households, we can
around and use it to suggest what sort of expect considerable variation in membership
social group we should reconstruct as resi- and demands on space, depending on how
dent in each house, but the relatively small many spouses - and their respective families
size, and particularly the standardization in - are incorporated into each household. The
size is itself strongly suggestive of nuclear or same applies to cultures with extended fam-
minimally-extended stem families - of about ily structures - where households usually
4-5 individuals or so.1 encompass the spouse of the household head,
From Sites to Communities 19
and their juvenile offspring, as well as the resent the replication of coherent sets of
families of variable numbers of married off- rooms within the larger architectural unit, as
spring. In such a situation, we should antici- is often clear with multi-family residences,
pate considerable variation in architectural but rather involve an expansion or scaling-up
house size, as indeed we see in Figure 2.2 for in individual room size (while retaining sim-
an ethnographically documented agricul- ilar features and layout), and the prolifera-
tural village on the left (where many house- tion of small rooms, used in increasingly
holds are composed of extended families: task-specific ways. In other words, the elabo-
Kramer 1982: 21), but not for Gournia, on the ration in scale (rather than duplication) of the
right. While a general trend can be seen in the same basic spatial organization. That this pat-
ethnographic case, linking larger houses with tern results from wealth differences, can be
family size, the relationship is not direct, explored in specific cases through compar-
because of the lack of synchrony between isons with the quality of construction and
social and architectural development - a materials, decorative elaboration, and quality
nuclear family may reside in a compound and quantity of finds. At a gross level, this
expanded to suit an extended family in the can also be suggested by the size distribution
recent past, while an extended family may be of the houses, with many small houses, and
limited by a variety of factors from expand- decreasing numbers of larger examples - typ-
ing their dwelling. ically what we would expect in a status pyra-
One might immediately object that any such mid. This is a very general pattern, which can
variation in the archaeological case could be be documented widely among both ancient
masked by the absence of information about and modern societies (Figure 2.3).2
the scale of occupation on the upper floor(s) of Bearing in mind differences in scale due to
the structures. However, this is also relevant to wealth differences, where we can isolate archi-
the ethnographic case, and the very different tecturally individual households in Minoan
pattern of variation still comes through in prehistory, there appears to be good evidence
ground floor areas alone. We can expect that for a long tradition of houses of approximately
such a degree of variability in spatial needs the same scale from the Early Bronze Age
will be manifest at the ground floor level as through to the Neopalatial period, which sug-
well as in total house floor area. gests the nuclear family as the basic social and
But estimating household composition residential unit in prehistoric Crete.
from house floor area is not quite that A relatively small number of sites which
straightforward, since the number of resi- were extensively excavated in the early
dents is not the only, or indeed always neces- decades of this century, provide the bulk of the
sarily the principal determinant of house 207 Neopalatial structures which I have felt
size. Commonly, cross-culturally, house size able to identify as distinct houses. At the pala-
also varies considerably with household tial centres, while interest has focused on the
wealth, which can be independent of house- exploration of the palaces themselves, a few
hold size (Kramer 1982: 116-38, 170-81; houses have been explored in the immediate
Home 1994: 157-60; Kamp 1987; Wilk 1983; vicinity of most palaces, which complement
Netting 1982). the picture available from the extensively
Such wealth distinctions appear to be iden- investigated non-palatial towns (Figure 2.4).3
tifiable archaeologicaHy, where, within a site Differences in the representation of houses
or culture, increases in house size do not rep- of different sizes at individual sites raises a
20 Urbanism in the Aegean Bronze Age
Figure 2.3 House size distributions for ancient cultures Figure 2.4 Neopalatial house size distributions by site.
of the Eastern Mediterranean and Middle (For data, see appendix.)
East. (For data sources, see note 2.)
new question. Palatial sites tend to have explore larger houses and ignore smaller
larger houses than non-palatial sites, but is ones, because of the nature of the finds antic-
this a real pattern which distinguishes ipated or actually being recovered?
between different communities, or types of I suggest the pattern incorporates both real
community, or is this simply the result of differences between communities, and excava-
excavation bias? Do excavators tend to tion bias. It is often clear from excavators' own
From Sites to Communities 21
descriptions that they have pursued certain To round out the architectural picture for
trenches in anticipation of 'rich' finds. At sites Neopalatial Crete, it is appropriate also to
such as Knossos, Malia and Zakros, excavation consider the size data for so-called 'villas'
has concentrated on relatively large houses and isolated rural structures (Figure 2.5).
close to the palaces, but smaller houses have
been revealed at some distance from the
palaces at Knossos (Stratigraphic Museum Neopalatial Community Organization
Extension: Warren 1981/1983; Acropolis
House: Catling et al 1979) and Malia (Quartiers So - can we now work out an average house
Alpha and Gamma: Demargne and Gallet de size, assume a household of five - six individ-
Santerre 1953; Quartier Theta: van Effenterre uals, and simply multiply up by our overall
and van Effenterre 1976). site sizes, to get population estimates? Alas
On the other hand, a comparison of sites no, there are a few other dynamics at work,
where considerable numbers of houses have which relate to the internal organization of
been uncovered, such as Gournia and communities. To work up from household
Palaikastro, suggests that there may have data to estimate community populations, we
been real differences in housing patterns need to understand the patterns of variation
between these sites - borne out by other dif- in households across a community, and situ-
ferences in construction style and finds - ate this within an understanding of the over-
which suggest that many of the houses exca- all spatial structure of the community.
vated at Palaikastro probably belonged to In many cultures, the residential density of
wealthier families. What we cannot know for a community varies with its nature - larger
certain is whether excavation has preferen- sites, such as towns and cities, fulfilling cen-
tially revealed the wealthy quarter of tral-place functions within the society, often
Palaikastro, or whether the whole town was,
in general, composed of larger houses than
are typical at sites such as Gournia and
Pseira. Again, I suspect perhaps a bit of both,
since smaller houses were investigated but
not reported in as much detail on the fringes
of the excavated area at Palaikastro (blocks X,
Y and A: Dawkins 1903-1904: 204, 214-15).
Overall, the evidence that we have suggests
that the palatial centres and other large com-
munities such as Palaikastro, had a consider-
able range of housing, for residents across the
entire social spectrum, whereas other sites,
such as Gournia and Pseira, had more uni-
form residential groups, perhaps with few or
no representatives of the wealthiest classes of
Minoan society. Given the likely difference in
position of different communities within
regional settlement hierarchies, this should Figure 2.5 Neopalatial structure sizes: community
occasion no surprise. houses, 'villas', and isolated rural 'farms'.
22 Urbanism in the Aegean Bronze Age
have higher residential densities than rural between different sites, from more localized
villages. In some traditional cities, particu- variations within individual sites. The exten-
larly high residential densities are found in sively investigated sites of Gournia and
the central, oldest quarters, as house-lots Pseira, for example, are generally character-
have been sub-divided over the centuries, ized by small houses (Figure 2.6). The less
leading to considerable variation in residen- extensive samples of houses at Malia and
tial density across the community. Central Zakros reveal a much higher proportion of
residences may also be socially desirable, and large houses, but as noted above, to a degree
smaller houses may be a response to the this probably represents the clustering of
higher value of plots in the popular areas of such residences close to the palaces - the
the city. On the other hand, in other pre- focus of most excavation. The smaller and
industrial cities, the city may be inversely more densely packed Protopalatial and early
zoned by status, with wealthy families able to Neopalatial houses of Quartiers Alpha,
maintain spacious houses in the centre of the Gamma and Theta at Malia may suggest that
community. In this case, despite larger num- smaller houses tended to dominate further
bers of resident servants, the spacious houses from the site centre. The excavated area at
may produce lower residential densities in Palaikastro gives a hint of such zoned organi-
the city's centre. Together, these conflicting zation, in that smaller houses are found away
possibilities indicate that we need to assess from the wide main street. Differences in
each ancient culture (or even community) in spacing can also be significant, and the recent
its own terms, to see how urbanization devel- investigations at Pseira demonstrate less
oped in that context, and identify its own dense crowding of houses on the fringe of the
specific residential patterns (Carter 1983). town (Betancourt and Davaras 1999), a pat-
Some ancient cities, such as Ur and Nippur tern which may also be represented by the
in Mesopotamia, reveal habitation quarters patchy distribution of surface sherds on the
with distinct characters (Woolley and periphery of the site at Palaikastro
Mallowan 1976; Stone 1987; Henrickson (MacGillivray et al. 1984).
1981), while others, such as Amarna, indicate At Knossos, while there has been less con-
at least some more general mixing of large tinuous exposure of residential areas, indi-
and small residences (Crocker 1971; Shaw vidual excavations are distributed fairly
1992; Borchardt and Ricke 1980). Such a pat- widely across the site. Close to the palace,
tern can be seen in detail at Pompeii and well-built and elaborately finished houses
Herculaneum, with houses of various sizes stand in isolation, though often fairly closely
inter-mixed, as well as revealing complex his- packed. At the Stratigraphic Museum
tories of changing house divisions and amal- Extension site, small houses are laid-out in
gamations (Wallace-Hadrill 1994; Ling 1997). packed blocks, whereas further out, as with
So, again, we need to work out, for the spe- Hogarth's Houses on Gypsades, houses are
cific culture and communities we are dealing more widely spaced. Finally, the small and
with, what sort of residential organization poorly constructed Acropolis House, seem-
prevailed. ingly stands on the fringe of the community,
There have been too few extensive expo- with no close neighbours.
sures of Neopalatial sites to establish a clear But in trying to understand community
pattern, and it remains difficult to distinguish organization, we are concerned with more
different overall patterns of behaviour than just variations in house size and density.
From Sites to Communities 23
It is clear from excavations at Knossos and some ancient urban sites (e.g. Pracchia et al.
Malia that there were large areas near the 1985; Stone and Zimansky 1994; Millon 1981;
cores of the palatial sites devoted to adminis- Calnek 1976; Laurence 1994; Tosi 1984; Carter
trative and public activities: the palaces 1983:150-70) would appear not to be a strong
themselves, but also neighbouring courts, structural component of Minoan urbanism.
and subsidiary, quasi-palatial or quasi-public What are the implications of the elements
structures. Equally structured space can be of spatial organization which do seem to
seen at smaller centres such as Gournia and exist, for our attempt to estimate the popula-
Ayia Triadha (Figure 2.7). Calculations based tion of Minoan towns? Taking Knossos as an
on Gournia, where we have a good overview example (Figure 2.8), the area immediately
of the use of different parts of the extensively surrounding the palace seems to comprise a
excavated core of the site, as well as an idea mix of elite residences with quasi-official
of its total extent, suggest that something like structures, such as the Northwest Treasure
3.5% of the town was administrative (the House (Evans 1928: 616-25, 637-47), any
palace), 22.5% was public (streets and the Neopalatial precursor to the Arsenal (Evans
town court), and 74% was residential. Rough 1928: 155; 1935: 668-69, 836-37), and the
calculations based on Knossos and Malia, 'grandstands' along the 'Royal Road'
where we also have some idea of the layout (Warren 1994). Further south, an open court
of the core, and estimates of the overall extent lies under the modern coach park
of each site, support broadly comparable (Hutchinson 1950: 210; Warren 1994), and the
divisions. 'Caravanserai' and 'Spring House', whatever
A variety of models has been proposed by their purpose, do not appear to be domestic
urban geographers attempting to understand (Evans 1928: 101-39; Schofield 1996). Of the
the organization of cities, though most are elaborate houses relatively close to the
particularly relevant to industrialized centres palace, the unique reception features of the
(Sjoberg 1960; Fox 1977; Southall 1998; Carter Royal Villa (Evans 1928: 396, 413), the House
1983: 171-83; Marcus 1983). Ancient cities of the Chancel Screen (Evans 1928: 391-96),
such as Pompeii demonstrate some degree of and further out, the House of the High Priest
economic zonation (Wallace-Hadrill 1995; (Evans 1935: 205-15), all suggest residences
Laurence 1994; 1995), though nowhere near for important individuals, as does the Little
as formalized as in later industrialized exam- Palace complex (Evans 1913-14; 1928: 513-44;
ples. While differentiated spaces can readily Hatzaki 1996).
be identified on Minoan sites, probably best Further out, more limited evidence from
illustrated by Protopalatial Malia (Van excavations and soundings would be consis-
Effenterre 1980; Poursat 1988), little formal tent with vaguely defined residential zones,
zoning is obvious. Rather, on the model of the dominated by different types of housing. At
Protopalatial maison-ateliers at Malia the core of the site, a zone of largely elite
(Poursat 1996), the scattered distribution of housing could support a residential density
workshops in the town of Zakros comparable to that of the excavated area at
(Chrysoulaki and Platon 1987), and the evi- Palaikastro - c. 250 persons per hectare. Close
dence for small-scale 'cottage industries' at to the palace, the palace itself and the open
Gournia, Poros and Kommos (Boyd Hawes et public areas and quasi-official structures in
al 1908; Dimopoulou 1998; Shaw 1996), the its immediate vicinity, would probably
larger-scale economic zonation identified at occupy about one-half of that area. In the
From Sites to Communities 25
middle zone, we might anticipate a concen- sites, such as Amnisos (Shaefer 1992),
tration of smaller houses, with a residential Archanes (Sakellarakis and Sapouna-
density comparable to that at Gournia - c. 450 Sakellaraki 1997), Tylissos (Hatzidakis 1934:
people per hectare. Finally, toward the fringe 70-72), and Kastelli Pediadha (Rethimiotakis
of the site, that density would be reduced - 1997), on the order of five to ten ha in size.
probably to half or less. Overall, juggling Cautiously referring to the range of observed
these figures in rough calculations, I would Neopalatial residential densities, we can sug-
suggest an overall population for Neopalatial gest populations of up to a couple of thou-
Knossos of some 14-18,000 individuals sand individuals for such sites. The regional
(Whitelaw 2000; in press). Averaged out over surveys conducted to date indicate that most
the whole area of the site, this would yield a Neopalatial communities were far smaller
global density of about 200-225 individuals than this, generally less than one to two ha,
per hectare. and it seems reasonable to suggest that they
had commensurately smaller populations.
Figure 2.9 Evidence for the extent of better-documented Neopalatial sites. A. Knossos; B. Phaistos; C. Zakros; D. Khania;
E. Kommos; F. Malia; G. Palaikastro; H. Gournia; I. Mochlos; J. Ayia Triadha; K. Myrtos Pyrgos; L. Kastellos
Tzermiado; M. Pseira; N. Petras; O. Tylissos; P. Kastelli Pediadha; Q. Archanes; R. Amnisos; S. Poros.
From Sites to Communities 29
Figure 2.10 Comparative estimates of site size, Late Bronze Age Aegean sites.
Figure 2.11 Eastern Mediterranean centres of the second millennium BC.
From Sites to Communities 31
contemporaries, has often been repeated. He study, rather than relying upon cross-cultural
was only able to make this case by dismissing formulae or specific analogies which are of
out of hand (1972: 238, 242-4) the empirical questionable relevance to Minoan Crete.6
data on which previous estimates of the sizes Finally, in examining different Neopalatial
of sites such as Knossos and Malia were sites, it is useful to recognize the differences
based (Hood 1958; Demargne and Gallet de in residential density which can be docu-
Santerre 1953: pi. 1; van Effenterre and van mented for different sites, rather than sub-
Effenterre 1963: 49-53). Subsequent field- suming them into an overall average density
work has shown these estimates to have been figure. Exploring these local differences in
reasonably accurate (Hood and Smyth 1981; residential practices may help us to recognize
Whitelaw 2000; in press; Muller 1990; 1991; the different characters of individual commu-
1992). It may be that Renfrew's position was nities in Neopalatial Crete, which are likely to
forced on him by his attempt to chart the have had different roles within larger-scale
emergence of states in the Aegean through regional settlement systems. In such a man-
Cycladic data, such that a small site such as ner, we can move beyond the study of indi-
Phylakopi needed to be definable as 'urban' vidual sites in isolation, and through
(Renfrew 1972: 238-40; C. Broodbank pers. integration with the emerging data from
comm.; see Renfrew and Wagstaff 1981 for regional survey, we can put the study of
the idea of Melos as a microcosm for moni- Minoan urbanism into its regional, as well as
toring Aegean-wide processes). developmental context.
In fact, any direct comparison with con-
temporary Eastern Mediterranean centres
(Figure 2.11), emphasizes that the Aegean Acknowledgments
urban communities were not rustic cousins
on the periphery of the civilized world, but I would like to thank the editor for the invita-
the urban centres of states, comparable in tion to participate in the Round Table (I hope
scale and almost certainly complexity, to I've satisfied my very specific brief), and
many of their Eastern contemporaries.5 express my appreciation to him and his col-
Finally, it is worth returning to the method- leagues (staff and students) and the other par-
ological perspective which this paper has ticipants, for making it such an enjoyable and
tried to develop. I have argued that there is effective occasion. Anthony Snodgrass's
considerably more relevant data, both for relaxed approach to chairing was particularly
Neopalatial house size and for community appreciated. Lisa Nevett helpfully commented
organization, than has previously been con- on a draft of the paper. Thanks, as ever, are
sidered in attempts to estimate Neopalatial due to Keith and Nong for their hospitality.
site populations. In addition, I have outlined
an explicit methodology for estimating popu-
lations, developed from a comparatively- Notes
based recognition of some of the factors
1. In estimating the population of Neopalatial Knossos,
affecting residential patterns; these are vari- Evans assumed 8 individuals for each household (1928:
able phenomena which need to be under- 562) while Hood and Smyth assumed 10 (1981: 10),
stood contextually. Such an approach allows though in neither case was the basis for this figure
explained. Recent studies of the demography of the fam-
us to develop an understanding of residential ily in pre-industrial societies support an average of c.
densities which is specific to the culture under four - five individuals per nuclear family (Laslett 1972;
32 Urbanism in the Aegean Bronze Age
Gallant 1991: 11-33). While it has been assumed that Adams and Nissen 1974: 28-30; Watson 1978; Sumner
there would also be servants or slaves resident in a 1979; Kramer 1980; Aurenche 1981; Postgate 1994; Garr
Minoan household (e.g. Marinatos and Betancourt 1995: 1987). It is also worth noting, based on compilations of
593), no concrete argument has been made to support recent urban residential data (Fletcher 1981; 1995; Storey
this, and in any event, it would seem unlikely that this 1997), just how rarely, world-wide, occupation densities
would be relevant for the majority of households within as high as 400 persons per hectare are reached.
a community. Dependent work-groups appear to be doc-
umented in the Linear B texts, and some craftsmen may
have dependent assistants (Chadwick 1976: 79; Lindgren
1973: 36-39; Uchitel 1984; Killer 1988), but as yet no clear Bibliography
evidence for similar individuals has been identified in
the Linear A records for the Neopalatial period, and ser- Adams, E.G.
vants or slaves are not obvious in Minoan figurative art. 1983 The architectural analogue to Hopi social orga-
nization and room use, and implications for
2. Sources for comparative house-size data are as fol- prehistoric northern Southwestern culture.
lows: New Kingdom Egypt: Crocker 1985; Arnold 1996;
American Antiquity 48: 44-61.
Tietze 1996; Mesopotamia: Battini-Villard 1999; Miglus
Adams, R.
1999; Castel 1992; Henrickson 1981; North Syria:
McClellan 1997; South Levant: Foucault-Forest 1996. 1965 Land Behind Baghdad. A History of Settlement on
the Diyala Plain. Chicago: University of Chicago
3. To explore the potential impact of upper floor space, Press.
each house plan was assessed for architectural evidence 1981 Heartland of Cities: Surveys of Ancient Settlement
of clear or likely staircases, and evidence for split-level and Land Use on the Central Flood Plain of the
terracing (particularly at Gournia and Pseira). Adding Euphrates. Chicago: University of Chicago Press.
additional floor space, where there was evidence for Adams, R. and H. Nissen
stairs, spread the house size frequency distribution, but 1972 The Uruk Countryside, 28-30. Chicago: University
did not change its overall shape significantly, of Chicago Press.
Arnold, F.
4. The extent of Phaistos in the Neopalatial period is not
1996 Settlement remains at Lisht-North. In M. Beitak
easy to estimate (La Rosa 1985: 48), though some sort of
(ed.), House and Palace in Ancient Egypt, 13-21.
activity (though not necessarily continuous) over some
Wien: Osterreichische Akademie der
60 ha, can be documented for the Protopalatial period
(Watrous et al. 1993: 225-27). Wissenschaften.
Aurenche, O.
5. The relatively small scale of the Aegean centres has 1981 Essai de demographie archeologique.
recently been emphasized by Manning, in his considera- L'example des villages du Proche Orient
tion of Knossos in the context of Fletcher's communica- ancien. Paleorient 7: 93-105.
tion-based model of the limits to urban growth Battini-Villard, L.
(Manning 2000; Fletcher 1995). However, the value of 1999 L'espace domestique en Mesopotamie de la Hie
such a comparison is unclear, since it completely ignores dynastie d'Ur a I'epocjue paleo-babylonienne. (BAR
the nature of the states involved, by comparing Knossos, IS 767) Oxford: Tempus Reparatum.
the capital of an archaic state, with the imperial capitals Bender, D.
of empires, which were, not surprisingly, often much 1967 A refinement of the concept of household: fam-
larger, spatially and demographically. ilies, co-residence, and domestic functions.
American Anthropologist 69: 493-504.
6. Following Frankfort's comparison of the residential Betancourt, P. and C. Davaras
densities of modern Aleppo and Damascus with esti- 1999 Pseira IV. Minoan Buildings in Areas B, C, D and
mates for ancient Ur, Tell Asmar and Kafaje (1950:103), a
F. Philadelphia: The University Museum
residential density of 400 persons per hectare has been
Blanton, R.
widely used to estimate populations for Near Eastern
and East Mediterranean sites, and occasionally cited by 1993 Houses and Households: a Comparative Study.
analogy for the Aegean, but rarely with any argument of New York: Plenum.
relevance (Renfrew 1972: 251; Wiener 1990: 131-33). By Borchardt, L. and H. Ricke
and large, Frankfort's figure is no longer used by 1980 Die Wohnhauser in Tell el-Amarna. Berlin: Mann.
Mesopotamian archaeologists (cf. Adams 1981: 349-50; Boyd Hawes, H., B. Williams, R. Seager, and E. Hall
Johnson 1987: 109; Wilkinson 2000: 247-49), as the com- 1908 Gournia, Vasiliki and Other Prehistoric Sites on the
plex patterns of variation in residential behaviour have Isthmus of Hierapetra, Crete. Philadelphia:
become increasingly appreciated (Adams 1965: 123; American Exploration Society.
From Sites to Communities 33
Renfrew, C. Sjoberg, G.
1972 The Emergence of Civilization, The Ci/dades and 1960 The Preindustrial City. Past and Present. London:
the Aegean in the Third Millennium B.C. London: Collier Macmillan.
Methuen. Sommer, R.
Rethimiotakis, G. 1969 Personal Space: The Behavioral Basis of Design.
1997 To minoiko 'kentriko ktirio' sto Kastelli Englewood Cliffs: Prentice Hall.
Pediadas. A Delt 47-48 (1992-93) A'; 29-64. Southall, A.
Rohn, A. 1998 The City in Time and Space. Cambridge:
1965 Postulation of socio-economic groups from Cambridge University Press.
archaeological evidence. In D. Osborne (ed.)/ Stone, E.
Contributions of the Wetherill Mesa Archaeological 1981 Texts, architecture and ethnographic analogy:
Project, 65-69. Washington D.C.: Society for patterns of residence in Old Babylonian
American Archaeology. Nippur. Iraq 43: 19-33.
Russell, J. 1987 Nippur Neighborhoods. Chicago: The Oriental
1958 Late ancient and medieval populations. Institute.
Transactions of the American Philosophical Society 1996 Houses, households and neighborhoods in the
48(3): 1-101. Old Babylonian period: the role of extended fam-
Sakellarakis, Y. and E. Sapouna-Sakellaraki ilies. In M. Mellink, J. de Roos, J. Roodenberg and
1997 Archanes. Minoan Crete in a New Light. Athens: K. Veenhof (eds.), Houses and Households in Ancient
Ammos Publications. Mesopotamia, 229-35. Istanbul: Nederlands
Sanders, W., J. Parsons and R. Santley Historisch-Archaeologischlnstituut.
Stone, E. and P. Zimansky
1979 The Basin of Mexico. Ecological Processes in the
1994 The Tell Abu Duwari Project, 1988-1990.
Evolution of a Civilization. New York: Academic
Journal of Field Archaeology 21: 437-55.
Press.
Storey, G.
Schacht, R.
1997 The population of ancient Rome. Antiquity 71:
1981 Estimating past population trends. Annual
966-78.
Review of Anthropology 10: 119-40.
Sumner, W.
Schaefer, J.
1979 Estimating population by analogy: an example.
1992 Amnisos. Berlin: Gebr. Mann.
In C. Kramer (ed.), Ethnoarchaeology: Implications
Schofield, E.
of Ethnography for Archaeology, 164-74. New
1996 Wash and brush up at the 'traveler's rest': the York: Columbia University Press.
Caravanserai reconsidered. In D. Evely, I. Tietze, C.
Lemos and S. Sherratt (eds.), Minotaur and 1996 Amarna, Wohn- und Lebensverhaltnisse in
Centaur. Studies in the Archaeology of Crete and einer aegyptischen Stadt. In M. Beitak (ed.),
Euboea Presented to Mervyn Popham, 27-33. House and Palace in Ancient Egypt, 231-37. Wien:
(BAR IS 638) Oxford: Tempus Reparatum. Osterreichische Akademie der Wissenschaften.
Severy, L.J. Tosi, M.
1979 Individual variation in perception, adaptation, 1984 The notion of craft specialization and its repre-
and consequent crowding. In M. Gurkaynak sentation in the archaeological record of early
and W LeCompte (eds.), Human Consequences states in the Tauranian Basin. In M. Spriggs
of Crowding, 57-65. London: Plenum Press. (ed.), Marxist Perspectives in Archaeology, 22-52.
Shaw, I. Cambridge: Cambridge University Press.
1992 Ideal homes in ancient Egypt: the archaeology Tuan, Y-F.
of social aspiration. CAJ 2: 147-66. 1977 Space and Place: the Perspective of Experience.
Shaw, J. London: Edward Arnold.
1996 Domestic economy and site development. In J. Uchitel, A.
Shaw and M. Shaw (eds.), Kommos I. The 1984 Women at work. Pylos and Knossos, Lagash
Kommos Region and Houses of the Minoan Town. and Ur. Historia 33: 257-82.
Part 2: The Minoan Hilltop and Hillside Houses, Van Effenterre, H.
379-400. Princeton: Princeton University Press. 1980 Le Palais de Malia et la Cite Minoenne. (Incunabula
Shiloh, Y. Graeca 76) Rome: Edizioni dell'Ateneo.
1980 The population of Iron Age Palestine in the van Effenterre, H. and M. van Effenterre
light of a sample analysis of urban plans, areas 1963 Fouilles Executees a Malia. Etude du Site
and population density. Bulletin of the American (1956-1957). (Etude Cretoises 13) Paris: Paul
School of Oriental Research 239: 25-35. Geuthner.
From Sites to Communities 37
1976 Fouilles Executees a Malia. Exploration des 139-88. (Ethnoarchaeological Series 1) Ann
Maisons et Quartiers d'Habitation (1956-1960). Arbor: International Monographs in Prehistory.
Volume 4. (Etude Cretoises 22) Paris: Paul 1994 Order without architecture: functional, social
Geuthner. and symbolic factors in hunter-gatherer settle-
van de Mieroop ment organisation. In M. Parker-Pearson and
1999 The Ancient Mesopotamian City. Oxford: Oxford C. Richards (eds.), Architecture and Order,
University Press. 217-43. London: Routledge.
Wallace-Hadrill, A. 2000 Beyond the palace: a century of investigation at
1994 Houses and Society in Pompeii and Herculaneum. Europe's oldest city, (abstract) BICS 44.
London: Princeton University Press. in press Estimating the population of Neopalatial
1995 In T. Cornell and K. Lomas (eds.), Urban Society Knossos. In G. Cadogan, E. Hatzaki and A.
in Roman Italy, 39-62. London: UCL Press. Vasilakis (eds.), Knossos: Palace, City and State.
Warren, P.M. London: The British School at Athens.
1981 Knossos: Stratigraphical Museum excavations, Wiener, M.
1978-80. Part I. AR 27: 73-92. 1990 The isles of Crete? The Minoan thalassocracy
1994 The Minoan roads of Knossos. In D. Evely, H. revisited. In D. Hardy, C. Doumas, J.
Hughes-Brock and N. Momigliano (eds.), Sakellarakis and P. Warren (eds.), Thera and the
Knossos. A Labyrinth of History: 189-212. Aegean World III. Vol. 1. Archaeology, 128-61.
Oxford: British School at Athens. London: Thera Foundation.
Wilcox, D.R.
Watrous, L.V., et al
1975 A strategy for perceiving social groups in
1993 A survey of the western Mesara plain in Crete:
puebloan sites. Fieldiana Anthropology 65:120-159.
preliminary report. Hesperia 62: 191-248.
Wilk, R.
Watson, M.O.
1983 Little house in the jungle: the causes of variation
1970 Proxemic Behaviour: a Cross-Cultural Study. The
in household size among modern Kekchi Maya.
Hague: Mouton.
Journal of Anthropological Archaeology 2: 99-116.
Watson, P.J.
Wilk, R. and R. Netting
1978 Architectural differentiation in some Near
1984 Households: changing forms and functions. In
Eastern communities, prehistoric and contem- R. Netting, R. Wilk and E. Arnold (eds.),
porary. In C. Redman, M. Berman, E. Curtin, Households. Comparative and Historical Studies of
W. Langhorne Jr, N. Versaggi and J. Wanser the Domestic Group, 1-28. London: University of
(eds.), Social Archaeology: Beyond Subsistence and California Press.
Dating, 131-57. New York: Academic Press. Wilkinson, T.
Whitelaw, T. 2000 Regional approaches to Mesopotamian archaeol-
1983 People and space in hunter-gatherer camps: A ogy: the contribution of archaeological surveys.
generalising approach in ethnoarchaeology. Journal of Archaeological Research 8: 219-67.
Archaeological Review from Cambridge 2(2): 48-66. Wooley C.L. and M. Mallowan
1983 The settlement at Fournou Korifi, Myrtos, and 1976 Ur Excavations. Volume VII: The Old Babylonian
aspects of Early Minoan social organization. In Period. London: British Museum Publications
O. Kryszkowska and L. Nixon (eds.), Minoan Ltd.
Society, 323-45. Bristol: Bristol Classical Press. Yanagisako, S.
1989 The Social Organisation of Space in Hunter- 1979 Family and household: the analysis of domes-
Gatherer Communities. Implications for Social tic groups. Annual Review of Anthropology 8:
Inference in Archaeology. PhD Dissertation, 161-205.
Cambridge University. Zubrow, E.
1991 Some dimensions of variability in the social 1974 Population, Contact, and Climate in the New
organisation of community space among for- Mexican Pueblos. (Anthropological Papers of
agers. In C. Gamble and W. Boismier (eds.), the University of Arizona 24) Tucson:
Ethnoarchaeological Approaches to Mobile Campsites, University of Arizona Press.
3
Keith Branigan
MINOAN TOWNS
reports suggest that sites which we would Nevertheless, with only about fifteen
classify as even potentially 'urban' are very urban sites even sampled by excavation, it is
few and far between. But two sites found in obviously difficult to confidently present an
recent surveys in eastern Crete suggest that urban hierarchy, and tempting to do so solely
such sites do remain to be found. The site of on the grounds of size. Rackham and Moody
Plakoures, between Pacheia Ammos and (1996: 89) took this course and concluded
Kavousi at 3-4 ha in size ranks above or along- there were four levels in the Minoan settle-
side Mochlos and Gournia (Watrous and ment hierarchy - hamlets, villages, small
Blitzer 1999: 906). Around the Stavromenos towns and large towns. On present evidence
hill at Katelionas is a settlement of about 5ha, the estimated sizes of Minoan towns do
overlooked by a villa-like structure on the hill. indeed fall into two very clear groups - those
It seems likely that the number of poten- above 20ha (Knossos, Malia, Palaikastro,
tially 'urban' sites identified in Minoan Phaistos, Galatas, and probably Monastiraki
Crete may slowly increase from around the and Khania, and those below lOha. In numer-
mid-twenties to the mid thirties. The first ical terms, the 'small' towns outnumber the
point we might make therefore is that Targe' towns about 3:1 at present, although it
Minoan Crete had few towns, perhaps is this group that is most likely to be
between two and three dozen, an average of increased by new discoveries and the true
one every 200 - 250km2 (Figure 3.1). But ratio may have been more like 5:1. In second
given the very mountainous terrain of much millennium Palestine, where there is a similar
of the island this is not surprising, and in (but less abruptly defined) dichotomy
fact it compares closely to the average den- between large towns and small towns, the
sity of towns in LBA Palestine, where there latter outnumber the former by about 6:1.
are about 70 towns in an area about twice Establishing a hierarchy by size alone, how-
the size of Crete. ever, would be to ignore the other defining
Figure 3.1 The location of certain and probable Minoan towns.
features of towns as outlined by Wirth and Crete is tri-partite like that proposed by Ilan
Trigger. Taking into account the evidence for for MBA Palestine (1996: 305). Ilan identifies
'service' provision, and particularly for Regional Centres and Gateways as the top
administrative services, we have to suggest a level, and then on a second level places both
slightly more complex structure. At the top sub-regional centres and loci of specialist
are the large towns with major administrative production or services. In fact, Ilan goes far-
/ceremonial buildings. This group would in ther by sub-dividing his top tier sites into
fact be the same as the Targe' towns of the two four groups - 1st, 2nd and 3rd order gate-
level hierarchy, with the possible exception of ways (all coastal sites or in the Rift Valley
Palaikastro which has stubbornly refused to controlling trade routes) and regional cen-
yield evidence of a 'palace', but which never- tres. His first order gateway is Hazor, which
theless sports paved principal roads with at 80ha is about the size of Knossos, and his
raised ribs, flanked by substantial houses second order gateways include Ashkelon
with ashlar facades, like the very biggest and which at 50ha is a fair match for Malia. It has
best palatial centres. The second level in this been common in Crete to separate Knossos
hierarchy includes the small towns with sub- from the remainder of the large urban sites
stantial central buildings which in their archi- and place it at the top of the Minoan
tectural features and furnishings emulate the hierarchy in the same way that Ilan places
principal palaces, of which the most obvious Hazor at the top of the Palestinian pyramid.
at present are Arkhanes, Gournia, and Petras. But with the evidence presently to hand we
A third level can be identified in the form of would do well to heed John Cherry's caution
small towns probably offering specialized (1984: 24) against a rush to judgement on this
services related not directly to administration matter.
so much as to marketing and exchange - Before leaving the topic of the urban hier-
coastal settlements such as Kommos, Mochlos, archy, we might briefly compare the size of
Pseira, Priniatiko Pyrgos and Zakro. Some of Minoan towns to that of towns in the con-
these towns have one or more buildings with temporary civilisations of Turkey and
ashlar masonry and a level of accommoda- Mesopotamia. It has become commonplace to
tion above that of the other buildings in the regard the towns of Minoan Crete as min-
settlement. A comparable group of small nows compared to those of the Near East
towns with one or more superior buildings (Renfrew 1972: 244; Dickinson 1994: 51).
seems to have existed inland too. Tylissos Undoubtedly Uruk, at 450ha, is in a class of
and Katelionas appear to be examples of this its own, but Nippur at 135ha and Boghazkoy
type. Whether there is a fourth level of urban at 120-160ha are also clearly much larger
settlement - small inland towns offering towns than Knossos, let alone the other
social and economic services but without Targe' Minoan towns. But thereafter, Minoan
any elite building as their focus is at present large towns are a match for those of the Near
debateable. The newly discovered site at Eastern states (Table 3.2). Ur at 60ha and
Plakoures might be an example. Without Khafaji at 40ha, Kultepe at 50ha, and Quatna
excavation such sites will be difficult to dis- at 65ha and Mari at 60ha are in the same
tinguish from a large village; even with exca- order of magnitude as Knossos, Malia, and
vation the distinction might be difficult to perhaps Palaikastro and Phaistos. It is also
establish. In essence then, the suggested salutary to note that two of the most impor-
hierarchy of urban settlement for Minoan tant trading centres (they are often called
42 Urbanism in the Aegean Bronze Age
cities) of the ancient Near East, Ugarit and Malia and Palaikastro, and Megiddo and
Byblos, were both much smaller than the four Jericho to Gournia and Arkhanes. Dever
or five largest Minoan towns. Small towns (1997: 180) has linked the smaller scale of
comparable to those of Crete also exist in urban settlements in Palestine to the 'much
these civilisations - Beycesultan (6ha), Alaja smaller-scale landscape' which is much more
(4ha), and even Tel Harmal (Syria, 1.7ha). If fragmented and less productive than Meso-
there is a distinction at this end of the hierar- potamia and Syria, and allows the threshold
chy it seems to be that the small towns make for 'truly urban configurations' to be low-
up a smaller percentage of the urban hierar- ered to about 6ha. Whether the similar scale
chy in Mesopotamia, Syria and Turkey and of urban settlements reflects a broadly simi-
that they are mostly rather larger than the lar geo-political situation in MBA Crete and
small Minoan towns. MBA Palestine is an interesting but arguable
In this respect, although Minoan Crete proposition.
might be said to stand comparison alongside
the imperial states of the contemporary Near
East, the closest comparisons in terms of Urban Organisation
urban hierarchy still lie with the towns of sec-
ond millennium Palestine. Here, Hazor com- The most striking feature of Minoan towns
pares to Knossos, Ashkelon and Beth Shan to when viewed in their contemporary context
Table 3.2 Comparative table of urban settlement size in Crete and the Near East in the 2nd millennium BC. (Sizes in
hectares)
URUK 450
NIPPUR 135
HATTUSAS 120
KNOSSOS 75 HAZOR 80
QUATNA 65
MARI 60 UR 60
MALIA 50 ASHKELON 50 KULTEPE 50
PHALISTOS 40 BETH SHAN 40 KAHFAJI 40
KABRI 40
PALAIK'o 30
GALATAS 25
MONAST'KI 20 RASSHAMRA 21
T. AJJUL 15
BYBLOS 9
ZAKRO 8
ARKHANES 8 LACHISH 8
POROS 8
BEYCE 6
GOURNIA 5
KOMMOS 4.5 MEGIDDO 4.5 ALAJA 4
TYLISSOS 4 JERICHO 3.8
MOCHLOS 2.5
A. TRIADHA 2.5
PETRAS 2.5
T. HARMAL 1.7
PSEIRA 1.5
Aspects ofMinoan Urbanism 43
theocracy behaved very differently to its con- appear to have an overwhelming or forma-
temporaries in the Near East. Where they tive influence on the shape and character of
expended much energy on the construction the town. Galatas may prove to be an excep-
of impressive temples which must have been tion to this rule. Nor do the palaces take up
urban focal points, the Minoan elite invested excessive amounts of space. The palaces at
much more modest amounts of energy Knossos, Malia, and Phaistos, all seem to
mostly in building and equipping small occupy about 2% of the total area of the
exclusive shrines within the palace walls or towns in which they stand. At Gournia,
in supporting remote peak-top sanctuaries Petras and Zakro the 'palaces' take up 4-5%
with clearly rural affinities. Apart from the of the urban space. These modest claims on
palaces, Minoan towns were not dominated urban space are matched by the relative
by the architecture of power, although pro- openness of the palaces; unlike many of their
topalatial Malia has a complex of major struc- counterparts in Anatolia and the Near East,
tures north of the site of the palace. Indeed, Minoan palaces do not stand inside walled
even the palaces are less dominant than we enclosures with fortified entrances. In this
might expect. They stand at the centre of the sense, at least, Minoan towns were more
settlement, but since in almost every case open and less restricted than most second
they seem to be implanted into an existing, if millennium BC settlements in the east
smaller, pre-palatial settlement, they do not Mediterranean.
Figure 3.2 A paved 'principal way' with raised central rib approaching the West Court at Phaistos.
Aspects ofMinoan Urbanism 45
Neither were they formally organized in many other purely domestic structures.
any way which is apparent in the archaeolog- There is certainly no reason to think that cer-
ical record. There were of course paved or tain quarters of Minoan towns were occupied
cobbled streets and lanes, some of which by anything like the tradesmen's guilds of
were flanked by clay water pipes. In general, medieval European cities.
however, the streets form no coherent pattern In analysing pre-industrial towns in west-
although some with paved surfaces and ern Europe Paul White (1984) identified three
raised central ribs were clearly identified as different types of town or city, which were
'principal ways' which usually begin or end characterized by different spatial organisa-
at the palace (Figure 3.2). In the case of tion. The Feudal City was divided into dis-
Knossos, Warren (1994: 201) has tentatively tricts, each occupied by both artisans and
argued that something approaching a road rural workers belonging to a faction led by
grid may have been laid out to the west of the wealthy rural landlords. Each district had its
palace in MM.IB or MM.IIA, but the evidence own social focus or 'palace'. In the Merchant
at present is inconclusive. Similarly, although City where craft and commerce were all
parts of the street system at Palaikastro sug- important and links with the countryside
gest something approaching a planned sys- were insignificant, the town was divided into
tem, other parts seem to lose direction and areas occupied by guilds and each serviced
regularity of width. The palaces lie at the by its own guildhall. In the Absolute City,
town centre and around them are found, for where religious and political power were in
the most part, the larger, better built and bet- the hands of a single family, the central area
ter furnished private houses. At Knossos, of the city was dominant and the elite clus-
Malia and Palaikastro the smallest and most tered around the rulers palace. The rest of the
simply built houses are found on the periph- mixed population of artisans and rural work-
ery. Although I have myself toyed with the ers lived around the periphery. In terms of
idea that there may have been quarters of a spatial organization the Minoan towns
town given over to craft activity (Branigan appear to most closely resemble the Absolute
1972: 758) the evidence for this remains tan- City in White's tripartite scheme. Whether
talisingly slight in Minoan towns. Perhaps this suggests similarities in the power struc-
the best evidence for the concentration of tures which lay behind the spatial patterning
craft activities in one area is to be found at is open to debate.
Malia and Mochlos. In quartier Mu at Malia
five or six workshops existed in a cluster,
making pottery, metalwork, stone vases, seals Demographic Significance
and perhaps bone artifacts. At Mochlos, the
discovery of buildings devoted to metal- Because Minoan towns reveal little evidence
working, pottery making and stone vase of anything approaching 'town planning' and
manufacture on the mainland contrasts with few monumental public buildings beyond the
the apparently entirely domestic nature of palaces, because they are not enclosed by
the houses on the island. But it would need impressive defences, and because beyond the
more extensive excavations on the mainland five or six largest towns we are looking at
to establish whether there was an industrial very small urban centres (between 2ha and
quarter here or just two or three buildings 8ha in area) it is tempting to regard the towns
involved in craft activity, surrounded by of Minoan Crete as under-developed and
46 Urbanism in the Aegean Bronze Age
may be able to arrive at some respectable estimated total area of the town. With a palace
generalized figure. Certainly there seems to of 8000m2 at Phaistos, we might on this basis
be a shift towards nucleated settlement from expect a town of about 40ha. In the case of
the EBA to the M/LBA, and from Vance Khania I have assumed it is a major settlement
Watrous' preliminary Mesara report (1993: but belongs in the lower echelons of this
227) villages and hamlets seem to outnumber group. I have then taken a figure of 4ha
farmsteads in the palatial period. I take the (slightly lower than the average of the known
balance as 2: 1 in favour of nucleated settle- small towns, which is around 4.5ha) and mul-
ments in my guesstimate, and so assume tiplied this by a total of 30 minor towns. This
about 800 villages and hamlets and 400 farm- assumes that there are still small town sites to
steads. I have then taken an average number identify, a justifiable assumption in that four of
of households for hamlets (with perhaps only the small towns on the list have only been dis-
5-10 families) and villages (with perhaps covered or their full extent only established
15-40 families) together as about 20. within the last ten years. Some possible future
Taking the widely agreed average figure of candidates for 'urban' status are listed in Table
5 persons per peasant family, this provides a 3.1. The net result of these calculations is a
total rural population of something over total of about 390ha of urban settlement.
82,000 people (Table 3.3). To convert this into an estimate of the total
To add to, and set against this figure, we urban population we have to estimate the
have to estimate the urban population. Here I likely number of inhabitants per hectare.
have taken the estimated size of the known Renfrew suggested a figure of 300 persons per
major settlements. The size of Phaistos is diffi- hectare for Minoan Bronze Age towns
cult to estimate. Watrous (1993: 225) has sug- (assuming a figure of 400 for Near Eastern
gested that it might be as large as 150ha, but towns), and Todd Whitelaw's differently
this assumes that widely spread deposits to based calculations for EM sites produced a
the west and south of the hill on which the figure equivalent to about 350 per ha
palace sits were part of a single continuous (Whitelaw, 1985: 337-39). However popula-
nucleated settlement. It seems unlikely that tion densities in Near Eastern towns have
Phaistos would be twice the size of Knossos been substantially revised downwards since
and three times that of Malia. There is nothing
Table 3.4 An estimate of the urban population of
in the size and architecture of the palace, nor palatial Crete
in the archival material to suggest that
URBAN POPULATION
Phaistos outshone Knossos and Malia so dra-
matically. A possible indication of its compar- Total area of major towns (inc average
allocated to Khania) 270ha
ative size is its provision of 'public' storage in
Total area of 30 minor towns with an
the koulouras along the edge of the west court. Average area of 4ha 120ha
The four koulouras at Phaistos could store Total area of urban settlement 390ha
about 75 tons of grain, compared to the 250 Urban population if average population
Density of 200 people per ha: 78,000
tons which could be stored in the contempo-
estimated total population 160,000
rary kourlouras at Knossos. Another possible
comparative indicator is the relative size of the Alternative scenario
Av urban density of 150 people per ha:
palace. At Knossos and Malia - the other Urban population: 58,500
major palatial towns of which we know much estimated total population 140,500
at all - the palaces take up about 2% of the (of which 42% town-based and 58% rural).
48 Urbanism in the Aegean Bronze Age
1972, to figures of around 150-200 per hectare proportion. We might compare it to Roman
(summarised in Manning 1999). At the same Britain - with its 100-120 towns - where the
time, a multi-factored approach to estimating urban population is widely agreed to have
both the population of Minoan Crete as a formed somewhere between 5 and 10% of the
whole and Knossos and Malia in particular total population (e.g. Millett 1990: 185).
(Firth 1995) has proposed revisions to both Interestingly, it is much closer to the situation
sets of figures. Firth has suggested (1995: in second millennium BC Palestine, where
35-56) that Renfrew's figures could be 'recali- Kempinski (1992) has proposed a total popu-
brated' by reference to the Minnesota lation of about 200,000 people, of which 60%
Messenia Expedition survey results, and that are urban and 40% rural.
the population of LM.III Crete on this basis Such a high proportion of town-dwellers
could be placed at around 110,000. He further raises all sorts of questions and speculations
argues (1995: 48-49) that suggested popula- for which we have space to mention only a
tion figures of 12,000 for Knossos and 10,000 few. Almost certainly a good number of town-
for Malia would have been unsustainable in dwellers must have been involved in agricul-
terms of basic food supplies in the Bronze tural production in the land around the
Age. If we apply the figure of 150 per ha to towns. This seems to be confirmed by excava-
our total area of urban settlement, we arrive at tions in the towns themselves which reveal
a figure of about 60,000 town dwellers in pala- only small numbers of houses where craft
tial Crete, with perhaps 11,000 at Knossos and production may have had anything more
7500 at Malia. than a casual role, and also the frequent
On the other hand, detailed study and sur- appearance of agricultural installations
vey at Pseira has suggested a total of about 60 (mostly grape or olive presses). Equally, the
houses in this small harbour town, which bulk of the work-force in the palaces -
might be translated into a population of whether clerks and administrators or menial
about 300 persons. Pseira probably occupied servants - probably lived in the town-houses
about 1.5ha of space, so that the population rather than the palace itself (again, a proposi-
density here might have been nearer 200 per tion supported by the small areas of domestic
hectare than 150. If that figure is used for occupation identified in the palaces).
Crete as a whole, then we would have an esti- It seems highly unlikely that the urban
mated urban population of around 80,000 populations could be supported entirely off
(Table 3.4). the produce of the land immediately around
There are obviously all sorts of caveats and the towns however, particularly in the case of
arguments one could introduce into these cal- the major towns (see Firth 1995: 48-49). The
culations, but even if (for example) we have reliance of the urban centres on the rural pop-
underestimated the rural population by a fac- ulation must have been even greater than we
tor of 30%, we are still left with a palatial have perhaps assumed in the past. Such
Crete where a very significant proportion of heavy dependence must have been reflected
the population lived in towns (the bulk of in the way in which the Minoan states orga-
them in just half a dozen large towns). The nized and controlled their hinterlands. The
percentage of 'town dwellers' would be relationship between town and country,
somewhere between about 40% at the low urban centre and hinterland, must have been
end of the calculations and up to 50% at close and crucial and fully justifies the atten-
the high end. This is a remarkably high tion devoted to it elsewhere in this volume.
Aspects ofMinoan Urbanism 49
Ucko, P., R. Tringham and G. Dimbleby (eds.) Watrous, L.V. and H. Blitzer
1972 Man, Settlement and Urbanism. London: 1999 The region of Gournia in the Neopalatial
Duckworth. period. In P. Bettancourt et al. (eds.), Meletemata
Wagstaff, M. and J. Cherry II. (Aegaeum 20), 905-10. Liege: University of
1982 Settlement and population change. In C. Liege.
Renfrew and M. Wagstaff (eds.), An Island Polity. White, P.
The Archaeology of Exploitation in Melos, 136-55. 1984 The West European City: A Social Geography.
Cambridge: Cambridge University Press. London: Longman.
Warren, P. Whitelaw, T.
1994 The Minoan roads of Knossos. In D. Evely, H. 1985 The settlement at Fournou Korifi, Myrtos and
Hughes-Brock, and N. Momigliano (eds.), aspects of Early Minoan Social Organisation. In
Knossos: A Labyrinth of History, 189-210. O. Kzrzyszkowska and L. Nixon (eds.), Minoan
London: British School at Athens. Society, 323^5. Bristol: Bristol Classical Press.
Watrous, L.V. et al. Wirth, L
1993 A survey of the western Mesara plain in Crete: 1938 Urbanism as a way of life. American J Sociology
Preliminary report. Hesperia 62,ii: 191-248. 44: 8.
4
Table 4.1 Surveys on Crete (partly based on Moody, Nixon, Price and Rackham 1998: 88; "indicates provisional data;
Knossos and Palaikastro are the only two as yet surveyed settlements; totals do not include Kythera)13.
1,100 km2 has been or is being surveyed, i.e. different than the general impression pro-
c. 12.5% and, as far as more or less published vided by Figure 4.2, with serious drops dur-
surveys are concerned, we only know the set- ing the 'black holes', followed immediately
tlement history of an area of about 437 km2 afterwards by a considerable upsurge.
that is, after all, about 5% of its surface. If we There is another drawback, however, since
take into account that 61% of the island is most of the older surveys have concentrated
between 0 and 400 m asl (with 26.6% between on areas that seem peripheral to the develop-
400 and 800 m asl, 12.3% over 800 m asl) we ment in the palatial centres. This is now being
observe that, relatively speaking, the Sitia remedied with the Western Mesara and Malia
area has more land below 800 m asl than the surveys. The great unknown is the Knossos
other three modern provinces. Most of the area: although the 'survey' by Hood (1958),
surveys indeed relate to land below 800 m and the updated Hood and Smyth (1981) ver-
(i.e. 87% of the island) and it is indeed doubt- sion, can give us an idea about the general
ful that, apart from some highland plateaux, trend of the settlement proper, we desperately
occupation was anything more than sparse in need information on the wider Knossos area
the mountains, which results in a somewhat before construction destroys all the evidence.
more positive picture'. The Sphakia survey Given all these limitations, it is obvious that
has, however, found traces of settlement at an speculations on modes of regional integration
altitude of +1000 m2 in the White Mountains, will need to be properly scrutinized.
which may imply that, in certain periods, the Whatever their deficiencies, however, surveys
highland formed an effective, integrated part remain the best way of obtaining a general
of some of the lowland polities. In our mod- idea about relative demographic changes, site
ern study area of Palaikastro this is certainly trajectories and diachronic relationships
still the case and the people of the plain use between cities and countryside.
the highland plateau of Magasa (situated at This is not the place to repeat the pros and
+550m) for a multitude of purposes3. cons of the unitary or peer polity model on
Aside from the fact that the picture pre- Neopalatial Crete or, if you wish, the exis-
sented here (Figure 4.2) is based on an unrep- tence at one or more periods of a single inte-
resentative sample, there are other problems grated political system as opposed to a
with the data. Degree of area coverage, site multitude of regional units (Cherry 1986;
determination, chronological attribution, Driessen and Macdonald 1998; Schoep 1999).
changes within a chronological phase and The present administrative organisation of
site extent (as well as their reciprocal influ- the island is largely based on that of the
ences) and a number of other features could Venetians who themselves took over the epis-
not always be defined properly in the various copal division of Late Roman times (Van
reports4. Most importantly, there are still Spitael 1981: 9; Bennet 1990: 205, fig. 4). It has
some 'black holes': EM III, MM III and LM II. to be admitted that the geography as well as
Although sites attributed to EM III and MM the elongated shape of the island begs for
III are mostly included in Figure 4.2, there such a division and, as Bennet (1990: 207) has
can be no doubt that something was happen- noted, it is indeed singularly difficult to
ing on the island in these respective administer the island from a single capital
'Intermediate Phases' that still eludes us (see except if this lies outside the island, such as
e.g. Watrous 1994; Haggis 1999). The result- Rome, Venice, Constantinople, Istanbul or
ing site curve, however, could look entirely Athens. Our contemporary (i.e. Bronze Age)
54 Urbanism in the Aegean Bronze Age
Figure 4.2 Crete: Settlement patterns from survey data. Column A = Neo/EMI. B = Prepalatial. C = Protopalatial.
D = Neopalatial. E = Postpalatial.
History and Hierarchy 55
written sources are not of much help either. clear-cut reconstruction. Zakros, for instance,
Near Eastern archives mention the geograph- has a larger court-centred building than
ical circumscription of the people Keftiu or Petras but both share other functions such as
Kaptara, for example, which can be compared storage and administration. Does size matter
with words such as Alashiya or Akhiyawa. and was Zakros a higher order centre than
Only twice, the ruler of the Keftiu is explicitly Petras? If so, what about Palaikastro, which
mentioned and I have suggested that this has as yet not revealed such a central build-
may reflect a short period of political unifica- ing but which in terms of the quality of
tion in LMII-IIIA1 (Driessen in press). I leave urbanization surpasses both Zakros and
aside our Cretan documentary evidence since Petras (Cunningham, this volume)? It is obvi-
this has recently been assessed by Use Schoep ous that only new excavations and surveys
(1999) who believes there is no conclusive can and will help to adjust this impression.
evidence for an overall administrative appa- It may therefore be better to make a tabula
ratus, at least during LM IB, corroborating rasa of all established ideas on Minoan state
some of the ideas expressed in Driessen and systems, largely a legacy of Sir Arthur Evans,
Macdonald (1998). We are left with the rest of and conduct a new exercise by looking at set-
our material evidence and to see whether or tlement distributions, their number and
not it reflects signs of integration. In the past, extent and this for a specific reason. The def-
architecture seems to have provided the best inition of the state, as correctly stressed by
indication (Driessen 1989) and indeed ashlar Flannery (1998: 15), should remain a task for
buildings obviously play a role in the recon- anthropologists and political scientists but
struction of the political landscape not only we can at least attempt to define its archaeo-
because of the elite message expressed by this logical correlates (cf. Cherry 1986; 1987). One
feature of conspicuous consumption but also of the essential characteristics of a state,
because they often repeat specific features of whether you call it pristine, early, archaic or
plan, design, construction and decoration whatever, is a variety of levels via which
(Driessen 1999:122). Crete is fortunate in con- goods are mobilized towards the centre and
serving a large sample of ashlar buildings thus a spatial hierarchy of settlement.
that may be identified as public because of 'Chiefdoms' are assumed to have two or a
scale, plan, location and other features. Many maximum of three levels, states at least four:
of these are court-centred and we are used to cities, towns, large villages and small vil-
dub them "palaces" although I prefer the lages. Although somewhat artificial, it pro-
term 'court-centred buildings'. These display vides a handy tool for a first explanatory
a functional and stylistic consistency both in attempt. The more levels we can detect, the
smaller and larger settlements. This could be more developed and complex the hierarchy
an argument in favour of a single integrated seems to have been. These are settlement
system since other, so-called "archaic states" hierarchies, however, and not administrative
(Marcus and Feinman 1998: 4-5) show simi- hierarchies which are much more difficult to
larities between the paramount and subordi- detect archaeologically on the terrain than
nate centres, because the lower order centres community sizes. There is a caveat, however.
would repeat, at a less grandiose scale, the Certain archaeological data, as we have tried
functions exerted by the capital. Such a sce- to suggest in the Troubled Island (Driessen and
nario would perhaps work for Crete but the Macdonald 1998), can reflect sudden changes
many gaps in our evidence do not allow a in political organization. Settlements and
56 Urbanism in the Aegean Bronze Age
architecture are much more static and will Greek (esp. S. Marinatos, N. Platon, C.
only be affected if the changes are really Davaras)19 and foreign archaeologists (A.J.
influencing social and political organization. Evans [1896], J.D.S. Pendlebury [1939;
Moreover, we should not expect a linear Pendlebury et al. 1934], T. Wroncka [1959], P.
development. It should be proven and not Faure [e.g. 1956; 1958; 1960; 1962] and K.
assumed that a polity, once it achieves state- Nowicki [e.g. 1991; 1992]) or those collected
hood, will remain static. The possibility by the Minoan Roads Project (Tzedakis et al
should be left open that a polity can demote 1989; 1990) and others. All sites located dur-
to a lower level of political development or ing these various explorations should, how-
be incorporated into a larger frame. I in fact ever, be included in any proper analysis and it
suspect that some of the features that will be is hoped that, when the Gazetteer of Bronze
highlighted in this paper are a result of Age Sites on Crete finally sees the daylight, it
demotion, breakdown, incorporation, reinte- will provide us with a better tool to proceed.
gration or upgrading of specific areas.
This said, it might be instructive to review
some of the surveyed areas on Crete briefly. Divergent Trajectories
First, I illustrate how settlement history differs
amongst the various regions and next I exam- Renfrew (1972: 233, fig. 14.2), in the Emergence
ine the presence of settlement hierarchy in the of Civilisation, argued for a continuous
various sub-zones. To run ahead of my argu- upward curve for Minoan settlement num-
ment, I want to stress spatial and temporal bers from the Early Minoan up to the end of
diversity of development on the island, and I the Late Minoan IB period20. A preliminary
would like to think that the best way to count of the site numbers on Crete (total of
explain this diversity is as a result of a varying 1061) given in Table 4.1, however, results in
impact of extra-regional stimuli. I have 88 Final Neolithic/Early Minoan I sites, 121
dubbed my paragraphs divergent trajectories Prepalatial (+37.5%), 383 Protopalatial
and regional dynamics. I have not included a (+216%), 315 Neopalatial (-17.75%) and 154
handful of surveys in this preliminary assess- Postpalatial sites (-51.11%), meaning that the
ment because the nature of the publication or curve already went down after the Proto-
the original set-up of the survey does not palatial period21.
allow an easy recovery of data. Absent are This, of course, is just a general trend, a
thus the surveys of the Sphakia area and the closer look at the different sub-zones results
islands of Gavdos15 and Pseira16, as are the in a more even-handed picture. I briefly sum-
unpublished surveys of Nerokourou (French marise the number of settlement sites known
1990: 80), the Pediadha and some others17, the and their chronological attribution.
Landbegehungen in the Rethymnon area In the Lasithi plain, Watrous (1982) identi-
(Schiering 1982) and in the southeast of the fied 15 FN/EM I sites. Only five, especially
island18. Apart from the data collected by slightly larger, hilltop sites at the edge of the
Captain T.W. Spratt, A. Taramelli and F. plain remain occupied during EM II-III but
Halbherr, I have also left aside the information this number increases to about 40 during the
collected during the explorations of Hood and Protopalatial period. This number plunges to
co. in the sixties (Hood 1965; 1967; Hood & about 15 in LM I and stays more or less the
Warren 1966; Hood, Warren & Cadogan 1964), same in the next period, LM IIIA2-B, after a
the individual site identifications by a score of possible hiatus during LM II-IIIA1. The drop
History and Hierarchy 57
in site numbers in the Neopalatial period is Three areas, three different settlement his-
attributed to emigration to prosperous tories. It gets worse though.
coastal centres (Watrous 1982: 15). The mate- The Malia survey, directed by Sylvie
rial from the Psychro Cave includes some LM Miiller, has found evidence for at least one
II, which suggests that visitors arriving on Neolithic site and a handful of EM sites. The
the plateau must have been able to stay increase in the number of sites during the
somewhere (Watrous 1996: 41, 52-53). During Protopalatial period to about 80 is staggering
LM I-III, the main centre may have been at especially considering the subsequent drop
Plati. The identity in settlement numbers in number during the Neopalatial (9+?) and
before and after the Middle Bronze Age is Postpalatial periods. The survey covered 40
interesting, suggesting that this is the norm km2 and in toto 87 sites were located (Miiller
and that the Middle Minoan boom results 1996: 1236; 1998: 548, 552).
from an external stimulus. The surveys by Hope Simpson of the
This inland area contrasts with coastal Kommos area (Shaw & Shaw 1995; Hope
Vrokastro where Hayden, Moody and Simpson 1995) and by Watrous and Vallianou
Rackham (1992) have surveyed an area of c. (1994) for the Western Mesara22 fortunately
50 km2, which includes 13 ecological zones seem to suggest a parallel settlement history.
with site density largest in the coastal area The Kommos survey covered an area of c.
and the river valleys up to the LM I period 17.5 km2 and no other Targe' (i.e. more than
(Hayden, Moody and Rackham 1992, 10.000 m2) settlement was found apart from
317-318). They located only three FN/EM I Minoan Kommos and Roman Matala (Hope
sites but the number of sites increases to 24 Simpson 1995: 326). The Western Mesara sur-
during the Prepalatial period, a number vey intensively covered 22 km2 around
which remains more or less the same after- Phaistos, a tiny fraction of the Mesara plain,
wards during the Protopalatial period. Here which, with its 362 km2, comprises two thirds
the settlement boom occurs in the of the best arable soil of the island. Both areas
Neopalatial period when the number almost together contain about 18 FN/EM I sites and
doubles to 46. After LM I, the settlement this number increases to about 30 in the
number drops a bit to 35 but still remains Prepalatial period. The settlement number is
high. highest during the Protopalatial period (at
Further along the coast, in the Kavousi least 25 for the Western Mesara, 32 for the
area surveyed by D. Haggis (1996), the pic- Kommos area). In the Neopalatial Period
ture seems again somewhat different: in an there is a small drop in both areas and a seri-
area of about 50 km2, Haggis located nine FN ous plunge during LM III with only half a
or EM sites with a drastic increase to about 60 dozen of sites remaining occupied in the
during the Protopalatial period. In Kommos area and slightly more in the
Neopalatial times, the overall site number Western Mesara.
drops slightly to c. 50 but afterwards the area Another recent survey, that of the Ayios
is almost abandoned with only six sites Vasilios Valley by Moody, Peatfield and
showing some signs of reoccupation. Markoulaki (1996: 95), succeeded in locating
'Equally striking is the marked increase in 88 Bronze Age sherd scatters (Tomlinson
settlement size and numbers in the subse- 1995: 64). There are about 15 FN/EM I sites.
quent LM IIIC and Protogeometric periods' Surprisingly the area seems to have been
(Haggis 1992: 408). basically abandoned afterwards up to the
58 Urbanism in the Aegean Bronze Age
Protopalatial Period when half a dozen sites The survey at Petras and Ayia Photia
were established, a number that doubled yielded two or three FN/EM I sites and a
again in the Neopalatial Period. similar number afterwards during the Pre-
Yet another, slightly different, picture and Protopalatial periods (Tsipopoulou
resulted from the Ayiofarango survey by 1990). Only during the Neopalatial period
Blackman and Branigan (1977: 68-69): several did the number slightly increase to half a
FN/EM I sites with more in EM II - EM III dozen after which it was halved again.
but a decline during the Protopalatial Period Finally, the Itanos survey in the north-eastern
and afterwards and the same was true on the tip of the island, directed by A. Schnapp-
coast nearby up to Kaloi Limenes (Blackman Gourbeillon (A. Schnapp [pers. comm.];
and Branigan 1975: 34-35). Moody's Akrotiri Greco et al 1996; 1997; Kalpaxis et al 1995), has
survey covered about 100 km2 and identified hitherto covered an area of c. 30 km2: they
about 245 sites. It also illustrates a continuous have found one Prepalatial site and about 20
upward site curve with about 18 FN, 56 EM, Neopalatial ones but as yet there is very little
147 MM and 150 Neopalatial sites (Moody evidence for a Protopalatial occupation of the
1987; 1990; Moody, Rackham and Rapp 1996). area, an impression that reinforces our ideas
Only afterwards a serious drop occurs to 76 about the Palaikastro countryside which
sites in the LM III period. seems only to have been used for extensive
A preliminary report of the Gournia sur- settlement from the Neopalatial period
vey mentions how an area of 24 km2 was sur- onwards (Driessen and MacGillivray 1989).
veyed and 154 prehistoric sites were We may conclude by referring to the results
identified. Preliminary numbers give seven of the recently published 1998 season of the
Prepalatial, 16 Protopalatial and 13 Kythera survey under direction of C.
Neopalatial sites (Watrous 1992: 607-608; Broodbank (1999: 200-209) that covered an
1999; pers. comm.)23, and, in recent paper, area of c. 13 km2: 7 Prepalatial and 4 Neo-
Betancourt (1999: Plate VHI.b) discusses palatial sites were located but no Protopalatial
material from three Final Neolithic/Early and Postpalatial sites24.
Minoan I sites. Of course, these numbers as such are not
The recently published Ziros survey found very meaningful since they can only be prop-
evidence for eight Final Neolithic sites after erly understood if the size of the settlements is
which the area was abandoned (Branigan also taken into account and processes of nucle-
1996; 1998; 1999). It was again settled during ation or dispersal. Some general comments are
the Protopalatial and Neopalatial periods possible though. For a start, it is obvious that
with half a dozen sites after which it was the main 'colonization' phase of the island is
again largely abandoned (one site remain- the Final Neolithic period when occupation
ing). The nearby Praisos area was also settled begins in all sub-zones of the island, both in
in the FN/EM I period (four sites) and also coastal and highland areas. This is the first
abandoned afterwards during the Prepalatial floruit for the entire island (cf. Strasser 1992;
period (Whitley, Prent and Thorne 1999: 233, Branigan 1999: 64)25. We may wonder whether
257; Whitley 1998). The Protopalatial period this move inland is a result of population pres-
is better represented than the Neopalatial sure or of factional competition in one or more
period but especially during the Postpalatial of the larger Neolithic centres.
phase, the area knew a boom in occupation In any case, something must have hap-
sites. pened in the course of the Prepalatial period
History and Hierarchy 59
to set off the divergent trajectories that can be process took place in these regions and a sim-
observed, from now onwards, in the various ilar process may have occurred on the island
sub-zones of the island. Four, rather remote, of Kythera (cf. Broodbank 1999: 212). The
inland areas (A. Vasilios, Lasithi, Ziros, south-east of Crete, including the Praisos,
Praisos) and perhaps even the Ayiofarango Palaikastro, Ziros and Zakros areas, are very
witness a serious drop in settlement num- much characterized by the presence of a large
bers. This is, apart from in the Ayiofarango, number of small but almost cyclopean struc-
not a result of nucleation, at least not within tures, situated at regular intervals along
the same sub-zone itself. It may, however, be road-terraces that connect the different
a result of the attraction which nearby coastal Protopalatial settlements (Tzedakis et al 1989:
agricultural areas presented, especially those 60, fig. 20 and fig. 22; 1990). Moreover, the
where palaces would be constructed in MM number of Protopalatial peak sanctuaries is
IB. In fact, it suggests already that larger cen- much higher in the far east than anywhere
tres were operational at this early stage, else on the island: 11 against 12 elsewhere on
something confirmed by the existence of the island (Nowicki 1991: fig. 7). How to
monumental buildings at various EM sites explain this coincidence of anomalies is
(Knossos, Malia, Palaikastro, etc.)26. another matter but I will further on suggest
During the Protopalatial period, the island that these features combined may explain the
again illustrates a combination of stagnation, special nature of the far east of the island.
boom and regression. Earlier abandoned or For the Neopalatial period, again the pic-
semi-abandoned inland areas are mostly re- ture differs: a drop in settlement numbers
colonized, suggesting a serious pressure on occurs in Lasithi and the regions around
agricultural land in the coastal zones. This is Gournia, Kavousi and Malia as well as in the
corroborated by the increase in settlement Western Mesara. Elsewhere, however, on
numbers in those coastal areas where palaces Akrotiri, in the Vrokastro area, the Ay.
are established such as Phaistos, Malia, per- Vasilios valley and in the east of the island,
haps Khania and possibly Gournia. Other around Petras, Palaikastro and Itanos, the
areas, and especially in the far east of the number of sites seems to rise slightly or con-
island (but also Kythera in the far west), do siderably. We assume that this difference
not follow this development and instead results from local environmental or geopolit-
show a drop or stagnation in site numbers. ical conditions and that another region posi-
When we first commented upon this feature, tively gained some of the loss of one region,
particularly where Protopalatial sites around especially since the drop is obvious in areas
Palaikastro were concerned (Driessen and in which court-centred buildings are situated.
MacGillivray 1989: 102), some authors (e.g. The great unknown here, yet again, is the
Dickinson 1994: 69 Haggis 1996: 393, n. 64) Knossos area. The picture is very blurred
protested by calling attention to develop- after the LM IB destructions when many
ments elsewhere on the island. I like to areas seem at least temporarily abandoned or
believe that our view is now substantiated by sparsely inhabited. The reoccupation or re-
some of the other surveys in the area and that colonization after LM II is especially obvious
a divergent trajectory did indeed exist. Since in some inland areas as Lasithi and Praisos
Petras, Palaikastro and Zakros were flourish- but some of the traditional agricultural
ing settlements in the Protopalatial Period, it coastal areas continue to be relatively densely
may well be that some kind of nucleation occupied as shown at Malia, Vrokastro,
60 Urbanism in the Aegean Bronze Age
Akrotiri and Gournia. Other areas, however, hectare (Hasan 1978; Postgate 1994; Whitelaw
are again basically abandoned such as this volume), we have the five categories seen
Kavousi, the Ayiofarango and the Ziros area. in Table 4.2.
Since some of these areas were also those that The drawback here is that only very few of
suffered mainly during the Early Minoan the published surveyed areas provide infor-
period, I assume that similar processes were mation on settlement extent and the exercise
at work. conducted here will need to be redone once
all the information is published. Until then it
is impossible to present proper rank-size dis-
Regional Dynamics tributions for the island or to examine how
the number of sites varies in each hierarchi-
It may be instructive at this point to have a cal level with each chronological period
look at site hierarchy and this is where it gets in each region (cf. Cherry 1987). What fol-
tricky. The existence and the size of 'public' lows are therefore simply some preliminary
structures at some sites evidently manifest observations.
the presence of some kind of 'power' at these For the Final Neolithic period, our data
settlements (Driessen 1999). The size of these are sparse but it appears that small, nucleated
structures requires both a considerable popu- villages and isolated farms of Level 5 co-
lation and a social cohesion and their mere existed throughout the island aside from
existence implies an energy input and hence Knossos where, according to a recent study
the presence of administrative personnel. (Manning 1999: 471), a Level 3 settlement or
Since many sites do reveal such buildings, it town of 5 ha may already have existed in the
can reasonably be assumed that there were Late Neolithic period. In view of the extent of
different levels of importance, i.e. a hierarchy the site during the Early Minoan period, this
of sites. Since we lack a sufficient number of seems somewhat exaggerated but it is not
excavated settlements within a single impossible that the sheer size of Final
region27, we are left with settlement sizes to Neolithic Knossos forced the migration of
detect different levels in spatial organization groups to other, not yet settled areas of the
at a given time. In her Akrotiri account, island, explaining the first wave of coloniza-
Moody (1987) has suggested a four-tiered tion alluded to above.
hierarchy but, following Near Eastern exam- During the Prepalatial period, some
ples, I thought it useful to add a category of regions witness the growth of larger, Level 4
sites that are larger than 25 hectares. If we settlements: Malia itself already had an
assume a density coefficient of 250 people per extent of about 2.58 ha and recently another
large site was identified a few kilometres the case of Malia, it can be argued that the site
east, near the Arkovouno hill (S. Miiller in grew through the incorporation of a popula-
Blackman 1997:109). Knossos is said to cover tion that was initially established elsewhere
4.84 ha (Whitelaw 1983: 339), Phaistos about on the plain (Arkovouno ?) or in the hinter-
1.5 ha (Watrous et al 1993: 224) and Watrous' land. Malia, at present, seems to have been
team has identified a series of small sites the largest Protopalatial polity of the island,
(campsites, hamlets) as well as 'the largest which is amply illustrated by the score of
EM I-II settlement in the Isthmus of public buildings at the site (e.g. Poursat
Hierapetra' at Halepa at the east end of the 1987). As shown by Sylvie Miiller (1996;
Pacheia Ammos bay which covers 2 ha 1997), the secondary settlements of this
(Tomlinson 1996: 45; BCH 120 (1996): period are located in three, almost concentric
1234-1235; L.V. Watrous pers. comm.). circles around the city. The success of
Blackman and Branigan (1977: 69) discovered Knossos is more difficult to explain, again
a 3.25 ha site in the Ayiofarango, 'possibly the because of the lack of surveys in its hinter-
largest Early Minoan settlement yet known'. land. In contrast to these areas, some other
In some areas, we seem to have a two- or regions such as Lasithi, Kavousi, Vrokastro
three-tiered hierarchy which people tend to and East Crete have failed to provide evi-
equate with 'chiefdoms', ranked societies in dence for a developed hierarchy in this par-
which chiefly families played a considerable ticular period. Other zones, as mentioned,
role. I think it is very likely that at least were largely abandoned during this phase.
Knossos and Malia, but most likely other This seems to suggest that these areas
places too, may already have developed fur- remained outside of the territories adminis-
ther at this point but this needs more archae- tered by the First Palaces and hence outside
ological corroboration. Malia at least already the mainstream development (Cadogan 1990;
seems to have some kind of central building 1995; Haggis 1996: 424; Knappett 1999;
and the same is probably true for Knossos Knappett and Schoep 2000). A fourth trajec-
and Palaikastro (MacGillivray and Driessen tory seems illustrated by the far east of the
1990: 399; Schoep 1999a). island (Figure 4.3): there is evidence for at
This, still limited, hierarchy heralds what least three Level 2 (if not Level 1) sites during
happens next in these areas since both the site the Protopalatial period, Petras, Zakros and
number and site extent show a massive Palaikastro, but as yet little evidence for the
increase in the Protopalatial period. Indeed, presence of sites that stood on a lower hierar-
the data make it clear that some regions such chical step apart from a very large number of
as that of the Akrotiri with the site of Khania Level 5c sites (Branigan 1972; Driessen and
(7.64 ha) and the Western Mesara with the MacGillivray 1989; Tsipopoulou 1995; 1996;
site of Phaistos (15 ha) (Watrous et al 1993: Tsipopoulou and Papacostopoulou 1997).
225) achieved a four-tiered hierarchy starting Road stations and peak sanctuaries form the
with Level 2 sites. The Malia and Knossos latter, at the lowest level. Nucleation appears
areas, on the other hand, may already have to have been the rule here, and, although I
known a five-tiered hierarchy since both evi- use the concept hesitantly and anachronisti-
dence Level 1 settlements of more than 25 ha cally, the east Cretan towns may have been
since 45 ha for Knossos and 60 ha for Malia some kind of city-states 'avant la lettre'. If
have been claimed (Whitelaw 1983: 339; correct, city and countryside may have been
Hood and Smyth 1981; Miiller 1997: 52)2X. In entirely integrated, perhaps in an achoritic
62 Urbanism in the Aegean Bronze Age
system, where farmers lived in town and not example. All hypotheses, however, depend
on farms. Such societies seem to have been on the position of Knossos and it would be
very much kinship based29. This achoritic sys- interesting to know precisely what is hap-
tem should then be considered against the pening in its hinterland during this phase
background of the mentioned 'watchtowers' and what its relations were with the major
and the presence of a high number of peak- settlements at Tylissos, Archanes, Amnisos
sanctuaries. There is obviously a great cul- and Galatas. The Knossos Survey (Hood and
tural uniformity at play here in the Smyth 1981: 10; Warren 1994: 209) suggests
Protopalatial period and I wonder whether an intensively settled area of 75 ha during the
the peak sanctuaries formed discrete but Neopalatial period31, which, if we accept a
obvious boundaries between the different density coefficient of 250 people per hectare,
components of some kind of confederacy, represents an increase for the city from the
linked via a road system. Protopalatial to the Neopalatial period from
The Gournia area, on the other hand, may 11,250 to about 18,750 people. The decline of
also have had a three or four-tiered hierarchy the Malia and Phaistos polities - both with
during the Protopalatial period and its main smaller palaces and surrounding settlements
settlement is now thought to have had an than in the Protopalatial period - with an
extent of about 21 ha30. There is as yet no accompanying loss of integration in their
proof that it attracted settlers from the rest of respective hinterlands may then suggest that
the Mirabello area where the site numbers Knossos, which boomed in this period, had
increase or stay the same and no hierarchy effectively taken the lead and that those sites
develops. Gournia indeed may well have that were situated nearby had been demoted
been a mini-state. into some kind of provinces, to all intents and
For the Neopalatial period, there is an purposes under control of Knossos. If so, the
islandwide establishment of small Level 5 Malia and Phaistos court-centred buildings
sites, usually identified as farmsteads or, would really represent smaller, subordinate
when there is obvious architectural elabora- versions of the capital's 'paramount palace'32.
tion, as 'villas' (Hagg 1997). At the same time, If this hypothesis is acceptable, it follows
I have the impression that some settlements that, at least during LM IA, Knossos was in
grow in size whereas some intermediate sites charge over the entire central Cretan area. It
disappear. In other words, the five-or four- thus acted as a real 'central place'33, with a
tiered hierarchy broke down into a three- well-integrated hierarchy in which the Level
tiered one. When such a process happens 2 towns (Phaistos, Malia, Ayia Triadha,
elsewhere in the ancient world, it usually Kommos, Tylissos, Archanes34, Galatas35, Plati36
suggests incorporation into a larger frame- etc.) would in turn be encircled by Level 3
work. Could this be happening on Crete? satellite settlements at regular and shorter
Some of the regions show an obvious, overall distances, in this case Vathypetro, Khannia-
reduction in settlement numbers and a Mitropolis, Pitsidia, Milatos etc. Myrtos-
reduction in the extent of the main settlement Pyrgos would, still following this recon-
during LM I. These same areas see an struction, represent some kind of boundary
increase in higher order centres (e.g. the station for the Knossos state in LM IA, explain-
Phaistos-Ayia Triadha-Kommos triangle ing its architectural elaboration with obvious
[Shaw and Shaw 1985]). This is especially Knossian features as I have discussed on
clear for the Malia and Phaistos areas, for another occasion (Driessen 1989: 21).
64 Urbanism in the Aegean Bronze Age
Elsewhere, especially in the far east and west and ideally, this preliminary outline of settle-
of the island, we can observe that the ment history and hierarchy should be sub-
Neopalatial settlements not only outgrow their stantiated with other archaeological data,
Protopalatial predecessors but these regions especially those related to the presence of
also illustrate an increase of site numbers and a high status artefacts, elaborate architecture
more developed hierarchy. In fact, they seem to and documentary information. Moreover, it
reflect a situation similar to that of the main must be verified with the settlement history
polities during the Protopalatial period and of some major sites such as Knossos, Malia
those in the far east may only now have and Palaikastro. Further, we should try to
crossed the early state threshold, illustrating establish whether areas with nucleated settle-
some kind of secondary state formation. Some ment patterns show traces of increased
areas, such as those around Gournia, Kavousi urbanization and from what moment
and Vrokastro and in the Lasithi, show a reduc- onwards 'central places' with their hierarchi-
tion in the number of Neopalatial settlements cal site arrangement but without signs of
but a more developed site hierarchy and obvi- urbanization form a feature of the Cretan
ous signs for nucleation in one or two larger landscape. What we first need, however, are
settlements. It is at this point, for instance, that proper intensive surveys of the Knossos,
several Protopalatial settlements along the Palaikastro and Zakros hinterlands to try out
north coast of Mirabello were abandoned and the validity of the hypotheses presented here.
we may assume that the Gournia polity inte- What is clear, however, from this first analy-
grated several areas37, perhaps as a reaction sis, is that different trajectories were being fol-
against pressure from the west (the Knossos lowed on the island and this from a very early
state). Plati too may well have taken control moment onwards. The evidence further seems
over the Lasithi plain. As far as this evidence is to suggest that already during the Early
concerned, I do not think it is sufficient to sug- Bronze Age regional powers existed in some
gest incorporation by Knossos. areas of the island and that these attracted
During the LM IIIA period, there are ini- people from outlying areas. The resulting
tially rather few larger settlements left on the demographic stress in the core settlement may
island and isolated farms and hamlets are also subsequently have led to the incorporation of
rare. The absence of a complex local hierarchy outlying areas, if only to ensure the supply of
may then suggest that Knossos was indeed, for agricultural produce to the centre. Territorial
some time, in overall control of the island, with claims may have led to conflict at the end of
a more developed supra-regional hierarchy.
the MM II period. Elsewhere on the island,
The progressive establishment of more settle-
both some sort of 'city-states' and more con-
ments on different levels illustrates, however,
servative kin-based systems may have been
that within the LM IIIA-B period, the political
the rule for quite a long time. Afterwards, dur-
situation also changed, opening the possibility
ing MM III and LM IA, it appears that some of
that this Knossos state disintegrated in differ-
the different, large, comparatively well inte-
ent political units (cf. Bennet 1985; 1987; 1990).
grated polities that existed during the Middle
Bronze Age in central Crete were incorporated
Preliminary Conclusions into a larger political framework and a territo-
rial state headed by Knossos. Elsewhere on the
Site numbers and sizes do not suffice to island we may only now witness the develop-
reconstruct a persuasive political geography ment of pristine states.
History and Hierarchy 65
A final note of warning, however. The paper, see Crumley 1979, 1987, 1995 and
regional dynamics and different trajectories Crumley & Marquardt 1987.
that can be observed make me wonder
whether, where Minoan Crete is concerned,
we are not trapped into established theoreti- Notes
cal models and definitions of political and
1. D. Haggis (pers. comm.) reminded me that the area
territorial organisation. The scarcity of fortifi- between 400 and 800 m is still largely unknown but may
cation systems, an absence of an overt royal eventually prove to be an important ecozone as shown
iconography and the difficulty of tying in the by a highland site such as Zominthos on the way to the
Idaean Cave and the defensible and pastoral sites identi-
Cretan situation to territorial organizations
fied e.g. by Faure 1962 and Nowicki 1991; 1992.
known from the Bronze Age Mediterranean, 2. One site in the Madhares area was even found at +
leaves me frustrated. Taken together, it may 1810-1840 m (L. Nixon in French 1993: 81); D. Haggis
imply that Crete, because of its insularity, had (pers. comm.) located a 2+ ha Protopalatial site at +600 in
the Triphti mountains whereas the Malia survey found a
developed a different type of territorial orga- protopalatial site at +801 m between Malia and Mochos
nization, perhaps largely ritually motivated, ('peut-etre les traces d'un bivouac de bergers', Miiller
which reduced intra-insular tension. 1991: 749).
3. I admit having 'fudged' some of the numerical data
where not obtainable: if a 'dozen' sites were mentioned, I
have taken this to represent '12'; when the site number is
Acknowledgments then said to be 'halved' I assume this to mean '6'.
Likewise, when a 'couple' of sites are mentioned, this is
*To Paul Faure, great explorer of Crete and interpreted as '2'. All numbers are, of course, provisional
and conditional. At the Knossos conference held in
great-grandfather of Danae. Heraklion in November 2000, Nikos and Marina
Panagiotakis presented the first result of a survey carried
This paper is part of the 'Topography of out in the Pediada, covering about 800 km2 in which about
Power' project carried out at the Universite 125 Protopalatial and 230 Neopalatial sites were discov-
Catholique de Louvain (FSR 2000). I would like ered. I thank Mr. Panagiotakis for his permission to men-
tion this work. Once published, the results will add
to thank Philip Betancourt, Cyprian
greatly to our understanding of the hinterland of Knossos.
Broodbank, Florence Gaignerot, Donald 4. For other problems, methodologies etc. of Cretan sur-
Haggis, Jenny Moody, Sylvie Miiller, Lucia veys, see Soetens and Gkiasta (in press).
Nixon, Nikos Panagiotakis, Alan Peatfield, 5. Moody 1987: chapter 6; I have used her numbers for
occupation and sacred sites.
Norbert Schlager, Annie Schnapp, Philip Van 6. L.V. Watrous (pers. comm.).
Peer, Didier Viviers, Vance Watrous and 7. Data based on personal communication with A.
James Whitley who have generously shared Schnapp and preliminary reports (Greco el al. 1996; 1997;
Kalpaxis et alii 1995).
information but who are in no way responsi- 8. Numbers of occupation sites based on the maps pub-
ble for some of my hypotheses. I also thank lished in Watrous 1982.
the members of the project (P. Fontaine, T. 9. Precise number not clear from the publication
(Whitley, Prent and Thorne 1999).
Cunningham, K. Vansteenhuyse, S. Soetens,
10. P.P. Betancourt (pers. comm.).
E. Druart) for comments and K. 11. Numbers based on maps published in Hayden,
Vansteenhuyse for producing the maps. Moody and Rackham 1992; since the three chronological
Many of the ideas expressed in this paper are groups given are EM-MM II, MM III-LMI and LM I/III-
EIA, I have counted the number of Pre-and Protopalatial
preliminary and their durability is largely sites twice.
dependent on the publication of surveys. For 12. Watrous et al. 1993, 225-228 mention 'a handful of
a recent account on Cretan surveys, see sites', 'this number doubles', etc. so the given numbers
only reflect a general trend.
Moody, Nixon, Price and Rackham 1998; for 13. Incidentally, this gives a site density of about 1 site
some of the ideas and methods used in this per 13.5 km2 in the Final Neolithic period, 1 per 9.61 km2
66 Urbanism in the Aegean Bronze Age
during the Prepalatial period, 1 per 3.35 km 2 in the 1992-93: 73, AR 40 for 1993-94: 81; AR 41 for 1994-95: 65
Protopalatial period, 1 per 4 km2 in the Neopalatial and AR 42 for 1995-96: 45.
period and 1 per 7.77 km2 in the Postpalatial period. 24. C. Broodbank (pers. comm.) reported that during the
14. Moody, Nixon, Price and Rackham 1989: 88. The 2000 campaign this picture was somewhat rectified with
Sphakia survey attempted to cover c. 470 km2 but it is not the discovery of two or three Protopalatial sites, one of
entirely clear whether the entire region was explored which very large, and many more Neopalatial sites.
intensively. After the 1987 and 1989 seasons, 15 km2 had 25. See also Warren 1984 for internal and external colo-
been explored and 34 prehistoric sites identified. Nixon nization for reasons of demographic pressure in later
(in French 1993, 81) reports that all in all 218 sites were Minoan periods.
discovered but no chronological attributions have as yet 26. For other processes at work during the Prepalatial
been presented that can be used for statistical analyses; period, see especially Dabney 1989, Dabney and Wright
see also Nixon 1996; Nixon, Moody, Price, Rackham and 1990 and Haggis (in press).
Niniou-Kindeli 1990; J. Moody and L. Nixon in French 27. The Knossos region, for instance, comprises a high
1990: 81-82. number of high-class sites including Archanes,
15. Although very few Minoan sites seem to have been Vathypetro, Tylissos, Amnisos etc, but without surface
identified, there is good evidence for Middle Minoan exploration it is extremely difficult to attribute a hierar-
occupation (Kopaka 1996: 70) chy to this collection. It is hoped that the Pediada survey
16. There is an abstract on the survey and some ideas on will rectify this situation.
landuse but no mention is made of the chronological 28. O. Pelon, J.-C. Poursat and R. Treuil in Aerial Atlas,
attribution of the sites (Betancourt and Hope Simpson 176, assume the Malia settlement to have covered 80 ha
1992; Hope Simpson and Betancourt 1990). For the but this seems to have been an educated guess before
Chrysokamino survey, see BCH 120: 3 (1996),, 1324. P.P. Miiller's survey took place.
Betancourt (pers. comm.) mentions the discovery of 300 29. Compare with Crumley 1995: 29, who distinguishes
sites on the island of Pseira (1.75 km2): 'the island is first between synchoritic settlements (where the population of
settled in the Final Neolithic. Its main period of use is the centre is supported by the surplus produced by a
LM I'. The Chrysokamino survey was included in the rural population), achoritic settlements (like the Greek
Kavousi survey (see below) but was even more intensive poleis) and epichoritic settlements where the centres are
and managed to locate 40 sites (1 km2). all but deserted except for a few specialists but the out-
17. The area around Achladia seems to have been exam- lying area supports a sizeable rural population.
ined in detail (Tsipopoulou 1995) and a survey in the 30. According to Watrous 1992: 608, Gournia measured
Nerokourou area is mentioned in French 1990: 80; appar- 700 x 300 m at its prime although Watrous 1999: 906,
ently surveys have also taken place in the areas around talks only about 4 ha. and 400 to 1200 inhabitants. The
Thronos/Sybrita (Rochetti 1994), Monastiraki and Aerial Atlas, 104, gives 2.5 ha excavated area (with the
Eleftherna. palace measuring c. 50 x 37 m, the court 40 x 15 m). L.V.
18. See especially Schlager 1987, 1991 and 1997 and Watrous (pers. comm.) mentions that many protopalatial
Blackman 1997: 117-118. Recent reports mention the dis- sites were found, a few of which are village sized.
covery of 44 Minoan and later sites in the communities of 31. We may wonder to what degree the size of the central
Ziros and Agia Triada but no information on their size or building reflects the size of the settlement: Knossos
chronological attribution is as yet available. palace measures 13,000 m2, Malia c. 9,350 m2, Phaistos c.
19. See the annual reports in Archaologischer Anzeiger 6,000 m2, Zakros about 3,000 m2 and Kommos c. 4,500 m2)
before World War II and those in Praktika tis Archeologikis (cf. Shaw and Shaw 1993:186, n. 151). Other authors give
Helenas and Archeologikon Deltion afterwards. different sizes e.g. Knossos palace and surroundings 4 ha
20. A similar stand is taken for the Vrokastro area by (Cadogan in Aerial Atlas, 129), Malia palace 8.900 m2
Hayden, Moody and Rackham 1992: 335: 'Activity grad- (Aerial Atlas, 176), Zakros palace c. 8000 m2 (Aerial Atlas,:
ually increased throughout the Bronze Age to culminate 298). Sakellarakis and Sakellarakis 1998: 77 assume a
in the LM I period; during LM III, the absolute numbers palace at Archanes of 14,000 m2 but this seems slightly
of sites decreased and the nucleation of settlement exaggerated; see especially Wiener 1990 for settlement
began'. sizes.
21. Compare with Renfrew 1972: 232, Table 14.111 where 32. And this may then also apply to Galatas. Compare
a total of 379 sites is considered (42 Neolithic, 111 with Warren 1985: 79, who reconstructed a Knossos
Prepalatial, 190 Protopalatial and 284 Neopalatial). polity with a territory of c. 1000 ha.; moreover, since he
22. For a long time, the eastern Mesara has remained calculates a population of 17,000, the site needs a consid-
terra incognita. This seems now gradually changing with erable catchment area.
new and exciting discoveries by the ephoreia as men- 33. This is a term for an administrative hierarchy so
tioned by P.M. Warren at the Knossos 2000 conference in well integrated that Tier 2 towns encircle the Tier 1 city
Heraklion. at very regular distances; in turn, Tier 3 settlements
23. See also the reports in BCH 117 (1993): 886; 118 (1994): encircle Tier 2 settlements at regular (and shorter) dis-
815; 119 (1995): 1020; 120 (1996): 1324-1325 and AR 39 for tances' (Flannery 1998: 18).
History and Hierarchy 67
34. Sakellarakis and Sakellarakis (1998: 139) argue for a Bonnefant, J.C.
very large settlement with another one at Vitsila, mea- 1972 La Crete. Etude morphologicjiie. Universite de
suring 18 km2 - surely this should be 18 ha. Lille: Service de reproduction des theses.
35. Approximate size: 7 ha (cf. BCH 120 (1996): 1332). Branigan, K.
36. Watrous (1982:15) suggests a settlement on two hills 1972 Minoan Settlements in East Crete. In P.J. Ucko,
of c. 2 ha. R. Tringham and G.W. Dimbleby (eds.), Man,
37. Watrous (1999: 908) believes that the abandonment Settlement and Urbanism, 751-59. London:
in the Gournia area only happened at the time of the Duckworth.
Santorini eruption. Hayden, Moody and Rackham 1996 Lamnoni - A Late Neolithic Landscape in East
(1992: 335) assume that the Vrokastro area formed part Crete. 8"' Cretological Congress: Summaries 21.
of the Gournia polity during the entire Minoan Palace Historical Society of Crete: Heraklion.
period. Unfortunately, the site numbers given for the 1998 Prehistoric and Early Historic Settlement in the
Vrokastro area are difficult to use because of the chrono- Ziros Region, Eastern Crete. BSA 93: 23-90.
logical attribution in two groups (MM III/LM I and LM 1999 Late Neolithic Colonization of the Uplands of
I /HI). Eastern Crete. In P. Halstead (ed.), Neolithic
Society in Greece: 57-65. Sheffield: Sheffield
Academic Press.
Broodbank, C.
Bibliography 1999 Kythera Survey: Preliminary Report on the
1998 Season. BSA 94: 191-214.
Aerial Atlas, J.W. Myers, E.E. Myers and G. Cadogan. Cadogan, G.
1992 The Aerial Atlas of Ancient Crete, Berkeley & Los 1990 Lasithi in the Old Palace Period. BLCS 37:172-74.
Angeles: University of California Press. 1995 Malia and Lasithi: A Palace State. Proc. 7"'
Bennet, J. Cretological Congress: 97-104. Rethymnon.
1985 The Structure of the Linear B Administration at Cherry, J.F.
Knossos. AJA 89: 231-49. 1986 Polities and palaces: Some Problems in Minoan
1987 The Wild Country east of Dikte: The Problem State Formation. In A.C. Renfrew & J.F. Cherry
of East Crete in the LM III Period. In J.T. Killen, (eds.), Peer Polity Interaction, 91-112.
J.L. Melena and J.-P. Olivier (eds.), Studies in Cambridge: Cambridge University Press.
Mycenaean and Classical Greek Presented to John 1987 Power in Space: Archaeological and
Chadwick, (Minos 20-22), 77-88. Salamanca: Geographical Studies of the State. In J.M.
Ediciones Universidad de Salamanca. Wagstaff (ed.), Landscape and Culture.
1990 Knossos in Context: Comparative Perspectives Geographical and Archaeological Perspectives,
on the Linear B Administration of LM II-III 146-72. Oxford: Blackwell.
Crete. AJA 94: 193-211. Crumley, C.L.
Betancourt, P.P. 1979 Three Locational Models: An Epistemological
1999 What is Minoan? FN/EMI in the Gulf of Assessment for Anthropology and Archaeology.
In M.B. Schiffer (ed.), Advances in Archaeological
Mirabello Region. In P.P. Betancourt, V.
Method and Theory, 141-73. New York: Academic
Karagheorgis, R. Laffineur and W.-D. Niemeier
Press.
(eds.), Meletemata. Studies in Aegean Archaeology
1987 A Dialectical Critique of Hierarchy. In T.C.
presented to Malcolm H. Wiener (Aegaeum, 20: 2),
Patterson and C. Ward Gailey (eds.), Power
33-9. Liege: Universite de Liege.
Relations and State Formation, 155-69.
Betancourt, P.P., and R. Hope Simpson Washington: American Anthropological
1992 The Agricultural System of Bronze Age Pseira. Association.
Cretan Studies 3: 47-54. 1995 Building an Historical Ecology of Gaulish
Blackman, D. Polities. In B. Arnold and D. Blair Gibson
1997 Archaeology in Greece 1996-1997. AR 43: (eds.), Celtic Chiefdom, Celtic State. The Evolution
1-125. of Complex Social Systems in Prehistoric Europe,
Blackman, D., and K. Branigan 26-33. Cambridge: Cambridge University
1975 An Archaeological Survey on the South Coast Press.
of Crete between the Ayio Farango and Crumley, C.L. and H.W. Marquardt
Chrisostomous. BSA 70: 17-36. 1987 Regional Dynamics in Burgundy. In C.L.
1977 An Archaeological Survey of the Lower Crumley and W.H. Marquardt (eds.), Regional
Catchment of the Ayiofarango Valley. BSA 72: Dynamics: Burgundian Landscapes in Historical
13-84. Perspective, 609-23. San Diego: Academic Press.
68 Urbanism in the Aegean Bronze Age
1965 Minoan Sites in the Far West of Crete. BSA 60: the Cretan Landscape, 52-59. Cambridge:
99-113. Cambridge University Press.
1967 Some Ancient Sites in South-West Crete. BSA Moody, J., L. Nixon, S. Price and O. Rackham
62: 47-56. 1998 Surveying poleis and larger sites in Sphakia. In
Hood, M.S.F. and D. Smyth W.G. Cavanagh and M Curtis (eds.), Post-
1981 Archaeological Survey of the Knossos Area. Minoan Crete. Proceedings of the 1st Colloquium,
London: British School at Athens. 87-95. London: The British School at Athens.
Hood, M.S.F. and P.M. Warren Moody,}., A. Peatfield and S. Markoulaki
1966 Ancient Sites in the Province of Ayios Vasilios, 1996 Report from the Aghios Vasilios Valley Survey.
Crete. BSA 61: 163-91. 8"' Cretological Congress: Summaries 95.
Hood, M.S.F., P. Warren and G. Cadogan Historical Society of Crete: Heraklion.
1964 Travels in Crete, 1962. BSA 59: 50-99. Miiller, S.
Hope Simpson, R. 1990 Prospection de la plaine de Malia. BCH 114:
1995 The Archaeological Survey of the Kommos 921-30.
Area. In J.W. and M.C. Shaw (eds.), Kommos I.I, 1991 Prospection de la plaine de Malia. BCH 115:
325-402. Princeton: University Press. 741-49.
Hope Simpson, R. and P.P. Betancourt 1992 Prospection de la plaine de Malia. BCH 116:
1990 Intensive Survey of Pseira island. AJA 94: 322. 742-53.
Kalpaxis, T, A. Schnapp and D. Viviers 1996 Prospection archeologique de la plaine de
1995 Itanos. BCH 119: 713-736. Malia. BCH 120: 921-28.
Knappett, C. 1997 L'organisation d'un territoire minoen. Dossiers
1999 Assessing a Polity in Protopalatial Crete: the d'Archeologie 222 (Avril 1997), 52.
Malia-Lassithi State. AJA 103: 615-639. 1998 Malia. Prospection archeologique de la plaine.
Knappett, C. and I. Schoep BCH 122: 548-52.
2000 Continuity and Change in Minoan Palatial Nixon, L.
Power. Antiquity 74: 365-371. 1996 Late Minoan Ill-Early Iron Age Sites in
Kopaka, K. Sphakia. 8"' Cretological Congress: Summaries.
1996 Epifaniaki erevna sti Gavdo. 8"' Cretological 102. Historical Society of Crete: Heraklion.
Congress: Summaries. 70. Historical Society of Nixon, L., Moody, J. and O. Rackham
Crete: Heraklion. 1988 Archaeological Survey in Sphakia, Crete. Echos
MacGillivray, J.A. and J. Driessen du Monde ClassiqueI Classical Views 32, n.s. 7:
1990 Minoan Settlement at Palaikastro. In P. Darcque 159-73.
and R. Treuil (eds.), L'Habitat e'ge'en prehistorique, Nixon, L., J. Moody, S. Price, O. Rackham and V. Niniou-
395-412. Paris: Ecole francaise d'Athenes.
Kindeli
MacGillivray, J.A. and L.H. Sackett
1990 Archaeological Survey in Sphakia, Crete. Echos
1984 An Archaeological Survey of the Roussolakkos
du Monde Classique/Classical Views 34, n.s. 9:
Area at Palaikastro. BSA 79: 129-59.
213-20.
Manning, S.
Nixon, L. et al.
1999 Knossos and the Limits of Settlement Growth.
In P.P. Betancourt, V. Karagheorgis, R. 1994 Rural Settlement in Sphakia, Crete. In P.
Laffineur and W.-D. Niemeier (eds.), Doukelis and L.G. Mendoni (eds.), Structures
Meleternata. Studies in Aegean Archaeology pre- Rurales et Socie'tes Antiques (Ann. Litt. de I'Univ.
sented to Malcolm H. Wiener (Aegaeum, 20: 2), de Besanfon, 508), 255-64. Paris: Belles Lettres.
469-80. Liege: Universite de Liege. Nowicki, K.
Marcus, J. and G.F. Feinman 1991 Report on Investigations in Crete VII. Studies
1998 Introduction. In G.M. Feinman and }. Marcus in 1990. Archeologia 42: 137-45.
(eds.), Archaic States, 3-13. Santa Fe: School of 1992 Report on Investigations in Crete VIII. Studies
American Research Press. in 1991. Archeologia 43: 113-19.
Moody, J. Pendlebury, J.D.S.
1987 The Environmental and Cultural Prehistory of the 1939 The Archaeology of Crete. London: Methuen.
Khania Region of West Crete: Neolithic through Pendlebury, J.D.S., M.B. Money-Coutts, and E. Eccles
Late Minoan III (PhD, University of Minnesota). 1934 Journeys in Crete, 1934. BSA 33: 80-100.
1990 The Continuity of Settlement and Landuse Postgate, N.
Systems in West Crete. In A.T. Grove, J. Moody 1994 How Many Sumerians per Hectare? - Probing
and O. Rackham (eds.), Stability and Change in the Anatomy of an Early City. CAJ 4(1): 47-65.
70 Urbanism in the Aegean Bronze Age
Figure 5.1 Map of eastern Crete showing sites mentioned in text: 1. Petras; 2. Achladia-Riza; 3. Ayios Georgios/Tourtoloi;
4. Zou; 5. Ayia Photia; 6. Kato Zakros; 7. Epano Zakros; 8. Palaikastro; 9. Vai (After Boquist).
suggest a three-tiered hierarchy of sites. Petras petrographic study (1997: 219-227) suggested
itself occupies the first tier, then the Villa' sites that in terms of distribution the territory
of Zou, Achladia-Riza, Ayios Georgios/ seems to be unified with two clear production
Tourtouloi, and Klimateria, then isolated centers, the main one at Petras itself and
farmsteads at Analoukas, Ayia Photia plain, another near Achladia, finds of a potters wheel
Siteia Airport, and Achladia-Platyskoinos. at Ayios Georgios / Tourtouloi and a kiln at
Aside from Klimateria which 'had a specific Zou must be taken into account (the former
function distinctly different from the other site was apparently not included in Day's
more distant rural villas' (1997: 210), the tier 2 study and the latter proved to be geologically
'villa' sites listed above are all themselves cen- indistinct from Petras). These sites had been
tral buildings within hamlets or villages (see seen as largely independent and self-sufficient
L. Platon 1997: 187-202; also discussion ses- (L. Platon 1997: 202). Furthermore, the situa-
sion p. 214 for a dissenting view). The rural tion of these 'villa' sites is similar to that of
villas also show evidence of storage (all three), Petras, particularly Ayios Georgios/Tourtouloi
industrial areas (all three), and localized pot- which is set atop a small hill; as Tsipopoulou
tery production (all three). Regarding the lat- states 'these 'villas' constitute the central
ter, it is important to note that while Day's entity in the intra-settlement hierarchy, thus
74 Urbanism in the Aegean Bronze Age
1987: 265). It may not be coincidence that the structures at Ta Skaria, Kouremenos, Ayia
two most likely explanations for the Triada, Ayios Nikolaos, and Angathia,
road/'watchtower' system are military whether houses or 'watchtowers' are too
and / or economic, the latter case being either small and too close to the main site at
for the transport of produce to the town or to Rousolakkos to be compared to the 'villas'
support the import/export activities of the around Petras; the remains at Kouremenos
harbor, perhaps in particular the export of and Angathia also belong to LMIIIB and are
timber. There is only one outlying site which thus outside consideration here.
shows possible signs of self-sufficiency, the Rousolakkos itself is by far the largest of
aforementioned 'villa' at Epano Zakros, for the three sites under consideration and is
whatever one makes of Chiromandres, it is usually ranked second (behind Knossos) or
certainly not a 'villa' in the sense of the Petras third (behind Knossos and Malia) in size and
examples or Epano Zakros. At Azokeramos, population for the Neopalatial period on
the pottery showed a mixture of Palaikastro Crete. The quarries at Ta Skaria are among
and Zakros fabrics (Day 1997: 225), suggest- the largest known for Minoan Crete (J.
ing that it lay on the border of these two terri- Driessen: pers. comm.), and in terms of house
tories and may not have belonged to either. sizes, abundance of ashlar masonry, profu-
Indeed, pottery sampled from the Zakros sion of elite architectural forms, and evidence
region showed 'such great differences in the for town planning the site is likewise remark-
technique of paste formation' that Day sug- able for the island as a whole, best compared
gests not only different production centers, to Knossos and perhaps Malia. Thus far no
but different 'cultural traditions of pottery central building has been found. The site is
making' - a sharp contrast to both Petras and densely occupied and as there is no sign that
Palaikastro (Day 1997: 225). this could be the result of geomorphological
The patterns for Zakros then can be stated: concerns, other explanations must be sought.
Branigan (1972: 756) suggested poorer 'ten-
• Nucleation at a central site without a sig- ants' filling in smaller unused lots; Driessen
nificant number of satellite settlements; and MacGillivray (1989:106) counter that this
• A system of constructed and controlled density of building may be the result of
access into the surrounding territory extended family or clan members adding on
requiring considerable resource expendi- to ancestral homes. A further explanation,
ture to service and/or defend the central though hypothetical, might be the presence
settlement. Though perhaps mandated by of a fortification wall at some point in the
an inhospitable topographical situation, site's development. Possible remnants of
what is most important to note here is that such a wall were noted in the 1983 survey
this system is not an adaptation of a (MacGillivray et al. 1984). The road/'watch-
'human' construct to a 'natural' one but tower' system mentioned above has been
rather a modification or transformation of traced as far as Palaikastro (MacGillivray and
the natural environment itself. Driessen 1990: 401). Much further study
remains to be done to understand fully the
Palaikastro nature of this system. However, even if it
Palaikastro also exhibits a nucleated settle- were discovered that this road system origi-
ment pattern. There is a 'villa' at Vai nated from Palaikastro and that Karoumbes
(Driessen and Macdonald 1997: 234), but and Zakros were in fact under the control of
76 Urbanism in the Aegean Bronze Age
Palaikastro, the point made above for Zakros As mentioned above, in comparison with the
would merely be extended, and would still other known palaces (Knossos, Phaistos,
hold true, as stated, for Zakros itself. Implicit Malia, Zakros and Galatas) and even with
in this hypothesis would be the discovery of some of the non-palatial (or not-as-yet pala-
a palace at Palaikastro along with other evi- tial) central buildings, such as Ayia Triada,
dence for such control. Kommos building J/T, Monastiraki, Khania,
The settlement patterns for Palaikastro are: and Archanes-Tourkogeitonia, the Petras
palace is small and lacks both certain palatial
• Nucleation at a central site; architectural features (such as a Lustral Basin,
• Considerable density of occupation, to a Minoan Hall, Light Well or Pillar Crypt) and
degree beyond that mandated by topo- elements of ostentation (such as extensive use
graphy. of gypsum or monumental orthostates).
Evidence of schist dadoes, mason's marks, cut
door jambs, columns, etc. (Tsipopoulou and
Spatial Organization Papacostopoulou 1997: 211) suggest that the
omissions may reflect choices, attitudes or
Petras perhaps functional differences and not simply
The Petras palace was built in MMIIA economic or technical limitations.
(Tsipopoulou 1999). EMU occupation is evi- The palace is situated atop a conspicuous
dent but obscured by later activity. The hill; that this hill was levelled, and in some
Cretan Hieroglyphic archive, found in an areas bedrock dug out to make room for the
MMIIB destruction layer, demonstrates that palace shows the importance of its location.
Petras was an administrative center at that That the location is defensible has been
time. Though the 'villas' mentioned above pointed out above; the presence of a massive
are said to have been built in MMIII, it cannot retaining wall and bastion may also indicate
be ruled out that there was earlier habitation defensive concerns in the first phase of its
at the same spots as they have not been fully existence (MMII). Another essential feature of
explored. Alternatively, these sites may be the the location of the palace, possibly more
result of processes of hinterland nucleation essential than pure defensibility, is its visibil-
subsequent to the establishment of the palace ity, both from the sea and the land around it.
at Petras, though the presence of fortified Given the horizontal restrictions of the hill-
and/or defensively located sites already in top, it may be that the palace was built up
MMI, such as Ayia Photia and Chamaizi vertically and the massive walls in the north
favors the former hypothesis. storage area could be an indication of this
The town of Petras is concentrated on a (Tsipopoulou 1999: 852). Such a building atop
series of four hills, particularly the main hill a hill itself coated with houses (an image
where the palace is located. The two houses familiar from Minoan iconography generally)
so far excavated and published to some projects a powerful message not only to trav-
degree (Tsipopoulou and Papacostopoulou ellers (Tsipopoulou 1999: 849) but perhaps
1997) give too incomplete a picture to speak more importantly to locals. The use of monu-
with much certainty about the spatial organi- mental building to visually transmit a unify-
zation of the town around the palace, ing social message is well documented
although it may be pointed out that they are (Fletcher 1995: 149 gives a material behavior-
terraced and seem to be fitted into the hillside. ist approach; also Moore 1996: 92-120). While
Variations on n Theme 77
Petras might not qualify as monumental remains if this earlier building might not have
building by some definitions (Trigger 1990: taken a significantly or even radically differ-
119 suggests that the resource investment in a ent form. Regardless, the imposition of the
monumental building must considerably new palace at Zakros must be seen as a highly
exceed that required by utilitarian concerns, significant event in the site's history and one
for example), from a functional standpoint it responsible for the arguments that Zakros
certainly does, if we consider the entire hill was 'the eastern port of Knossos' (Warren
with the palace atop as providing a visual 1999: 902; also Wiener 1987: 265). This con-
representation of authority and of identity. struction could not have begun any earlier
To summarize: than the beginning of LMIA; and if L. Platon
is right, it was not completed until at least the
• The town of Petras is adapted to fit onto the beginning of LMIB.
hills below and around the palace, subordi- In its initial phase, the palace seems to have
nate to and supporting the central building. been rather permeable, with entrances from
• The palace itself, set atop the hill and ris- the northeast, north, southwest and possibly
ing two or perhaps even three (?) stories, northwest (Driessen and Macdonald 1997:
is placed so as to project a unifying (and 237). The boundary between the palace and
possibly threatening) visual message. the town to the north has been so difficult to
locate that the excavator, N. Platon, felt that
Kato Zakros these houses closest to the palace must have
In sharp contrast to Petras, at Kato Zakros the been incorporated into it. This is no longer
palace is located at the lowest point of the set- believed to be the case (Chrysoulaki and L.
tlement, surrounded and overlooked by the Platon 1987: 80) but remains a powerful illus-
town, which 'seems to dominate the palace' tration of the physical relationship of palace
(Chrysoulaki and L. Platon 1987: 77). The date to town. These factors have led to the opinion
for the establishment of the LMI palace at that the palace was inserted into a pre-exist-
Zakros has been traditionally given as early in ing town of quite different character
LMIA (Driessen and Macdonald 1997: 235). (Chrysoulaki and L. Platon 1987: 80; L. Platon
Recently, L. Platon (1999: 680; also paper 1999: 678).
delivered at the 'Knossos: Palace, City, State' Although the houses of the town are in a
conference, Iraklion, 2000) has argued, based seemingly dominant position to the palace,
on recent ceramic study, that the palace was they are notably smaller and show fewer
not constructed until the beginning of LMIB. examples of elite architecture than might be
While it may still be too early to accept this expected, given the size and quality of the
latest date, it must in any case be considered. palace itself. This is not to say that there are
The precise nature of the Protopalatial no elite indications in the town of Zakros,
Building at Zakros remains uncertain. rather that in comparison with the other
Limited tests below the central court revealed excavated palatial centers, the dichotomy in
earlier court surfaces (L. Platon 1999: 680), but terms of elite signifiers and manifestations of
of course the dimensions and dates of these wealth between the palace and its immediate
are unknown. If the road/'watchtower' sys- neighbors is far greater at Zakros than at any
tem was built from Zakros there must have other site. This is of course true for the larger
been a central authority and, probably, a cen- Central Cretan palatial sites, but is also true
tral building to go along with it. The question when compared to Petras, which shows a
78 Urbanism in the Aegean Bronze Age
relative harmony or balance between town numerous small rooms, used chiefly for storage,
are arranged around a large room with a bench
and palace in those regards.
and a central support, with direct access from the
The town itself lacks intelligibility, a measure street. (Chrysoulaki and L. Platon 1987: 78).
of the extent to which the layout of a town or
city can be understood and hence navigated This central room also usually contains a
from what is immediately visible in a given staircase, where one is evident. The lack of
spot (Hillier 1996: 313-327). With the single intelligibility of the grid and integration of
exception of Harbor Road, the street grid at the houses in the town of Zakros suggest that
Zakros, such as is preserved (Figure 5.2), lacks this street grid was primarily if not exclu-
any degree of axiality and is continuously iso- sively used to service the local inhabitants
vistically blocked. Within this system of low and that this local population may have been
intelligibility the houses themselves are gener- markedly homogeneous. This is supported
ally well integrated according to Hillier's defi-
by the exceptional Building G (Chrysoulaki
nition (1996: 36). As Chrysoulaki states,
and L. Platon 1987: 80) which does have a for-
The majority of the houses at Zakros have the fol- mal entrance (with portico) and a vestibule
lowing interior arrangement on the ground floor: which could serve as a transitional space
Variations on a Theme 79
street and accessible only through a succes- through the settlement rather than between the
sion of two to three vestibules or corridors. domestic buildings, and while wide enough
This need for greater segregation may be a for considerable traffic, it still lacks 'pooling'
consequence of the greater size and density spaces where social interaction could occur.
of the settlement at Palaikastro; however, it Rooms within the houses are not laid out axi-
may also reflect differences in the social ally. This fundamental difference in the struc-
matrix. It is possible that Zakros, with its per- turing of space, coupled with the placement of
meable palace and attached town with simi- main rooms away from the street, might pro-
larly permeable houses, was a more duce a disjunction between public and private
homogenous and isolated area, with a far space. This disjunction itself would necessitate
lower occurrence of non-resident interaction. transitional areas (vestibules, corridors) in the
An examination of the spatial layout of the houses. Furthermore, it might heighten resi-
towns themselves supports this. dents' awareness of the street system as a dis-
At Palaikastro the street grid itself is remark- crete spatial entity. Leaving a house and
ably straight and regular showing a high stepping out into the street, one would have a
degree of axiality, (Figure 5.4) in contrast to clear sense of being in a unified spatial envi-
Zakros. The street system facilitates movement ronment. There is no reason not to believe that
Variations on a Theme 81
from the current archaeological record at an extroverted ritual space - the time needed
Palaikastro. One of these is storage. While to make (presumably regular and communal)
there are storage spaces in the houses, proba- pilgrimages to the peak sanctuary would
bly to the extent of sufficing for the basic serve as a bonding agent. Conspicuously
dietary requirements of the inhabitants of a absent, and expected (though of course not
particular house, the amount of space present proven) is monumental public building,
for food storage is totally inadequate to pro- which likewise provides a clear isovistic mes-
vide for the surplus necessary to conduct the sage across the increasingly larger and more
substantial economic transactions and public complex interaction-communication field
works evidenced at the site (see Knappett (Fletcher 1995: 134-151).
and Schoep 2000: 366 for their discussion of Another missing element at Palaikastro is
the importance of accumulation). The street some form of administrative archive, and
system was almost certainly such a public again given the evidence for economic activ-
work, and as mentioned above, the quarries ity indicated by the finds, and quality of the
at Ta Skaria are among the largest yet known buildings, the scanty documents turned up
for a Minoan site. That such activity was car- so far in the houses themselves are clearly
ried out voluntarily, or under coercion, by insufficient.
local families seems unlikely. There is a decided lack of public open
Secondly, the site displays significant signs space suitable for group interaction at
of stress incurred when population density Palaikastro. Even in settlements where each
and settlement size approach certain thresh- domestic unit contains a large open court-
olds of tolerance having to do with frequency yard, which is not the case here, such public
of interaction and difficulty of communica- interaction spaces are usually found. Given
tion (as outlined by Fletcher 1995). The place- the iconographic evidence suggesting the
ment of central rooms away from the street, importance and even ritualization of large
the serial segregation of internal space, the public gatherings in Minoan civilization, one
street system, which would facilitate large would expect to find such a space in a town
scale communication, and the use of more as large and densely settled as Palaikastro.
durable materials (such as ashlar) all point to Another paradox of the site as excavated is
such stress. Indeed, the habit of rebuilding the evidence on the one hand of social strati-
one or two walls, or even only portions of fication (the presence of elite forms) without
walls in ashlar is usually seen as a desire for clear examples of some of the strata, espe-
prestige restricted by economic considera- cially at the very top.
tions, while the practical, functional aspects One way to identify indicators or manifes-
of such construction, namely that it blocks tations of social stratification is to look at rel-
noise, heat, and smells more effectively, has ative building sizes. Of the 36 buildings
been ignored. clearly discernible, ranging from 48 to 486 sq
Given these signs of stress and reactions to m, 21 fall between 130 and 180 sq m; six are
it, we can perhaps expect to see others. One between 200 and 300 sq m and only four are
of the predicted responses has to do with the less than 100 sq m (Driessen and Macgillivray
need to project a unifying social message. 1989: XXVII; I use only the most securely
This can perhaps be seen on an individual identified houses). In terms of location the
scale with the use of a common repertoire of section of main street bordered by block M to
elite signifiers. It is also one of the effects of the north and D to the south, which is the
Variations on a Theme 83
widest section of the street and which was dwellings themselves. What representations
believed to have had ashlar facades on both or suggestions we do have of a ruling class or
sides, should be the 'best address' in town; entity can be summarized as follows:
however, Building 1, and Block X, located at
the furthest reaches of town from this seg- • Public works, such as the street grid, the
ment are both large ashlar buildings. quarries, the cult building on Petsophas,
Likewise Building 2, which though only par- and perhaps the road system in the sur-
tially excavated, shows an impressive facade rounding territory.
of shaped limestone blocks. Other outlying • Prototypes or primary examples of the
houses, which do seem smaller, such as elite symbols previously noted; in particu-
Blocks Y, S, K and L, produced bronzes, and lar the Palaikastro Hall. Such a local inno-
in the case of Block S two ivory figurines. vation of a new elite architectural form
As far as elite architectural forms go, we strongly suggests that the original exam-
find represented at Palaikastro Lustral Basins ple or at least the largest and most osten-
(2 examples), Palaikastro Halls (3 examples), tatious example remains to be discovered.
one example of a Minoan Hall, ashlar
masonry (11 buildings), mason's marks, fres- Since all of these indicators point to the exis-
coes (10 buildings), cut door-jambs (found in tence of a central building at Palaikastro,
at least 7 buildings) and pier and door parti- almost to the point of demanding such a
tions (3 examples) (Driessen and Macgillivray structure, we should perhaps consider what
1989: XXVI-XXVII; updated with figures from form it might take. Aside from the canonical
the recent excavations). Interestingly there is palace, there are examples, notably Ayia
no one building at Palaikastro that stands out Triadha, of central buildings that are not
over the others, rather, these elite signifiers palaces. Furthermore, and particularly in the
are spread through a number of structures of latter case, there is still the question of
roughly equal apparent status. whether this building would likely be the
In sum, though indications of elite status location of a ruling 'agent' from outside (i.e.
are widespread throughout the excavated Zakros) or itself the seat of power for the
remains, the suggestion is that what we have town and territory around.
so far represents a fraction of the full extent of
the town and specifically represents a frac-
tion inhabited by members of an elite, though Synthesis
not ruling class. There is no indication that
any of these houses and their inhabitants The three sites in consideration show marked
could individually or collectively represent differences in settlement patterns and spatial
the authority or administration needed to organization. All three show signs of organic
create and operate such a town. Even if the or local development, evidenced by internally
houses thus far excavated were home to a consistent growth patterns, with the exception
kind of oligarchic elite who ruled by council, of the palace at Zakros, which may indicate an
one would still expect some sort of instance of outside interference. Yes or no, it
'Bouleuterion'. Furthermore, many of the was certainly the most significant event at the
stress responses demonstrated in the exca- site prior to the final LMIB destruction. This
vated district are private - having to do with construction event happened no earlier than
the structure and composition of the early LMIA and no later than early LMIB.
84 Urbanism in the Aegean Bronze Age
At Petras, although the palace is destroyed demonstrations of status and power which is
and rebuilt with modifications, there seem to island-wide. A member of the elite, living in
be no changes of a scale comparable to the Palaikastro, Zakros, or Petras is fully versed in
initial establishment of the palace and the architectural vernacular of power and able
archive in MMII until, perhaps, its final to imitate, elaborate, or innovate within what
destruction, also at the end of LMIB. can be considered a Central Cretan or more
Though the LMIB period may have been precisely Knossian repertoire. The range of
one of vigorous activity and change at variation suggests certain freedom of room to
Palaikastro (Area/Building 6 was levelled, a maneuver within a nonetheless coherent
wall put up around it and two new wells dug framework of elite status indicators, whose
(MacGillivray et al. 1998:226); part of Building overall coherence, stability and ubiquity indi-
5 was converted into a shrine (MacGillivray, cate at least an 'ideological' state that encom-
Driessen and Sackett 2000: 42); cult practice passed the whole island.
may have changed and new local pottery We must seek new ways to characterize the
styles were developed) it certainly shows interrelationship of sites on Crete and aban-
nothing like the Knossian connections of don previous notions regarding the nature of
Zakros. Quite the opposite, in fact - the events the link between the cultural and the politi-
at Palaikastro seem more consistent with a cal, as suggested recently by Knappett (1999)
greater degree of autonomy or some internal and Knappett and Schoep (2000). If Knappett
shift. is right, and 'ideology was at least as impor-
If the establishment of the LMI palace at tant as (political) economy in the initial emer-
Zakros was a result of some Knossian inter- gence of states on Crete' then might this not
vention there is no sign that such an interven- have remained the case?
tion had any effect on Palaikastro or Petras. The organizational patterns examined
The only possible sign of Knossian (or, in this above which show such divergence are pri-
case possibly Malliot) influence on such a marily concerned with social production -
scale at Palaikastro would be the Minoan Hall control and distribution of assets, exploitation
uncovered in Building 6. However, this fea- of natural resources and social and political
ture is no later than MMIIIB (Macgillivray et organization at the local level. What of social
al 1998: 255) at the end of which period the reproduction? If social production is about
building is destroyed in an earthquake. the survival, growth and maintenance of indi-
Not only is there no sign of either Petras, viduals within a society, social reproduction is
Zakros or Palaikastro wielding any sort of eco- about the perpetuation of the ideological
nomic or political authority over each other, basis for defining that society. As Hillier
there is no indication that in the hypothetical states, 'Social reproduction, we might say,
case of Knossian control of one of these sites requires symbolic forms of space, social pro-
(for which the best case is of course Zakros in duction instrumental forms of space.' (1996:
LMI) such control could have extended beyond 222). While the 'tools' used in each of the three
the limits of that site's own immediate territory. territories to establish and maintain social
The evidence for divergence and diversity is cohesion and economic function were used in
clear. The similarities, which can be summed different ways and with different results,
up under the general heading of 'minoanness', these 'tools' themselves seem to have all come
are also clear. There is a coherent sense of cul- from the same genotype. Knappett, picking
tural identity inclusive of representations or up from Southall, introduces the concept of
Variations on a Theme 85
1971 The Discovery of a Lost Palace of Ancient Crete. Tsipopoulou, M. and A. Papacostopoulou
New York: Scribner. 1997 'Villas' and villages in the hinterland of Petras,
Sackett, L.H., M.R. Popham and P. Warren Siteia. In R. Ha'gg (ed.), The Function of the
1965 Excavations at Palaikastro VI. BSA 60: 248-315. 'Minoan Villa', 203-14. Stockholm: Skrifter
Trigger, B. utgivna av Svenska Institutet i Athen.
1990 Monumental Architecture: a thennodynamic Tzedakis, Y, S. Chrysoulaki, S. Voutsaki and Y. Venieri
explanation of symbolic behaviour. World 1989 Les routes minoennes: rapport preliminaire:
Archaeology 22:119-31. defense de la circulation ou circulation de la
Tsipopoulou, M. defense. BCH 113: 43-76.
1991 Recenti scoperte de epoca minoica nel golfo di Tzedakis, Y, S. Chrysoulaki, Y. Venieri and M. Argouli
Sitia. Seminari CNR 1990: 105-21. 1990 Les routes minoennes: le poste de Choiromandres
1997 Palace-Centered Polities in Eastern Crete: et le controle des communications. BCH 114:43-65.
Neopalatial Petras and Its Neighbors. In W.E. Warren, P.
Aufrecht, N.A. Mirau and S.W. Gauley (eds.), 1999 LMIA: Knossos, Thera, Gournia. In P.P.
Urbanism in Antiquity. From Mesopotamia to Betancourt et al. (eds.), Meletemata: Studies in
Crete, 263-77. Sheffield: Sheffield Academic Aegean Archaeology Presented to Malcom H.
Press. Wiener (Aegaeum 20), 893-902. Liege: Universite
1999 Before, During, After: The Architectural Phases de Liege.
Of The Palatial Building At Petras, Siteia. In P.P. Wiener, M.
Betancourt et al. (eds.), Meletemata: Studies in 1987 Trade and Rule in Palatial Crete. In R. Hagg
Aegean Archaeology Presented to Malcom H. and N. Marinatos (eds.), The Function of the
Wiener (Aegaeum 20), 847-54. Liege: Universite Minoan Palaces, 262-66. Stockholm: Skrifter
de Liege. utgivna av Svenska Institutet i Athen.
6
In the last two decades, a need to contextual- these 'first palaces' following their destruc-
ize the Minoan 'palaces' has caused a shift tion at the end of MM II (around 1700 BC).
from an archaeology of palaces and urban Although the 'first' and 'second palaces'
centres to an archaeology of their hinterlands. seem to have functioned within different
This shift is most clearly evidenced in the socio-political frameworks, their objectives
boom in intensive and extensive surveys of seem to have been broadly similar, namely a
specific regions in Crete, which have now concern with the accumulation of goods, the
covered a considerable surface of the island bureaucratisation which oversaw this accu-
(Driessen in this volume). This increasing mulation and capitalisation or 'putting
interest in the wider social, political and eco- resources to work' (Knappett and Schoep
nomic landscape and more specifically in the 2000: 365-71). This paper will focus on two of
relationship between centre and hinterland these three aspects, accumulation and bureau-
closely follows earlier developments in cratisation. To oversee accumulation, the
Mesoamerican and Near Eastern archaeology Protopalatial centres of Knossos and Phaistos
(Adams and Nissen 1972; Johnson 1973,1987; were making use of regional administrative
Schwartz and Falconer 1994, Blanton, systems and it is no mere coincidence that the
Kowalewski, Feinman and Appel 1981). creation of the first effective writing systems
At the beginning of the Protopalatial on Crete is situated within the wider context
period (c. 2000 BC), the construction of cen- of the emergence of the palaces (Schoep 1999:
tral buildings with monumental character 268). Direct evidence for the accumulation of
directly testifies to the centralized mobilisa- goods can be found in the administrative doc-
tion of labour, materials and construction uments and the presence of storage space in
techniques and indirectly to the emergence of the 'first palaces', while changes in settlement
a new form of sociopolitical complexity. Such patterns and patterns of land-use in the hin-
central buildings were constructed in differ- terland may also provide more indirect evi-
ent parts of the island (Knossos, Malia, dence for exploitation by the centre.
Phaistos and Petras) at approximately the By presenting case-studies from the
same time and each seems to have exercised Protopalatial (Knossos and Phaistos) and
some form of influence over a hinterland (cf. Neopalatial (Ayia Triada and Khania) periods,
Knappett 1999). These central buildings have this paper will highlight two important
long been labelled 'palaces', largely on anal- changes in the application and distribution of
ogy with the even more monumental writing for administrative purposes which
Neopalatial structures which succeeded occurred between these two periods: the first
88 Urbanism in the Aegean Bronze Age
involves a change from regionally located and which were spatially restricted to urban cen-
differentiated administrative systems in the tres and more specifically to their central
Protopalatial period to a single administrative buildings, the so-called 'first palaces' (Figure
system in the Neopalatial period. The second 6.1). This spatial restriction of written docu-
involves changes in the distribution of script ments makes the distinction between centre
and administration, which in the Protopalatial and hinterland or the rural and urban sector
period was restricted to palatial centres but straightforward, a situation quite unlike that
during the Neopalatial period became wide- of the Neopalatial period. The administrative
spread over several administrative levels. documents thus constitute a valuable source
for an assessment of centre/hinterland rela-
tionships. Here, the main question to be
The Protopalatial centres at Knossos and addressed, is how and to what degree the
Phaistos and their hinterlands hinterland was affected by the emergence of
the first 'palatial' centres and more specifi-
In the Protopalatial period, writing is almost cally the presumed need for agricultural sur-
exclusively found on archival1 documents, pluses to support both the ruling elite and
Figure 6.1 Find places of administrative documents in the Proto- and Neopalatial periods
Managing the Hinterland 89
other groups not directly involved in food Knossos, will be used to explore the possiblity
production at the centre. Administrative doc- that the external differences between the
uments provide a valuable window on the administrations correspond to structural dif-
commodities in which the centres were inter- ferences involving not only different book-
ested, while changes in settlement patterns keeping procedures, but also different
may reflect changes in the organisation of economic strategies.
production (cf. Cherry 1984: 32), caused per-
haps by the need to secure greater agricul- Administration at Phaistos
tural surpluses following the emergence of The main document type attested at Phaistos
the 'palatial' centres. In this way the rural is the direct object sealing, of which c. 6000
concerns of the administrations of Knossos were found in a context which seems to con-
and Phaistos will be explored and compared stitute the remains of an archive (room 25)
to relevant data from settlement patterns in and which is either in situ, has fallen from an
their respective hinterlands, in order to upper story or is part of a fill. The sealings
define and compare the strategies used by were found with small numbers of other doc-
these respective centres to accumulate and ument types, such as tablets, roundels and
bureaucratize. direct object sealings. There seems to be no
doubt that these documents had been kept
A View from the Centre: The Rural Concerns of for archival purposes, especially if one con-
the Protopalatial Administration siders the absence of the storage containers,
In the Protopalatial period, Knossos and which the direct object sealings had been
Phaistos were using different scripts—respec- sealing, such as pithoi, wooden chests and
tively Cretan Hieroglyphic and Linear A—to other perishable containers (bags) (Fiandra
keep track of incoming and outgoing goods. 1968). Room 25 is not the only location where
Not only are the script and possibly the lan- administrative documents were found,
guages different at these centres, but also the although it presents the largest number and
types of documents used to organize the variety. Direct object sealings were also found
administration (Schoep 1999: 266-68). The in other rooms (VIII, XXI, LV,VIII, XI, XXI,
range of archival documents from Knossos XXII, XXV, XXVII, LII, LV) (Sbonias 1995:
includes four-sided bars, medallions and cres- 168-69), often in direct association with
cents, whereas those from Phaistos comprise pithoi. In addition, a cache of flat-based nod-
direct object sealings (in their thousands), ules and roundels was found in room LI and
page-shaped tablets, roundels and hanging isolated tablets occurred in rooms LIII-LV
nodules (Figure 6.2). Although the total num- and XXVIII (Schoep 1995 [1997]: 38-43).
ber of documents (102) from Knossos is much Although the majority of direct object seal-
higher than the number of documents from ings sealed door pommels and wooden
Phaistos (61) (Olivier 1994), it must be kept in chests, clay containers were also sealed
mind that the total quantities of documents (Fiandra 1968: 383-97, Militello 2000: 225).
involved are very small, especially in compar- This pattern of sealing allows several infer-
ison with Near Eastern or even Linear B con- ences to be made. The direct object sealings
texts. In the following section contextual seem to be connected with the checking and
analysis of the different ways in which docu- storing of goods in particular storage rooms
ments were used in the Linear A and Cretan as well as the overseeing of their distribution,
Hieroglyphic administrations at Phaistos and on which occasion the sealing was removed
90 Urbanism in the Aegean Bronze Age
Figure 6.2 Range of document types used in conjunction with the Cretan Hieroglyphic and Linear A scripts
and kept for bookkeeping purposes (cf. dent seal-owners (cf. Weingarten 1994: 181)
Ferioli and Fiandra 1990: 222). Since the con- and from this it follows that the majority of
tainers comprise both pithoi and wooden the documents were sealed on the spot. The
chests, it is not unlikely that agricultural com- sealings from containers are therefore most
modities and non-agricultural commodities likely to concern outgoing rather than
were involved (e.g. raw materials and manu- incoming commodities (cf. Fiandra and
factured goods). The system was straightfor- Ferioli 1990: 222).
ward and the number of seal-impressions per This system of direct object sealings seems
sealing could have indicated the quantities or to have been used in conjunction with docu-
the number of measures involved (Godart ments written in Linear A. However, in con-
1990: 147); however, it cannot be excluded trast to direct object sealings, written
that the quantities were standardized2. documents never occur in contexts where
The seal impressions on the direct object bulk commodities were stored. This could
sealings provide information about the num- suggest that they were administering a dif-
bers of seals that were in use and indirectly ferent type of transaction from the direct
about the people involved in this activity. object sealings. Unfortunately, their low
The sealing system at Phaistos is for the main number and fragmentary state precludes an
part intensive, implying that the thousands identification of the type of transaction, but
of impressions were made by a small num- the numerals could point to transactions
ber of seal stones3. This pattern has been involving small quantities of wine (PH 25)
interpreted by Weingarten as reflecting resi- and perhaps a type of grain (PH 1) and thus
Managing the Hinterland 91
would suggest the distribution of these com- few impressions. This type of non-intensive
modities rather than their collection (cf. sealing pattern has been identified with non-
Killen 1969, Schoep forthcoming 1). That resident seal-owners (Weingarten 1996: 302,
written sealed documents were also adminis- 308). The Cretan Hieroglyphic inscriptions on
tering non-agricultural products is suggested this type of document contain a wide variety
by a collection of roundels and flat-based of sign-groups and a number of different
nodules found together with raw materials logograms, some of which can be identified as
and tools in a possible workshop context in grain, olives and wine. The lack of any
room LI (Branigan 1987). In the same area of numerals or other such indication of quantity
the palace, Rooms XXVIII and IL (PH 6 and suggests that crescents were either attached to
PH Zg 47) may have functioned as general the goods or accompanied them in some other
potter's stores, containing over 25% of all the way. It is a fair assumption that the crescents
vessels found in the palace (Branigan 1987: fulfilled a similar function to the Linear B
245-249) and it cannot be excluded that these hanging nodules and represent goods
were also a subject of the administration4. brought in from the hinterland to the centre
Thus, although there is evidence for the dis- (Piteros, Olivier and Melena 1990). This
tribution of commodities from palatial stor- would imply that goods, including agricul-
age rooms in different parts of the palace at tural produce, were mobilized from the hin-
Phaistos, the administrative documents do terland, an action for which no administrative
not provide any information as to how these evidence is available at Phaistos.
commodities were procured or where they Further indication of the direct involve-
came from5. Consequently, the administrative ment of the central administration at Knossos
records at Phaistos do not seem to yield any in its hinterland is provided by the (perfo-
direct information concerning the mobilisa- rated) four-sided bars (Figure 6.2). The num-
tion of agricultural commodities from its hin- bers on this type of document are sometimes
terland. This conclusion is surprising, since very large, including hundreds and even
the mobilisation of agricultural surplus is thousands. Although the bars contain several
considered to be one of the main functions of entries no logograms are recorded, so pre-
the 'first palaces' (Branigan 1987; Halstead sumably individual entries refer to the same
1981)". commodity. As to what commodity can be
booked on these bars, two possiblities may be
Administration at Knossos suggested: 1) people: the omission of the
Contextual analysis of the administrative doc- logogram when booking a commodity in the
uments from Knossos indicates quite a differ- later Linear A and Linear B administrations is
ent picture from that outlined for Phaistos, most common in connection with the
since at Knossos the administrative docu- logogram for people (Schoep forthcoming 1);
ments do seem to provide evidence for the 2) sheep: the structure of the inscriptions on
collection of goods7. The main sealed docu- the four-sided bars bears resemblance to the
ment type at Knossos is the 'crescent', a clay Linear B sheep tablets (cf. Olivier 1994-1995:
nodule shaped around a string (Figure 6.2). In 266), in that they almost exclusively book
contrast to direct objects sealings, crescents even numbers. If people are booked this
were attached indirectly to the object that was could suggest the existence of census records,
sealed and moreover exhibit a different seal- which could have formed the basis for taxa-
ing pattern, with many different seals making tion, whereas if sheep are booked this could
92 Urbanism in the Aegean Bronze Age
indicate the existence of a specialized staple for the large-scale collection of agricultural
finance by the Cretan Hieroglyphic adminis- commodities, archaeological evidence for
tration at Knossos. At any rate, irrespective of such a mobilisation is evidenced by the
whether people or sheep are booked, the pres- ample storage space in the 'first palace' and
ence of these large numbers implies that the the construction of the kouloures, which are
administration was concerned with a region hard to interpret as anything other than stor-
of considerable size and thus testifies to the age for a specialized surplus (Strasser 1997,
centralized exploitation by the centre at Halstead 1997). The Early Minoan I and II set-
Knossos of its hinterland; It is perhaps worth tlements in the Mesara are spaced equally (c.
noting that there is also abundant evidence 8-10 km) and although the best locations
for the presence of exotic materials and goods were settled first, a filling-in of the space
at Knossos in the Protopalatial period between the existing settlements can be noted
(Warren 1995), while at Phaistos these seem during EM III/MM IA (Sbonias 1995: 13).
to be lacking (Branigan 1987, Carinci 2000). This pattern seems to suggest that each vil-
To conclude, it may be noted that whereas lage had a claim to the agricultural land
the documents from Knossos provide indica- immediately around it (cf. Murphy 1997: 27).
tions for its direct exploitation of the hinter- It may therefore be assumed that during EM
land, this does not hold true for the II and EM III/MM I A, agricultural produc-
documents from Phaistos, where documents tion would have been organized at a local
seem to focus primarily on the distribution of scale, with individual villages probably being
goods. This of course does not imply that largely self-sufficient when it came to food
Phaistos was not mobilising any goods, but it supplies. Other items such as seal stones,
could suggest that this mobilisation was orga- stone vases, daggers, pottery and perhaps a
nized and administered in a different way. variety of other goods may have been pro-
Surpluses may be mobilized in different cured through a wider exchange network.
ways, such as for example, tax, direct produc- At present, it remains unclear as to what
tion and/or exchange (Halstead 1999). If real, degree the construction of the 'palace' at
these differences would seem to indicate that Phaistos in MM I affected this situation. The
the administrations at Phaistos and Knossos MM I-II period in the area to the west of
may have exploited their hinterlands in dif- Phaistos surveyed by Watrous et al. seems to
ferent ways, which perhaps reflect different be characterized by a growth in the number
strategies of integrating the hinterland. If so, and size of settlements (Watrous et al. 1993)
then one might reasonably expect these dif- (Figure 6.3). However, in the Asterousia sites
ferent strategies to find some expression in were abandoned (Branigan and Blackman
the settlement patterns of these different 1977:198), which would seem to suggest that
regions. at least here surplus production was not pro-
cured through capital investment in the
improvement of marginal land (Halstead
A View from the Hinterland: Settlement 1992; Manning 1994: 236). There is no evi-
Patterns and Land-use dence for the foundation of new settlements
in the central Mesara in MM I-II and in gen-
Phaistos and the Mesara eral settlement growth may have been static
Although the administrative documents (Branigan 1970: 127). Sbonias notes that in
from Phaistos do not contain clear evidence MM IB, competition between villages
Managing the Hinterland 93
appears to have diminished considerably suggest the direct involvement of the centre
(Sbonias 1995: 132-133). The static growth in in the production of large-scale surpluses
MM I-II settlements in the central Mesara within an area to the east of Phaistos. The
(Sbonias 1995:16) may be contrasted with the data available at present instead suggest con-
apparent growth in settlements in the area tinuity of local intensive agriculture, which
immediately around and to the west of could not have yielded such surpluses. If
Phaistos. If this is real and not due to the dif- agricultural surpluses were collected from
ferent degree to which the two regions have this area, they would have been restricted in
been researched, it could suggest that the quantity. There is, however, some evidence to
impact of the 'first palace' at Phaistos upon suggest that the palace of Phaistos may have
settlement patterns may have been limited. been more actively involved in extensive
One might even tentatively suggest that pat- agricultural exploitation, in the area west of
terns of land-use remained more or less the Phaistos. Also worth considering in this con-
same and were not affected significantly by text is the site of Monastiraki in the Amari
the need of the centre to mobilize agricultural valley, which features a monumental central
surpluses. The continued distribution of building with ample storage space and abun-
farmhouses and small villages in the land- dant links with Phaistos (Matz 1951, Kanta
scape would rather suggest the widespread 1999: 387-93). Not only is the pottery very
continuation of intensive local agriculture, similar to that of Phaistos (Kanta 1999), albeit
which need not have produced large-scale of a lesser quality than the finely decorated
surpluses for the centre (Halstead 1992). ware at Phaistos (Walberg 1983: 91), but also
During this same period, nucleation has an administrative link is suggested by the
been noted for settlements such as Koumasa hundreds of direct object sealings, some of
and Platanos and this may indirectly testify which were impressed by sealstones similar
to some sort of extensification of agriculture or even identical to those at Phaistos (Godart
(Halstead 1987: 83). However, since the num- 1999: 40). As at Phaistos, the sealing system is
ber of abandoned settlements in MM I-II is intensive and on present evidence, particu-
quite low (Sbonias 1995: 16) and moreover, larly the absence of written documents8, it
since the nucleation at sites such as Platanos seems fair to argue that the administration at
and Koumasa predates the emergence of the Monastiraki was secondary to that at
palace in MM IB (Sbonias 1995: 32), any con- Phaistos. The first phase at Monastiraki dates
nection between a nucleated settlement pat- to MM IB/IIA (Walberg 1983: 89) and it may
tern and the emergence of the 'first palace' at conjecturally be suggested that an expansion-
Phaistos is difficult to sustain and cannot ist strategy by Phaistos with the aim of secur-
therefore be used as evidence for the direct ing a larger hinterland may have led to the
involvement of that centre in the extensifica- establishment of this centre in the Amari. The
tion of local agriculture for the production of absence of a developed settlement hierarchy
surpluses. in the Amari in the Prepalatial and
And so, if the differences in settlement pat- Protopalatial periods would seem to corrobo-
terns are real, it would seem that, following rate this hypothesis suggesting that any
the construction of the 'first palace' at expansionist plan by Phaistos would have
Phaistos, the westernmost part of the western met less resistance in this area than if it had
Mesara fared differently from the area further been implemented within the Mesara itself9.
east. The archaeological evidence does not Besides the possiblity for the mobilisation of
94 Urbanism in the Aegean Bronze Age
Figure 6.3 Area surveyed by the Western Mesara Survey (after Watrous et al. 1993, fig. 8)
agricultural produce, Monastiraki is also sit- less conclusive. Cadogan has tentatively recon-
uated on a natural route to the north coast, structed the hinterland of Protopalatial
which may have played a role in the move- Knossos as stretching as far east as Gouves, as
ment of goods along this north-south axis10. far south as Ayia Varvara and possibly as far
west as Tylissos, including part of the
Knossos and Central Crete Myloptamos district (Cadogan 1994: 66-68).
When one applies the same approach to the Protopalatial Knossos itself has been estimated
Protopalatial centre at Knossos, the results are to cover c. 45 ha (Whitelaw 1983: 340) and it is
Managing the Hinterland 95
clear therefore that this centre was consider- goods was centralized, rather than decentral-
ably larger than Phaistos and consequently ized. This centralized mobilisation of goods,
required even larger agricultural surpluses to presumably from what are different locations
feed its population; this may also be implied by in the hinterland, indicates an important dif-
the larger size of the kouloures (Branigan ference from Phaistos. It was suggested that
1987). Although the administrative documents because settlement patterns to the east of
from Knossos provide convincing evidence for Phaistos seem to have changed very little in
the exploitation of its hinterland, the picture MM I-II, land-ownership and land-use are
obtained from the settlement patterns in the not likely to have changed significantly.
presumed hinterland is at best vague. We are Expansion into the Amari valley may there-
confronted in the first place by insufficient fore have been at least partly connected with
knowledge of the hinterland itself, which is the possibility that no large agricultural sur-
mainly due to the small number of excavated pluses were acquired from the part of the
sites (Figure 6.4). Furthermore, although Hood Mesara to the east of Phaistos. If Monastiraki
and Smyth (1981: 8-10) note that the extent of were indeed a secondary order centre of
Knossos town expanded greatly in MM I-II, no Phaistos, then this decentralisation contrasts
surveys have been conducted in its alleged sharply with the administration at Knossos,
hinterland, with the result that it is unclear as where all control may have been concen-
to what degree MM I-II regional settlement trated at the centre. It would seem therefore
patterns were affected by this. Known that external administrative differences
Protopalatial sites, which presumably lay in between Knossos and Phaistos (i.e. script,
this hinterland are Poros-Katsambas, Skalani, document types) do also correspond to dif-
Archanes, Vitsila, Ayios Myron, Kato Vathia ferences in the economic exploitation of their
and Dalia, louktas, Ayia Marina, Krousonas, respective hinterlands.
Gouves, Amnissos, Skamni (area of Gonies
Maleviziou) and Tylissos, however very few of
these sites have actually been excavated. The Concerns of Neopalatial Administration:
Excavations at Tylissos and Archanes have the Cases of Ayia Triada and Khania
provided the best evidence, but the sketchy
nature of the remains do not allow any defini- In contrast to the Protopalatial period, in the
tion of the nature of these Protopalatial settle- Neopalatial period written administrative
ments and their relationship to Knossos. The documents are no longer restricted to the
reservoir at Archanes dates to the Protopalatial palatial centres, but now occur in settlements
period and the quality of its construction is of varying sizes, in central buildings as well
unparalleled in the Phaistos hinterland. as in other contexts (Schoep 1996 [1999]). This
Functional differentiation between sites is per- implies that the clear administrative distinc-
haps more pronounced: Poros-Katsambas is tion, which could be made in the
already a harbour of Knossos and there are Protopalatial period between the urban and
perhaps more specialized ritual sites, such as rural sector, becomes more complicated. At
Anemospilia, louktas and other peak sanctuar- Ayia Triada, for example, although the Linear
ies (Nowicki 1994: fig. 8)". A tablets imply that the Villa reale' was
To conclude, the crescent-shaped nodules administering to a hinterland of some size
from Knossos seem to suggest that the collec- (Schoep forthcoming 1), it is no longer clear
tion of agricultural and perhaps also other whether the settlement was a rural outpost of
96 Urbanism in the Aegean Bronze Age
Figure 6.4 Map of Protopalatial Sites known in Central Crete (adapted from Cadogan 1994). 1. Tylissos, 2. Kavrochori,
3. Gazi, 4. Giophyrakia, 5. Poros, 6. Prassa, 7. Gournes, 8. louktas, 9. Archanes-Phourni, 10. Archanes village, 11.
Vathypetro, 12. Kastelli (Vitsila), 13. Vorou, 14. Kato Vathia, 15. Anopoli, 16. Ayios Myron, 17. Krousonas, 18. Gonies
(Skamni), 19. Amnissos, 20. Gouves, 21. Ideori Antron
a larger urban centre or whether it was an ments were found, whether private, public or
independent urban centre. In order to illus- religious, 2) the types of document attested,
trate this increased administrative complex- 3) their contents and 4) their palaeography.
ity, I will focus on the cases of Ayia Triada
and Khania. Different criteria can provide Private Administration and Central
information as to the nature of the adminis- Administration
tration conducted in a particular location: 1) There seems to be little doubt that the 'villa'
the nature of the buildings in which docu- at Ayia Triada housed a central archive and
Managing the Hinterland 97
and KI-RO are directly opposed deal with AB land owned by the central administration
120, which is either barley or wheat. On HT (Schoep forthcoming 1). The fact that differ-
15, a deficit of AB 120 is booked and the fact ent commodities were booked in different
that the numerals '570' and '684' are multi- administrative contexts, which are sugges-
ples of '57' and 400 is one greater than a mul- tive of another provenance or type of contri-
tiple of 57 (Figure 6.5), could suggest that this bution, suggests that the interests of the
commodity was also taxed. (Killen 1969) central administration at Ayia Triada were
The existence of a type of tablet, on which selective and that the collection of large-scale
incoming goods are balanced against outgo- agricultural surpluses was specialized.
ing goods, has important implications since it The extent of the hinterland of Ayia Triada
hints indirectly at the existence of target remains unknown and it is not clear what the
records for expected contributions for wheat relationship between Ayia Triada, Kommos
or barley, olives and another unidentified and Phaistos was. The quantities of grain
commodity. Whether these reflect contribu- booked on the Ayia Triada tablets may pro-
tions from land owned by the central admin- vide an indication as to the amount of arable
istration or land that was privately owned land needed for its production. It has been
cannot be determined, but since the ratios estimated by Palaima, using Halstead's fig-
between the numerals are suggestive of ures, that the total of barley (AB 120) in the
taxation, this could perhaps imply that the Ayia Triada tablets would have required 830
land was not owned directly by the central ha or roughly 8km2 arable land (Palaima
administration. 1994: 319, Halsteadl981), a relatively small
It is interesting to note that other agricul- area. With regards to possible evidence for
tural commodities such as wine and figs, the involvement of the centre in extensive
which also seem to have been collected at a agriculture in order to produce large-scale
large scale, are booked in a different contex- surpluses of grain, it is interesting to note
tual position. This would seem to suggest that a hundred oxen (AB 23 BOS) are perhaps
that they were acquired in another way or booked on a tablet from the villa (HT 34).
from a different source, such as, for example, The Ayia Triada tablets also testify to the
Managing the Hinterland 99
existence of a dependent workforce, part of Triada are thought to derive from non-local
which was receiving allocations of food- clays, which, if correct, would imply that
stuffs. The fact that they were receiving food- documents were travelling from settlements
stuffs could suggest that they were to Ayia Triada (Hallager 1996: 212-13, Schoep
employed by the central administration, per- 1999a: 214-217).
haps on lands owned by the centre12. If these
foodstuffs were leaving the magazines of the
'villa', this could suggest that these work- Summary
forces were located at or very near to the cen-
tre. The tablets also seem to deal with people An assessment of the rural concerns of the
who were possibly located further away as urban administration at different centres
some groups of people are associated not forms an interesting perspective through
with rations but with large quantities of which to consider the changing relationships
figs or wine, in contexts which could between centres and their hinterlands. In the
suggest contributions to the centre, and are Protopalatial period, the distinction between
thus likely to be located in the hinter- centre and hinterland is clear with adminis-
land rather than at the centre (Schoep tration confined to central buildings. A com-
forthcoming 1). parison of the administrative documents
Contextual analysis of Linear A tablets from two different centres, Knossos in north-
therefore suggests the existence of a hinter- central Crete and Phaistos in the Mesara, sug-
land administered by the central administra- gests the existence of administrative
tion at Ayia Triada, which, according to the differences between these centres involving
quantities of grain booked in the tablets, different scripts (Linear A versus Cretan
need not to have been very extensive. Hieroglyphic); and different typologies of
Unfortunately, since toponyms cannot be documents; in addition these administrative
identified with certainty, the tablets do not differences appear to correlate with differ-
provide any information regarding the extent ences in economic strategies and spatially in
of the hinterland of Ayia Triada. The sign- the way in which each centre exploited its
group PA-I-TO, which is the Linear B name hinterland. This suggests that not all 'palatial'
for Phaistos, occurs on two tablets of which centres were structured in the same way. For
one books grain (HT 120) and another a the Neopalatial period, it is harder to assess
group of people (HT 97). No written docu- centre-hinterland relationships because of an
ments, which could shed light on the nature increasing administrative complexity, which
of the administration at Ayia Triada have is evidenced by the proliferation of Linear A
been found either at Neopalatial Phaistos or at several levels, and because of the difficul-
Kommos. Despite the lack of information in ties involved in relating this administrative
the tablets, however, it is clear that Ayia complexity to settlement hierarchy. Urban-
Triada participated in a wider network of rural relationships now operated at different
contacts, which extended beyond the Mesara. levels and scales, varying from the small-
For example, a pithos from the 'villa' is scale and possibly private, as suggested by
inscribed with the sign-group AB the evidence from Khania, to the large-scale
58-67-53-59/SU-KI-RI-TA, which is the and public, as at Ayia Triada. Thus, whereas
Linear B toponym for Sybrita in the Amari in the Protopalatial period, the presence of
valley. In addition, some nodules from Ayia administrative documents is closely linked to
100 Urbanism in the Aegean Bronze Age
political supremacy, this cannot so clearly be which could point towards the collection of goods, such
as bags or rush matting.
argued for the Neopalatial period.
7. The absence of direct object sealings is striking, but it
is not clear whether this is due to the hazard of discov-
ery or whether some other document type, e.g. the
medallions, fulfilled this function in the Knossos
Acknowledgments administration.
8. With the exception of the logogram for wine on a
I am very grateful to Keith Branigan for direct object sealing.
9. For a survey of the area around Apodoulou see
inviting me to represent the 'administrative
nwXo-yuipyn 1987: 125-160.
point of view' at the Sheffield Round Table. I 10. There has been some discussion about whether
would also like to thank Paul Halstead, Todd Kommos in the Protopalatial period was actively used as
a harbour for trade with the East and Egypt (Shaw 1998).
Whitelaw and Cyprian Broodbank for valu-
Carinci (2000) has pointed out recently that this is
able comments on the paper delivered dur- unlikely, considering the limited number of exotica at
ing the Round Table. Thanks are also due to Kommos, Phaistos and all the sites in the hinterland. It
P. Tomkins for correcting the English and for must be noted that at Phaistos hardly any objects testify-
ing to external trade have been found.
comments and support. Finally, I would also 11. The distribution of Kamares ware seems to have
like to thank the Reseach Foundation been more widespread than assumed at first and the dis-
Flanders (F.W.O.), who funded a 12 month tinction made by Walberg between palatial and provin-
cial Kamares cannot be sustained (Day and Wilson 1998:
stay at the University of Sheffield in 352).
1999-2000. 12. The generic logogram for people is ligatured to dif-
ferent signs, which seem to indicate administrative dif-
ferences between different groups of people. It is
therefore not unlikely that different workforces were
Notes employed by the central administration. Because of the
resemblance of the ligatured sign which the logogram for
1. Inscriptions on vases and seal stones are not archival people is holding, to a tool (e.g. hoe), it is tempting to
because they were not intended to be kept in the interpret the logograms A 570, A 571 and A 571 as labour-
archives, although they probably served an administra- ers on the land, but this of course must remain conjectural
tive function. It is noteworthy that inscriptions on vases (Schoep forthcoming 1).
and seal-stones were always in Cretan Hieroglyphic and
occur mainly in the north-central and north-east part of
the island. Only in the Neopalatial period was writing
first used on a large scale for purposes other than admin- Bibliography
istration, with inscriptions now occuring on a variety of
supports, such as libation tables, architecture, metal Adams, R. McC and H.J. Nissen
objects etc. 1972 The Uruk Countryside. Chicago: University of
2. It is interesting to note in this respect that in this Chicago Press.
period, conical cups were mass-produced in three sizes Blackman, D. and K. Branigan
(large, medium and small) (Fiandra 1973). 1977 An archaeological survey of the lower catchment
3. Of these 327 different seals, 44 (or 13%) account for area of the Ayiofarango Valley. BSA 72:13-84.
70% of all sealings.
Branigan, K.
4. Tablet PH 6 does not contain any logograms and thus
1987 The economic role of the first palaces. In R.
may not have dealt with agricultural commodities.
Hagg and N. Marinates (eds.), The Function of
5. In Alalakh, the absence of documents concerning
incoming commodities seems to have been due to the the Minoan Palaces, 245-49. Goteborg: Paul
fact that these commodities were stored elsewhere Astroms Forlag.
(Magness-Gardiner 1994: 44). 1970 The Tombs ofMesam. London: Duckworth.
6. One possibility is that the seal-stones, which have Blanton, R.E., S.A. Kowalewski, G.M. Feinman and J.
been used only once or twice (c. 283, cf. note 2), reflect a Appel
non-intensive sealing patttern, which could suggest non- 1981 Ancient Mesoamerica: a Comparison of Change in
resident seal-owners. It would be interesting to know Three Regions. Cambridge: Cambridge University
whether these were sealing a specific type of support, Press.
Managing the Hinterland 101
claims about the settlement must involve a east slope and south slopes. In the ensuing
degree of abstraction and be generalizing; this sections, I refer to a number of topological
is imposed by the idiosyncratic and piecemeal districts on the Kadmeia, which generally fol-
nature of archaeological exploration, as well low her morphology. Every effort has been
as a hiatus in the publication of preliminary made to classify consistently individual sites,
reports between 1984-1992. I trust that the although the areas serve as conventional ana-
Archaeological Service has reported all sub- lytical units, not as rigid compartments
stantial Middle Helladic remains, excavated reflecting the structure of the settlement.
before and after this period.
That said, I would like to start with a brief
note on the Theban landscape, since the loca- Extent and density
tional analysis of the data presupposes some
familiarity with topological features of the Most areas of the Kadmeia show evidence of
Kadmeia. domestic and mortuary use in Middle
Helladic times: the north and northeast
slopes of Ag. Andreas- thereafter 'Ag.
Topography Andreas' (Demakopoulou 1976a; 1979a, c;
Aravantinos 1981; 1982a); the area east of
The modern city of Thebes stands out from Pouros (Spyropoulos 1970b; 1971a, b;
the surrounding low hills as a built-over, Spyropoulos and Chadwick 1975) and south-
pear-shaped plateau, with the beds of east of Pouros (Touloupa et al. 1966b, c;
Strophia and Dirke rivers running along the Faraklas 1968b; Demakopoulou 1975a;
east and west foothills respectively (Figure Aravantinos 1982c; Andrikou 1994); the ridge
7.1). The insight that the original relief would over the east and south slopes (Touloupa et al.
have looked more rugged than today belongs 1965b; Faraklas 1968c; Symeonoglou 1973;
to Antonios Keramopoullos, who suggested Demakopoulou 1975b; 1976b; loannidou
that the prehistoric Kadmeia comprised at 1973) and the slopes themselves (Touloupa et
least four distinct hills (Keramopoullos 1909: al. 1964; 1965a, c; Spyropoulos 1969a; 1971c;
107-10). To the south, Ag. Andreas features Demakopoulou 1979b; 1980; Sampson 1980;
even today the highest elevation. To the west, Aravantinos 1982b; 1994a, b; Andrikou 1993).
the conical-shaped Pouros tou Kavallari may A series of neighbouring plots along the north-
have been just as prominent before its top west slope, above Gourna, have also preserved
was levelled. A lower hilltop apparently substantial evidence of Middle Helladic use
existed east of Pouros; it is now razed and (Touloupa et al. 1966a; Spyropoulos 1969b;
occupied by the modern town market. A 1970a; Demakopoulou 1973^b).
deep ravine, that still demarcates the north- Even the precipitous west part of the
west edge of the Kadmeia (Gourna), sepa- Kadmeia (Touloupa et al. 1966d) and the
rated the higher areas from two small knolls foothills within the area bound by the rivers, to
to the north (the archaeological museum site northwest, northeast and southeast, were occu-
and the 'Ampheion'). A narrow ridge run- pied (Faraklas 1968a; Demakopoulou 1973-4e;
ning just above the east slopes of the 1975c; 1978; Sampson 1981; Aravantinos 1983;
Kadmeia seems to have enabled access to the Piteros 1983). Moreover, the excavations west
central part of Thebes from the north. We can of the museum have recently demonstrated
envisage the main routes ascending from the that this area was used as a burial site
Plotting Fragments 105
Figure 7.1 Topographical map of Thebes (based on Symeonoglou 1985). A. Ag. Andreas, B. Pouros tou Kavallari, C
Area east of Pouros (including the levelled House of Kadmos hill-top), D. NW slope, E. NW foothill (Gourna area), R
Museum area, G. Towards Ampheion knoll, H. Towards Kastellia, I. East foothill, J. East slope, K. East ridge, L. SE
slope, M. Area S/SE of Pouros, N. Towards Ismenion, O. Towards Kolonaki-Ayia Anna.
106 Urbanism in the Aegean Bronze Age
(Aravantinos, pers.comm.). Previously, it was not so well sampled areas on the other (aver-
generally assumed that the entire north part of age overall REU=5.5 per area).
the Kadmeia was not used in this period. The uneven sampling of the Kadmeia pre-
Although the new excavations demonstrate vents us from comparing frequencies of plots
that this was not the case, it is striking that a with Middle Helladic remains across these two
single plot with Middle Helladic remains has categories. However, it allows us to make some
been reported within the central block defined inferences about the frequencies of plots with
by Pindarou, Oidipodos, Epameinondou and Middle Helladic remains within each category,
Antigonis streets (Spyropoulos 1970b). in this case the relatively well-sampled areas.
Another lacuna in the southwest sector of the Definitive conclusions are, of course, contin-
modern grid, i.e. south of Oidipodos and west gent on actual densities of remains within indi-
of Epameinondou street, may reflect a steep vidual plots. Let us focus on the areas where
extremity of the prehistoric Kadmeia. Very few archaeological exploration has been more sys-
prehistoric deposits have been reported there. tematic (average overall REU=15.5 per area).
On the basis of domestic remains mainly, Almost 60% of excavated plots along the cen-
the extent of the settlement has been esti- tral east slope of the Kadmeia have preserved
mated to reach 8 ha (Symeonoglou 1985). Middle Helladic remains (Figure 7.2). Further,
This approximation rests on the assumption nearly 40% of REU's in Ag. Andreas and
of a fortification, though, no trace of which south/southeast of Pouros and 30% of exca-
has survived (see below). An updated distri- vated plots east of Pouros, at the east ridge and
bution of REU's with domestic remains the east foothill have produced evidence for
points to a higher figure, perhaps around 20 Middle Helladic use. About 20% of excavation
ha. This does not include more peripheral units at the southeast and northwest slope,
areas, for example the museum site and the finally, have produced such evidence. I am not
northwest and east foothills; in addition, we trying to suggest that these decreasing figures
correspond exactly with the density of the
should allow for the possibility that isolated
Middle Helladic settlement in the given areas.
dwellings existed beyond the Kadmeia, as a
The possibility of a very uneven preservation
single, child burial near Kastellia might sug-
of sites by area should not be ultimately
gest (Demakopoulou 1973-4d).
excluded but it is feasible that the highest per-
If the distribution of excavated remains
centages at the east slope, Ag. Andreas and
hints at the spread, it does not reveal much
south/southeast of Pouros reflect a denser
about the density of the settlement. A rough
occupation there.
approximation of relative densities in various
I would like to proceed to an examination
areas could be arrived at by examining dif-
of types of use (domestic or mortuary) on the
ferences in the number of REU's with Middle
Middle Helladic Kadmeia, since this can
Helladic remains throughout the Kadmeia
enhance the resolution of the picture I have
(percentages of overall REU's in each area, so far portrayed.
regardless of presence or date of finds). But
we need to draw a distinction between areas
representing a larger sample of excavated Spatial Analysis of Domestic and Mortuary
space on one hand (average overall Use
REU=15.5 per area; Ag Andreas, southeast
slope, the areas southeast, south and east of Comparatively more REU's, among those situ-
Pouros, the east and northwest slopes) and ated in the better excavated districts of the
Plotting Fragments 107
Kadmeia, have yielded houses and domestic slope, while even less excavations have pro-
deposits along the east slope and south/south- duced domestic remains at the northwest slope
east of Pouros. Ag. Andreas and the east ridge (Figure 7.3). Houses and domestic deposits
follow closely. Fewer such plots are known east have also turned up in a few plots west of
of Pouros (near the centre) and the southeast Pouros, the northeast slope, as well as the
Figure 7.3 Survey of REU's with houses, domestic deposits and tombs.
108 Urbanism in the Aegean Bronze Age
northeast, northwest, and the southeast tombs and houses where these features seem
foothills. Their small number seems to be con- to coexist. In this respect, it is beneficial to see
gruent with the fewer excavations undertaken which parts of Middle Helladic Kadmeia have
there by the Archaeological Service, although it yielded domestic burials. It is important to
is likely that these extremities were not as stress that the distinction between 'intramural'
densely inhabited as other areas further uphill. and 'extramural' tombs does not apply, since
Most REU's with Middle Helladic tombs are no trace of a Middle Helladic circuit has ever
located at the east slope, a pattern which cor- come to light (see below). Thus, when I speak
responds to the attested frequency of domestic of 'domestic burials', I mean the tombs associ-
use there (Figure 7.3). The area south/south- ated with houses regardless of their location.
east of Pouros and Ag. Andreas follow; com-
paratively plots with tombs have been less
frequent. Plots with evidence of mortuary use Houses
have been frequent at the northwest slope,
where those with domestic remains have been Most REU's with domestic finds (aside plots
rare. The frequency of REU's with tombs at the with deposits only) are known to contain
southeast slope, the east ridge and east foothill buildings showing rectangular or trapezoidal
is lower, and in the case of the east ridge, it is plans (Figure 7.4). Apsidal houses have
again in contrast with the occurrence of plots also come to light (Touloupa et al. 1966a;
with domestic remains. The same applies to Demakopoulou 1973a; 1976a; Andrikou 1994),
the area east of Pouros, where tombs have sometimes side-by-side to rectangular build-
been reported rather infrequently, unlike ings.1 REU's with apsidals seem to be com-
houses and domestic deposits. In short, both paratively more common in Ag. Andreas.
plots with domestic remains and tombs have This might be again the result of a denser
more frequently turned up southeast/south of occupation or simply preservation. Plots with
Pouros, along the east slope and at Ag. non-apsidal houses have also been cited
Andreas. It might not be irrelevant that, as I along the northeast, northwest and southeast
have argued above, these areas may have been foothills of the Kadmeia, west of Pouros and
more densely occupied. Tombs have, of at the northeast slope. On the other hand,
course, been reported in the not so well-sam- traces of 'wattle-and-daub' huts have been
pled areas too, and at Kastellia. Interestingly, erroneously reported near Ag. Andreas
'pre-Mycenaean' sherds have been found in a (Filippaki et al. 1967).2
couple of chamber tombs at the Late Helladic The published excavation data give us a
cemetery of Kolonaki-Ayia Anna, c. 200-600 m rough idea about building standards in
south of Ag. Andreas (Keramopoullos 1917: Middle Helladic Thebes. The foundations
146,196). built on stereo or Early Helladic strata, can be
With the exception of the east foothill and quite deep (Symeonoglou 1973). The socles
Kastellia, tombs have come to light in the are of unworked rubble or slabs, their width
vicinity of houses. Does this suggest that ranging between 0.55-0.80 m, and support a
domestic space was undifferentiated from mudbrick elevation. They can be clay-coated
burial space in the south and central part of (Aravantinos 1983). Trodden earth floors are
the Kadmeia? To be able to approach this quite common, but red or white clay floors
question, it is important to attain a better are not unusual (Touloupa et al. 1966b;
understanding of the relationship between Demakopoulou and Konsola 1975; Andrikou
Plotting Fragments 109
1993; 1994). Storage clay-lined pits and the Middle Helladic and spans the transition
smaller refuse pits are located inside or out- to the Late Helladic period. Built of monolithic
side the domestic units. The roofing material slabs, it can be quite large (Touloupa et at. 1964;
seems to have included timber logs (Faraklas Faraklas 1968a; see below).
1968c). Hearths and ovens must have been Likewise, plots with burial pits are more
more frequent than cited (Faraklas 1966; common along the east slope, but not as much
Demakopoulou and Konsola 1975; cf. Konsola as those with cists (Figure 7.5). Apparently,
1981: 113). the spread of REU's with burial pits over the
Kadmeia is more limited, though Middle
Helladic pits have been reported at the south-
Graves east and northwest edges of the citadel.
Construction-wise, the pits are opened in
It can be asserted that REU's with rectangular stereo or earlier strata, their walls lined with
cists are in general more common compared to mudbricks or clay plaster (Demakopoulou
plots with burial pits or pithoi. They are fre- 1978; Sampson 1981). They are covered with
quently reported in most parts of the soil or stone slabs (Demakopoulou 1975b;
Kadmeia, especially along the east and north- 1980). Further, plots with burial pithoi have
west slopes (Figure 7.5). They are also present been more common at Ag. Andreas. Although
at the southeast ridge and foothill, the north- they have been reported rather infrequently,
east slope and the museum area. In most their presence at the northeast and northwest
cases, the cists are stone-lined, covered with extremeties and Kastellia (Spyropoulos 1970a;
slabs and have floors of pebbles, clay or stereo. 1971c; Demakopoulou 1973-4d) may reflect a
A more refined version appears at the end of more widespread use than the one attested.
110 Urbanism in the Aegean Bronze Age
Housing the living and the dead higher frequencies occur at the east and
northwest parts of the Kadmeia; they form a
A survey of REU's with types of burials across rather distinct batch towards the right part of
the Kadmeia (Figure 7.6) indicates that plots Figure 7.7, culminating at the east slope. It is
with children burials only are more frequent difficult to know for a fact whether single
closer to the centre of the citadel, namely Ag. tombs are 'non-domestic' burials, instead let
Andreas, east ridge and the area south/south- the focus be on the occurrence of evidenced
east of Pouros. Plots with both adult and chil- domestic burials as specified previously. The
dren burials are seemingly not confined to distribution of REU's with domestic burials is
specific areas. In addition to the areas included quite widespread on the Kadmeia. Aside from
in Figure 7.6, they have also been reported the districts included in Figure 7.7, they are
at the northwest and southeast foothills also reported at the southeast and northwest
(Demakopoulou 1978; Piteros 1983). It should foothills (Demakopoulou 1978; Aravantinos
be mentioned that tomb types can generally be 1983). Clearly, though, they occur more fre-
correlated with the age of burials; while cists quently southeast/south of Pouros.
can be used both for children and adults, pits Domestic burials are not reserved for chil-
and pithoi tend to accommodate children buri- dren and infants. This probably accounts for
als, though there are exceptions (Piteros 1983). the fact that a correlation between tomb types
REU's with tombs only, have been reported and domestic burials is not manifest (since, as
in all well-sampled areas, but relatively I have suggested above, there is a trend to
Plotting Fragments III
associate certain types of tombs with children Figure 7.8) and are more spaced apart, com-
burials). Domestic burials are indiscrimi- pared to the 'domestic graveyard' mentioned
nately placed in cists, pits and pithoi. All three earlier. It is possible that the east slope ceme-
known types have been located under floors tery, which spans the end of the Middle
(Symeonoglou 1973; Demakopoulou 1973^ia; Helladic period and the transition to the early
1975b; Aravantinos 1983) and between houses Late Helladic, constitutes a new development,
(Demakopoulou 1975a; 1978). An infant's cist known to have occurred at other Mainland sites
is reportedly encased in a house wall in this period (Dickinson 1977: 33, nt. 23).
(Symeonoglou 1973). Whether the result of a booming town on the
Turning to a brief discussion of the nature of Kadmeia, a growing investment in mortuary
burial concentrations on the Kadmeia, it is display, or sanitary/religious considerations,
worth mentioning that the distinction between the gradual segregation of the settlement from
domestic and non-domestic burials does not burial sites must have been under way in late
necessarily imply the absence or presence of Middle Helladic Thebes; remoter burial sites
graveyards respectively in certain areas. For (i.e. the Kolonaki-Ayia Anna chamber tomb
example, a dense graveyard has been unearthed cemetery) were evidently in use already as
between and underneath dwellings, which are early as the Late Helladic I period.3
contemporary with the tombs, at the northwest
foothill (Demakopoulou 1978). The overlap-
ping graves are compressed between the Communal Works?
houses and suggest a combined domestic/
mortuary character over a substantial span of To date, there is no preserved evidence of
time. It is likely that such 'domestic cemeteries' large-scale terracing or community projects
were intended for families, resident in the of this kind. The assumption that the settle-
nearby houses. The concentration of tombs ment was surrounded by a mudbrick
between them might represent land tenure enceinte (Symeonoglou 1985) is grounded on
and / or religious beliefs pertaining to the depo- the distribution of domestic remains versus
sition of the dead. burial sites, excavated up to 1973; also, by a
On the other hand, the east slope of the "continuing proximity" of the settlement to
Kadmeia could tell a different story. Although Ag. Andreas and Pouros, presumably
plots with both domestic and mortuary imposed by defensive considerations.
remains are quite frequent there (Figure 7.3), Symeonoglou made a good case for the den-
they rarely show an association between sity of occupation near these hills, based on
houses and tombs (Figure 7.7). A cemetery the number of sites that were known to him at
comprising at least twenty-two sizeable and the time. Yet, a preoccupation with defence
well-built cists (thirteen depicted in Figure 7.8), may not have been the sole factor determining
with monolithic cover slabs, is located under the location of the settlement (Dickinson 1977:
the palatial complex of the Late Helladic period 32); a denser occupation near Ag. Andreas and
(Touloupa et al. 1964; 1965a, c; Demakopoulou Pouros need not have been the effect of physi-
1973^c; 1980; Aravantinos 1994a, b). Similar cal constraint caused by a fortification.
cists, have been excavated further to the east Moreover, new excavation data do not
along the feet of the Kadmeia (Faraklas 1968a; always conform to the clear-cut distinction
cf. Sampson 1981). Most tombs in this cemetery between domestic 'intramural' sites and
share roughly the same orientation (east-west, burial 'extramural' sites (the assumption
Plotting Fragments 113
Figure 7.8 Part of the cemetery at the east slope of the Kadmeia, in Pavlogiannopoulou plot (after Touloupa et al. 1965a).
being that 'intramural' burials are usually of plots with children burials only have been
children). For example, fragmentary houses more frequent at Ag. Andreas, the east ridge
have been more recently come to light close and the area south/southeast of Pouros
and beyond the conjectured east fortification (Figure 7.6). But, it is also the case that 'domes-
line (Demakopoulou 1980; Aravantinos tic' burials, designated as such on the basis of
1982a; 1983; Piteros 1983; Andrikou 1993). If context rather than a preconceived 'intramu-
we exclude the neighbourhood at the remote ral' location, are not reserved for children.
northwest foothill (Demakopoulou 1978), it is
not clear what the criteria are for drawing a
dividing line between 'intramural' and 'extra- 'Big Houses', Monumental Tombs and
mural' houses at the east part of the Kadmeia. Prestige Goods
On the other hand, the location and content
of graves does not suffice to trace the contour Finally, I move on to consider possibilities of
of a missing fortification. I have shown that social differentiation in Middle Helladic
114 Urbanism in the Aegean Bronze Age
Thebes. In accord with other Mainland sites, to the excavator, it is associated with a large,
pronounced qualitative differences in domestic stone-built structure (reportedly a double
architecture and household goods are not hearth) situated just a few metres to the east.
apparent. It is towards the end of the Middle The orientation and layout of the 'megaron',
Helladic period and the transition to the Late which dates to the transitional period proba-
Helladic that more elaborate dwellings appear. bly, appears to have been roughly deter-
The date and nature of fragmentary 'pre- mined by an underlying building, dating to
Mycenaean' structures under the House of the end of the Middle Helladic period.
Kadmos (Keramopoullos 1909) is dubious.4 Other dwellings stand out, not so much
Nevertheless, the excavations in Tzortzi plot because of their construction, but because
east of Pouros have brought to light a note- they contained household goods made of
worthy dwelling (Faraklas 1966; Figure 7.9). imported materials. For instance, an apsidal
It is situated partly underneath the Late house at Ag. Andreas, dating to the Middle
Helladic palatial complex, (not in itself a Helladic II period, yielded a plaque with
compelling argument for its significance). diamond-shaped, incised decoration. Accord-
The surviving, rectangular plan is of the ing to the excavator, it is made of ivory (not
megaron type and comprises two axially mere bone) and served as a furniture inlay. This
arranged rooms; the west one seems to be is an interesting find in the light of the Late
larger (9 x 5.5 m). The house is quite well- Helladic inlay industry in Thebes and the pala-
built with thick stone socles (0.70 m) and tial interest in ivory and other exotic materials.
occupies in its present state 72 m2. According On the other hand, a bone object, probably a
pommel or furniture accessory, originates from tomb was looted, presumably in antiquity.
the same plot, though it is unstratified An ellipsoid shaft above it, a possible result
(Demakopoulou 1973-4a; Demakopoulou and of this operation, contained mixed Bronze
Konsola 1975; cf. Konsola 1981: 138). Another Age pottery, human bones and four gold
house at the northwest foothill yielded bone pieces of jewellery with papyrus and spiral
ornaments, namely a perforated inlay and a decorations, as well as gold beads and a gold
key-shaped pendant (Demakopoulou 1978). rounded ornament (Figure 7.10). The dis-
In most cases, prestige goods originate turbed fill of the tomb itself yielded a gold
from mortuary contexts, and date from the bead, whereas gold wire was found during
end of the Middle Helladic onwards. Graves early soundings (Faraklas 1967; Spyropoulos
with such artefacts can be of Middle Helladic 1972; 1973; 1981; cf. Demakopoulou and
II date, however (Demakopoulou and Konsola 1981: fig. 26; Faraklas 1998:193-205).
Konsola 1975). Gold (Touloupa et al. 1965a) A proposed Early Helladic date (Spyropoulos
and bronze jewellery (Spyropoulos 1969b; 1981) has not been well-received, though the
Demakopoulou and Konsola 1975), glass and site was apparently used in Early Helladic
carnelian beads (Touloupa et al. 1965a), times. The current consensus is that the
weapons in a combination of materials megalithic cist dates to the Middle Helladic
(Demakopoulou 1980), have been reported in period (Demakopoulou and Konsola 1981:
cists at the east slope, Ag. Andreas and the 23, 56; Symeonoglou 1985). Its construction is
northwest slope. Adornments in gold, silver, congruent with this notion. The pieces of
amethyst, sardium and faience are known to gold jewellery find parallels at the Mycenae
have accompanied pit and pithoi burials at Shaft Graves (Dickinson 1977: 97-8) and sug-
Ag. Andreas (Demakopoulou and Konsola gest a late/transitional Middle Helladic date,
1975; Demakopoulou 1979a; Aravantinos if we accept that they originated from the
1982a). In a similar vein, the more elaborate tomb.s
tombs in terms of construction and size date
from the end of the Middle Helladic period
onwards. As I mentioned earlier, the east Conclusions
slope cemetery mostly comprises well-built
and sizeable cists. Other such examples have Middle Helladic Thebes emerges as a more
been excavated elsewhere (Faraklas 1968a; extensive settlement than previously sus-
Demakopoulou 1973-4b, c; Demakopoulou pected. My analysis adduces support to the
and Konsola 1975; Aravantinos, pers.comm). notion that the south and east parts of the
The partly man-made 'Ampheion' hill Kadmeia would have been more densely
seems relevant in the context of increasing occupied (cf. Konsola 1981: 168). This has
mortuary elaboration. A prehistoric tumulus been interpreted, more or less, as the effect of
of sun-dried, rectangular mudbricks would a fortification, but I have tried to show that
have stood at least 2 m high on top of the nat- the criteria for tracing the postulated circuit
ural outcrop of soft stereo. A single grave is become increasingly hazy as the ongoing
situated in the centre of the tumulus; over 2 exploration of the Kadmeia brings to light
m long, the cist is built of large conglomerate new evidence. The contrast between domestic
slabs and features a massive limestone cover. and burial space need not have been as pro-
Levelled stereo forms the floor, at the south nounced, at least not in the early and middle
part of which two rectangular pits exist. The part of the period. Domestic burials were
116 Urbanism in the Aegean Bronze Age
Figure 7.10 Gold jewellery from the Ampheion (adapted from Demakopoulou and Konsola 1981).
widespread on the Kadmeia and were not not in the presence of organization (a com-
reserved for children. They can form exten- mon 'urban essential') that they differed from
sive clusters in direct association with domestic graveyards, but in the nature of
dwellings. It is towards the end of the Middle organization. It is perhaps through the grad-
Helladic period that more discrete graveyards ual transformation of kin-based identities to
seem to come into sight, a development community based ones that we should seek to
which forecasts the establishment of remote, explain changes in the Theban settlement. The
dedicated cemeteries in the Late Helladic I. increasing mortuary deposition of wealth and
These discrete graveyards did not necessarily the appearance of salient burial and domestic
constitute the 'first organized cemeteries'. It is structures are remniscent of changing social
Plotting Fragments 117
strategies of display and identity as recognized 1994 OiKOJtE&o noi)XiojtoijA.oiJ (OT 335, 0665
in other Mainland sites and seem to mark this EjtafjLEivwvSa 79). ADelt 49 Bl: 276-78.
Aravantinos, V.
transformation. 1981 O66g AipKT|g 23 (oiKOjteSo Kwv. rKeXr]). ADelt
36 Bl:188-89.
1982a OSog AipKT|g 25 KCU Ei)pi8ii<r|g (oiKOJie8o Aipcbv
Acknowledgments 5/raiKOu). ADelt 37 Bl: 165.
1982b 6665 IleXomSou 26 KCXI Oi8ijto8og (oiKOireSo
I am grateful to Prof. K. Branigan and Dr. P. Mt|Tp6jtoA.r|g 0T)|3wv, jtpcb-r|V Acptbv Ei)6r|).
Halstead for inviting me to give a talk at the 5th ADelt 37 Bl: 165-67.
Round Table in Sheffield (on the Late Helladic 1982c O66g Oi8iJTo8og 30 (oixojreSo AOUKCX AOIJKCX).
ADelt 37B1:169.
settlement), as well as for the generous exten-
1983 O66g Oi8ijr;oSog KXU mpoSog II. OIKOVOJJIOD
sion of time leading to the submission of the (OIKOJTESO E. Km M. XpiaToSoi)X.ou). ADelt 38
final draft. Many thanks are due to Dr. V. Bl:129-30.
Aravantinos and the IX Ephorate of Prehistoric 1986 The fortified EH II building at Thebes: some
and Classical Antiquities for their support and notes on its architecture. In Hagg, R., and
Konsola, D., (eds.), Early Helladic Architecture
kind permission to refer to recent excavations and Urbanization, SMA 76: 57-63. Goteborg:
in Thebes. I am also indebted to Dr. S. Voutsaki Paul Astroms Forlag.
for suggestions and comments and to M. Hild 1987 Mycenaean place-names from Thebes: the new
for statistical advice. Further I should like to evidence. Minos 20-22: 33-40.
1994a O66c; OeXomSoD 28 (apxeio juvaid&a)V
thank Mrs. Branigan, O. Peperaki and M.
Tpa(ji|JUKTjg B). ADelt 49 Bl: 271-74.
Katapotis for their hospitality in Sheffield. 1994b OjtA.oOr|Kr| (OIKOJTESO A. nauXo^iavvojtovAoiJ,
666g nsXom&oi! 28). ADelt 49 Bl: 274-76.
Dakouri-Hild, A.
Notes 2001 The House of Kadmos in Mycenaean Thebes
reconsidered: architecture, chronology and
1. The exact temporal relationship between apsidal and
context. BSA 96: 80-108 (in print).
rectangular dwellings within the Middle Helladic period
In prep The application of Geographic Information
is unclear.
2. It was later pointed out that these remains should have Systems (GIS) in the study of Mycenaean
been dated to the Early Helladic period, because the Thebes (in preparation).
Middle Helladic pottery associated with the structures Demakopoulou, K.
was intrusive (Symeonoglou 1985: 241; cf. Demakopoulou 1973-4a O66g llivSapov* Km AipK^g. Adelt 29 B2: 433-36.
and Konsola 1975: 45, nt. 3). 1973-4b O66q KotS|xou 58. Adelt 29 B2: 439-40.
3. Three Late Helladic I vessels from tomb 2 at Kolonaki- 1973-4c O66c neXojtiSov 28. ADelt 29 B2: 441.
Ayia Anna are currently on display at the archaeological 1973-4d O66q AoryXapiSoD 14. ADelt 29 B2: 439.
museum of Thebes (cf. Demakopoulou and Konsola 1973-4e O66g Tae(3a Kai BotipSovpiJTa. ADelt 29 B2:
1981: 48). 437-38.
4. However, the site has produced unstratified Minyan 1975a O66g 'Ena^eLvcovBa 58 (oiKOJteSo II. Me\ETiov).
pottery originating from the terrace fills and the mud- ADelt 30 Bl: 128-30.
brick elevations (Dakouri-Hild 2001). 1975b O66g neX.ojri8oiJ 25 (oiKOjteSo ATJ(JL. STEtpa).
5. D. Konsola (1981: 140-41) does not exclude the possi-
ADelt 30 Bl: 131-33.
bility that jewellery is of Early Helladic date, drawing
parallels to the Poliochni treasure. 1975c O66g Toe^a 24 (oiKOJteSo N. SitoupXr)). ADelt
30 Bl 130-31.
1976a OiKOJteSo II. ZovAdjxo^ou. ADelt 31 Bl: 125-26.
1976b OiKOjreSo ST. KOU N. MavioaXr]. ADelt 31 Bl:
Bibliography 121-25.
1978 OiKOJteSo SCOT. Kopojtoi)\ri. ADelt 33 Bl: 108-12.
Andrikou, E. 1979a O66q IleXomSovj 5 (oLKOiteSo A. IIaD\o'YLav-
1993 OixoJieSo AT]|JL. Aajjutpou (^covia o8tbv AipKr|g 8 vojrovAovj). ADelt 34 Bl: 165.
Kai aSiE^oSoi) ST. Bpv£aKT], OT 314). ADelt 48 1979b O66g AvTL^ovrjg 14 (oUojieSo A. AOXJKOD).
Bl: 173-76. ADelt 34 Bl: 163-65.
118 Urbanism in the Aegean Bronze Age
1979c O66g riiuSdpou 13 (oiKOJieSo F. Kai K. AeSe). 1981 FldpoSog o8oi> loKdoTTig (OT 156) (oiKOJieSo
ADelt 34 Bl: 165-66. a88\9cbv AVT. Kai Ain(jLOO0. MardXa). ADelt 36
1980 6665 ZeyyivT] 3. ADe/f 35 Bl: 215-17. Bl: 190-91.
Demakopoulou, K. and D. Konsola Spyropoulos, T.
1975 Aatyava JipanoEXXaSiKOi), (juEOOEXXaSiKOt) Kai 1969a OiKOJteSov KODVOT. Aoiipou (jrapd TTJV
WT£pO£XXa8iKOi) oiKiO|jioi) OTTI 0r||3a. ADelt MTiTpoitoXiv 0-ripcbv). ADelt 24 Bl: 180.
30 A: 44-89. 1969b OiKOiteSov d8eXcpcov Aa-/8eXevT] (666g
1981 ApxaioXo-yiKO Movoeio TT|g 0r||3ag. Athens: 'EjtafJLEivcovSoi)). ADelt 24 Bl: 180-82.
T.A.P. 1970a OiKOJteSov dSeX^cov Aa^SeXevTi. ADelt 25 Bl:
Dickinson, O. 211-12.
1977 The Origins of the Mycenaean Civilisation, SIMA 1970b OiKOJteSov Ap. OaaovXojrouXov (666g
49, Goteborg: Paul Astroms Forlag. AvTL^6vr]g Kai 'EjiaiJueivcovSov). ADelt 25 Bl:
Faraklas, N. 218.
1966 KaSjJLEioi). OiKOJie8ov A. KOU S. T£cbpT£r| 1971a OiKOiceSov ScoT^pLou-AovyeKOD. ADelt 26 Bl:
(ILvSdpou KCXI AvTi/yovTjs). ADelt 21 Bl: 177-80. 195-202.
1967 'Epeuva TOIJ n^fiov TOIJ Afx^eiou. ADelt 22 Bl: 1971b OLKOJte8ov XapaXdjjLJtovg SraiJidTTi. ADelt 26
229-30. Bl: 202.
1968a Avao-Katpr) oiKOJte8ov 'I. ADelt 23 B2: 211. 1971c OiKOJTe8ov Evayy. Bo-yKXri (666g IIeXojri8ov).
1968b OiKOiteSov N. 'Iwdvvou. ADelt 23 B2: 207-208. ADelt 26 Bl: 207.
1968c OiKOJtESov II. AeovtapTj. ADelt 23 B2: 208-10.
1972 Aijupeiov. ADelt 27 B2: 307-308.
1998 0ti(3aiK(x. Apxaio^oyiKri EiprpEpig: 135 (1996).
1973 A|ji<pEiov. ADelt 28 Bl: 248-52.
Athens: Archaeological Society of Athens.
1981 A(jL9£Lov. 'Epeuva Kai, MeXETT) TOT) MvTuxeioD TOT)
Filippaki, V. et al.
AIJL^EIOD ©TiBwv. Sparta: (privately published).
1967 OiKOJtESov A. ©eoStbpov. ADelt 22 Bl: 226-27.
Spyropoulos, T. and J. Chadwick
Galanaki, I.
1975 The Thebes Tablets II., Minos suppl. 4.
2001 Structural changes in settlement patterns in
Salamanca: Universidad de Salamanca.
Thebes, Boeotia, during the Shaft Grave period. In
Symeonoglou, S.
Symposium of Mediterranean Archaeology (SOMA),
1973 Kadmeia I. Mycenaean Finds from Thebes, Greece.
23-25 February 2001 (unpublished paper),
loannidou, A. Excavation at 14 Oedipus st., SIMA 35, Goteborg:
1973 OiKOJieSov A6. STo^cm]. ADelt 28 Bl: 274, 276. Paul Astroms Forlag.
Keramopoullos, A. 1985 The Topography of Thebes from the Bronze Age to
1909 FI OiKia TOW Kd8|jLOU. AE: 57-122. Modern Times, Princeton: Princeton University
1917 ©Tipaixd. ADelt 3, Athens: Estia. Press.
Konsola, D. Touloupa, E. et al.
1981 ripo(aA)Kr|vaiKr| 0r|(3a: XcopOTa^iKt) Kai 1964 AvaOKacpr) oiKOJtESou AT]JJL. Fla^Xo-yiav-
OiKi0TLKr| Aidp9pa)ar|. Athens: Karobemista. vojtouXov. ADelt 19 B2: 197.
1984 H IIpcbi|jiri AoTiKOJtoir|ar| oiovc, npa)ToeX- 1965a KaSjxEiov. OiKOJTE8ov Ar|p,. IlauXo-yiavvojiouXou
XaBiKOiig OiKLOjjLOijg. Athens: Athanassopoulos, (Il£Xom8ou 28). ADelt 20 B2: 233-35.
Papadamis and Zacharopoulos. 1965b OiKOJtESov A6. KopSaT^, 666c Oi8i,jto8og dpi9.
1988 0r|3a: n;p(jai|x8g ^doeig oiKiOTiKr|g 8^eX.i^r|g. In 14. ADelt 20 B2: 235-37.
Proceedings of the lsl International Boeotian 1965c OiKOJtE8ov 'Itodvvou AiaKOu, 666g Zeyyivr],
Conference (1986), vol. A,a: 65-73. dpiO. 3. ADelt 20 B2: 239.
Piteros, C. 1966a OLKOJT£8ov riavvoJioi)Xou-AT])jLr|TpaKOJioi)XoD
1983 O66g OiSurcoSog 1 (oiKOJte8o A. Aid"/Ka). ADelt (666g 'Ejta|jL£LVcbv8ou). ADelt 21 Bl: 183-87.
38 Bl: 131-34. 1966b OiKOJteSov STajx. 2Ta|xctTTi. ADelt 21 Bl: 189-91.
Sampson, A. 1966c OiKOJtESov nriXiSri. ADelt 21 Bl: 192.
1980 OSog neXomSoiJ (oiKOJteSo M. AOUKOU). ADelt 1966d OiKOjtESov 2Taupojtoi)Xou. ADelt 21 Bl:
35 Bl: 218-20. 192-93.
8
tombs and houses are curiously intercalated An approach to the study of courts and
at a number of Mycenaean settlements. But squares
our knowledge of the configuration of
houses and open space in what excavators The analysis to follow attempts to draw from
often call the 'lower' towns, outside the the design of squares inferences about their
citadels, is frustratingly limited. All the role. This approach is open to objection. People
same, putting aside the rather special case of do not always behave as architects and plan-
Gla, there is no indication in mainland ners would have them. Thus on the planned
Greece of paved squares contemporary with high-rise estates of the 1960s (AD) spaces envis-
the Piazzale dei Saccelli or the large court in aged by their architects as friendly, communal,
front of the Great Stoa at Ayia Triadha, on social areas, are today said to be stalked by
Crete. The latter follow Minoan antecedents crime and violence, and shunned by residents.
and thus belong to a different cultural In a similar vein note that important assemblies
tradition. can gather in any open space, it need not be
There are, naturally, fundamental differ- enclosed by walls or clearly demarcated within
ences between the medieval and Bronze Age a town (though the Homeric agora, as it hap-
towns, for example the sort of market pens, was quite formalized, see Wees 1992: 29,
exchange crucial to the medieval town does and agoras of the 8th century BC have been
not apply under the bureaucratic administra- recognised at Megara Hyblaia and Dreros, see
tion centred in the palaces to be discussed Vallet 1973 and de Polignac 1995: 8-11). In
here (though it would be a mistake to rule out response I would say that here the theme is not
commercial exchange from Mycenaean the everyday interaction of individuals, but the
Greece see, for example, Wiener 1987, the meeting of ruler and ruled, always a formal
papers in Gale 1991, Shelmerdine 1997: 567). occasion, and usually one that is stage-
Nevertheless Heers has stressed that the managed. To anticipate, the Mycenaean courts
monumentalization of the grand piazzas in were the result of an evolution, which shaped
the Italian cities was an expression of civic them to their end, and they were an expression
polity, and quite distinct from the provision of a self-conscious idiom aimed at providing an
of markets. Indeed there are many cases from architectural set for the exercise of power.
different cultures where expenditure lav- As a means of broaching the question it is
ished on the religious/political centre was proposed to look at the following characteris-
not matched in the markets: recall, for exam- tics:
ple, the ritual precincts in the Aztec sites of
Tenochtitlan and Tlatelolco with their mas- - access
sive monuments, and beside them the huge - size
but simple markets. - orientation
Here the more specific question is asked - focus
how are we to understand the courts and - perspective
open spaces in Mycenaean palaces. Political - visibility
assembly, religious procession, military - appointment
muster, economic transaction, organisation of - frontage
labour, supply and circulation might all form
part of the answer. What can archaeology tell These terms are jargon and require some
us? attempt to turn them into English. Access
Empty Space? Courts and Squares in Mi/cenaean Toivns 121
raises the question of how easy or difficult it not extend so far as the Court did (Miiller
might be for an individual to enter the court. 1930: 134). Indeed, although he eventually
Thus Wright (1994; cf. Kilian 1987a, 28 rejected the hypothesis, Miiller entertained
'Megaron als Endglied einer architektonis- the theory that the west portico was part of
chen Steigerungskette') has emphasized 'cen- an earlier construction, later incorporated
tredness' as an organising principle in into the Great Court when that was first laid
Mycenaean settlement form, and the impor- out. This theory makes good sense, and
tance of monumental entrances marking a Miiller's arguments for rejecting it are not
progress through a series of boundaries to the overwhelming. The south and east portico
innermost seat of authority. Size can set a would seem also to go with the winding cor-
limit to the capacity of a square and reflects ridor, 36, itself an integral part of the East
its monumentality. Orientation and focus are Wing at Tiryns, now dated to the later LH IIIB
characteristics of the shape and sense of phase (Kilian 1987a: 28). Finally the Great
direction given to the space. Is it broad, or Court at Mycenae also seems to be approxi-
long and narrow? Is it directed towards an mately contemporary with the other two, and
obvious focal point? Visibility and perspective replaced some earlier disposition. Thus Wace
mean, in the first place, is someone standing observed that traces of an earlier north wall
in the space easily seen by others, and in the implied a different arrangement which pre-
second, does someone standing in the space ceded the megaron and Court (Wace 1921-23:
have a clear view of their surroundings. 195). The date of the formation of the Court at
Appointment can include furnishings, such as Mycenae is bound up with the construction
an altar or a podium, and decoration, such as of the Grand Staircase (Wace 1921-23:
painted plaster. Frontage refers to the build- 179-86). Indeed a massive conglomerate
ings which look onto the space. threshold (unfortunately displaced in the
nineteenth century) marked the passage from
the 'anteroom' (Mylonas 1966: 66 suggests
The Great Courts that the 'anteroom' was an unroofed court),
at the top of the Staircase, to the Court. The
The Development of the Great Court creation of the Staircase involved the destruc-
The Great Courts (those fronting the megara) tion of the Pillar Basement. Now the con-
at Mycenae, Tiryns and Pylos are creations of struction of the Pillar Basement was no
the thirteenth century BC (cf. Kilian 1987a: 33: earlier than LH IIIA2, to judge from the
'die grossraumigen Megara eine Bauform, sherds illustrated by Wace (1921-23: pi. 31
die erst in der SH IIIB-zeit hinzukam'; j-o; though there was clearly some later dis-
Shelmerdine 1997: 558-59). The Palace at turbance of these levels). This in turn gives a
Pylos was built as a single unit, perhaps at terminus post quern after LH IIIA2 for the
the beginning of LHIIIB (Blegen and Rawson building of the Grand Staircase (lakovidis
1966: 422). The Court at Tiryns belongs essen- places the destruction of the Pillar Basement
tially with the megaron and the lesser propy- in LH IIIB: 1983: 61). Wace thought there was
laea to LH IIIB. Miiller established that the an earlier wooden staircase, which preceded
floors stratified below the Court were not the Grand Staircase, associated with the
levels of its earlier phases, but belonged to Great Court and the megaron, in which case
other structures: they were associated with they would have been built towards the end
walls on a different orientation, and they did of LH IIIA. A case could be made for the
122 Urbanism in the Aegean Bronze Age
Grand Staircase and the Great Court going the propylaeum at Mycenae some 50 m to the
together, and in that case they would date to NW of the Court makes the approach there
LH IIIB. Perhaps the earliest clear example of more circuitous and the effect in some ways
a court in front of a megaron (it measures c. less bold. Indeed Wace saw the main entrance
14 x 14 m = 196 m2; Atkinson et al. 1904: to the Court as marked by the massive con-
55-58), is that at Phylakopi. The megaron has glomerate threshold in the SW corner ('T' on
been dated to LH IIIA1 (Barber 1992: 19), but Figure 8.1), as mentioned above. This would
the court has a different orientation from the link the primary approach, via the Grand
building, and may belong to an earlier phase, Staircase, with the lower extension to the
perhaps to the underlying Minoan edifice; citadel, and probably with the Cult Centre.
the well in the court at Phylakopi, if contem- The route from the Lion Gate evidently led
porary, hints at a very different usage from through the propylaeum to the West Entrance
that of the Great Courts. At Ayia Triadha, to the South Corridor (45/40), then probably
although the Piazzale dei Sacelli was repaved by way of a dog-leg passage to corridor 49 and
at (about) the time the megaron was built, in thence into the Great Court (Figure 8.1: routes
LM IIIA2, the area immediately to the east of indicated by arrows; lakovidis 1983: 57-60;
the megaron was occupied by a stoa, which Kilian 1987b: 103 underlines the uncertainties
effectively blocked the open space in front. It due to poor preservation). In all cases access
appears that megaron P, also LM IIIA2, was closely regulated; a double door closed
opened onto the Agora, but, of course, this the 2.86 m wide opening at Tiryns. The
open space mainly serves the Stoa. In addi- approach at Mycenae was particularly
tion to the other reasons put forward (Catling labyrinthine. In all three the entrances are
1976-77: 29), perhaps Mansion II at the deliberately off-axis to the megaron.
Menelaion was reoriented in H IIIA1 also to Neither at Mycenae nor at Pylos was the
take advantage of the top of the knoll to form Great Court accessible to wheeled vehicles,
a large courtyard in front of it. Unfortunately and the same was probably true of Tiryns.
the remains on the summit were too eroded The new evidence for a fortification wall at
for us to get a clear picture of the arrange- Pylos (Zangger et al. 1997: 606-13; Blackman
ment of the court there. The SW Building at 1998: 54-55) is probably to be linked with the
Pylos, unlike the megaron, was not preceded early fortifications, and not relevant to the
by a symmetrical enclosed courtyard, and LH IIIB palace. Even so this most open of the
court 16 at Tiryns, before the smaller Great Courts may be thought, like the others,
megaron, is also quite different in disposi- to have stood at the centre of a sequence of
tion. In brief, then, on present evidence the protected entrances (cf. Wright's comment
Great Courts were new in LH IIIB, and stand above; Blegen interpreted the collapsed
at the end of a long evolution. masonry to the south-west of the propy-
laeum as the remains of a gate-way, see
below). At each barrier, we may assume,
Design of the Great Courts entrance was ever more restricted. In all three
cases there are minor, indirect ways into the
The main access at both Pylos and Tiryns was Court, not to mention access from the
through propylaea, double and single colum- Megaron itself, but these were subsidiary.
nar respectively. These command and regulate The size of the unroofed area of the Courts
the approach to the courts. The positioning of varies: Tiryns (approx. 15.2 x 20.3 = 308 m2),
Empty Space? Courts and Squares in Mycenaean Towns 123
very roughly twice the area of Mycenae Mycenae would have had the effect of draw-
(approx. 12 x 15 = 180 m2) and Pylos is just ing the eye along the main axes of the square.
over half that again (7.3 x 12.9 = 97 m2). This There is evidence that ceremonial was con-
range, combined with the evidence for ducted in the porch of the Megaron (see
restricted access, argues that their capacity below) which also, of course,gave entrance to
was not the important consideration, but and from the centre of power. But the orienta-
monumentality. In all cases the main focus is tion, with the Court's longer axis at right
the megaron, which faces one as one enters the angles to the megaron's, has the effect of
Court. Hirsch (1977: 45) has pointed out how crowding the view. The facades of the megara,
the grid pattern of the painted floor at two storeys high, must have measured a good
Figure 8.1 Plan of the palace at Mycenae (after Kilian 1987: 102 fig. 1).
124 Urbanism in the Aegean Bronze Age
eight-ten metres, and at no more than 11-15 m the open participating in any action.
from the viewer would have towered above Mycenae is exceptional in having no portico,
them. Again there was a conscious aim to and probably no propylaeum. However,
overawe. The dominant, almost claustropho- Wace pointed out that the indent in the north
bic crowding of the viewer's perspective is wall of the Court at Mycenae probably
confirmed by the 'isovista' of someone stand- marked a window, looking from the adjoin-
ing on the threshold at the entrance into the ing corridor out onto the court. Indeed, on
Court. The angle of view subtended from the the basis of three conglomerate column
facade of the megaron (anta to anta) is at bases, which had fallen into the court, he
Tiryns 30°, at Mycenae 45° and at Pylos 50°. suggested that there may have been a gallery
Moore has observed that 'Visual perception in or above the South Corridor (Kilian 1987b:
changes at 18 degrees, when an object first 103 with n. 20 appears sceptical). This would
takes on monumental nature, 27 degrees when effectively parallel the ground floor porti-
it fills the viewer's range of vision, and 45 coes lining the Great Courts at Tiryns and
degrees at which details are visible' (Moore Pylos. Indeed more generally he observed
1996: 100-101 and fig. 3.4, quoting Higuchi 'into it looked windows from the apartments
1983). The fact that the Courts are oriented so and corridors on the west, north and east ...'
as to more than fill the viewer's range of (Wace 1921-23: 189). At Pylos, on Blegen's
vision is no accident, but a deliberate psycho- restoration, a gallery above the porch looked
logical device on the part of the architects who onto the Court (connecting on the interior
designed the buildings. with the balcony above the throne). This is
The effect was intensified by the enclosed entirely compatible with the iconographic
space. On entering the Court one's perspec- evidence of Aegean art. Spectators, often
tive was restricted, indeed Wace thought it women, are frequently shown looking down
strange that the magnificent views of the onto events as varied as the marine proces-
plain of Argos, visible today from the Great sion, in the Thera Ship Fresco, and the battle
Court, should have been shut off in the scene on the Silver Siege rhyton. They look
Mycenaean period (Wace 1921-23: 189; 1949: out from windows, balconies and roof
75; Mylonas 1966: 63-64; lakovidis 1983: 60; terraces.
even if one restores a balustrade to the south, In their appointment of the Great Courts the
the view is still effectively blocked by the builders aimed at refinement. Wace had rea-
Grand Staircase). The same applies to Tiryns son to praise the high quality of the cement
and Pylos, one cannot look out from the which formed the floor at Mycenae and at
court. Thus whilst modern sensibilities Tiryns, whilst the stucco at Pylos consisted of
might require a room to have a view, the four successive coats (Blegen et al. 1966: 64).
Great Courts are designed precisely not to At a late stage this was covered at Mycenae,
have one. If one effect of enclosing the space all over by a layer of painted stucco, with
by surrounding it with buildings was to rein- decorative motifs set in squares bounded by
force focus on the main edifice, another was red lines; the squares were coloured yellow,
to make the square an arena overlooked by blue or red. Although preserved only by the
spectators, what I have termed visibility. The north wall it is clear that the whole surface of
porticoes around the square can demarcate the Court was thus painted (Wace 1921-23:
the space so that, on occasion, the onlookers 192-94). The grid pattern painted on the floor
standing within the colonnade view those in would have the effect of drawing the eye
Empty Space? Courts and Squares in Mi/cennenn Towns 125
along the main axes of the square (Hirsch Lesser and Greater Propylaea, emphasized
1977: 45). Similar painted floors were found by the angled alignment of the latter (Figure
in various rooms and corridors at Pylos, 8.2: 2). Thus, although the court measures
Tiryns and Mycenae, not least, of course, in approximately 21 x 31 m maximum, the
the megara themselves; and outside the effect is rather of a processional way crossing
palaces only in the mansions at Gla. Moreover the open space. The perspective from within
the fine ashlar masonry at Mycenae was plas- this larger space may have been less claustro-
tered and painted with a rosette and triglyph phobic than within the Great Court, and it is
design similar to the NE corner of the possible that on 'state occasions' people
megaron porch, and reminiscent of the gyp- would congregate in the southern part of the
sum and glass relief from the porch at Tiryns square. On the other hand there is little to
(Miiller 1930: 139-43). Blegen noted traces of suggest that the square was overlooked by
plaster on poros blocks facing the court at windows: evidently the north wall of 45 was
Pylos (Blegen et at 1966: 63, 65 the anta to the blank, whilst the portico, 39, tucked away to
megaron painted). In these cases the refine- the south of the Great Propylaea seems to be
ments characteristic of the most important a minor feature, with no clear view. The
interiors were lavished on a courtyard. Court quite probably served other functions,
Certainly the floor fresco must have been for the movement of goods into the palace
even more subject to wear through weathering stores and workshops. Indeed the single
than those on interiors (note Rodenwaldt's rooms set between the two Propylaea, and
observation that painted floors can hardly opening onto the Forecourt are reminiscent of
have had a long life (1912: 235 n. 2), and Rooms 7 and 8 and Room 57 at Pylos. They
needed constant renewal). The altar in the may have served similar functions as guard-
Great Court at Tiryns is unique, and may go room and archive (though there are other
some way to explain the larger size of that possible candidates at Tiryns, such as the
square. However Papadimitriou noted the group of rooms 38).
presence of an alabaster slab and low rounded Access into the Outer Court at Tiryns, as into
altar in the porch of the Megaron at Mycenae, all those we have considered thus far, was
probably used for libations (Papadimitriou restricted, even if it did not form part of the
1955: 230-31; Hagg 1990: 180; Shelmerdine palace proper (Miiller 1930: 167). Although
1997: 578). rather different in size (approximately 24.1 x
11.1 m), shape and arrangement the Outer
Court is, in fact, similar to the Forecourt both
Outer courts in that the dominant focus was the Great
Propylaea and because its orientation was
The Forecourt at Tiryns is entered through essentially linear, following the processional
the Great Propylaea (Figure 8.2: 3), whose way from the Propylaea to the defended
width rivalled that of the megaron itself, and entrance system. The exact restoration of its
which marked a grand entrance into the eastern side is far from certain (Miiller 1930:
palace precinct. Indeed the gateway evi- 74). The portico normally suggested would
dently formed the major focus for one stand- have provided, as in the other courts, a shel-
ing within the court, and the orientation was tered area from which any procession would
dictated less by the forecourt's (irregular) be visible. Evidently, the galleries were also
shape and more by the axis between the reached from here, and produce to be stored in
126 Urbanism in the Aegean Bronze Age
them could have been unloaded in the court fused with the court's stucco floor, beside the
and manhandled down. sump which drained the court (op. cit. 229-30,
This combination of the ceremonial and the figs. 217.4, 272.3-4). In terms of appointment it
practical is also to be found at Pylos. Court 58 is tempting to identify this table as an altar
(the outer court: Figure 8.3, 58) was paved and link it with the circular altar in the Great
with plaster for a distance of at least 7 m Court at Tiryns (also located by a drain) and
(Blegen et al. 1966: 227-28). The ashlar facade with the altar in the porch to the megaron at
and propylaeum of the palace would have cer- Mycenae. Unfortunately the areas to the south
tainly acted as a focus, whilst the poros foun- and east of the palace were severely eroded,
dation built onto the SW fagade of the palace and consequently characteristics such as ori-
may have formed 'a rostrum ... or reviewing entation, perspective and visibility cannot be
stand, where royalty could sit to observe cere- analysed. Nevertheless Blegen found slight
monies on gala occasions' (Blegen et al. 1966: evidence for a colonnade to the south-west,
229; Figure 8.3 just above the caption 'Main and he suggested a mass of collapsed masonry
Building'). Blegen also reports fragments of a to the south-east may have formed a gateway
circular marble table, half a meter in diameter (Blegen et al. 1966: 228). Wright has put
and originally with inlaid decoration, found together a powerful case that a number of
Figure 8.3 Plan of the palace at Pylos (after Blegen and Rawson 1966).
128 Urbanism in the Aegean Bronze Age
changes towards the end of the life of the frontage of the main court, its proportions
palace at Pylos, increasingly closed off access (ratio of length to width: 9.2 x 13.5 m) give a
(Wright 1984, though with an emphasis on similar orientation and its size (124 m2) is not
economic rather than ceremonial reasons). The inconsiderable, though much smaller than
recent cleaning under F. Cooper, on behalf of the Great Court. Moreover the progression
the Minnesota Pylos Project, has revealed a from Court 30 to Court 16 imitates that from
monumental stepped entranceway, filling the 2 to 4. However, the main access is very dif-
south quadrant beyond the SW Building. This ferent (following an indirect passage down
has two phases, the earlier had a flight of shal- the long, dark, winding corridor 36), and
low broad stone steps, built of slabs. The later may argue that the public functions of the
phase consisted of a stucco ramp and a more Great Court played no part in the role of
elaborate L-shaped stairway. The approach Court 16. It has, if not a true propylaeum, a
therefore was indirect and would involve a door set in antis (Miiller 1930: 160); but the
zig-zag course to reach the propylaeum door gives onto the east portico, so that the
(Blackman 1998: 55-56, figs. 78-79). A broad focus of one entering the court is distracted
similarity with the outer court at Tiryris seems from the megaron's facade by the portico's
not implausible. In both cases the areas columns and ceiling. Once standing in the
undoubtedly served ceremonial, as the care court one's perspective is, as usual, closed off;
lavished on their appearance makes clear, and perhaps even more than the Great Court this
their articulation through a sequence of is a space one looks into, rather than out
entrances suggests that the ceremonies were from. The scaled-down symmetry is, there-
processional, with more static moments fore, probably a reflection of relative status
marked by the altars and 'reviewing stand'. rather than parallel function.
Our understanding of Court 63 at Pylos is
complicated by the uncertainties surrounding
Lesser courts the phasing in this part of the site, a question
which is currently under investigation. The
Kilian has argued that the three following comments here are more than usually
buildings are to be identified as the official hostages to the fortunes of future research.
residences of the Lawagetas: the Lesser The main hall complex (Rooms 64-65) together
Megaron at Tiryns, the South-West Building with Court 63 represent an original feature.
at Pylos and the House of Columns at Evidently, earlier buildings were demolished
Mycenae (Kilian 1987a). Dickinson has when the Hall and Court were laid out. In a
expressed reservations (1994: 154). Can an later phase, after their construction, there were
analysis of their courts, according to our cri- alterations and extensions: a staircase added
teria, shed light on their public aspect? Of to the NW side and a complex, including a
course so restricted an inquiry passes over room and a 2m terrace, apparently impinged
other very important considerations, but, as onto what was originally a larger Court 63 to
will be seen, the courts of these complexes do the SE (Blackman 1997: 50). Blegen's Key Plan
not show that degree of similarity we have marks the entry system into the court, defined
found in the design of the Great Courts. by Rooms 60-62, with hatching, suggesting
Court 16 at Tiryns almost mimics the Great that this too was a secondary feature added
Court nearby, but with significant differ- after the original layout. The recent discovery
ences. The north and south porticoes echo the of a multiphase staircase to the south and
Empty Space? Courts and Squares in Mycenaean Towns 129
evidently partly below Rooms 60-62, makes it courts considered up to now. Thus the court of
even clearer that there is a long, complex his- the House of Columns differs very consider-
tory to development of this part of the site. It ably in approach and access, size, focus, orien-
would appear that there was a period when tation and appointment from the Great
access to Court 63 was more open, both from Courts. Evidently there were similar small
the south via the newly recognised staircase, courtyards in the southern part of the citadel
and from the SE, from Court 58; though the at Tiryns (Miiller 1930: 126, 'zwei in sich
superstructure of the staircase must still have geschlossene Baugruppen lagen, von denen
enclosed the ramp up from Court 58. The indi- jede ihren eigenen Hof haben musste, also
rect approach, via Ramp 59 and Corridor 61 offenbar Wohnungen, und zwar gewiss von
seem to mark a reconstruction which aimed to Leuten, die zur Hofhaltung gehorten/).
enclose 63 and cut off any view from outside Domestic courtyards are, of course, not con-
the Court of the open hall 64. Not quite as sidered here, as we are interested in the public
impenetrable as Court 16's at Tiryns, this indi- areas of palaces.
rect access is comparable. Unfortunately In brief, there are certain features which link
Rooms 89-90 (of Iron Age date) have obliter- Tiryns 16 and Pylos 63, particularly their large
ated any trace of the frontage to Court 63, but it size, focus and (lack of) visibility, and are suf-
would seem the Court although enclosed was ficient to suggest that the Tiryns court merits
not colonnaded. Thus, rather in contrast to consideration as a candidate for a banqueting
Tiryns 16, the emphasis is less on visibility, area. The House of Columns, on the other
those outside the court looking onto those hand, does not meet the argument nearly so
within it. On the other hand the deep Hall 64 well, and to that extent fits Kilian's hypothesis
is the only focus from the Court, if not equally linking the three buildings less effectively. It
visible from all parts of it. The shape of 63 is may be that the 'Lawagetas' House' at
irregular, and less clearly oriented than the oth- Mycenae was located on the brow of the
ers we have considered, but its size, approxi- acropolis hill, an area whose eroded and frag-
mately 225 m2, is impressive, especially as Hall mentary remains still defy a definitive recon-
64 can almost be added to it and 88 may well struction (see, most recently, Kilian 1987b).
have been open too. The suggestion that the
open space here was used for banqueting
(Davis et al. 1998: 84), makes excellent sense of Grave Circle A
the context and lay-out of this area.
The House of Columns at Mycenae (Wace Even though Schliemann's enthusiasm for
1949: 91-97) presents us with even greater Paley's hypothesis was misplaced his instinct
problems of interpretation because of its poor that this space was designed for gatherings is
preservation. It is plain that the main doorway surely correct. Indeed Gates has recently
gives onto a dark passage which opens into revived the idea (1985, esp. 271-72; for reasons
the west colonnade. Again in some ways sim- which will become apparent, I do not think it
ilar to the access to Tiryns 16. Wace restored a was a place of political assembly). Since
megaron on the north side of the central court, Wace's pioneering analysis, the history of the
which places the main focus and orientation of Grave Circle has been made clearer thanks, in
the court out of sight to anyone entering. The particular, to the work of Pelon (1976), Gates
court measures just over six metres by about (1985) and Laffineur (1987, 1990). In almost
seven and a half: a fraction of the size of the every detail it differs markedly from the
130 Urbanism in the Aegean Bronze Age
Courts discussed above. Controlled through show of their power in public. It is plain that
the Lion Gate, nevertheless access to the Grave the Mycenaeans were masters of the architec-
Circle was more direct than to most of the ture of power (cf. Wright 1987). The Great
other courts we have considered. On the other Courts were the culmination of a tradition
hand when the circle of slabs was first erected, which saw its ever more sophisticated
in LH IIIB, there was no entrance at all (Gates expression. But the Great Courts very
1985: 270, Laffineur (less certain) 1990: 202- emphatically were not public: access was
203); and rather later in its history the way carefully controlled and their capacity was
into the entrance was obstructed by an exten- limited. Certainly there was, in the citadels, a
sion to the Granary. Inside the slabs it mea- conscious articulation of the courts and open
sures 25-26 m across (Felon 1976: 117), so the spaces, but one that was very different from
Circle encloses roughly 500 m2, in size larger that found in, say, the Central Courts and the
than any of the Great Courts. Its perimeter West Courts of Minoan Crete. In both cases
was, probably, dictated by the original circle there was a hierarchy, the peak of which is
of graves (see the careful discussion in the palace itself. In the case of the Mycenaean
Laffineur 1987, though note the dissenting palaces the megaron lay at the centre, the hub
view of Gates 1985), but quite a large number of political power, and the place where the
of people could congregate there. Whilst, on business of state was formally conducted,
the standard reconstruction, the grave stelai before a very restricted audience. Thereafter
would have stood prominent within the one seems to pass through a series of concen-
Circle, they did not form a single focus, and tric circles. Certain, at least ceremonial, activ-
the circular shape imposes no clear orientation, ities would take place in the only slightly
as even the entrance is not especially promi- larger arena of the Great Courts: to judge
nent. The perspective from within is much from the altar at Tiryns, this might include
more open. Although we cannot be certain sacrifice. Beyond this, in the outer courts at
how high the cyclopean circuit wall reached, Tiryns and Pylos access was still controlled,
the fine view, so missed by Wace from the but less strictly than in the inner circles.
Great Court, could be enjoyed from Circle A. Evidently economic transactions, the storage
Ceremonial conducted inside Circle A was of produce and craft activities would also be
visible from the walls, the roof of the Granary, found in these areas. Whereas the more cere-
the Ramp and buildings beyond, but in a monial aspects of royal power seem to have
rather casual fashion in marked contrast with concentrated on a processional progress,
the 'arena' of the Great Court above. Evidence rather than a static display. Such an interpre-
for the essentially cult usage of the Circle tation is suggested in the first place by the
comes from the large numbers of phi and psi focus, for example at Tiryns, on the gateways
figurines found by Schliemann (1880: 129, cf. which regularly mark the route of the proces-
71-74 and 88; cf. Wace 1921-23: 104-105).
sional way. In the second place this argument
helps explain the exception of Mycenae,
Conclusions where forecourts such as those at Pylos and
Tiryns, have not been located. At Mycenae
This essay started with the hypothesis that the two functions, the processional and the
open spaces within Mycenaean towns would redistributive, were separated. Storage areas
act as places where those in command might (for example the Granary, by the Lion Gate,
exercise their authority, or at least make a and the magazines by the Postern Gate) and
Empty Space? Courts and Squares in Mycenaean Towns 131
craft areas (such as the workshops to the west the frequent chariot scenes on pictorial kraters,
of the House of Columns lakovidis 1983: whose iconography was probably borrowed
63-64) were distributed in various parts of the from wall painting. Crouwel has written
site. The ceremonial progress from the Great 'Unlike the wall paintings from mainland
Court proceeded either by way of the Grand Greece, the numerous Mycenaean vases with
Staircase to, among other destinations, the Dual chariots rarely, if ever, show the vehicles
Cult Centre, or, by way of the Propylaeum in a military setting. From their earliest, rela-
ultimately to the Ramp and the Lion Gate tively accurately drawn, representations
(Wace 1921-23: 209-13). onwards the chariots are usually depicted as
The search for a public meeting place, where taking part in processions of some kind, mov-
populace and rulers might come together, has ing slowly and often accompanied by human
failed. Almost certainly future research holds figures on foot' (Crouwel 1981: 133, see also
surprises for us, and in particular the very 135-39 on the civil uses of dual chariots, and
imprecise picture we have of the Tower towns' 142 on civil uses of Rail chariots and chariot
associated with the great citadels leaves great racing).
scope to find important squares or other pub- Moreover, one thing we can say about the
lic places. Nevertheless I would like to hazard lower town at Mycenae is that it was tra-
a view which suggests an alternative interpre- versed by built roadways, which then contin-
tation. What has come through very clearly ued out into the surrounding countryside (for
from this discussion is the emphasis and focus a collection of recent references see Appendix
on procession. In particular, as Wright has 2). It has been argued traditionally that the
already pointed out, the series of monumental roads were constructed for chariots (Crouwel
gateways, which form such a marked 1981: 29-31 summarizes earlier views; see
sequence through the Mycenaean palatial com- also Jansen 1997 5-8 for a discussion of the
plexes, seem to give a rhythm and emphasis to military theory), but whilst one purpose
the progress from Megaron to the outer world might be military, they could also have served
(and vice versa). Procession is also a constantly other purposes: Lavery (1990: 1995) and
repeated theme in Mycenaean wall painting, Jansen (1997) have pointed to their impor-
indeed Immerwahr has identified the proces- tance for the movement of grain, in particular,
sion of all female votaries or priestesses as a and of building materials, and here I wish to
Mycenaean innovation, distinct from the point up their significance for processions.
Minoan, and reflecting a change of religious Jansen has, in a similar vein, suggested that
emphasis (Immerwahr 1990: 114-5). Likewise the network of roads from Mycenae acted as a
Shelmerdine has recently underlined the pro- sign of the political reach of the citadel. Great
cession as an important aspect of Mycenaean monuments, such as the Lion Gate and the
religious observance (Shelmerdine 1997: Treasury of Atreus, look onto the major routes
577-80). At times procession on the walls and would themselves have formed points on
seems to accompany the processional way, at a route where the populace could gather. I
least there is a notable example from the should not wish to build an edifice of infer-
vestibule to the megaron at Pylos (more gener- ence beyond what the evidence can bear. If,
ally on this see McCallum 1987 (non vidi)). however, the procession and the royal
There were processions on foot and there were progress are accepted as an important, per-
processions by chariot, to judge from the haps the important way in which rulers and
female charioteers in the Tiryns Boar hunt, and ruled came together in Bronze Age Greece,
132 Urbanism in the Aegean Bronze Age
1994 The spatial configuration of belief: the archae- Zangger, Eberhard, Michael E. Timpson, Sergei B.
ology of Mycenaean religion. In S. E. Alcock Yazvenko, Falko Kuhnke, and Jost Knauss.
and R. Osborne (eds.), Placing the Gods: 1997 The Pylos Regional Archaeological Project.
Sanctuaries and Sacred Space in Ancient Greece, Part II: Landscape Evolution and Site
37-78. Oxford: Oxford University Press. Preservation. Hesperia 66: 549-641.
9
The goal of this contribution is to summarize grid squares over the area of highest density
recent work by the Pylos Regional (Davis et al. 1997: 403-404, 427-30). From the
Archaeological Project (PRAP) in the vicinity study (chiefly by Cynthia Shelmerdine and
- both immediate and further afield - of the Sharon Stocker) of the over 35,000 ceramics
Bronze Age palatial centre at Ano Englianos, collected we were able to chart the site's
commonly referred to as the 'Palace of expansion through time (Davis et al. 1997:
Nestor' (Davis et al. 1997; Zangger et al. 1997 429, fig. 12; Bennet 1999: 12, fig. 2.3).
[extensive preliminary reports]; Davis et al. There was virtually no Early Bronze Age
1999; Davis 1998). The first section explores material in the immediate vicinity of the
the 'suburban' archaeology of the site, chart- palace. Indeed the earliest material on the
ing the growth of the so-called Lower Town Englianos ridge seems to have been confined
around the central palatial structures them- to the Deriziotis Aloni site that belongs to the
selves. The second section draws compar- latest phase of the Early Helladic (EH)
isons with the growth of nearby settlements, (Blegen et al. 1973: 219-24; Stocker 1995).
suggesting possible implications for relations Within the Middle Helladic (MH), the site
between these sites and that of Ano expanded considerably, reaching a maximum
Englianos. Because much of our discussion size in terms of the presence of diagnostic
has appeared elsewhere, we have restricted it MH material within collection units of
to a minimum. 5.48ha. By Late Helladic (LH) I-II, the early
Mycenaean phase, datable material was pre-
sent over 7.08ha. Within the following LHIII
Bronze Age Pylos phase, we ran into some difficulties in nar-
row chronological definition. Definitively
The Englianos ridge was one of the prime tar- LHIIIA material was distributed over 2.36ha;
gets of PRAP's programme of prehistoric definitively LH IIIB material over 4.6ha.
research in the area. Our major goal was to Although this might give the impression of a
elucidate from surface remains the extent of decrease in size in these periods, in fact mate-
the Late Bronze Age 'town' that was known rial that could be defined generally as LHIII
to exist around the palatial structures (e.g. (and therefore likely to contain both LHIIIA
Blegen et al. 1973: 3-68). The area was tract- and IIIB) extended over 12.4ha, a signifi-
walked in 1992 to determine the extent of the cantly larger area than that of the LHI-II
densest material, while in 1993 and 1994 total period. On this basis, we suggest that the
artefact collections were made in 468 20-m total area of the site in LHIIIB, including the
136 Urbanism in the Aegean Bronze Age
palatial structures themselves (c. 2ha) was southwest of the palace structures (Zangger
perhaps 14-15ha. et al. 1997: 606-13, esp. fig. 40).
We are constantly reminded that 'size mat- Thirdly, because we know (or think we
ters' but, in attempting to understand the know) that Bronze Age Pylos ended up as the
expansion of a complex site like that at Ano centre of political and economic power in the
Englianos, structure is perhaps more signifi- area, we tend to take its earlier development
cant. It is easy to view the changes in size for granted. It is worth reminding ourselves,
documented by PRAP from the point of view first of all, that the overall site has a long and
of the end product, the final palace and town. complex history. Even the central palatial
In such interpretations, growth is constructed structures changed substantially from early
ideologically towards this end, the inevitable to late LHIIIB, with perhaps three phases
progress of Bronze Age Pylos to become the (Shelmerdine 1997: 545, summarizing recent
largest site in Late Bronze Age southwest investigations by F.A. Cooper and M.C.
Messenia. Nelson). These were not just changes in form,
We would like to make three points in rela- but also in approach (Wright 1984: figs. 1 and
tion to such a 'story'. Firstly, the overall site 2; cf. Davis and Bennet 1999: 110). Similarly,
represented by the densest surface material the direction of approach to the site might
does not merely comprise undifferentiated have changed, as the overall topography and
habitation, but contains an important the discovery of a monumental stairway to
sequence of burial structures that were situ- the SW building suggest (Blackman 1997-98:
ated in different locations relative to the set- 56, fig. 79). In addition, areas changed func-
tlement centre in different periods (Bennet tion through time. It seems, for example, that
1999: 13, fig. 2.4). These are the so-called the 'Grave Circle' went out of use early and
'Grave Circle' constructed in late MH to the may have been overrun by settlement, unlike
southwest of the area of the later palace cen- the monumentalized open space in front of
tre; Tholos IV constructed in early LH I (per- Tholos IV. Perhaps the construction of Tholos
haps one to two generations later) to the III at some distance from the palace was
northeast; and Tholos III constructed in LHII prompted as settlement grew still further,
nearly a km distant along (i.e. to the south- although it may also have facilitated funerary
west) the Englianos ridgetop at a location display, now focused on the palatial elite, for
called Kato Englianos. whom, it seems, the tholos was the exclusive
Secondly, a wall was constructed around preserve by LHII.
the highest point of the citadel, apparently in
late MH/early LHI, with its gateway aligned
on Tholos IV, implying a relationship Pylos and its Neighbours
between whatever the wall encircled and the
tomb at roughly the same time (Blegen et al. We should also bear in mind that, at the end
1973: 4-18). Between the wall and the tomb, it of MH, Bronze Age Pylos was one of a num-
appears, was open space (Blegen et al. 1973: ber of communities of roughly comparable
64-68; Zangger et al. 1997: 604-605), perhaps size. Two other PRAP sites in particular offer
a plaza of sorts, associated with funerary dis- an instructive comparison with each other
play when the tomb was in use. It is possible and with Pylos itself (see also Bennet 1999:
that PRAP has traced the wall further, if it is 13-17; Shelmerdine 2001). After Bronze Age
represented by a geophysical anomaly to the Pylos, Koryfasio Beylerbey (II) and Gargaliani
Not the Palace of Nestor 137
Or dines (Kl) are the two most substantial Lolos 1987: 172-78). Fine and coarse ware
sites in the PRAP study area. In LHI-II, Pylos ceramics are of comparable number and
was already the largest in area of the PRAP quality at the two sites, adding to the impres-
sites, but it was just one of several substantial sion that they were both full and flourishing
communities in western Messenia. The grow- settlements, though Beylerbey is perhaps
ing success of these early Mycenaean powers only half the size of Pylos (3.32ha) by LHI-II
is measurable in several ways, such as settle- (Bennet 1999: 14). The LHI-II pottery at
ment expansion, the construction of elite Beylerbey is concentrated in the centre and
tombs, and a marked increase from MH in northeast parts of the site, as the MH material
the amount of datable pottery (Bennet 1999; had been. Because of the difficulty in distin-
Shelmerdine 2001). guishing early Mycenaean from Mycenaean
coarse wares, however, it is not possible to be
Table 9.1 PRAP statistics for Pylos, Beylerbey and clear about the relative proportion or place-
Ordines (after Bennet 1999; Shelmerdine 2001) ment of these during early Mycenaean.
Site Area in ha In the LHIII period Beylerbey grows little
in area, but the amount of pottery decreases
EH MH EH/-// EH///
by nearly two-thirds. It is also diffused over
Pylos 5.48 7.08 12-14
Beylerbey 1.64 3.32 3.52 a larger area of the site, though still predom-
Ordines 0.6 0.92 2.1 inantly across the center. Other finds are few
and help little in defining the nature of the
Site Datable potsherds
settlement. Twenty nine chert and eight
EH MH EH/-// EH/// obsidian lithic artefacts were scattered
Pylos 33 671 2251 8052 around much of the site, particularly in the
Beylerbey 5 98 656 266
Ordines 12 33 95 152 east-central part where LHI-II and LHIII
ceramics predominate, but, as surface finds,
Site Potsherds per ha few could be dated. A human figurine frag-
EH MH EH/-// EH/// ment (SF 0018: Davis et al. 1997: 453) and a
Pvlos 122.5 288.6 575.6 sealstone of Younger's Mainland Popular
Beylerbey 59.8 197.6 75.6 Group (SF 0091: Davis et al. 1997: 453, pi. 90f)
Ordines 55 103.3 72.4
are certainly LH IIIA-B, and a grinding stone
and a spindle whorl/dress weight (Davis et al.
At both Pylos and Koryfasio Beylerbey (PRAP 1997: 446, fig. 16: 6, 453) complete the MH-
site II), the presence of a noticeable amount LH picture. The decrease in pottery from
of MHIII and LHI ceramics indicates the tim- LHI-II to LHIII is a notable feature of the
ing of this growth spurt more precisely than site's history. One can argue that it suggests a
at other sites within the PRAP study region. diminution of the settlement's importance,
A tholos tomb was constructed at the same perhaps reflecting its demotion from compet-
time (late MH) at each site, the earliest exam- ing power to second-order center, after Pylos
ples in Greece of this new elite monumental emerged as the dominant center of the
tomb form. The one at Pylos is the so-called region, as it appears to have done in LHIIIA.
'Grave Circle' (Blegen et al. 1973: 134-76), Despite its apparent diminution in the LHIII
while Beylerbey is probably to be associated period, Beylerbey remains the only site
with the Osmanaga tholos tomb (Hope known in this vicinity with a comparable
Simpson and Dickinson 1979: 130-31 [D5]; presence to that of Bronze Age Pylos. This
138 Urbanism in the Aegean Bronze Age
fact, along with its geographical position Implications from Textual Evidence
south of Pylos and its coastal perspective,
suggests that it could be the location of a-ke-re- Despite their discrepancies in area, Pylos,
wa, a taxation center and coastal lookout point Ordines and Beylerbey are the only PRAP sites
attested in the Pylos tablets (Davis et al. 1997: where numbers of potsherds are in triple dig-
424-27, fig. 11). As noted above, such a guess its during the Mycenaean period; other sites
depends on whether size does matter. It became or remained smaller still, and those
seems likely that it does, and also that higher immediately to the east of Mt. Aigaleon, which
concentrations of finds go along with higher we have followed Chadwick in recognizing as
degrees of importance. The usual caveat must the provincial boundary of the polity, are vir-
however apply, that assumptions based only tually abandoned in this period (Davis et al
on surface material can carry only limited 1997: 423-24). If the identifications of Ordines
weight (cf. Zangger et al. 1997: 575-76). and Beylerbey with taxation centres are valid,
Another site with a different but equally one could speculate that Pylos, as the control-
interesting history in relation to Pylos is ling administrative authority, was able to
Gargaliani Ordines (PRAP site Kl), which lies encourage certain sites to flourish at the
a little way inland from the coast on the high expense of others. The incentive would be to
southern edge of the Langouvardos River create a Network of Preferred Providers, as
gorge (e.g. Bennet 1999: 11, fig. 2.2). This is American insurance companies call it, a net-
the only site within PRAP's study region work we see in the tablets as the nine 'taxation
apart from Pylos itself to exhibit steady centres' of the so-called Hither Province. The
growth from Neolithic/EH to LHIII. It is also continuing growth of Ordines might indicate
the only known Mycenaean site north of the that it was one of the settlements that emerged
palace with a coastal view. Ordines is always in the Mycenaean period as beneficiaries of
much smaller than Beylerbey, covering the palatial administrative system. Beylerbey,
0.92ha in the early Mycenaean period and though larger than Ordines in area, is reduced
doubling to 2.1 ha in the Mycenaean period to a comparable density of ceramics per ha.
(Bennet 1999: 16-17). LHI-II pottery is found This too could have been a direct result of the
all over the site, while in the LHIII period the concentration of authority at Pylos, and per-
heavier concentrations are in the southern haps even a desire on the part of the centre to
part. Lithics are numerous (69 chert, 11 obsid- limit the strength of its former competitor and
ian, 1 quartz), especially in the northern part, current subordinate.
with a heavy cluster of obsidian and espe- An important implication of this view of
cially chert in the northwest sector. Six MH- sites as 'taxation centres' is that the references
LH grinding stones (saddle querns) were on Linear B documents, particularly to the nine
found in the south-central part of the site, places of the Hither Province, are not to dis-
two at least of LH date (Davis et al. 1997: 453). tricts, but to place-names representative of dis-
The only other Bronze Age find is a LH IIIA- tricts. It is tempting to see these as the chief
B clay spindle whorl or dress weight. The settlements within the Pylos polity. However,
prominence and strategic location of this site if we look behind the specific documentary
are unmatched in this vicinity, and it is a mentions, then some inconsistencies emerge.
good candidate for the taxation center pe-to- In the first instance, not all such place-names
no in the northern part of the Hither Province are of equal status. Some, notably in the
(Davis et al. 1997: 424-27, fig. 11). Further Province list of seven or eight places
Not the Palace of Nestor 139
appear to be explicitly district rather than set- 'taxation centre' in the assessment and contri-
tlement names (e.g. a-te-re-wi-ja, 'Atrewia', 'the bution texts. It is also not clear that 'taxation
land of Atreus': PY Ma 244 [see Ruiperez and centres' are always significant settlements.
Melena 1990: 115; Melena 1994-95: 274]). For example, in the case of one such place-
Equally, some names appear to be interchange- name, ro-u-so, we have the document that
able, as in the case of ro-u-so (PY Jn 829.10) gives the total assessment for its district (PY
which alternates with e-ra-to (PY Cn 608.9; Vn Ma 365) and another giving the breakdown
20.9), or the name *e-ro (PY Jn 829.19) which is of contributions of the type of cloth denoted
apparently 'replaced' by the district names a- by the logogram *146 for it and the eight
te-re-wi-ja (Ma 244) and e-sa-re-wi-ja - also other place-names in its district (PY Mn 456:
based on a name, either a title or a personal Killen 1996). ro-u-so contributes a smaller
name (Ma 330). More significantly, those texts amount of *146 than others within its area,
that deal with 'taxation centres' deal only with something that goes against our expectations,
a selective range of commodities, no doubt if such 'taxation centres' are major places
those generally available across the polity and within their districts. We should note, how-
not subject to micro-regional variation, notably ever, that ro-u-so clearly has an importance in
in the standard suite of commodities, includ- the palace-organized manufacture of textiles
ing a basic cloth type (*146) and animal skins from its mention in the Pylos Aa / Ad records,
(*152), assessed in the Ma tax documents. which suggest a female workforce of 86, sec-
It is interesting to compare the mentions ond only to Pylos itself in the Hither Province
made of some place-names in contexts (Chadwick 1988: 85).
beyond their listing in the standard range of If we are reading the variation in archaeo-
assessment or contribution texts. The logical prominence of sites correctly, then the
'residue' of references are not equivalent and shifting pattern of relations with the centre
might tell us something about different func- that it implies and the possibility suggested
tions or different relationships between them by the Linear B data that sites of apparently
and the palatial centre. For example, pe-to-no lesser status become the focal points for tax
only occurs in assessment or contribution collection, then a possible implication is that
texts, while a-ke-re-wa enjoys a range of addi- relations between the centre at Englianos
tional mentions (including personnel, land, and other sites were mediated through indi-
and livestock). One can argue that these ref- viduals. These would either be anonymous
erences are consistent with the identifications officials (such as the ko-re-te and po-ro-ko-re-te
suggested above relating to the sites of explicitly identified as responsible for
Ordines and Beylerbey: Ordines (if it is acquiring the bronze mobilized in PY Jn 829)
Linear B pe-to-no} appears to acquire signifi- or named members of the elite. In this inter-
cance only within the LHIII period, whereas pretation, therefore, Pylian expansion would
Beylerbey (if it is Linear B a-ke-re-wa) appears be less about bounded territorial gain than
to have a more complex history and therefore about the negotiation of relations with key
relationship with the palatial centre. A simi- individuals at sites that came to have a
larly complicated situation appears to arise in strategic importance in the centre's strategy
the case of the place-name pa-ki-ja-ne, clearly of mobilizing commodities to support its
an important cult centre, probably close to political economy.
the palace, perhaps in the vicinity of Hora In the broader context of urbanism in the
Volimidia (e.g. Chadwick 1972:109), but also a Aegean Bronze Age, we hope to have
140 Urbanism in the Aegean Bronze Age
suggested ways in which detailed archaeo- Davis, J.L., J. Bennet and C.W. Shelmerdine
logical data can be brought into dialogue 1999 The Pylos Regional Archaeological Project: the
prehistoric investigations. In P. Betancourt, V.
with textual data to outline the overall devel- Karageorghis, R. Laffineur, and W.-D. Niemeier
opment of power relations among settle- (eds.), Meletemata. Studies in Aegean Archaeology
ments through time in the Pylos region. Just Presented to Malcolm H. Wiener, I, (Aegaeum 20),
as it seems that not all sites within the LHIIIB 177-84. Liege: Universite de Liege.
Davis, J.L., S.E. Alcock, J. Bennet, Y.G. Lolos and C.W.
Pylian polity had identical histories, we Shelmerdine
would like to suggest that comparative 1997 The Pylos Regional Archaeological Project.
research into the particular histories of other Part I: overview and the archaeological survey.
Hesperia 66: 391-494.
regions dominated by Mycenaean centres
Hope Simpson, R. and O.T.P.K. Dickinson
might well reveal differences that have impli- 1979 A Gazetteer of Aegean Civilisation in the Bronze
cations for our conception of the Mycenaean Age, Vol. I: the Mainland and Islands. SIMA 52.
world as a unity. Goteborg: Paul Astroms Forlag.
Killen, J.T.
1996 Administering a Mycenaean Kingdom: some
taxing problems. BICS 41: 147-48.
Bibliography Lolos, Y.G.
1987 The Late Helladic I Pottery of the Southwestern
Peloponnesos and its Local Characteristics. SIMA-
Bennet, J.
PB 50. Goteborg: Paul Astroms Forlag.
1999 Pylos: the expansion of a Mycenaean palatial
Melena, J.L.
center. In M.L. Galaty and W.A. Parkinson
1994-95 133 joins and quasi-joins of fragments in the
(eds.), Rethinking Mycenaean Palaces: New
Linear B tablets from Pylos. Minos 29-30:
Interpretations of an Old Idea, 9-18. Los Angeles:
271-88.
Cotsen Institute of Archaeology.
Ruiperez, M.S. and J.L. Melena
Blackman, D.J. 1990 Los griegos micenicos. Madrid: MELSA.
1997-98 Archaeology in Greece 1997-98. AR 44: 1-128. Shelmerdine, C.W.
Blegen, C.W., M. Rawson, W. Taylour and W.P. Donovan 1997 Review of Aegean Prehistory VI: the palatial
1973 The Palace of Nestor at Pylos in Western Messenia, Bronze Age of the southern and central Greek
Volume III: Acropolis and Lower Town, Tholoi and mainland. AJA 101: 537-85.
Grave Circle, Chamber Tombs, Discoveries outside 2001 The evolution of administration at Pylos. In S.
the Citadel. Princeton: Princeton University Press. Voutsaki and J.T. Killen (eds.), Economy and pol-
Chadwick, J. itics in Mycenaean palace states, 113-28.
1972 The Mycenaean documents. In W.A. Proceedings of the Cambridge Philological
McDonald and G.R. Rapp, Jr. (eds.), The Society Supplement. Cambridge: Cambridge
Minnesota Messenia Expedition: Reconstructing a Philological Society.
Bronze Age regional environment, 100-16. Stacker, S.R.
Minneapolis: University of Minnesota Press. 1995 Deriziotis Aloni: a small Bronze Age site in
1988 The women of Pylos. In J.-P. Olivier and T.G. Messenia. Unpublished MA thesis, University
Palaima (eds.), Texts, Tablets and Scribes: studies of Cincinnati.
in Mycenaean epigraphy and economy offered to Wright, J.C.
Emmett L. Bennett, Jr., 43-95. Minos Supplement 1984 Changes in form and function of the palace at
10. Salamanca: Universidad de Salamanca. Pylos. In C.W. Shelmerdine and T.G. Palaima
Davis, J.L. (ed.) (eds.), Pylos Comes Alive: industry and adminis-
1998 Sandy Pylos. An Archaeological History from Nestor tration in a Mycenaean palace, 19-29. New York:
to Navarino. Austin: University of Texas Press. Archaeological Institute of America/Fordham
Davis, J.L. and J. Bennet University.
1999 Making Mycenaeans: warfare, territorial expan- Zangger, E., M.B. Timpson, S.B. Yazvenko, F. Kuhnke
sion, and representations of the Other in the and J. Knauss
Pylian kingdom. In R. Laffineur (ed.), Polemos. Le 1997 The Pylos Regional Archaeological Project.
contexte guerrier en Egee a I'age du Bronze, 105-20. Part II: landscape evolution and site preserva-
Aegaeum 19. Liege: Universite de Liege. tion. Hesperia 66: 549-641.
10
Among Aegean prehistorians, there has been In a volume devoted to urbanism in the
a long-term interest in urbanism that seems to Aegean Bronze Age, then, of what relevance
be traceable to the impact of Gordon Childe's is the present paper about the countryside
writings concerning the 'urban revolution' in and, specifically, the rural hinterlands of
the Near East, and to the corollary belief that Mycenae? Anthony Snodgrass provided the
comparable processes of urbanization must justification, in his closing remarks to the
have been part and parcel of the rise of com- 1992 Corfu conference on Rural Structures
plex societies in the Aegean. We are hesitant and Ancient Societies. He wrote, 'I think it is
to apply the term 'urban' or even 'proto- a point that cannot be stressed too much that
urban' to any place, at any period of Aegean the word rural is meaningless except in the
prehistory (except, perhaps, in Crete). The context of an urban counterpart. There is no
point has been made often enough before: but such thing as rural settlement until there is
size does matter, and questions of scale are urban settlement. The two are logically coun-
significant. Even if the calculations require terposed' (Snodgrass 1994: 483). Turning this
substantial recalibration (see Branigan and around, it is difficult to see how one can
Whitelaw, this volume), one of the most strik- study urbanism as a phenomenon, without
ing illustrations in Renfrew's The Emergence of taking on board the disposition of popula-
Civilisation (1972: 236-44, fig. 14.8) still retains tion, settlements, and resources within the
much of its force—the one displaying plans of wider region that helped sustain such high-
some of the most important Minoan and est-order central places, and over which
Mycenaean sites alongside their Near Eastern they presumably exercised some measure of
counterparts, showing that many are tiny by political authority.
comparison, and some even of a different This paper, thus, concerns the Argolid in
order of magnitude altogether. The word the Mycenaean era; and its focus is not the
'urban' conjures up for the modern reader a site of Mycenae itself (whether or not we
set of notions and preconceptions that are choose to apply to it the term urban),
largely irrelevant or inappropriate for many but rather some of the lands which—in
ancient world contexts, with the result that Schliemann's (1878: 28) memorable phrase
there is often something of a semantic mis- —probably lay 'under the sceptre of
match between conceptual terminology and Agamemnon'. In this regard, there is an
empirical data (e.g. for Aegean prehistory, obvious, but nonetheless useful distinction
Konsola 1984; Ha'gg and Konsola 1986; to be drawn between three distinct spatial
Cosmopoulos 1998). frames of reference:
142 Urbanism in the Aegean Bronze Age
(1) Study of the central place itself (i.e. the renowned map of the Mycenae area more
nature, size and functions of Bronze Age than a century ago (Steffen 1884), and it will
towns). be invaluable in much the same way that the
(2) Study of the peri-urban, peripheral, outer Knossos Area Survey helped fill out the urban
zones, or suburbs. and peri-urban context of the palatial struc-
(3) Study of the urban centre's rural hinter- tures at its core (Hood and Smyth 1981). So far
lands (which might well shade off into as we are aware, however, no detailed and
exploitable territory beyond its direct systematic counting, collection or mapping of
political control, as with the eschatia on the the density and distribution of artefacts of dif-
fringes of the chora of a colonial Greek ferent dates was attempted as part of the
city). Mycenae Survey, which seems to us a missed
opportunity, considering the success of large-
In the case of Mycenae, the first of these has site surveys of this sort in other recent pro-
been the primary focus for investigations jects, notably in Boeotia (Bintliff and
since the 1870s. It could fairly be claimed that Snodgrass 1988; Snodgrass and Bintliff 1991),
we now have a reasonably sound under- the Corinthia (Alcock 1991), the Cyclades
standing of the site's development and inter- (Cherry et al. 1991: 199-216, 265-84), and
nal organization, if not necessarily of all its Crete (MacGillivray and Sackett 1984; Moody
functions—let alone how it was experienced et al. 1998; Whitley et al. 1999). Thus it is at
by those who ruled or resided or worked present only in Messenia, from new work
there, or who visited it from elsewhere, or undertaken around the Palace of Nestor and
what it symbolized (for provocative sugges- along the Englianos ridge (Davis et al. 1997:
tions about these latter aspects, see especially 427-30, fig. 12; Bennet 1999; Bennet and
Wright 1987). Shelmerdine, this volume), that we have any
Whether the Cyclopean fortification walls possibility of assessing changes in the size or
bounding the citadel represent the most nature of the community at a site which
appropriate division between the core and developed later in Mycenaean times as a cen-
peri-urban areas is not clear, but this is a ques- tral place with a palace.
tion on which some light may be shed by the An interesting aspect of the Mycenae
as-yet-unpublished Mycenae Survey, begun Survey has been the renewed attention paid
in 1991 by Elizabeth French with Spiros to the surviving (but fast disappearing) traces
lakovides, under the aegis of the Athens of the system of several roads which provided
Archaeological Society. This was not a survey the essential link between the citadel and the
in the usual sense of the word, but rather a adjacent built-up areas on the one hand, and
detailed, computer-based sites-and-monu- on the other the open country of its sustaining
ments record within the 'Greater Mycenae' rural hinterlands (Figure 10.1). There are sig-
core area, resulting in the precise mapping of nificant differences between the maps of this
both old and hitherto unrecorded cultural fea- road network as proposed by Mylonas (1966:
tures—over 200 tombs, many hundreds of 86-87), Hope Simpson (1981:15-17, fig. 2, pis.
other structures, traces of roads, quarries, etc. 2b, 6), Lavery (1990), and Jansen (1994; 1997),
(for preliminary accounts, see French 1992: and we must also take account of the re-sur-
16-17; 1993: 18; 1994: 16; Tomlinson 1995: 12). vey of the roads that run through parts of the
This work was avowedly an attempt to pro- area studied by the Berbati-Limnes Survey
duce a cyber-age version of Steffen's (Wells 1996: 130-33, figs. 4-10; Wells et al.
'Under the Sceptre of Agamemnon' 143
1993: 57); Crouwel (1981: 29 n. 5) surveys Late reaches of the Argive Plain, thus leading into
Bronze Age roads that are known in a number the Tretos and Kelossa passes, which provide
of other areas of Greece. Three of the four access to the Phliasian Plain, and the Nemea
routes for which well-documented traces sur- and Kleonai valleys (Jansen 1997: 4, map 1).
vive, especially within the peri-urban zone This area north of Mycenae has always
close in to Mycenae itself, lead to the seemed rather empty of traces of Mycenaean
north—presumably heading for various settlement (Figure 10.2). One very striking
points in the southern Corinthia. One, in feature is the total absence of tholos tombs
particular—Jansen's Highway 3—originating and other built graves in this area, both in
at the Lion Gate, runs directly north along the Early Mycenaean times and at the acme of
western flank of Mt. Profitis Elias, using Mycenaean power in LH IIIA-B; even cham-
Cyclopean bridges to cross a number of ber tombs are scarce (Cavanagh and Mee
ravines, before descending into the upper 1998: figs. 5.1, 5.2, 6.1, 6.2, 6.3). Extensive-
Figure 10.1 Argolid and Corinthia with major sites and reconstructed courses of Mycenaean roads; dashed lines indi-
cate roads in use before LH IIIB, solid lines roads in use in LH IIIB and LH IIIB/C. (Drawing by Rosemary
J. Robertson; course of roads after Lavery 1990; 1995.)
144 Urbanism in the Aegean Bronze Age
Figure 10.2 Tombs in the Argolid and Corinthia in the early Mycenaean period (top) and LH IIIA/B period (bottom).
(Drawing by Rosemary J. Robertson; data for burials from Cavanagh and Mee 1998.)
'Under the Sceptre of Agamemnon' 145
style reconnaissance throughout the Corinthia to conduct an intensive survey which could
in the 1960s and 1970s turned up very little provide information on the size, density,
(Wiseman 1978). Zygouries, in the valley of nature, and distribution of both early and
Kleonai, about 7 km north-northwest of developed Mycenaean settlement for one
Mycenae, is a well-known site, but it is no strategically located valley system near
Mycenae (Blegen 1928). What was going on Mycenae (Figure 10.3). In fact, the southern
in these northern hinterlands—which Bintliff, boundary of the survey area lies at the edge
for one, thinks fell squarely under the politi- of the Argive Plain, on the very doorstep of
cal authority of Mycenae (Bintliff 1977: 91-93, Mycenae, only c. 4 km further to the south-
fig. 15B; cf. Hope Simpson and Dickinson east. Yet within this region of some 85 sq km
1979: maps 3, 4, A)? Might we not expect to Hope Simpson and Dickinson reported in
be able to monitor the process of the forma- their gazetteer only two Late Bronze Age
tion of the kingdom centred on Mycenae also sites (1979: 67-68, sites A70 and A71)
in its rural hinterlands? Surely, the reverbera- —Tsoungiza (since re-excavated by NVAP:
tions of Mycenaean power would be felt in see Wright et al. 1990: 618-38) and Phlius
the countryside too? (with just a single kylix stem and a steatite
One ambition of the Nemea Valley whorl); a third site (A72, Agia Irini), which
Archaeological Project (NVAP), which took has some material of late MH/ early LH date,
place between 1983 and 1989, was precisely lies just outside the survey area, on the east-
Figure 10.3 A view towards Mycenae from near the southern boundary of the NVAP survey area.
146 Urbanism in the Aegean Bronze Age
ern edge of the Phliasian Plain. The only be a consequence of historical and political
other Mycenaean material previously known variation' (Cavanagh 1995: 81). A number of
in this region is that from Stephen Miller's recent studies have used survey data produc-
excavations in the Sanctuary of Zeus at tively in this way to consider inter-regional
Nemea, although in no place has a pure variation in the prehistory of the Peloponnese
Mycenaean stratum yet been encountered (e.g. Cavanagh 1995; 1999; Mee 1999; Rutter
there (Klein 1987). The Nemea survey has 1993: 747-58, table 1; Shelmerdine 1997:
filled out this picture considerably, and in 550-54, table 2; cf. the earlier, largely pre-sur-
what follows we present an outline sketch of vey regional comparisons of Dickinson
the results, as they pertain to the later Bronze [1982]), and of Crete (e.g. Watrous 1994:
Age. (For an earlier study of Neolithic settle- 698-99; Driessen 1999; this volume). What is
ment within the area of the NVAP survey, see most relevant for present purposes, however,
Cherry et at. 1988; for an overview of the pro- is that—in a few areas—the amount of survey
ject and of all periods in this survey, see effort has been sufficient to start building up
Wright et al 1985; 1987; Wright 1990; Cherry a local regional picture based on several sets
et al. n.d.) of survey data, albeit collected at different
At the time, this was perhaps the first field- times, by different teams, with different goals
work, conducted in a reasonably systematic in mind.
and intensive manner, to produce detailed This is at last beginning to be the case for
information about the pattern of rural settle- Mycenae's hinterlands (Figure 10.4). Here we
ment and artefactual 'offsite' activity in the have, in addition to the Greater Mycenae
vicinity of a centre of Mycenaean civilisation. Survey, published or partly published inten-
Since then, others have taken place, and some sive surveys of the Nemea region to the north-
are even published in final form (notably in west of Mycenae, and the Berbati-Limnes area
Messenia [Davis et al. 1997; Davis 1998] and immediately to its east and northeast. New
Laconia [Cavanagh et al. 1996])—part of the surveys are currently under way by J. Maran
huge spurt of regional survey activity in the in the Phliasian Plain, adjacent to the EH and
Aegean over the past two decades or more LH sites at Petri and Aidonia, and by T.
(Cherry in press; in prep.). These surveys are Gregory and others in the eastern Corinthia.
now making it possible to look at regional and A good deal of work has been done in and
inter-regional settlement trajectories and paths around the Argive Plain itself, especially by
to complexity in some interesting new ways. H.-J. Weisshaar (1990) and the late K. Kilian,
This can be at a relatively gross level—for for which Zangger's (1993; 1994) extensive
example, Alcock (1993) for Hellenistic and geoarchaeological studies provide a useful
Roman Greece as a whole, Halstead (1994) on context. Farther afield to the south and south-
the north-south divide in the prehistoric east lie the areas of excellently conceived
period, or Bintliff (1997) on the long-term shift regional projects on Methana (Mee and
of demographic and political focus from Forbes 1997) and in the southern Argolid
southeast to northwest. Or it can be at a more (Jameson et al. 1994), though whether these
intermediate scale, such as 'between areas areas fell within the ambit of Mycenae's
such as Messenia, Laconia, the Argolid, direct, or even indirect, control remains a
Boeotia and Central Greece, where the land- debatable point. When the results of all of this
forms and productive potential are compara- work are fully available, they will comprise a
ble, and where differences are more likely to vast databank of information for studying
'Under the Sceptre of Agamemnon' 147
Figure 10.4 Archaeological surveys in the Argolid and Corinthia; NVAP (Nemea Valley Archaeological Project), EKAS
(Eastern Korinthia Archaeological Survey), AEP (Argolid Exploration Project). (Drawing by Rosemary J.
Robertson.)
relations of different types and scales, and as this regard. Third, we consider, albeit very
they changed over time, between the centre at sketchily, in what respects the Nemean data
Mycenae and outlying zones. conform to, or differ from, the picture pre-
In the remainder of this paper, we can sented by some of the other nearby projects
touch only briefly on several relevant topics. just noted. Finally, some preliminary and ten-
First, we provide a summary of the data from tative suggestions will be offered about the
the Nemea survey for early and developed possible nature of the connections between
Mycenaean settlement and landuse patterns. the emergent centre at Mycenae and these
The interpretation of these data engages outlying areas.
directly with some recent claims that the sur-
viving surface archaeological record for any
of the prehistoric periods in Greece seriously Mycenean Nemea
under-represents the density and distribution
of sites that once existed; and so, secondly, a Work in the valley of Ancient Nemea and
few comments (again drawing directly on adjacent areas during the 1980s resulted in
information from Nemea) are warranted in virtually total high-intensity survey coverage
148 Urbanism in the Aegean Bronze Age
of the valley itself, parts of the Tretos Pass, demands care, taking into account not only
the valley of Xerokambos / Tourkovrisi to the findspots (on- and off-site), but also quantity
south of the town of New Nemea, as well as and type of artefacts, the spatial extent of
the whole of the acropolis of ancient Phlius their distribution, and the mode of surface
on the west, and territory towards ancient collection employed. We return to this point
Kleonai to the east and Mycenae to the south- below.
east (Figure 10.5). Field methods are not our LBA artefacts (mostly pottery) were found
focus here: suffice it to say that we have gen- at about 25 sites, in addition to Tsoungiza and
erated essentially two main types of data for the adjacent Sanctuary of Zeus, as well as in
considering Mycenaean settlement and land- much smaller quantities at several dozen off-
use patterns. The first consists of the distribu- site locations, widely distributed throughout
tion and density of artefacts counted or the study area (Figure 10.6). At many of these
collected during close-spaced fieldwalking of sites, only a few sherds of Mycenaean pottery
some 5,400 individual tracts, with a median were found—a density, that is, not much
area of almost exactly 1 ha. The second com- higher than at off-site locations producing
prises material picked up, using a variety of Mycenaean material. Eight sites yielded more
collection systems, at locations defined, on substantial material (Table 10.1): they are
the basis of local variations in artefact den- widely dispersed, all at relatively low eleva-
sity, as 'places of special interest' or sites. Of tions, and most on or near natural routes of
course, some Mycenaean artefacts collected communication.
during initial tract-walking are in fact from Not surprisingly, sites where the most
locations that were later defined as sites; but Mycenaean material has been found are also
in many other instances, they were found in the largest in terms of the area over which the
very small quantities and in isolation, and Mycenaean finds are distributed—but since
thus may not have anything to do with set- virtually all 100 or so sites defined in this sur-
tlement at the locations where they were vey are multi-component, overall site size is
found. This might also be true of cases in an unreliable basis for estimating the size of a
which just a few Late Bronze Age (LBA) arte- site in any specific period represented among
facts were collected on sites where the vast the finds from it. Without going into the
majority of finds belong to other (usually, details of how we have estimated the area
later) periods. Making sense of the overall over which finds specifically of LBA date were
patterns of Mycenaean material thus recorded, nor of why we think the estimate of
Site El. (m) Setting Min./Max. size Collection method Phases represented
Figure 10.5 Topographic map of the Nemea Valley, with tracts defined by the Nemea Valley Archaeological Project
superimposed. (Drawing by Rosemary J. Robertson.)
150 Urbanism in the Aegean Bronze Age
size in one case (site 925) is greatly exagger- ha or less category, which in turn are far more
ated, the conclusion is that all these conspicuous in the landscape than the several
Mycenaean components seem to be smaller dozen other locations where Mycenaean mate-
than 2 ha, some much smaller still. rial is scanty, but at least some of which may
None of them rivals Tsoungiza. Here we very well reflect rural settlement at the lowest
have the combined evidence from tract-walk- level, and other non-residential activities.
ing around that site's periphery, from gridded Assuming (as seems virtually certain) that the
surface collections, and from our excavations, site of Mycenae itself represents the highest-
to estimate its size, conservatively, as c. 7.5 order level of settlement in this part of the
ha—clearly much larger than any other site northeastern Peloponnese, there thus appears
we know of in this region (Wright et al. 1990; to be a four-level hierarchy.
Cherry et al. n.d.). There exists other evidence, About sites comparable to Tsoungiza in this
too, suggesting that the dominant focus of northern and northwestern chunk of
Mycenaean settlement was at this southern Mycenae's hinterland we cannot be very cer-
end of the Nemea valley. Mycenaean finds tain. Zygouries must be one. Phlius may be
have been found in substantial quantities in another, judging from the thin but pervasive
three different parts of the Sanctuary of Zeus, scatter of MH and LH finds on the acropolis
though never in pure strata and generally ridge and in its foothills revealed by the very
heavily worn; topography alone precludes systematic, large-site survey carried out by
the possibility that all these finds eroded from Alcock (1991: 458, fig. 12; cf. Blegen 1925: 32;
nearby Tsoungiza, and nothing about them Vermeule 1987: 134); but the overburden of
supports the suggestion that a Mycenaean historical strata there is problematic. Ayia
cult place preceded the later historical sanctu- Irini, a low hill on the west bank of the River
ary. It is also striking that tracts producing off- Asopos 2.5 km west-northwest of modern
site Mycenaean finds are much more New Nemea, is not as large (c. 1.5 ha), but has
abundant at the head of the valley—south of produced substantial material of LH IIIA and
Tsoungiza and the modern village of Archaia IIIB date (Miller 1990: 53). And further afield
Nemea, and east and northeast of the across the Phliasian Plain lies Aidonia, with its
Sanctuary of Zeus—than in any other part of rich chamber tombs (Demakopoulou 1996);
the survey area (Figure 10.6). The evidence but such wealth, at this spatial remove from
could well suggest something like an artefac- the centre, may well indicate a political rival,
tual 'halo' built up around Tsoungiza, quite rather than a subordinate, of Mycenae—at
possibly as a result of local short-term activi- least in the period before the zenith of
ties, in-field manuring, dumping, erosion, Mycenae's power in LH III.
and no doubt numerous other factors besides. This brings us to chronology, and here the
There exist other interpretative difficulties evidence is striking. MH finds from survey are
too (see below); but what emerges from all this extremely rare in the Nemea area, at least
evidence is the notion that in Mycenaean before the late MH/LH transition (i.e. Shaft
times there existed a fairly clear size-hierarchy Grave times; for this period in the Argolid as a
of sites in this region. Tsoungiza and the area whole, see Dietz 1991). That this is not just a
in its immediate proximity at the southern end surface ceramic visibility problem of the kind
of the Nemea valley is dominant, and it highlighted by Rutter (1983) is indicated by the
appears to be at least three or four times larger excavations at Tsoungiza, which seems actu-
(perhaps more so) than any of the sites in the 2 ally to have been abandoned after EH III until
'Under the Sceptre of Agamemnon' 151
Figure 10.6 Mycenaean finds and soil units in the Nemea Valley region. (Drawing by Rosemary J. Robertson; map of
soil units courtesy of A. Demitrack.)
152 Urbanism in the Aegean Bronze Age
the Shaft Grave period. A similar pattern of the extent to which alluvial soils may today
emerges from the finds at the Sanctuary of be masking prehistoric land surfaces (for a
Zeus (Klein 1987). The Early Mycenaean summary, see Demitrack, in Wright et al. 1990:
period is quite well represented throughout 587-91, fig. 3, table 1). Demitrack initially
the study region by finds made both on- and defined and sequenced soil units with refer-
off-site (Figure 10.6), but in each type of collec- ence to soil characteristics, stratigraphy, and
tion far fewer artefacts can be dated to this artefacts found in exposed sections or col-
period than to LH IIIA2-IIIB. There is a good lected from sites situated on top of the soil
deal of site continuity as well: six of the seven (Figure 10.6). Three Holocene units were
sites where later MH finds are known continue identified: HI, the oldest, deposited some
into the next phase, while LH artefacts have time between Early Neolithic and Early
been identified at nearly 20 additional sites. Helladic II; H2, being post-classical, and in
Not surprisingly, much of the survey pottery part post-Roman; and H3, which is quite
recognizable as Mycenaean cannot be closely modern. Subsequent mapping of the distribu-
dated, but three things are worth mentioning: tions of artefacts collected from sites and
first, that every major stage of the Mycenaean tracts confirms this picture. In the main valley
period can be recognized among these surface of the Nemea River, no site where Mycenaean
collections; second, that in both tract and site finds were collected lies on the H2 or H3 soil
collections, far fewer artefacts can be dated to unit; nor is any tract where Mycenaean finds
the late MH, Shaft Grave or Early Mycenaean were collected situated in its entirety on the
periods than to developed LH IIIA2-IIIB; and H2 or H3 unit. Excavated portions of
third, that the sites with significant amounts of Tsoungiza lie on marl and colluvium, whereas
material were evidently used for long periods the Sanctuary of Zeus is entirely covered by
of time, and both early and late Mycenaean the H2 soil unit. Naturally, this means that
pottery has been found at most of them (Table there are some parts of the lowermost valley
1). In short, the data suggest the development bottom where prehistoric remains could also
of a long-lasting 4-level site hierarchy, spread be buried beneath the H2 or H3 soils, espe-
widely throughout the study region, begin- cially at its southern end—but this in fact is
ning in the Shaft Grave era, after several cen- precisely the area where we seem to have the
turies of near-abandonment of the area. best evidence for LBA finds. In other words,
while not wishing to sweep a potentially seri-
ous issue under the carpet without due con-
Evaluation and Comparison sideration, we do not at present regard
alluviation to be a significant source of mask-
Can we regard the picture outlined above as ing and distortion of the overall distributional
reliable? How does it compare with other pattern of Mycenaean finds in the NVAP sur-
nearby regions? And what light might it shed vey area taken as a whole.
on the sequence and causation of Mycenae's The second point relates to arguments
development as the controlling central place advanced in a number of articles by John
for this part of the northeastern Peloponnese? Bintliff, and most fully in a recent paper by
On the first issue—data reliability—only Bintliff et al. (1999). These authors have sug-
two points can be mentioned here. One con- gested that even the most systematic and
cerns the mapping, by Anne Demitrack, of intensive surface surveys in Greece have not
soil units in the study area, and the evaluation proved sufficiently sensitive to detect the sorts
'Under the Sceptre of Agamemnon' 153
of artefacts that are all that typically remain on finds little sign of MH settlement. In fact, 'not
the surfaces of extremely small prehistoric a single sherd of Middle Helladic pottery was
sites; they rightly point out that it is often only identified in the survey, neither in the valley
from the gridded surface collection, on hands nor in the Limnes area' (Wells 1996:121), pos-
and knees, of sites of later (historic) date that sibly reflecting wholesale depopulation
traces of prehistoric occupation emerge, and resulting from catastrophic soil erosion at the
so they conceive of post-prehistoric sites as end of the Early Bronze Age. Recovery seems
fortuitous 'windows' onto a largely hidden to have been slow: the Early Mycenaean
prehistoric landscape, whose extent and den- period is better represented by past excava-
sity have thus been greatly underestimated. tions (at the Berbati tholos tomb and the pot-
This charge is, potentially, a very damaging ter's workshop on the Mastos, both LH
one, but in reality we incline to think it has II/IIIA1) than by anything among the new
been presented in a somewhat exaggerated survey finds, which do not indicate intensive
form. Obviously, search intensity does make a occupation until LH IIIA2-IIIB1. Eventually,
difference: NVAP found prehistoric compo- at the height of the Mycenaean palatial age,
nents at nearly 40% of its sites, and maybe settlement expansion onto marginal soils, the
some of these would indeed have gone alto- labor-intensive construction of roads and
gether undetected without the minute (more controversially) terraces, and the lack
scrutiny given to these locations. On the other of elite tombs and structures all plausibly
hand, routine tract-walking was quite success- suggest the annexation of Berbati-Limnes by
ful in locating additional prehistoric finds in Mycenae as an important strategic and eco-
many other places: Neolithic in 18 tracts, Early nomic asset (Schallin 1996: 170-73).
Bronze Age in 64, Late Helladic in 72 (of which It will be intriguing to see if surveys now
almost two thirds [63%] were from tracts not under way in the Phliasian Plain and the
later associated with sites). This makes us eastern Corinthia similarly produce evidence
wonder whether the appropriate metaphor is of what Dickinson (1982) some years ago
not so much sites as windows, but rather as referred to as the 'underdeveloped mainland'
blankets that block our view of prehistoric in the Middle Bronze Age, and of a late phase
remains by obscuring them with a thick man- of colonisation, coinciding roughly with the
tle of later material—and were these sites not Shaft Grave period, and linked with the new
present, tract-walking would still be reason- political power emanating from Mycenae.
ably effective in signalling the presence of pre- Table 10.2 (after Cavanagh 1995: 83, table 2,
historic material at these locations. which is itself based on Rutter 1993: 748, table
In any event, some sort of 'archaeological 1) provides a wider comparative framework,
source criticism' (Alcock 1993: 49-53) of raw by presenting counts of MH and LH sites
survey data is needed before we can embark known from a number of intensive surveys in
on building meaningful comparisons involv- southern and central Greece. It reveals obvi-
ing results from a number of different pro- ous regional variability, and no doubt also
jects, as we suggested earlier is now just suffers from some 'noise' arising from the use
beginning to become possible. Yet even com- of different survey methods; but the tripling
parison at a gross level is instructive. Thus, of site numbers in the LH period is very strik-
for example, the Berbati-Limnes survey of the ing, as is the fact that this increase seems
area immediately east of Mycenae, and even greatest in surveyed regions of the northeast-
closer to it than the Nemea region, likewise ern Peloponnese—Mycenae's ambit. Rutter
154 Urbanism in the Aegean Bronze Age
Table 10.2 Increases in numbers of sites of the Middle processes of political centralization, eco-
Helladic and Late Helladic periods, based on intensive
surveys in southern and central Greece. (After
nomic growth, architectural change, fiscal
Cavanagh 1995: table 2, with additional data from the innovation, etc., taking place at the centre
Asea Survey courtesy of J. Forsen and B. Forsen.) (Mycenae) had an impact that, unsurpris-
Survey (sq km) MH sites LH sites Increase ingly, is stamped on the history of the devel-
opment of the region as a whole, and of its
Laconia (70) 10 9 xO.9
Boeotia (21) 9 8-17 x 1-2
subsidiary rural settlements. Seen from the
Oropos (15) 2 3 x 1.5 perspective of Nemea, there are few overt
Pylos (12) 2-4 6 x 1.5-3 signs of the spectacular developments and
Methana (10) 3-4 5-8 x2 foreign relations evidenced at Mycenae, just a
Skourta Plain (32) 3-6 14-18 x3-4
Nemea (50) 2 8 x4 few kilometers away; even the largest settle-
S. Argolid (44) 5 27-37 x5-7 ment we know of, Tsoungiza, is not so much
Berbati (25) 0 19 infinite a major political or economic centre, as a
Asea (33) 1-3 4-5 x3
Total 37-44 103-129 x 2.3-3.5 hamlet without centralized organization
(Wright 1990). In studying such processes,
UMME 58-107 168-195 x 1.6-3.4 perhaps we should focus not only on the tra-
ditional framework of external contacts
between far-flung areas of the Mediterranean
(1993: 781) sees evidence of a broader picture
and Europe in the Shaft Grave era, but
of late MH settlement expansion, and in fact
equally on purely local-level relationships.
regards the resettlement of the Nemea region
Why Mycenae would have an interest in
as 'but one instance of a far more widespread
seeing the Nemea region resettled is not too
episode of colonization of the interior. Such
difficult to imagine. Recent discussions of
resettlement may have occurred not only in
Mycenaean state-formation (e.g. Halstead
other valleys of the northeastern Peloponnese
1995; Bennet and Galaty 1997: 84-87) have
but perhaps also in valley systems as far away
emphasized the role that manipulation of sur
as Attica, Phocis and the southeastern
pluses is likely to have played for the early
Argolid'. While the major expansion of settle-
Mycenaean elite, in consolidating their
ment seems to be a function of the mature
authority by establishing control over access
Mycenaean state in LH II-III, the critical
to exotic goods, in establishing reciprocal rela-
period of change is nonetheless in the early
tions with peers, and in ensuring the loyalty
Mycenaean period—a time when, for the first
of supporters. Increased agricultural produc-
time since the EH period, there is evidence for
tion may have been fundamental to ambitious
growth in site numbers, settlement hierar-
local leaders at Mycenae seeking to dominate
chies, and for social stratification on the main-
their competitors. By harnessing labour, by
land, most notably obvious in the form of the
bringing additional land under production in
Shaft Graves and related developments at
marginally occupied areas of its hinterland,
Mycenae.
by settling and controlling the Nemea Valley
and perhaps other similarly vacant areas,
Mycenae and Nemea Mycenae could have precipitated a quantum
leap in the productive capabilities of the
The relevance of such developments to the polity as a whole. Those in power would have
general theme of urbanism in the Aegean profited greatly by exacting a surplus agricul-
Bronze Age is, we hope, self-evident. The tural product from farmers of the area. These,
'Under the Sceptre of Agamemnon' 155
incidentally, are precisely the sorts of argu- conditions would probably have made agri-
ment that have been deployed to account for culture in the valley bottom problematic, and
parallel and contemporary developments in settlement in the area correspondingly less
the Berbati-Limnes area to the east of attractive. It is certainly possible that one of
Mycenae (Wells et al. 1993: 58; Schallin 1996: these early episodes of flooding correlates
170-73). with the occupational hiatus in the MH
One piece of evidence may offer a little cir- period. There exists plenty of ethnohistorical
cumstantial support to this line of reasoning. evidence attesting to flooding and depopula-
We have prepared maps, based on aerial pho- tion of the area in more recent, Ottoman
tographs, to show the extent of cultivation in times. Re-establishing agriculture in the val-
the Nemea area at various points over the ley after such a period of abandonment was
past half century (1948, 1970, 1980; these no mean task: it required the assistance of
maps can be viewed at the website: Cherry et French engineers and the use of industrial
al. n.d.). Mycenaean artefacts have been equipment to drain its floor of stagnant and
found in virtually every part of the Nemea malarial waters in the 1880s (Miller 1990: 12,
region that was then in cultivation—indeed, 14, 96). Excavations in 1997-98 within the
that was ever in cultivation in the latter half of Sanctuary of Zeus have produced important
the 20th century. Perhaps, then, the scale of new data on the history and course of the
land use here during the Mycenaean period Nemea River, including its channel during
(more than 50 sq km) was not dissimilar to the later Mycenaean period and clear evi-
that in the 20th century, when the area was dence of an artificially cut riverbed, more
farmed by a population of around 700-800, than 50 feet in width, of Early Christian date
and on much the same types of land, more (S.G. Miller, in Blackman 1999: 24-26, fig. 29).
suitable for olive and vine cultivation than As Miller comments, 'we can now say that
for grain (S.B. Sutton, in Wright et al. 1990: the Nemea Valley is a bowl that does not
594-603, table 3). This would be wholly in want to drain naturally, and that a river has
keeping, of course, with what we know of the existed only at those times when man has cut
developed Mycenaean agricultural system in one. Indeed, the name Nemea derives from
other regions, such as Pylos, where the direct the Greek verb nemein (to graze) that may
textual evidence is better. refer to an area that was too swampy for cul-
But if early Mycenaean resettlement in the tivation and suitable only for grazing'.
Nemea Valley took place in a landscape Like the well-known case of the Copaic
essentially vacant for the previous half mil- Basin in Boeotia, therefore, the valley does not
lennium, what accounts for the hiatus in drain well and can easily become clogged,
occupation? Geoarchaeological data raise one resulting in flooding, and in need of human
intriguing possibility. A series of five auger intervention. But intervention by whom, in
cores drilled across the valley revealed evi- the mid-second millennium BC? Was
dence of several prolonged periods of flood- Mycenae, the closest power-centre, capable of
ing of the valley floor during the early and organizing such an engineering program so
middle Holocene (A. Demitrack, in Wright et early in its history? The answer is at present
al. 1990: 591, fig. 3; cf. Atherden et al. 1993). purely speculative, but in light of recent dis-
These events cannot yet be dated at all pre- coveries we should not underestimate the
cisely, but sluggish drainage, stream flood- extent to which Greeks of the Bronze Age
ing, and likely year-round swampy were able to shape their environment.
156 Urbanism in the Aegean Bronze Age
Drainage of the Copaic basin was not the only 110), 'in the north-east Peloponnese Mycenae
Mycenaean engineering marvel: we can now seems to have quickly overcome all possible
add the construction of the Tiryns dam and rivals [and] was the earliest established centre
canal (Zangger 1994: 204-211, figs. 9-13), evi- of real power on the mainland'. In short, the
dence for hydraulic works in the Pheneios shadow of Agamemnon's sceptre, as it were,
basin of Arcadia (Salowey 1994) and else- seems to have fallen across these lands rather
where (Knauss 1991), or the discovery of early.1
partly artificial ports at the Palace of Nestor
(Zangger et al. 1997: 613-23; E. Zangger, in
Davis 1998: 69-74) and possibly in other Notes
places besides (Zangger et al. 1999). If Nemea
were to be added to this list, it would only 1. Catherine Morgan's excellent discussion (Morgan
1999: 347-65) of relationships between the Corinthia and
reinforce our sense of how closely the devel- Mycenae appeared in print too late for us to take it into
opment of the region was bound up with that account.
of the centre at Mycenae.
Finally, we suggest that the distribution of
monumental burials also may offer insights Bibliography
on social and political organisation (Mee and
Cavanagh 1984), and, in this region specifi- Alcock, S.E.
cally, could be read as suggesting that 1991 Urban survey and the polls of Phlius. Hesperia
60: 421-63.
Mycenae already dominated Nemean lands 1993 Graecia capta: The Landscapes of Roman Greece.
before its sovereignty over the Argive Plain Cambridge: Cambridge University Press.
itself was firmly established. Before the 14th Atherden, M, J. Hall and J.C. Wright
century BC, tholos tombs in the northeastern 1993 A pollen diagram from the northeast
Peloponnese. The Holocene 3 (4): 351-56.
Peloponnese were built at a variety of Argive Bennet, J.
locations (some quite distant from Mycenae), 1999 Pylos: the expansion of a Mycenaean palatial
but later their spatial distribution contracts center. In M.L. Galaty and W.A. Parkinson
markedly, with only those in the immediate (eds.), Rethinking Mycenaean Palaces: New
Interpretations of an Old Idea, 9-18. Los Angeles:
environs of Mycenae remaining in use. But,
Cotsen Institute of Archaeology, UCLA.
as we remarked earlier, tholoi are conspicu- Bennet, J. and M.L. Galaty
ously absent, not merely in the Nemea area, 1997 Ancient Greece: recent developments in
but in the entire Corinthia: indeed, after a Aegean archaeology and regional studies.
Journal of Archaeological Research 5: 75-120.
century of concerted exploration, not a single
Bintliff, J.L.
such tomb of any date has been found 1977 New approaches to human geography.
(Cavanagh and Mee 1998: figs. 5.2, 6.2; see Prehistoric Greece: a case-study. In F.W. Carter
also Figure 10.2). This, admittedly, is negative (ed.), A Historical Geography of the Balkans,
evidence; but it might imply that Mycenae's 59-114. London: Academic Press.
1997 Regional survey, demography and the rise of
power extended over most of the Corinthia complex societies in the ancient Aegean: core-
by the 15th century, when this form of burial periphery, neo-Malthusian, and other interpre-
first achieves prominence elsewhere in the tive models. Journal of Field Archaeology 24:1-38.
northeast Peloponnese, and that already in Bintliff, J.L., P. Howard and A.M. Snodgrass
1999 The hidden landscape of prehistoric Greece.
early Mycenaean times there were no rival
JMA 12: 139-68 (with further discussion and
elites able to challenge Mycenae. As debate, and a rejoinder, in JMA 13 [2000]
Dickinson argued some years ago (1977: 108, 100-123).
'Under the Sceptre of Agamemnon' 157
Bintliff, J.L. and A.M. Snodgrass Touchais (eds.), Argos et I'Argolide: topographie
1988 Mediterranean survey and the city. Antiquity et urbanisme, 41-56. Athens: Ecole francaise
62: 57-71. d'Athenes.
Blackman, D. Crouwel, J.
1999 Archaeology in Greece 1998-1999. AR 45: 1981 Chariots and Other Means of Land Transport in
1-124. Bronze Age Greece. Amsterdam: Allard Pierson
Blegen, C.W. Museum.
1925 Excavations at Phlius, 1924. Art and Archaeology Davis, J.L. (ed.)
20: 23-33. 1998 Sandy Pylos: An Archaeological History from Nestor
1928 Zygouries: A Prehistoric Settlement in the Valley of to Navarino. Austin: University of Texas Press.
Cleonae. Cambridge, Mass: Harvard University Davis, J.L., S.E. Alcock, J. Bennet, Y.G. Lolos and C.W.
Press. Shelmerdine
Cavanagh, W.G. 1997 The Pylos Regional Archaeological Project.
1995 Development of the Mycenaean state in Part I: overview and the archaeological survey.
Laconia: evidence from the Laconia survey. In R. Hesperia 66: 391-494.
Laffineur and W.-D. Niemeier (eds.), Politeia: Demakopoulou, K. (ed.)
Society and State in the Aegean Bronze Age. 1996 The Aidonia Treasure: Seals and Jewellery of the
(Aegaeum 12), 81-87. Liege: Universite de Liege. Aegean Late Bronze Age. Athens: Ministry of
1999 Revenons a nos moutons: surface survey and the Culture.
Peloponnese in the Late and Final Neolithic. In J. Dickinson, O.T.P.K.
Renard (ed.), Le Peloponnese: arche'ologie et histoire, 1977 The Origins of Mycenaean Civilisation. SIMA 49.
31-65. Rennes: Les Presses Universitaires. Goteborg: Paul Astroms Forlag.
Cavanagh, W.G., }. Crouwel, R. Catling and G. Shipley 1982 Parallels and contrasts in the Bronze Age of the
(eds.) Peloponnese. OJA 1: 125-38.
1996 Continuity and Change in a Greek Rural Landsape: Dietz, S.
The Laconia Survey II. London: British School at 1991 The Argolid at the Transition to the Mycenaean
Athens. Age: Studies in the Chronology and Cultural
Cavanagh, W.G. and C. Mee Development in the Shaft Grave Period.
1998 A Private Place: Death in Prehistoric Greece. SIMA Copenhagen: National Museum of Denmark.
125. Jonsered: Paul Astroms Forlag. Driessen, J.
Cherry, J.F. 1999 Regional dynamics: divergent settlement tra-
in press Archaeology beyond the site: regional survey jectories in Protopalatial and Neopalatial Crete.
and its future. In R. Leventhal and J. Unpublished paper delivered at the University
Papadopoulos (eds.), Archaeology in the of Cambridge, 1999.
Mediterranean: The Present State and Future Scope French, E.B.
of a Discipline. Los Angeles: Cotsen Institute of 1992 Archaeology in Greece 1991-92. AR 38: 3-70.
Archaeology, UCLA, 1993 Archaeology in Greece 1992-93. AR 39: 3-81.
in prep Review of Aegean Prehistory IX: the impact of 1994 Archaeology in Greece 1993-94. AR 40: 3-84.
regional survey on Aegean prehistory, ca. Hagg, R. and D. Konsola (eds.)
1970-2000. A]A. 1986 Early Helladic Architecture and Urbanization.
Cherry, J.F., J.L. Davis, A. Demitrack, E. Mantzourani, SIMA 76. Goteborg: Paul Astroms Forlag.
T.F. Strasser and L.E. Talalay Halstead, P.
1988 Archaeological survey in an artefact-rich land- 1994 The North-South divide: regional paths to
scape: a Middle Neolithic example from complexity in prehistoric Greece. In C. Mathers
Nemea, Greece. A]A 92: 159-76. and S. Stoddart (eds.), Development and Decline
Cherry, J.F., J.L. Davis and E. Mantzourani in the Mediterranean Bronze Age, 195-219.
1991 Archaeology as Long-term History: The Landscape Sheffield: J.R. Collis Publications.
of Northern Keos in the Cyclades. Los Angeles: 1995 From sharing to hoarding: the Neolithic foun-
Institute of Archaeology, UCLA, dations of Aegean Bronze Age society. In W.D.
n.d. The Nemea Valley Archaeological Project Niemeier and R. Laffineur (eds.), Politeia:
Archaeological Survey: Internet Edition. Society and State in the Aegean Bronze Age.
http: // classics.lsa.umich.edu / NVAP.html. (Aegaeum 13), 11-22. Liege: Universite de Liege.
Cosmopoulos, M.B. Hood, S. and D. Smyth
1998 Le Bronze Ancient 2 en Argolide: habitat, 1981 Archaeological Survey of the Knossos Area.
urbanisation, population. In A. Pariente and G. London: British School at Athens.
158 Urbanism in the Aegean Bronze Age
Watrous, L.V. Wright, J.C., J.F. Cherry, J.L. Davis and E. Mantzourani
1994 Review of Aegean prehistory III: Crete from 1985 To erevnitiko archaiologiko programma stin
earliest prehistory through the Protopalatial koilada tis Nemeas kata ta eti 1984-1985. AAA
period. A]A 98: 695-753. 18:86-104.
Weisshaar, H.-J. et al. 1987 Das Nemea Valley Archaeological Project:
1990 Tiryns: Forschungen und Berichte XL Mainz: 1984-1985. Kolloquium zur Agaischen
Philipp von Zabern. Vorgeschichte, Mannheim, 20-22.2.1986 (Schriften
Wells, B. with C. Runnels (ed.) des Deutschen Archaologen-Verbandes Aegaeis IX),
1996 The Berbati-Limnes Archaeological Survey 98-112. Mannheim: Deutscher Archaologen-
1988-1990. Stockholm: Paul Astroms Forlag. Verband.
Wells, B., C. Runnels and E. Zangger Wright, J.C., J.R Cherry, J.L. Davis, E. Mantzourani, S.B.
1993 In the shadow of Mycenae. Archaeology 46 (1): Sutton and R.F. Sutton Jr.
1990 The Nemea Valley Archaeological Project: a
54-58, 63.
preliminary report. Hesperia 59: 479-559.
Whitley, ]., M. Prent and S. Thorne
Zangger, E.
1999 Praisos IV: a preliminary report on the 1993
1993 Argolis II: The Geoarchaeology of the Argolid.
and 1994 survey seasons. BSA 94: 215-64.
Berlin: Mann.
Wiseman, J.R.
1994 Landscape changes around Tiryns during the
1978 The Land of the Ancient Corinthians. SIMA 50.
Bronze Age. AJA 98: 189-212.
Goteborg: Paul Astroms Forlag. Zangger, E., M.E. Timpson, S.B. Yazvenko, F. Kuhnke,
Wright, J.C. and J. Knauss
1987 Death and power at Mycenae. In R. Laffineur 1997 The Pylos Regional Archaeological Project.
(ed.), Thanatos: Les coutumes funeraires en Egee a Part II: Landscape evolution and site preserva-
I'age du bronze. Aegaeum 1: 171-84. Liege: tion. Hesperia 66: 549-641.
Universite de Liege, Histoire de 1'art et Zangger, E., M.E. Timpson, S.B. Yazvenko, and H.
archeologie de la Grece antique. Leiermann
1990 An early Mycenaean hamlet on Tsoungiza at 1999 Searching for the ports of Troy. In P. Leveau, F.
ancient Nemea. In P. Darcque and R. Treuil Trement, K. Walsh and G. Barker (eds.),
(eds.), L'habitat egeen preliistorique. BCHSupp 19: Environmental Reconstruction in Mediterranean
347-57. Athens and Paris: Ecole francaise Landscape Archaeology, 89-103. Oxford: Oxbow
d'Athenes. Books.
11
Exploring the Patterns of Power in the Bronze Age
Settlements of Northern Greece
Stelios Andreou
During the period that the politico-economic when complex forms of social organization
relations usually identified with southern such as urbanism or states were attained has
Aegean 'urbanism' emerged, long stretches of the danger of raising these forms to the level
northern Aegean lands, were occupied by of historical necessities. At the same time,
societies which appear to have 'resisted' the however, social processes, which may aim at
adoption of analogous organizational forms avoiding or upturning the development of
(Andreou, Fotiadis and Kotsakis 1996). such forms among small-scale communities,
Stratified social structures, exploitative elites, are underplayed (Bender 1990).
fast demographic growth, nucleated centers I shall examine some aspects of community
and bureaucratic controls of economic trans- life in one area of the northern Aegean, where
actions were salient features of the 'urban' small-scale societies had an enduring pres-
centers. Instead, the contemporary 'small ence. The aim is to elucidate some of the char-
scale societies' look as if they underscored acteristic structures of social life in the north,
social structures of equality, emphasized resi- which made unwelcome the adoption of
dence in small village sites or hamlets and urban life there, during the period of
lacked any specialized political mechanism Mycenaean 'urbanism' in the south. I shall
and institutions. A common feature of Aegean focus my discussion on the cultural, social
'urban' formations was their brief life and and political parameters of human settlement
their low ability to survive. It is worth point- in central Macedonia during the Late Bronze
ing out the almost total lack of continuity they Age and more specifically, during the period
display in terms of organizational structures from ca 1400 to ca 1100 BC Figure 11.1.
with developments in the same areas during My discussion concentrates on the lowland
the early centuries of the 1st millennium BC. part of the area, a landscape of coastal plains,
On the other hand, contemporary 'small scale river valleys, inland basins and hills. The
societies' exhibit remarkable stability and archaeological record comprises information
endurance for long periods of time without from intensive survey, extensive reconnais-
however being monolithic and static. In view sance, small soundings and extensive excava-
of the brief and discontinuous presence of the tions in three settlements. It is by far richer
Aegean 'urban' formations and their limited than that of any other area or period of
geographical distribution, one might think Macedonian prehistory and implies some
that they might not have been an altogether significant changes regarding human rela-
expected consequence of regional develop- tions inside and between communities.
mental trajectories. Stressing the periods (Andreou, Kotsakis and Fotiadis 1996;
Exploring the Patterns of Power in the Bronze Age Settlements of Northern Greece 161
Wardle 1997). Considerable gaps however, ments in the area increased slowly and more
still exist, primarily in terms of detailed or less steadily from the 3rd millennium
sequences and 'on site' data, which delimit onwards until the EIA. During the same
the ability to test propositions concerning period settlement sizes remained small, rarely
socio-political and economic processes and exceeding 1 ha. and only in the latter phase
the interpretation of cultural attitudes. few larger population aggregations may have
started forming. The western Langadas basin
displays a more detailed picture, allowing
The Settlement in the Landscape some insight into the process of settlement
expansion during the LBA. Figure 11.2.
Research has shown that the number of settle- (Andreou and Kotsakis 1994: 21; Grammenos,
162 Urbanism in the Aegean Bronze Age
Besios and Kotsos 1997: 13-51; 87-88.) five kilometers apart. They were ancient set-
Excavation and survey suggest that human tlements with longer or shorter life spans
occupation in the area, at the beginning of the going back to previous eras and continuing to
LBA was sparse. It was limited to no more the next. All, except one, were situated on rel-
than four mounds, standing around eight to atively flat ground on the lower terrace
Figure 11.2 Late Bronze Age sites in the western Langadas basin.
Exploring the Patterns of Power in the Bronze Age Settlements of Northern Greece 163
system that surrounds the lower plain. They Nevertheless, the more limited alluvial fans
had immediate access to major alluvial fans, and the process of continuous erosion make
very good for cultivation, given the appropri- the area less advantageous for cultivation
ate amount of rain, but also access to water compared to the region directly accessible to
retentive soils more suitable in times of arid- the old established settlements a few kilo-
ity. Their location offered security in terms of meters away (Morrison 1992: 258). On the
subsistence, providing good prospects for the other hand, new habitation sites could have
intensification of production through the taken advantage of the often rugged terrain
increase of labor and encouraging at the same for protection and defence if the need arose.
time the diversification of cultivation Possibly, this process of settlement expansion
(Andreou and Kotsakis 1994: 20-21; Morrison and the filling up of the landscape resulted
1993: 275). In fact, archaeobotanical evidence from continuous splitting of established com-
from Assiros Toumba, one of the sites of this munities when the community reached a
early group, indicates a regime of diversified threshold after which further growth was
farming, which was based both on the inten- unwelcome. Alternatively, it could occur
sive use of gardens as well as the more exten- when participation in the old community was
sive use of fields (Jones 1992). The efficiency considered disadvantageous for economic or
of the practices used is convincingly demon- ideological reasons by some of its members.
strated by the agricultural wealth of the Late In fact, during the same period a gradual
Bronze Age storerooms of the site (Jones, restriction of the occupied area has been
Wardle, Halstead and Wardle 1996; Wardle observed in individual settlements (Andreou
1987: 326-29; Wardle 1988: 460-62; Halstead and Kotsakis: 1987: 80-81).
1994: 202, 206). Before we turn to the examination of the
During the course of the period the number on-site evidence for habitation during the
of settlements doubled, and possibly tripled Late Bronze Age, some discussion is neces-
at the very end, reaching a level of density sary of some general features of the settle-
unprecedented in previous phases, but still ments during this period. As opposed to
lower than that seen in contemporary situa- earlier and later periods, intensive survey has
tions further south (Grammenos, Besios and confirmed the view that habitation during the
Kotsos 1997: 13-51; Andreou, Fotiadis and LBA in central Macedonia was restricted in
Kotsakis 1996: 578; Andreou and Kotsakis well-bounded, steep sided and highly visible
1999b: 40-41). The new sites form a fairly dis- mounds (tell, toumba) with base diameters
tinct group. They are very small mounds, rarely exceeding the 100 m. Figure 11.3.
located often on top of precipitous hills and (Andreou and Kotsakis 1999a: 40-41; Wardle
usually at a distance of five to three kilome- 1997: 96). It is well known that settlement
ters from the already established settlements. mounds are a regular feature of many areas of
They seem however, to occupy a zone strictly SE Europe and the Near East since the begin-
outside the limits of the area used by the pre- ning of the Neolithic and their formal charac-
vious group of sites, on the intersection of the teristics, formation processes, ecological and
upper terrace system and the surrounding symbolic aspects have been repeatedly dis-
hills. The area provides again opportunities cussed in recent years (Sherratt 1983;
for diversification, particularly through the Halstead 1984; Andreou and Kotsakis 1987;
proximity of mountain and forest, allowing Miller-Rosen 1986; Chapman 1990; Chapman
for hunting, fruit collection and herding. 1994; Kotsakis 1999; Halstead 1999).
164 Urbanism in the Aegean Bronze Age
Of all the regions of Macedonia, mounds tively built walls of diverse construction and
continue during the late Bronze Age primar- sizes regularly mark the contour lines of the
ily in its central part and only a few examples tell. These walls may retain and divide habi-
are known from areas immediately to the east tation at different levels of the mound or may
and west (Wardle 1997: 96). Despite the fact serve other functions such as defence and
that researchers have pointed out differences display; moreover, they highlight the limits
in form between the Neolithic and the Bronze of occupation creating a clear boundary
Age mounds of the area, the later are often between the inside and the outside. It has
considered as a uniform phenomenon. The been pointed out that difference in materials,
uniformity nevertheless, implied by the small construction method and size of these walls
fluctuations in size and the present form of possibly indicate the emergence of loose site
the Bronze Age mounds is disrupted when hierarchies during this period (Wardle 1980,
one looks at the on-site evidence obtained 261; Kotsakis and Andreou 1989; Andreou
through excavation or careful inspection. and Kotsakis 1999b). There is some evidence
This evidence suggests a variety of formation that prominence was particularly sought
processes at work and it has been suggested after by LBA communities. More often site
that these processes are connected to an array heights fluctuate between 5 and 10m, but the
of conscious activities related to the use and deposits of some Late Bronze Age mounds
the structuring of settlement and regional may rise occasionally over 15 m. above the
space during the LBA (Andreou and Kotsakis surrounding ground. This was only partly
1987: 63-64). Form and height are certainly a the result of long-term use. It was also
function of length of occupation with attained through vertical rebuilding with a
repeated rebuilding of individual houses in a minimum amount of toppling of previous
confined area, reuse of earlier walls and the walls and through the erection of massive
employment of mud - primarily mud bricks earthworks on the edge (Andreou and
in LBA Macedonian mounds - as the essen- Kotsakis 1987: 75-77). On the other hand, sit-
tial building material. Furthermore, collec- uating mounds on natural knolls and hilltops
Exploring the Patterns of Power in the Bronze Age Settlements of Northern Greece 165
secured a physically bounded setting. At the not be transgressed easily. On the other hand,
same time it was a means that rapidly natural hills and knolls could sustain claims
enhanced the visibility of the settlement, par- of permanence and social continuity, despite
ticularly in a period of progressive deforesta- the lack of physical continuity (cf. Chapman
tion (Bottema 1982). It appears consequently, 1990). The expansion of settlement to the
that during the Late Bronze Age, more than marginal ecological zones during this period
in any previous period, spatial circumscrip- was perhaps not unrelated to the additional
tion and prominence were vital issues for advantage these zones presented for the ful-
communities. They were objectives that were fillment of the ideological requirements of
attained through individual house rebuilding habitation, despite possible undesirable
and collective architectural activities, effects on continued existence. Whatever the
through the employment of technology or the reasons, this move to marginal zones was far
exploitation of topography. from smooth for the newly founded settle-
It has been repeatedly remarked in the con- ments and created, as we shall see below, dis-
text of discussions related to the earlier parities between old and newly founded
Southeast European mounds, that these types settlements on the regional level.
of settlement apart from being the places of
habitation of the living, were simultaneously
the material expression of continuity The communities
between the living and their ancestors.
Genealogical continuity was established The excavated mound of Kastanas was one of
through the persistent architectural and the newly occupied sites early in the LBA
domestic activity on the specific location of (Hansel 1989). It was located on a small
the material remains of the previous occupa- island not far from the coast of a lagoon,
tion (Chapman 1991: 155; Kotsakis 1999: 68). which in the Bronze Age extended into the
Thus, the in situ rebuilding of houses, so well present lower part of the Axios river (Schultz
attested in the LBA Macedonian mounds was 1989). The LBA foundation took advantage of
the symbolic means for the expression of the an Early Bronze Age mound that had been
social identity of their households, (cf. deserted for over two centuries. Despite the
Kotsakis 1999; Halstead 1999). On the other limited area exposed, the detailed reports
hand, the circumscription in space and the provide interesting information about the fre-
prominence of the LBA tells, emphasized by quent shifts in the configuration of habitation
the collective rebuilding of perimeter walls and the development of the social and eco-
was a strong reference to a common past, par- nomic strategies of the new community in the
ticularly important for the reproduction of unstable setting near the mouth of the river.
the community in the regional level during a The small, randomly placed, mud brick
period of settlement expansion. Communities buildings of the earlier LBA phases (17-16)
were probably stating claims to parts of the could not have housed groups larger than
space necessary for their continued existence. nuclear families, which were occasionally
The nearly neat spatial division between the sharing yards and some external facilities.
old settlements of the plain and new mounds Indeed, the evidence indicates a low level of
on the surrounding hills in the Western self-sufficiency and basic domestic activities
Langadas basin may in fact imply the exis- taking place inside and out of the houses. It is
tence of a regional genealogy, which could significant that in the following phases
166 Urbanism in the Aegean Bronze Age
(15-14), fewer and more spacious establish- the ancestors through rebuilding was the cru-
ments were successively taking over the cial factor for the social identity of the house-
space of the earlier houses. The large hold, new larger houses appropriated the
'megaron' of phase 14b particularly, which space of the old in a situation of continuous
occupied the greater part of the excavated competition and exercise of social power.
area, exhibited a significant capacity for stor- Judging from the increased amounts of deco-
age, a diversified supply of provisions and a rated drinking vessels in the houses of phases
far greater than before, scale and frequency of 14b and 12 occasional episodes of collective
food preparation and consumption. In addi- consumption of food and alcoholic drinking
tion, compared to earlier phases, the deposits were used to strengthen the cohesion of these
display a considerably more elaborate mate- larger groups. During these occasions
rial culture and a greater variety of activities. Mycenaean wheel-made pottery, imported
The excavators suggest that the big house of and locally produced, started to be displayed,
phase 14b actually provided for a larger implying the ability of local household or
crowd than earlier houses (Becker 1995). descent group heads to participate in regional
Even larger and more efficient groups follow- networks of exchange. (Jung in press). Despite
ing a more diversified farming regime (Kroll the growth of labour and the intensification of
1984) are suggested by the establishment of production, the stability of the community of
more tightly arranged, large and stable com- Kastanas was disrupted seriously several
plexes near the end of the 11th century BC times during its lifetime. The important disad-
(Hansel 1989). vantages related to the marginal setting of the
The sequence of the Late Bronze Age levels new LBA communities may not be totally
of Kastanas reveals some important aspects of irrelevant to these events. It should be added
LBA expanding habitation The original com- that despite the more compact plan of the later
munity comprised small, spatially distinct, but phase, spatial organization never reached the
marginally self-sufficient groups with very complexity displayed by the two other con-
low ability for production of surpluses and temporary sites that are examined below.
very sensitive to any fluctuations in the labor The mounds of Assiros and Thessaloniki
force or the resources. Several strategies were (Wardle 1988; Andreou and Kotsakis 1996)
used by prehistoric communities to counteract belong to the group of the old settlements.
this fairly widespread problem (Halstead They were located in zones with rich resources,
1989: 68-80). The dominant social strategy amenable to diversification and intensification,
however, in the community of Kastanas - as in if adequate labor power were available. These
the other tell communities of LBA central communities were very different from the con-
Macedonia - was to intensify production temporary Kastanas in terms of complexity in
through the increase of the labour force of the the use of settlement space.
household. This could be accomplished either Both were surrounded during successive
through the temporary or permanent control phases of the Late Bronze Age by systems of
of post-marital mobility of the junior members perimeter walls with large dimensions,
of the family or through the incorporation of which imply leadership, and the investment
weaker households by more successful ones of considerable collective labor. Inside these
(Netting 1990, 39-^0; Blanton 1994: 5-6). In the perimeter walls the settlements display large,
context of the mound settlement, where the roughly rectangular complexes with over a
establishment of residential continuity with dozen separate spaces each, amounting to
Exploring the Patterns of Power in the Bronze Age Settlements of Northern Greece 167
c. 200 m2 of floor area. The buildings were issues that certainly cannot be resolved with
tightly packed on the highest terraces of the satisfaction at this stage. In terms of internal
sites. The narrow lanes between the com- arrangement, 28% of the c. 200 m2 area of
plexes were the only free spaces in the settle- building A in Thessaloniki during phase four
ments. Assiros particularly, during phases (c. 1200 BC), was used for large scale storage,
9-6 exhibits a formalized plan with buildings as the ca 15 pithoi found in adjacent spaces
arranged in a strict order in parallel rows. show (Figure 11.5). (Andreou and Kotsakis
Furthermore, spaces inside the buildings 1996: 374). In addition however, clay vats and
seem to repeat an architectural module of a pithos or two were dispersed in most of the
2x4m or 4x4m (Wardle 1996). It should be remaining rooms of the complex, along with
pointed out that the same settlement plan a variety of other domestic activities concern-
was being rebuilt during four successive ing food preparation and cooking, eating,
building phases for over two centuries drinking and domestic industries (Karadimou
(Wardle 1996). The buildings of Assiros were 1998). It is not possible to know if this pat-
thus turned into genuine 'ancestral homes'. tern was repeated in all the buildings of the
The seven partly excavated complexes of the settlement. At the moment, there are indica-
Thessaloniki Toumba, display a less regular tions that two more complexes owned facili-
plan, but indicate a similar persistence in ties of large-scale storage. I would like to
space for a period of c. 150 years Figure 11.4. argue that the large complexes of Toumba
The function of these large complexes in were probably housing large descent groups
Assiros and Thessaloniki and the character of or complex households. The produce of their
the residential groups that occupied them are combined labour was stored in the storage
Figure 11.4 Settlement plan of the toumba of Thessaloniki. Phase 4 (12 century BC.)
168 Urbanism in the Aegean Bronze Age
Figure 11.5 The distribution of activities in building A of the toumba of Thessaloniki in phase 4 (based on Karadimou
1998).
rooms probably under the control of the valuables (Wardle in press.; Andreou and
group leader. In view of the evidence from Kotsakis 1996). But overall, the lack of pres-
Thessaloniki, it is conceivable that the rich tige objects in settlements is more impressive
storerooms found at different parts of the site than their presence. One could argue for their
of Assiros during phase nine and possibly deposition in tombs, but the latter are even
eight, may represent a comparable commu- more conspicuous for their absence, and this
nity organization. may not be a matter of accident. Finally, the
Inside these establishments pottery and stability and coherence of the large descent
other implements displayed types which fol- groups and the conformity of their members
lowed the local traditions. Few items only to the social rules that were presumably set
refer to the imitation of practices of foreign by the group's leading personalities, were
elites, to the south or to the north and indi- probably secured through collective events
cate the ability of household heads to invest during which wine was possibly consumed
part of the stored surplus into long distance and Mycenaean drinking vessels were dis-
exchange networks for the acquisition of played (Andreou forthcoming).
Exploring the Patterns of Power in the Bronze Age Settlement? of Northern Greece 169
Thessaloniki and Assiros are examples of was a period of intensive social and cultural
two old communities, which developed suc- activity on the regional, the community and
cessful social and economic strategies, based the household level. This activity was related
on the local values and taking advantage of to the rearrangement of human relations
the favorable local resources and eventually inside communities and to the restructuring
reached a level of considerable stability and of human presence in the landscape. The set-
complexity. The collectively built walls, tlement mound was one of the central factors
among other practical functions, emphasized that defined the configuration of human rela-
the success of the community and its long- tions during the period. Another one was the
term history in the area. It is not easy to social strategy of LBA households to increase
define at the moment the details of the power in size, in order to combat risks to self-suffi-
structures that were involved in their erec- ciency and survival from the disruptive
tion. They demarcated the community effects of the fluctuations in labour power
towards the outside, but to some extent they common among farming communities. With
confined it to the inside signifying the safe the intensification of labour, production
limits for the expansion of its power. The set- could be increased and satisfactory surpluses
tlement areas in the meantime became the could be produced, given the variety, the
loci of antagonism between resident groups. quality, and the quantity of the resources in
The extremely crowded pattern on the top of close proximity to the Macedonian settle-
the mounds indicates the force exercised by ments. It is possible that this new regime
corporate groups competing for the control of required also a readjustment of the tradi-
the precious ancestral space (cf. Chapman tional small-scale intensive system of farming
1990). There are no clear manifestations of rit- and the evidence for a faster rate of defor-
ual acts performed at any site, but on the estation in Central, compared to Eastern and
other hand, the formalized plan of the settle- Western, Macedonia may be an indication of
ment of Assiros and the strict regulation of a more extensive system of cultivation (cf.
space perhaps point to the ritual power of Halstead 1994: 200-202; Bottema 1882).
community heads to restrain excessive ambi- Mounds rose in the area during the Late
tions. At Thessaloniki, there are indications Neolithic as an ideological mechanism to
for the existence of size differences between emphasise the importance of the independent
buildings. Some indications also exist that household, which was developing at the time
groups were differentiated according to the in competition to other households. This was a
level on which houses were located (Andreou new form of social relations arising in opposi-
and Kotsakis 1996. Kotsakis and Andreou tion to the relations of reciprocal communality
1993). It is difficult however, to find evidence that characterized the flat extended Neolithic
at the moment, in either site, pointing to the sites present in the area. The new form of social
material manifestation of a chief or a group of relations was connected to a reorganization of
chiefs, of local or regional range. production towards a more efficient and versa-
tile system, amenable to intensification
through a new emphasis on the diversification
Conclusions of farming (Kotsakis 1999: 72-74). The scarcity
of information about the details of habitation
Increasing information points to the fact that during the intervening period between the
the Late Bronze Age in central Macedonia Late Neolithic and the beginning of the Late
170 Urbanism in the Aegean Bronze Age
Bronze Age does not permit a detailed under- their household heads to mobilize enough
standing of the circumstances under which labour. Presumably, there were many occa-
central Macedonian households started imple- sions for the development of regional
menting the social strategies that facilitated alliances which would bind strong with
their growth in size and labour power. It weaker sites; marital exchanges were neces-
should be pointed out that the new strategies sary and probably frequent events, moving
were embedded in the traditional values and people and labour up and foodstuffs down
the cultural practices related to the ideology and also entailing long-term ties, hospitality
and symbolism of the tell and the emphasis to and mutual dependence. Regular participa-
the ties with the ancestors. A new symbolic tion of needy regional neighbours in feasts
emphasis however, was directed to the com- taking place in the old sites, where some of
munity which, as a group, by seniority and the surplus was being redistributed could
perhaps other forms of power, could claim par- have been a regular practice, which strength-
ticular resources from other communities. The ened bonds, created dependencies and
size and diversification of these resources were secured services. It is conceivable, although
crucial for the successful investment of the hard to support archaeologically, that loose
growing labour power of the households. regional hierarchical networks could have
Thus, the erection of walls around the settle- developed on this basis, particularly among
ment was not intended simply as a means of neighbouring communities (Wardle 1988:
demarcating the community; it was now 462; Andreou and Kotsakis in press;
turned also into a field of expression for antag- Andreou, Fotiadis and Kotsakis 1996: 585).
onism and the display of power. Impressive Inside communities there would have been
works like the ones excavated in Thessaloniki many sources of tension among more and less
and Assiros were the result of this process. successful groups. Competition for ancestral
In the course of the Late Bronze Age, rela- space, which was crucial for the social repro-
tionships inside and between communities duction and the further growth of groups, was
were readjusted. Several situations of strong as the crowded tops of Assiros and the
inequality between members became more Toumba of Thessaloniki show. There is some
pronounced than before, but many remained evidence that prestige goods circulating in
unresolved, hidden behind traditional val- inter-regional exchange networks were being
ues and attitudes, creating several sources of employed by some household or descent
tension. group heads in the intracommunal antago-
Inequalities were expressed in a more visi- nisms and it is plausible, that something simi-
ble form on the regional level than inside lar was happening with Mycenaean type
communities, where the ritual expression of pottery, although there is no clear supporting
the bonds with the ancestors restrained their evidence yet. Finally, it is possible that during
articulation. Competition for good land was this antagonistic process some weaker and spa-
probably an important source of tension tially marginal descent groups would decide to
between communities. The comparison fission and move to the surrounding hills. The
between Kastanas and the other two sites and gradual shrinkage of mound tops (Wardle
the frequency of disrupting events in the for- 1980: 231) in the course of the LBA, may in fact
mer, probably demonstrate the advantages in indicate a slight decrease of population.
terms of stability and variety in the resource Tensions would emerge also inside the
base of the old sites and also the ability of households or descent groups. The switch
Exploring the Patterns of Power in the Bronze Age Settlements of Northern Greece 171
from multi-room buildings with common horoorganosis. In Amitos. Timitikos tomos gia ton
kathigiti Manoli Androniko, 57-86. Thessaloniki:
storerooms in phases nine and eight to dis-
Aristotelio Panepistimio Thessalonikis.
persed storage in phases seven and six at 1994 Prehistoric rural communities in perspective: the
Assiros, without any other major changes in Langadas Survey Project. In P.N. Doukellis and
the plan of the buildings, could be a sign of L.G. Mendoni (eds.), Structures rurales et socie'te's
antiques, 17-25. Centre de Recherches d' Histoire
direct challenge to the power of the descent
Ancienne., Besancon: Universite de Besancon.
group heads (Wardle 1989: 462). 1996 I proistoriki toumba tis Thessalonikis. Palia kai
None of these tensions, antagonisms and nea erotimata. AEMTh 10: 369-89.
inequalities, however, was resolved during 1999a Counting people in an artifact poor landscape.
The Langadas case, Macedonia, Greece. In ].
the Late Bronze Age, in a way that would Bintliff and K. Sbonias (eds.), Restructuring Past
allow one group of the community to gain Population Trends in Mediterranean Europe (3000
excessive power over the rest. Instead, it BC-AD 1800), 35-43. Oxford: Oxbow Books.
appears that efforts were taken, in some cases 1999b 'Mikinaiki periferia', 'Mikinaiki parousia', i
toumba tis Thessalonikis, mia thesi tis epokhis
through the formalization of settlement and tou khalkou sti Makedonia. In F. Dakoronia
individual building plans, to stress the values and M. Papakonstantinou (eds.), / perifereia tou
which were related to the independence of mykinaikou kosmou. Praktika tou 1"" Diethnous
individual households. Diepistimonikou Simposiou, 107-17. Lamia.
Andreou, S.
From the Late Neolithic until the end of the forthcoming Looking South through a Wineglass: Mycenaean
Bronze Age the mounds remained the primary, pottery in Northern Greece.
and in the Late Bronze Age the only, foci of Andreou, S., M. Fotiadis and K. Kotsakis
1996 Review of Aegean prehistory V: the Neolithic
social activity in the central Macedonian land-
and Bronze Age of Northern Greece. AJA 100:
scape. They were the places where identities 537-97.
were' created, relationships were negotiated Becker, C.
and various forms of power were employed. 1995 Zur Reconstruction von Aktivitatsmustern in
spatbronzezeitlichen Haushalten, untersucht am
During the Neolithic, the tells were the sym-
Fundmaterial aus Kastanas (Nordgriecheland).
bolic manifestations of the independent and Prahistorische Zeitschrift 70: 96-114.
competing households and they remained as Bender, B.
such through the Late Bronze Age despite 1995 The dynamics of nonhierarchical societies. In S.
Upham (ed.), The Evolution of Political Systems.
important changes in the relationships
Sociopolitics in Small-Scale Sedentary Societies,
between household members and between 247-63. Cambridge: Cambridge University
households. In the meantime, a more elaborate Press.
way of life had developed. Important changes Blanton, R.E.
1994 Houses and households. A comparative study. New
in the symbolic content of mounds and a rear- York and London: Plenum Press.
rangement of the organizational principles of Bottema, S.
mound habitation became possible during the 1982 Palynological investigations in Greece with
Early Iron Age when cemeteries emerged as special reference to pollen as an indicator of
human activity. Paleohistoria 24: 257-83.
the loci where new social identities could be Chapman, J.
created, and human relations renegotiated. 1990 Social inequality on Bulgarian tells and the
Varna problem. In R. Samson (ed.), The Social
Archaeology of Houses, 49-91. Edinburgh:
Edinburgh University Press.
Bibliography 1991 The creation of social arenas in the Neolithic
and Copper Age of S. E. Europe: the case of
Andreou. S. and K. Kotsakis Varna. In P. Garwood, D. Jennings, R. Skeats, J.
1987 Diastaseis tou khorou stin kendriki Makedonia: Toms (eds.), Sacred and Profane, 152-71. Oxford:
Apotyposi tis endokoinotikis kai diakoinotikis Oxford University Committee for Archaeology.
172 Urbanism in the Aegean Bronze Age
1989 Excavations at Assiros Toumba 1988. A prelim- graphical perspective. In Etaireia Makedonikon
inary report. BSA 84: 460-63. Spoudon, Aficroma ston N. C. L. Hammond,
1996 Change or continuity: Assiros Toumba: the 81-113. (Parartima Makedonikon 7) Thessaloniki:
transition from Bronze to Iron Age. AEMTh 10: Etaireia Makedonika of Spoudon.
443-57. in press Metalworking in Late Bronze Age Central
1997 The prehistory of Northern Greece: a geo- Macedonia. AEMTh 13.
Appendix 1
Todd Whitelaw
Isolated structures:
Chalinomouri Farmhouse 198
Cheiromandres Farmhouse 143
Chondros Rousses 91
Chondros Tourkissa 149
Karoumes 1 Farmhouse 135
Karoumes 2 Farmhouse 85
Kato Mesara Farmhouse 129
Kokkino Froudi Farmhouse 108
Koumoi Farmhouse 331
Stous Skarveli Farmhouse 260
Villas:
Achladia Villa 289
Ano Zakro Villa 322
Kannia Villa 424
Klimataria Villa 486
Makryialos Villa 590
Nirou Chani Villa 542
Flakes Villa 258
Pyrgos Villa 251
Sklavokambos Villa 396
Tourtouli Villa 527
Tylissos Villa A 575
Tylissos Villa B 365
Tylissos Villa C 451
Vathypetro Villa 872
Zominthos Villa 1115
Zou Villa 351
Small Palaces:
Gournia Palace 1514
Petras Palace 1064
Zakros Palace 2825
Appendix 2
William Cavanagh
Chryssoulaki, S. La very, J.
1990 L'urbanisme minoen. A. Le reseau routier 1990 Some Aspects of Mycenaean Topography. BICS
urbain. In P. Darcque and R. Treuil (eds.), 37: 165-71.
L'habitat egeen prehistorique, 371-80. [BCH Supp: 1995 Some 'New' Mycenaean Roads at Mycenae:
19]. Paris: CNRS. Euruaguia Mykene. BICS 40: 264-5.
Crouwel, J. H. Levin, M.H.
1981 Chariots and Other Means of Land Transport in 1981 The Roads to Gla. The Pennsylvania Gazette 80,
Bronze Age Greece. Amsterdam: Allard Pierson 2: 25-29.
Museum. McDonald, W.A.
French, L. 1964 Overland Communications in Greece during
1993 Archaeology in Greece: Mycenae Survey. AR LH III, with Special Reference to the Southwest
39:18 Peloponnese. In E.L. Bennett (ed.), Mycenaean
Jansen, A. Studies: proceedings of the Third International
1997 Bronze Age highways at Mycenae. Classical Colloquium for Mycenaean Studies held at
Views: Echos du monde classique. N.S. 16: 1-16. 'Wingspread', 4-8 September 1961, 217-240.
Kardara, Chrysoula P. Madison: University of Wisconsin Press.
1971 The Isthmian Wall (A Retaining Wall for a
McDonald, W.A. and N. Wilkie (eds.)
Road?). AAA 4: 85-89.
1992 Excavations at Nichoria in Southwest Greece
Kase, Edward W.
II: The Bronze Age Occupation, (esp. 389-98).
1973 Mycenaean roads in Phocis. A/A 77: 74-77.
Minneapolis: University of Minnesota Press.
1973 A Surface Exploration in Search of Mycenaean
Mylonas, G.E.
Roads in Nomos Fokhidhos and Nomos
1966 Mycenae and the Mycenaean Age, 86-8. Princeton,
Fthiotidhos. PhD Dissertation, Loyola Univer-
NJ: Princeton University Press.
sity of Chicago. University Microfilms, Ann
Miiller, S.
Arbor.
1991 Routes minoennes en relation avec le site de
Kase, Edward William et al. (eds.)
1991 The Great Isthmus Corridor route: explorations Malia. BCH 115: 545-560.
of the Phokis-Doris Expedition 1. (Publications Schallin, A-L.
in ancient studies, University of Minnesota. 1996 The Berbati-Limnes Archaeological Survey:
Center for Ancient Studies no. 3). Dubuque, the Late Helladic period. In B. Wells (ed.), The
Iowa: Kendall/Hunt Publishing Co. Berbati-Limnes Archaeological Survey 1988-1990,
Krigas, Eleftherios J. 123-75. Stockholm: Svenska Institutet i Athen.
1987 Amota and rota: road-transport in Mycenaean (Reference to Mycenaean roads on pp. 131-3.)
Arcadia. Kadmos 26: 74-83. Simpson, R. Hope
Kiipper, M. 1981 Mycenaean Greece 15-17, 27, 61, 81, 143. Park
1996 Mykenische Architektur: Material, Bearbeitungs- Ridge: Noyes Press.
technik, Konstruktion und Erscheinungsbild. Tomlinson, R.A.
Espelkamp, Internationale Archaologie vol. 25. 1995 Archaeology in Greece: Mycenae Survey. AR
57-8. 41: 12.
Langdon, Merle K. Tzedakis, Yannis, Stella Chrisoulaki, Sophia Voutsaki,
1994 A Cyclopean Bridge and Rutted Road in the and Yanna Venieri.
Thriasian Plain. SMEA 34: 51-60. 1989 Les Routes minoennes: Rapport preliminaire.
Recent References to Bronze Age Ronds in the Aegean 181