ArticlePDF Available

Aegean Prehistory as World Archaeology: Recent Trends in the Archaeology of Bronze Age Greece


Abstract and Figures

This article surveys archaeological work of the last decade on the Greek Bronze Age, part of the broader discipline known as Aegean prehistory. Naturally, the literature is vast, so I focus on a set of topics that may be of general interest to non-Aegeanists: chronology, regional studies, the emergence and organization of archaic states, ritual and religion, and archaeological science. Greek Bronze Age archaeology rarely appears in the comparative archaeological literature; accordingly, in this article I place this work in the context of world archaeology, arguing for a reconsideration of the potential of Aegean archaeology to provide enlightening comparative material.
Content may be subject to copyright.
Aegean Prehistory as World Archaeology: Recent
Trends in the Archaeology of Bronze Age Greece
Thomas F. Tartaron
Published online: 20 November 2007
Springer Science+Business Media, LLC 2007
Abstract This article surveys archaeological work of the last decade on the Greek
Bronze Age, part of the broader discipline known as Aegean prehistory. Naturally,
the literature is vast, so I focus on a set of topics that may be of general interest to
non-Aegeanists: chronology, regional studies, the emergence and organization of
archaic states, ritual and religion, and archaeological science. Greek Bronze Age
archaeology rarely appears in the comparative archaeological literature; accord-
ingly, in this article I place this work in the context of world archaeology, arguing
for a reconsideration of the potential of Aegean archaeology to provide enlightening
comparative material.
Keywords Archaeology Greece Bronze Age Aegean prehistory
The present review updates the article by Bennet and Galaty (1997) in this journal,
reporting work published mainly between 1996 and 2006. Whereas they charac-
terized trends in all of Greek archaeology, here I focus exclusively on the Bronze
Age, roughly 3100–1000 B.C. (Table 1). The geographical scope of this review is
more or less the boundaries of the modern state of Greece, rather arbitrarily of
course since such boundaries did not exist in the Bronze Age, nor was there a
uniform culture across this expanse of space and time. Nevertheless, distinct
archaeological cultures flourished on the Greek mainland, on Crete, and on the
Aegean Islands (Figs. 1,2), interacting with one another and the wider Mediter-
ranean world. Although these broader contacts form an essential part of the Bronze
T. F. Tartaron (&)
Department of Classical Studies, University of Pennsylvania, 201–202 Logan Hall, Philadelphia,
Pennsylvania 19104-6304, USA
J Archaeol Res (2008) 16:83–161
DOI 10.1007/s10814-007-9018-7
Age story, I mention only a few examples of the important archaeological work
done in Anatolia, Cyprus, the Levant, northern Africa, the Balkans, and the central
Mediterranean, which bears upon interaction with the Aegean area. The Greek
Bronze Age forms part of the longer timeframe of the discipline of Aegean
prehistory, and archaeologists of the Bronze Age typically consider themselves
Aegean prehistorians. The term ‘‘Aegean’’ is used widely in place of ‘‘Greek’’ to
acknowledge that many who lived in the region in the Bronze Age had no linguistic
or ethnic connection to recognizably Greek populations.
Research on the Aegean Bronze Age is a vast enterprise, and considerations of
space preclude a comprehensive reporting of all deserving fieldwork and
interpretive study. Inevitably, this review is selective and idiosyncratic, reflecting
my own experience as well as issues that seem to me to have particular relevance for
archaeologists working in other world areas. I have chosen to examine a limited
number of topics of general interest in the hope that non-Aegeanists will recognize
in them similar problems to which we may all share approaches, if not common
solutions. These topics are chronology, regional studies, the emergence and
organization of archaic states, ritual and religion, and archaeological science. A
central claim of this survey is that Aegean prehistory is an underutilized resource for
Table 1 Chronological table of the Aegean Bronze Age, using a modified low chronology
Crete (Minoan) Mainland (Helladic)
Pottery Phase Calendar dates Pottery Phase Calendar dates
Prepalatial Early Minoan
(EM) I
3100–2700 Early Helladic
(EH) I
EM II 2700–2200 EH II 2700–2200
EM III 2200–2100 EH III 2200–2000
Middle Minoan
2100–1900 Middle Helladic
(MH) I
Protopalatial MM IB 1900–1800
MM II 1800–1700 MH II 1850–1700
Neopalatial MM III 1700–1600 MH III 1700–1600 Shaft Grave Era
Late Minoan
1600–1480 Late Helladic
(LH) I
LM IB 1480–1425
Final Palatial LM II 1425–1390 LH IIA 1500–1440 Mycenaean
LH IIB 1440–1390
LM IIIA1 1390–1370 LH IIIA1 1390–1370
Postpalatial LM IIIA2 1370–1300 LH IIIA2 1370–1300
LM IIIB 1300–1190 LH IIIB 1300–1190
LM IIIC 1190–1070 LH IIIC 1190–1070
Subminoan 1070–1000 Submycenaean 1070–1015
The relative merits of low and high chronologies are discussed in the text. All dates B.C.
84 J Archaeol Res (2008) 16:83–161
comparative studies in world archaeology. A growing trend toward convergence in
theoretical and methodological discourse around the world makes the study of
prehistoric Greece more comprehensible and relevant to similar discussions
occurring elsewhere. The Aegean Bronze Age witnessed the emergence, expansion,
and collapse of two major complex societies (Minoan and Mycenaean), and the rich
data sets produced in more than a century of archaeological investigation are crying
out for use by other prehistorians. There are real similarities and differences
between Aegean and other societies in scale, organization, and historical trajecto-
ries, and in closing this review I offer a brief example of how a recent opening of
communication has been illuminating to experts on Maya and Mycenaean
archaeology, with great promise for fruitful interaction in the future. Aegean
prehistorians are consumers of world archaeological literature, but they are also
innovators in many areas such as regional archaeology, archaeometry, and the
integration of textual and archaeological data. My hope is that the admittedly
selective treatment below suggests places to begin exploring the rich Aegean
In view of the expected audience, there is a bias toward English-language sources
that in no way should be read as a disparagement of flourishing non-Anglophone
Fig. 1 Map of the Aegean and surrounding areas showing regions and other geographical references
mentioned in the text
J Archaeol Res (2008) 16:83–161 85
archaeological research in the Aegean. Archaeologists in the Greek Archaeological
Service spend their careers mainly in rescue mode, rarely experiencing the luxury of
a proper research project. Without their dedication, the battle against rampant
destruction of Greece’s archaeological heritage would be lost.
Chronology and Thera revisited
There are many new developments in Bronze Age chronology, but discrepancies
persist between traditional and more recent, mainly science-based, chronological
frameworks. We may expect the effort to resolve these differences to be a prominent
feature of Aegean prehistory for years to come. As Bennet and Galaty (1997, pp.
82–84) explain, an absolute chronology for the Aegean Bronze Age relied
traditionally on pegging extensively developed pottery sequences to the well-
established Egyptian historical chronology through the frequent appearance of
Aegean objects in archaeological contexts in Egypt and vice versa. More recently,
Knossos Malia
Ay. Konstantinos
Aidonia Nemea
Lavrion mines
Fig. 2 Map of the Aegean area showing important sites mentioned in the text
86 J Archaeol Res (2008) 16:83–161
both text-based and scientific dating methods have been refined. Egyptian
astronomical dates, based on textual references to the heliacal rising of Sirius
(Sothic dates) and the phases of the moon (lunar dates), have been refined by new
calculations. These dates generally buttress the traditional Egyptian chronology, but
they contain similar margins of error because of the ambiguity of such observations
and the complex movements of celestial bodies over long periods of time (Krauss
2003; O’Mara 2003; Wells 2002; Wiener 2003b).
A general radiocarbon framework for Aegean Bronze Age chronology is now
firmly established (Manning 1995,1999), but is by no means universally accepted
as more accurate or precise than the traditional chronology in key phases,
particularly the latter part of the Bronze Age in the second millennium. Critics point
to persistent anomalies that are largely inherent to the method. Among these are
interlaboratory measurement variability [where there is steady improvement
(Boaretto et al. 2003)], regional and diachronic variation in the absorption of
radiocarbon (Olsson 2003), and perhaps most troubling, serious issues of bias and
convergence introduced in the calibration of raw radiocarbon dates (Wiener 2003a,
pp. 380–387). The most promising developments in scientific dating have come
from dendrochronology (for basic principles, see Nash 2002). The Aegean does not
enjoy the advantage of excellent preservation of long-lived tree species that
characterizes northern Europe and the southwestern United States, nor has it been
common practice to collect wood or wood charcoal for dendrochronological
analysis (Kuniholm 2001; Newton and Kuniholm 1999). Yet the Aegean
Dendrochronology Project, now directed by S. Manning, who has taken the place
of the retired P. Kuniholm (, accessed 3 March
2007), has made tremendous strides in constructing a robust tree-ring sequence for
the eastern Mediterranean using Anatolian oak, pine, juniper, and other species
(Kuniholm et al. 2005; Manning et al. 2001,2003). Currently, the sequence is
continuous from c. 2657 to 649 B.C., but it is a ‘‘floating chronology’’ because it is
not anchored precisely to any marker event, and it does not yet overlap with the
sequence extending to the present. Nevertheless, based on a series of high-precision
radiocarbon wiggle-match analyses, these collaborators claim that the dendro-
chronological sequence is ‘‘near-absolute,’’ with 3rerror ranges of +16/-7 or better
within the Bronze and Iron Ages (Manning et al. 2003). The full impact of these
findings will not be realized for some time, but a preview of the implications may be
seen in a new series of tree-ring dates published for four charred building timbers at
the site of Assiros in Macedonia, northern Greece (Newton et al. 2005). There, the
association of timbers with a Protogeometric pottery vessel—marking the beginning
of the post-Bronze Age period—in a secure deposit has yielded dendrochronolog-
ical and radiocarbon dates that would move the traditional date for the end of the
Bronze Age earlier by as many as 75 years. If verified, this result would squeeze the
last phases of the Bronze Age (Late Helladic [LH] IIIC and ‘‘Submycenaean’’)
while lengthening that immediately following the Bronze Age (Protogeometric).
Newton and colleagues (2005, p. 188) even suggest that the changes might be
accommodated by moving back the beginning of LH IIIC, now dated no earlier than
1200 by coordinating Egyptian records with widespread destructions at Mycenaean
palaces, into the 13th century B.C. The findings from Assiros are sure to be
J Archaeol Res (2008) 16:83–161 87
controversial, and indeed a first blistering critique (Keenan n.d., pp. 12–14) ques-
tions both the quantitative methodology and the comparability of Greek and
Anatolian trees.
Evidence that a rapprochement is not imminent can be seen in the still-ongoing
debate surrounding the date of the cataclysmic eruption of the volcanic island of
Thera (Santorini) sometime in the mid-second millennium (see Bennet and Galaty
1997, p. 83). The dating of this event is crucial because it occurred at the transition
between pottery styles labeled Late Minoan IA and IB, squarely in the Cretan
Neopalatial period, and it holds important consequences for cross-dating material
throughout the eastern Mediterranean. The traditional date of this ceramic
transition—and by extension the eruption—at around 1550 B.C. or a bit later had
been challenged by interpretation of ice cores, frost-ring dendrochronology, and
determinations, suggesting a date in the second half of the 17th century B.C. for
material thought to originate in the eruption. The consensus that Bennet and Galaty
saw as inevitable in favor of the earlier date has failed to materialize, however, as
grave concerns persist about the scientific data [see especially Bietak’s (2004)
critical review of Manning (1999)]. It is now acknowledged that the tephra from
Greenland, originally dated to 1628, cannot derive from the Thera eruption. A new
candidate for Theran tephra in the GRIP and North GRIP cores, redated with
reference to the Dye 3 core from southern Greenland to 1645 ±7 B.C. Hammer
2000; Hammer et al. 2003), has met with similar skepticism regarding its origin,
possibly Alaskan rather than Theran (Keenan 2003; Wiener 2003a, pp. 373–376,
2003b, 2006). Radiocarbon dates are of limited help because they suffer from
unfortunate oscillations and plateaus in the calibration curve at just this period.
Some success for wiggle matching of Anatolian tree-ring dates to
C determina-
tions in that range has been claimed (Manning et al. 2003), but larger sample sets of
internally well-dated tree rings are needed. The most recent radiocarbon reassess-
ment, using a new calibration curve and three laboratories, of 127 samples from
Akrotiri on Thera and other Aegean sites conventionally dated between 1700 and
1400 B.C. yielded a date for the eruption between 1660 and 1613 B.C. at a 95%
confidence level (Manning et al. 2006). Another group independently dated an olive
branch, said to have been found buried alive in the pumice of the Thera eruption, to
1627–1600, again at a 95% confidence level (Friedrich et al. 2006). Yet those
favoring the later chronology remain entirely unconvinced, insisting that scientific
dates that diverge from the Egyptian historical chronology by 100 years or more are
impossible (Bietak, quoted in Balter 2006; Wiener 2006). If accepted, the early
dates would cause a lengthening of the period marked by Late Minoan (LM)
IB-style pottery to between 125 and 175 years, considered by many to be highly
unlikely (T. Brogan, personal communication 2006). It should be noted, however,
that there has been no systematic attempt to subject Egyptian sites and materials to
radiocarbon evaluation, and this has allowed Egyptologists to avoid confronting any
discrepancies that may result. Thus, as of this writing, there is substantial support
for both low and high (or ‘‘short’’ and ‘‘long’’) chronologies, including a modified
long chronology that tracks recent developments in scientific dating (Manning et al.
2002; Rehak and Younger 2001, pp. 389–392, 467). This debate reveals a peculiar
methodological rift that can be attributed to a long reliance on historical calendars
88 J Archaeol Res (2008) 16:83–161
for absolute dating, one likely to puzzle archaeologists in other world areas, who
have always relied on scientific methods for chronometric dates in prehistory. But it
is healthy that scientific techniques have forced a thorough reevaluation of the bases
of our chronological schemes, and it seems inevitable that advances from that
direction, most likely when radiocarbon dates are ‘‘locked in’’ by dendrochronol-
ogy, will one day permit the synchronization that is at present so elusive.
Regional studies and ‘‘Mediterranean myopia’
The state of Aegean regional studies in the mid-1990s was well characterized by
Bennet and Galaty (1997, pp. 96–99; see also Galaty 2005), but there are new
developments and once again, a consensus that they perceived—this time on
methodology—is under siege. In recent years, Greek regional studies have come in
for vigorous criticism from observers within Mediterranean archaeology and
elsewhere (Blanton 2001; Osborne 2004a, 2004b; Terrenato 2004). The failings that
Blanton describes as ‘‘Mediterranean myopia’’ include a localism that hinders data
comparability from region to region, resulting in a lack of integration of results into
regional syntheses; an obsession with methods over historical reconstruction; a
focus on intensive, siteless surveys rather than large-area, site-based ‘‘full-
coverage’’ surveys, at tremendous cost for too little area covered; and an alleged
environmentally deterministic landscape archaeology. Stanish (2003) claims that
Mediterranean and Americanist regional studies constitute very different research
traditions. He notes that settlement archaeology in the Americas routinely
incorporates surface survey and excavation, as features of a methodology that
arose from explicit ties to the theoretical framework of cultural ecology and an
emerging processual paradigm. He views the differences in Mediterranean practice
as resulting from a lack of those same theoretical ties. Osborne (2004a, p. 89), a
historian, concludes that survey has had little impact on Greek archaeology.
How much of this criticism is valid and why should regional studies in Greece be
apparently so different from those in the Americas and elsewhere? To understand
the current situation, a brief historical digression is necessary. The Minnesota
Messenia Expedition (MME), the grandfather of all Aegean regional studies,
explored a large region (approximately 3,800 km
) in southern Greece in the 1950s
and 1960s (Bennet and Galaty 1997, pp. 96–97; McDonald and Rapp 1972). In a
revealing introductory chapter, McDonald makes clear the project’s debt to
Steward’s cultural ecology and cites specifically the influence of contemporary
regional studies directed by Braidwood, Adams, MacNeish, and Sanders (McDonald
1972). The research design was remarkably interdisciplinary for its time, very much
in line with the scope of settlement pattern studies as they unfolded in the New
World. In the same year as the MME publication, Renfrew’s unabashedly
processual The Emergence of Civilisation gave impetus to the next generation of
regional studies in Greece (Barrett and Halstead 2004; Bennet and Galaty 1997,p.
77; Cherry 2004, pp. 1–4; Renfrew 1972), which have until very recently remained
true to those processual theoretical roots (Davis 1994). The early projects often
J Archaeol Res (2008) 16:83–161 89
integrated survey with excavation of one or more key sites (Jameson et al. 1994;
McDonald and Wilkie 1992; Wright et al. 1990).
Already in the 1970s, however, Aegean regional studies had begun to take a
different and idiosyncratic path. Perhaps the two most significant developments
were the widespread adoption of intensive, siteless survey methods and increasing
legal restrictions on regional-scale fieldwork. The forceful advocacy of hyperin-
tensive survey methods in a constant stream of publications by a handful of leading
archaeologists, notably Cherry, Davis, Bintliff, and Snodgrass, undeniably shaped
survey practice. These archaeologists were influenced by the New Archaeology’s
interest in sampling and statistical methods, and by early Anglo-American
experiments with siteless methods (Cherry 1983; Cherry et al. 1991; cf. Foley
1981; Plog et al. 1978; Shennan 1985; Thomas 1975). Quite apart from this,
however, legal and administrative realities have played a major role. The laws
governing foreign archaeological work in Greece have become increasingly
restrictive (a summary of the recent law 3028/2002 may be viewed at, accessed 5 March 2007) (see
Cherry et al. 1991, p. xv; Davis 2004, pp. 22–23; Tartaron et al. 2006a, pp.
464–466). Currently, each foreign archaeological school or mission is allotted only
six permits for archaeological work annually. With the resulting logjam of appli-
cations, no project can expect a permit longer than around three seasons. Further,
survey and excavation cannot be combined in a single project, except in the highly
unlikely event that a project could consume two permits. Field seasons are limited to
six weeks per year, and a project’s survey area must not exceed 30,000 stremmata
(just 300 hectares); thus, survey on the scale of the Basin of Mexico or Valley of
Oaxaca is for the time being unthinkable. The consequences of these policies for
regional archaeology are sobering: holistic archaeological approaches to regions
that are practiced around the world, integrating surface survey with plowzone
experiments, excavations of key sites, long-term replication experiments, and other
such studies, are rarely possible. The reasons for this restrictive posture are com-
plex, but they include a sense that foreigners often come to Greece with colonial
attitudes (see Atwood 2005 for a New World parallel), and exasperation over the
severe burden that site management and storage of massive amounts of cultural
material have placed on the Greek Archaeological Service (for different perspec-
tives, see Cherry 2003; Doumas 2001; Kardulias 1994).
These conditions illuminate many aspects of Blanton’s critique. The map of
survey coverage in Greece contains mainly smaller-scale, regionally noncontiguous
surveys, although there are exceptions in a few vast surveys and in recent efforts to
join survey territories. Archaeologists in the Aegean are indeed obsessed with high-
resolution methods of siteless survey; with few other options, they have focused
increasingly on refining the data extracted from the surface alone. Fine-scale
geomorphological analysis in tandem with intensive survey has become a defining
characteristic of this approach (Francovich et al. 2000; Given et al. 2002; Tartaron
et al. 2006a, pp. 466–470). These innovations are sophisticated and should be of
interest beyond the Mediterranean. Yet as Blanton observes, there has been a real
sacrifice in terms of regional coverage, problem orientation, and comparative
perspective. Stanish (2003, p. 167) makes the key point that in the Americas the
90 J Archaeol Res (2008) 16:83–161
methodology of settlement pattern survey is explicitly directed toward comparative
approaches to regional problems of anthropological interest. In recent times, this
broader outlook has been lacking in Greece, resulting in a patchwork of small
surveys using diverse methods, hindering data comparability and ultimately the
illumination of supraregional phenomena. Happily, there is a concerted effort
underway to utilize survey data in comparative studies (Bintliff 1997; Cavanagh
1995; Cherry and Davis 2001; Cherry and Parkinson 2003; Cunningham 2001;
Cunningham and Driessen 2004; Driessen 2001; Halstead 1994; Mee 1999; Moody
2004; Wright 2004a), facilitated by the appearance of several preliminary and
‘final’’ reports on surveys carried out from the 1970s to the early 1990s (Cavanagh
et al. 1996,2003; Cosmopoulos 2001; Davis 1998; Davis et al. 1997; Haggis 1996;
Jameson et al. 1994; Mee and Forbes 1997; Runnels et al. 1995; Watrous et al.
2005; Wells and Runnels 1996; Wiseman and Zachos 2003). It should be obvious
from the foregoing discussion that differences in the Aegean and Americanist
traditions resolve in part to different sets of options available to the archaeologist,
rather than a sharp philosophical or theoretical divide. If restrictions on the scope of
fieldwork in Greece were relaxed, one immediate effect would be fundamental
changes in research designs that would bring regional projects more in line with
practice in the Americas. Yet much scope for fruitful dialogue would remain. For
example, Blanton’s advocacy of the so-called full-coverage survey practiced in
Mesoamerica (Blanton 2001; Fish 1999; Fish and Kowalewski 1990) will seem no
panacea to many Aegean archaeologists. The term itself is misleading in that it
entails neither inspection of nor even interest in the totality of the surface of a given
region, and this is sampling by another name. Further, it appears to some as a broad-
brush method in which ‘‘sites’’ are unproblematic, readily identified constructs and
all that is not recognized as a site is ignored or decontextualized as a grab sample
(Cherry 2004, pp. 11–12). An alternative point of view, widely held in the Aegean,
emphasizes the wealth of information that resides in nonsite patterns of presence/
absence and density, in clear contradiction to Blanton’s (2001, p. 629) character-
ization of the ‘‘comparatively trivial matter of off-site artifact densities.’’ Whatever
one’s position, we must acknowledge that these methods produce genuinely
alternative readings of the archaeological landscape. While it is certainly the case
that the expansion of survey in full-coverage mode yields new insights and may
correct prior interpretations (Feinman and Nicholas 1999, pp. 172–173; Finsten and
Kowalewski 1999), it is equally true that high-resolution, siteless survey greatly
modifies interpretations drawn from coarser investigative methods. Consider one
brief example from the Eastern Korinthia Archaeological Survey (EKAS) in
southern Greece. Kromna, a historical-period settlement on the coastal plain east of
the ancient city of Corinth, was investigated in part by high-resolution siteless
survey (Caraher et al. 2006; Tartaron et al. 2006a, pp. 494–513). These methods
revealed consistent, low-density background scatters of material for four periods
that likely would have been missed entirely by extensive or site-based methods.
These results, enhanced by geographic information systems (GIS) analysis, changed
radically our thinking about settlement patterns in the region by peopling the
landscape in periods that were thought to be depopulated, and by documenting a
shifting spatial focus of activity over time at this large settlement. It seems obvious
J Archaeol Res (2008) 16:83–161 91
that both full-coverage and siteless surveys, with their different levels of resolution
and distinct units of discovery and analysis, play important roles in revealing the
fullness of human presence in any region, not as either/or propositions, and offer a
fruitful point of departure for mutually beneficial discussion across these traditions.
A final point in Blanton’s critique is that an environmentally deterministic
perspective, discarded decades ago in the Americas, has been adopted in Mediter-
ranean landscape archaeology (Blanton 2001, p. 629). In Greece, landscape
archaeology did emerge initially from a processual tradition that emphasized the
interaction of human societies (as aggregate entities) and the natural environment to
ensure subsistence and adaptive success (e.g., Kardulias 1994, p. 10: ‘‘the interactive
biological and cultural aspects of human existence within an environmental context’’).
But following their colleagues in the Americas and western Europe, survey
archaeologists working in Greece developed varied responses to neo-evolutionism
and ecological/demographic/technological determinism that in no way foregrounded
environment as the decisive factor in social organization or cultural change, such as
world systems theory (Kardulias 1999a, 1999b; Kardulias and Yerkes 2004), peer
polity interaction (Renfrew and Cherry 1986), and annales history and other forms of
structure/contingency modeling (Bintliff 1991,1999; Knapp 1992; Sutton 2000). One
benefit of so many small surveys has been an appreciation for the surprising variability
of human culture across time and space, which in turn has been instrumental in
stimulating resistance to deterministic explanations in Greece as elsewhere (Trigger
1989, pp. 329–347). In operationalizing the concept of landscape in regional studies,
Aegean archaeology has adopted not so much the kind of environmental fetishism that
Blanton seems to imply, as a focus on paleoenvironment as a means to contextualize
ancient societies and an emphasis on geomorphology as a way to better understand the
formation of the surface archaeological record.
By the late 1980s, the influence of (mainly British) postmodernists and
postprocessual archaeologists was felt (e.g., Bender 1993; Bradley 1994,2000;
Cosgrove 1985; Hodder 1987; Tilley 1994). Their writings emphasized the individual
and experiential aspects of living in the world, and their notion that there is no single,
objectively observable landscape is now commonplace. Recent literature across the
spectrum of Aegean prehistory shows the gradual incorporation of these ideas into
archaeological analysis and interpretation, but more rarely into research designs for
regional studies. A challenge going forward will be to learn to incorporate these ideas
in the research design phase and to collect data in such a way as to facilitate addressing
significant anthropological questions (Cherry 2003, pp. 158–159). Some initial steps
toward that goal have been taken by the Sydney Cyprus Survey Project (Given and
Knapp 2003). Given (2004) explores the possibility of recreating the ‘‘ideational’’
landscapes (Knapp and Ashmore 1999) of a rural Cypriot population by combining
information from targeted surface survey, geomorphology, prior excavations,
historical documents and oral histories, ethnoarchaeology, and GIS analyses,
including viewsheds. In this case, Given attempts a phenomenological approach to
changing perceptions of the world between the Roman period, characterized by an
extensively cultivated and industrial landscape, and later Medieval and Ottoman
times, when settlement was highly nucleated in small villages and sharply bounded by
surrounding zones of cultivation. This approach is successful in part because written
92 J Archaeol Res (2008) 16:83–161
documents and oral histories can verify certain emic perceptions, but what is most
striking is that unlike previous landscape studies in the postmodern vein, Given
explicitly utilizes off-site artifact distribution data to make his case.
Some Mediterranean archaeologists have called for fundamental changes in
regional archaeology, including the abandonment, or at least rethinking, of
hyperintensive survey as the main method of surface archaeology (Fentress 2000;
Terrenato 2004). In both methodology and research orientation, Aegean regional
studies stand at a crucial crossroads, with several difficult issues to confront. (1)
Will the recent spate of publications that attempt to utilize siteless data in regional
and supraregional syntheses deliver the goods, showing the viability and compa-
rability of these data across many diverse surveys? Attempts to create and maintain
a formal framework for data sharing and standardization have yielded only modest
results (Dunn n.d.), but the Collaboratory for GIS and Mediterranean Archaeology
(CGMA:, accessed 3 March 2007) is set to release a pan-
Mediterranean GIS with metadata and some primary data on survey projects in this
vast region (Galaty 2005, p. 318). (2) The microlevel surface record is messy, and
the sources of distortion and bias are well documented (e.g., Given 2004; Terrenato
2004), but do we simply turn away from this messiness because it is easier to do so,
disregarding the complexity of the archaeological landscape at this scale? (3)
Should we reorient regional research away from traditional processual concerns
(Fish 1999, pp. 205–207) toward a more relativistic and humanized approach fueled
by postmodern conceptions of landscape (Anscheutz et al. 2001)? Terrenato (2004,
p. 47) recommends that we generate internally consistent, empirically based nar-
ratives that would lead to relative rather than absolute comparisons of data that do
not rely on complete or near-complete recovery. Along the way, some narratives
would find wider acceptance than others. (4) Finally, is settlement pattern archae-
ology sustainable in the long term, in the face of rampant destruction of the
plowzone by modern development and the generally negative attitude of the Greek
authorities toward spatially extensive fieldwork? Should we stay the (intensive)
course, or focus instead on full-coverage surveys, maximizing time and coverage to
inventory the largest possible amount of territory while there is still time?
Rethinking archaic states in the Aegean
One of the most important developments of the last decade has been an explicit
effort to deconstruct the dominant paradigms of state formation in the Aegean,
particularly the concept of the Minoan and Mycenaean ‘‘palace’’ as a monolithic
and pervasive locus of royal power. With no small influence from postmodern
points of view, there has been a radical break with the normative and monumental
conceptualization of the palace as elite residence and centralized administrative and
redistributive center (Haggis, personal communication 2004). Some have called for
the abandonment of such terms as ‘‘palace’’ and ‘‘king’’ in the Minoan world,
finding little archaeological or documentary evidence to sustain them (e.g., Driessen
2002, Schoep 2002a, b). This critique is more muted for the Mycenaean world,
J Archaeol Res (2008) 16:83–161 93
where the Linear B texts detail a complex, pyramidal social hierarchy with a king
(wanax) at the apex, but even there much new work has called into question the
pervasiveness of a palace’s control—political, economic, and social—over every-
day life in outlying communities (e.g., Galaty and Parkinson 1999a; for an excellent
overview of Linear B administration, see Palaima 2003).
In the middle of the third millennium B.C., precocious developments toward
complexity occurred in the Cycladic Islands, on Crete, and in parts of the Greek
mainland (Broodbank 2000; Renfrew 1972). A recent landmark publication on this
phenomenon is Broodbank’s An Island Archaeology of the Early Cyclades (2000),
in which the author presents a new synthesis of 100 years of archaeological data and
a spatial (network) analysis to describe and explain the patterns of interconnection
within the Cycladic island group and with the wider Aegean world. Drawing
comparative material from island archaeologies in the Pacific and elsewhere,
Broodbank seeks to show the common experiences of island life but also concludes
that the Cyclades followed a unique historical trajectory. The work should have
appeal beyond the Aegean because of the way Broodbank blends a meticulous
empiricism with a postmodern landscape perspective.
In the Cyclades and on the mainland, however, this complexity is regarded as a
‘false start,’’ unraveling with destructions and abandonments in the latter centuries
of the Early Bronze Age before achieving what we would call state-level society.
Thus, there can be no question of continuity to the later Mycenaean palaces, the
origin of which cannot be found prior to the end of the Middle Bronze Age in the so-
called Shaft Grave Era. On Crete, however, no irrefutable discontinuity exists, and a
current matter of discussion is whether a continuous evolution of complexity
culminating in the Middle Minoan (MM) palaces began already in Early Minoan
(EM) II. The debate revolves around several elements of material culture into which
some read social complexity and others do not, as well as general attitudes toward
evolutionary explanations. For example, Schoep (1999,2001,2006) has argued for
emerging complexity in EM II in the form of monumental structures and seals and
sealings that may attest to organized administrative activity. Others disagree,
pointing out that the very small number of seals and sealings that can be positively
attributed to this early period are found in burials and in other patently
nonadministrative contexts, and, furthermore, other apparent signs of social
differentiation are absent in burials, dwelling sizes, etc. Weingarten (1986,1990)
argues that the system of administrative sealings was imported along with the
concept of the palace from Anatolia in the Middle Minoan period. A possible
exception, still debated, appears at Mochlos, where contrasts in elaboration of EM
burials and the presence of seals in domestic contexts of EM II may connote
meaningful differences in social status (compare Soles 1992, pp. 255–258 and
Watrous 2001, pp. 173–175, 191–192).
The trend to read the record as one of discontinuity is partly a reaction to
evolutionary arguments that are not well supported by empirical evidence (Watrous
2001, pp. 174–179). There has been a problem of up to three ‘‘lost centuries’’ in
EM III, which from the evidence of regional surveys and excavations was
characterized by site abandonments and a largely deserted countryside (Watrous
2001, pp. 179–182; Watrous et al. 2005), but this gap appears to be closing with
94 J Archaeol Res (2008) 16:83–161
good EM III–MM IA deposits at several sites, including Mochlos and Gournia
(Brogan, personal communication 2006). More recently, Schoep and Knappett
(2004) have offered something of a compromise position, arguing that a gradual,
‘slow-boiling’ heterarchical social competition developed many traits of complex-
ity before a hierarchy ‘‘exploded’’ onto the scene in the first palaces of MM II.
The palaces of Minoan Crete
Turning to Minoan Crete, the ‘‘palatial model,’’ as constructed by its critics,
presents a monolithic and relatively static picture of a limited number of small,
essentially similar palace ‘‘states’’ across Crete. These states are characterized by a
centralized and hierarchical political authority; economic and political power is
centralized in the palace and embodied in an individual ruler (‘‘king’’). The palace
serves as a residence for the ruler and his relatives and retainers, and as a center for
the administration of an economy concerned with mobilization and storage of
subsistence goods and the production of luxury items. The polity maintains political
and economic control over a regional territory, which it exploits and to which it
transmits elements of style and ideology. The palaces are recognizable by their
exceptional size within their settlements; by their distinctive architectural form, a
square or rectangular central court surrounded by wings containing complexes of
rooms, halls, and staircases leading to upper stories; and by architectural elaboration
including ashlar masonry, use of gypsum to decorate facades, frescoes, light wells,
pier-and-door partitions, and specialized ritual spaces such as ‘‘lustral basins’’ and
‘pillar crypts.’’ This court-centered complex served as a vehicle for the presentation
of elite status, ideology, and solidarity.
The fundamental problem with this picture is that its main features were
formulated at the beginning of the 20th century by Arthur Evans, the first non-Greek
excavator of Knossos, partly from the archaeological remains of the final (LM II–
III) phase of the palace and partly in his imagination (Hitchcock and Koudounaris
2002; Klynne 1998; Papadopoulos 2005). Several scholars have noted that Evans
constructed a Minoan world reminiscent of the late British Empire and its Victorian/
Edwardian sensibilities: the Minoans were elegant, nature-loving, and hierarchical,
with kings and queens reigning over a peaceful empire made safe through naval
might, i.e., a thalassocracy (Hamilakis 2002a; MacGillivray 2000; Schoep 2002b,
pp. 102–103). Evans was deeply influenced also by prevailing evolutionary views of
culture as organic, unilinear, and directional, progressing from a simple past to a
complex (European) present (Hamilakis 2002a, pp. 3–4). Subsequently, these
notions were perpetuated by comparisons with the palace- and temple-centered
administration of Near Eastern Bronze Age cities and with the patently hierarchical
organization of Mycenaean palace centers revealed in the Linear B archives.
According to Hamilakis (2002a), an anachronizing, evolutionary perspective has
persisted through Renfrew’s (1972) application of neo-evolutionist ideas borrowed
largely from American processualism (see Schoep and Knappett 2004; Whitelaw
2004), more recently in modified form (e.g., Earle 1997; Stein and Rothman 1994),
leaving us with embedded terminology, theories, and methods that continue to shape
J Archaeol Res (2008) 16:83–161 95
our narrative of Cretan Bronze Age society. This kind of critique is a familiar one,
very much in tune with recent reviews in other world areas (e.g., Yoffee 1993,
2005). As with Yoffee’s (2005) analysis, there is a hint of a straw man in that the
archaeologist who subscribes to all points in the model may not exist, but clearly
many of these assumptions remain with us and need to be addressed.
One problematic result of the influence of these ideas has been that since Evans’
time the model, generated from the final (Postpalatial) palace at Knossos, has been
applied uncritically to the earlier palaces of the Neopalatial and Protopalatial
periods across Crete. In the final palace at Knossos, under the apparent control of a
Greek-speaking (probably Mycenaean) elite, the Linear B archive details a
hierarchical social and economic polity, and here the model has some merit. But
a mechanistic evolutionary development from earlier to later palaces, assuming
formal and functional similarity and continuity between the more archaeologically
visible later palaces and their scantily preserved (notably, Protopalatial) predeces-
sors, has been challenged by recent archaeological evidence that sheds new light on
form, function, process, and variability across space and time. The controversy
about ‘‘palace’ as a valid term is also fueled by recent discoveries of ‘‘court-
centered’’ buildings at Petras, Galatas, and Kommos, calling into question the
distinction between palace and nonpalace, and diminishing the uniqueness of the
palaces architecturally.
In a series of articles drawing mainly on the site of Malia, Schoep (2002a, b, 2004,
2006) has decoupled the Protopalatial palace centers from earlier and later
manifestations of political organization. She demonstrates that the first (Protopalatial)
palaces lack many ‘‘palatial’’ architectural features, which are instead found first in
elite, nonpalatial structures in those same settlements (Schoep 2002b, 2004,2006, pp.
39–42). Furthermore, elite pottery styles such as polychrome-painted Kamares ware,
once believed to have been produced and consumed restrictively by palace elites at
Knossos, are now known to have been imported and consumed in a wide range of
palatial and nonpalatial locations on the site (Day and Wilson 1998). The evidence of
sealstones and sealings tells a similar story: their earliest use predates the first palaces,
and through the Protopalatial period they are found in elite burials and nonpalatial
structures as well as in the palaces (Schoep 2006, pp. 44–48).
This evidence has been taken to mean that the palaces in the Protopalatial period,
whatever their function, did not house an overarching central authority. Many have
proposed the abandonment of the term ‘‘palace’’ in favor of more neutral
designations such as ‘‘court compound’’ or ‘‘court complex.’’ Inevitably, this also
has led to the demotion of the king (Driessen 2002). The absence, in both the
Protopalatial and Neopalatial periods, of ‘‘royal’’ burials, representations of royal
personages in frescoes and other media, or plausible mentions of rulers or dynasties
in the (still undeciphered) Linear A and Cretan Hieroglyphic texts persists in spite of
much new work. The search for alternative agents of power has led many to
consider heterarchical models and the role of factions (Brumfiel 1994,1995;
Crumley 1995; Feinman 1998) in the emergence and maintenance of Minoan
political systems (Hamilakis 2002b; Schoep 2006; Wright 2004b). Although these
agents are usually assumed to have been elites living outside the palaces, their
formation, composition, bases of power, and reasons for competition often remain
96 J Archaeol Res (2008) 16:83–161
unclear (Day and Relaki 2002, pp. 224–228). Control of ritual and ideology is
frequently invoked as a means by which factions distinguish themselves from other
groups within a society. One such strategy involves the creation of ‘‘high culture,’’ a
set of aesthetic objects and values imbued with connotations of a legitimizing
cosmic order in which the group possessing them takes its rightful place as agent of
social and political authority (Baines and Yoffee 2000; Brumfiel 2000). As one
example, Schoep (2006, p. 51) cites elite, extrapalatial control of attached
craftspersons at Protopalatial Malia, who used innovative techniques to create
objects of high culture such as luxury pottery with applique
´decoration. Following
Helms (1988,1993; see also Broodbank [2000] on the power of esoteric knowledge
among maritime traders), Schoep suggests that factions acquired and maintained
ascendancy through knowledge of and access to objects and ideologies from distant
places. In late Prepalatial and Protopalatial times (EM III–MM II), extrapalatial
elites may have exercised some control over expanding trade with Egypt and the
Levant, and by EM III local imitations of Egyptian objects may have conferred
prestige. At the same time, new technologies arrived in Crete, including the sail, use
of written script, the fast potter’s wheel, faience making, and architectural
innovations, which initially were restricted in access.
If the role of palaces as loci of political power in the Protopalatial period has
been challenged, what were their functions and who controlled them? Most
alternative interpretations cast the palace as a site for ritual performances such as
feasting, dancing, processions, sacrifices, and communal meals (Day and Wilson
1998,2002; Driessen 2002). The Protopalatial palaces at Knossos, Malia, and
Phaistos contain open western courts, accessible from outside the palace, and
enclosed central courts, to which access was more restricted. From this perspective,
the western courts were spaces of inclusion that witnessed large-scale communal
ritual emphasizing the solidarity of the community and the munificence of the
sponsoring elite, while simultaneously dispatching ideological and religious
symbols to reinforce the social hierarchy. In the central court, elites created a
controlled space, decorated with evocative symbols such as double axes and horns
of consecration, for rituals of solidarity and legitimation for a selective audience that
some might view as an ascendant faction. Two reconstructed frescoes from later
(Neopalatial) Knossos, the Sacred Grove and Dance Fresco and the Grandstand
Fresco, appear to depict ritual performances in the western and central courts,
respectively. If the first palaces were in fact ritual rather than residential complexes,
it will be necessary to imagine the bulk storage facilities and food-processing
equipment as intended for ritual feasting (Christakis 2004), and there must be more
cogent explanations for the way that control of the palace—a large and highly
complex architectural compound, after all—was negotiated and maintained by
competing elite groups. Were these shadowy elites members of factions, defined as
spontaneous groups formed by leaders toward specific ends and then disbanded once
the outcome of their purpose was achieved or confounded? If so, can this model
answer Betancourt’s (2002) contention that only through continuity and longevity of
authority is it reasonable to envision the management of such crucial activities as
long-distance trade, successive monumental building projects, and the bureaucracy
J Archaeol Res (2008) 16:83–161 97
associated with palace storage? Indeed, if we envision a world of fluid political
associations, how and why were the palaces built in the first place?
Another way to interpret the Protopalatial palaces is to picture a central authority
deliberately pursuing an inclusive and decentralized ‘‘corporate’’ strategy (Blanton
et al. 1996; Driessen 2002, pp. 11–12; Feinman 2000; Feinman et al. 2000;
Parkinson and Galaty 2007). Within this dual-processual framework, such an
authority maintains its position and promotes compliance through inclusive
practices such as communal ritual and labor projects, unifying ideologies, and the
suppression of economic differentiation and personal aggrandizement in arenas such
as art and burial (evoking Renfrew’s [1974,2001] ‘‘group-oriented chiefdom’’).
Archaeologically, the corporate strategy may present a distinctly heterarchical
signature, and certainly the contextual details of the Protopalatial palace centers
might be interpreted in this light. Schoep (2002a), however, rejects the idea that this
was a single, palatial authority since innovation and exercise of power seem to issue
from outside the palaces, and a redundancy of activities across the sites suggests to
her multiple competing groups.
The Neopalatial period would appear to present a stronger case for central
authority focused on the palaces, notably on the palace at Knossos (Fitton 2002, pp.
133–135). An unprecedented homogeneity of material culture characterizes much of
the island, including architectural forms and techniques, pottery styles, iconography,
artisanal products, and administrative practices using Linear A script (Gates 2004;
Rehak and Younger 2001, pp. 392–441; Schoep 1999,2001). Many of these features
and styles seem to emanate from Knossos, and no contemporary site can match
Knossos for size and splendor (Branigan 2001; Cadogan et al. 2004; Whitelaw
2001a). This preeminence has often been understood as the extension of a Knossian
political hegemony over much of the island, for which further evidence is adduced
in the virtual absence of fortifications on Crete, a kind of enforced ‘‘pax Knossiana.’
From this point of view, palaces controlled international trade, military matters, and
ritual activities in palatial courts and peak sanctuaries (Wiener 1987,1990,1999).
Many who reject the notion of a single, autocratic king envision instead an oligarchy
of elite families at Knossos with common theocratic and mercantile interests
(Betancourt and Marinatos 1997; Chapin 2004; Weingarten 1999).
From the same evidence, others infer a very different political world in which
numerous independent, mostly small ‘‘peer polities’’ (Renfrew and Cherry 1986)
engaged in intensive interaction and competition internally, regionally, and
internationally. The cultural influence of Knossos is acknowledged, but the
similarities to Knossian style and technique are attributed to emulation without the
implication of political control, i.e., Wiener’s ‘‘Versailles effect’’ (Wiener 1986,p.
17). Furthermore, the shared language of culture provided a unified ideology that
formed the basis for competition among factions in a heterarchical political
landscape (Hamilakis 2002b; Schoep 2002a; Vansteenhuyse 2002). Adams (2004,
2006) stresses that hierarchy and heterarchy can exist side-by-side: at Neopalatial
centers like Knossos and Malia, she argues for a paramount central authority in the
palace but intense competition among second-tier elites. She does not view these
elites as members of factions at Knossos, because they remained economically
dependent on the palace, but at Malia the variety of ceremonial practices and signs
98 J Archaeol Res (2008) 16:83–161
of economic self-sufficiency are consistent with a factional model (Adams 2004, pp.
New data from excavations and particularly from regional surface surveys have
added greater nuance, permitting cultural and political influence to be separated and
providing alternatives to the notion of island-wide political control from Knossos.
Recently, two prominent scholars independently reached the conclusion that in LM
IA Knossos led a central Cretan state, while in eastern and western Crete small
states remained independent (Driessen 2001; Warren 2002). In addition to spatial
variability, there also were important diachronic changes within the Neopalatial
period. Macdonald’s (2002) careful contextual analysis of the architectural and
artifactual remains at the palace of Knossos reveals three building phases, which he
calls the new (MM IIIB), frescoed (LM IA), and ruined (LM IB) palaces. In his
view, the MM IIIB palace was conceived and built as a single great program by an
elite group with the consensus and participation of the broader community. After a
devastating earthquake late in MM IIIB severely damaged this palace, it was rebuilt
in LM IA on a less ambitious scale with changes that effectively excluded access for
the wider community. This frescoed palace incorporated monumentalized frescoes
with a symbolic program probably meant to convey social distance and control to
the community and to foreign visitors. In this transformation it is plausible to see an
elite group reacting to post-earthquake social disturbances, and in the terms of the
dual-processual framework, moving away from corporate toward exclusionary
‘network’’ strategies, exemplified also by changes in funerary forms and practice
to increasingly private burial (Soles 2001), nucleation and expansion of palatial
authority in the economic realm, and increased gift exchange among elites on
Crete and beyond (cf. Feinman 2000, table 3.2). Parkinson and Galaty (2007, pp.
120–122) note, however, there are still strong elements of a corporate orientation,
such as the continuing absence of portraits of individual rulers.
Subsequent disturbances in LM IB may have been triggered by the eruption of
Thera and the earthquakes that preceded it (see above). Some perceive LM IB as a
time of political fragmentation and a return to localized political authority following
these natural and human-induced destructions, which itself ends in a horizon of
destructions at all major sites except Knossos (Driessen and Macdonald 1997).
Thus, although the Neopalatial period is often presented uniformly as the apogee
of Minoan civilization, the emerging story is one of considerable variability over
time and across the island as the fortunes of sites and regions waxed and waned. The
nature of social and political relations depends on the time and place one examines,
and a salutary trend of the last decade has been the growth of systematic contextual
analyses involving the examination and comparison of whole assemblages rather
than individual objects such as imports (Adams 2006; Brogan, personal commu-
nication 2004).
The results of surface surveys support broad outlines of both island-wide changes
and pervasive regional variability (Cunningham 2001; Cunningham and Driessen
2004; Driessen 2001; Haggis 2002). Using data from the Kavousi Survey in eastern
Crete, Haggis (2002) asserts that economic (and by implication political)
interference from the palaces can be seen in the regional surface record as a
change from integration to connectedness. He describes the countryside in
J Archaeol Res (2008) 16:83–161 99
Prepalatial times (EM III–MM IA) as a heterarchical arrangement of clusters of
farmhouses and small hamlets surrounding discrete areas of arable land and
perennial water sources. These clusters shared agricultural resources, applied
diverse subsistence strategies, and interacted intensively with other clusters. Haggis
argues that such systems are well integrated, that is, characterized by a ‘‘multiplicity
of linkages between individuals, sites, and the landscape itself’ (Haggis 2002,p.
123), and are perhaps the most stable adaptations to Aegean landscapes. In contrast,
the Neopalatial (MM III–LM IA) pattern reflects nucleation under palatial influence,
with a clear site-size hierarchy, where settlements are no longer oriented to arable
zones but instead to trade routes, ports, and rivers for irrigation. Palatial systems
display high connectedness through regional political hierarchies rather than local
relationships based on traditional land use and other social and economic
interactions. As a result, they break down traditional structures and tend to be
poorly integrated and inherently unstable. The syntheses of settlement pattern data
across Crete that have begun to appear (Cunningham 2001; Cunningham and
Driessen 2004; Driessen 2001; Knappett 1999) support a process of political
expansion and centralization in central Crete focused on Knossos in the Neopalatial
period, but this appears to be an aberration against a background of multiple
divergent trajectories and a strong sense of local autonomy at most times and places
in the Cretan Bronze Age.
The debates surrounding the social and political organization of the Cretan
Bronze Age partly reflect a generational divide: older scholars who came of age
with the comparative moral certainties of the post-World War II and Cold War
world order naturally hold different worldviews than younger colleagues emerging
in a highly ambiguous post-Cold War, postmodern world. This new generation of
scholars has its own, possibly anachronistic biases about the past, a fact
acknowledged and even embraced (Hamilakis 2002a). But there is also the simple
fact that the archaeological record, now vastly more voluminous than even ten years
ago, is inevitably more complex. With this enlarged database, it is possible to
demand more exacting criteria for defining and explaining concepts such as the
‘palace.’’ It is interesting that the deconstruction of the palace concept occurs in the
Aegean at a time when it is receiving increasing attention in the New World
(Christie 2003; Evans and Pillsbury 2004; Inomata and Houston 2001). Similar
questions of identifying the architectural attributes, residents, and ranges of
activities and functions that define a palace are being posed, and there is much to be
learned in the variety of perspectives cross-culturally. In contrast to the New World,
however, the concept of the palace was never stigmatized as inappropriate or
offensively colonial in the Aegean because in European nationalist narratives,
Greece held the honor of being the first ‘‘European’’ civilization, thus having
produced the first ‘‘Western’’ palaces.
Emergence of complexity in Mycenaean Greece
One of the enduring mysteries of Mycenaean civilization is how and why signs of
social complexity appeared suddenly in Middle Helladic (MH) III, exemplified by
100 J Archaeol Res (2008) 16:83–161
the spectacularly rich shaft graves at Mycenae and the tholos tombs of Messenia,
from a Middle Helladic background of poorly furnished graves and decidedly
simple settlements (Rutter 2001, pp. 124–147, 151–155); subsequently a small
number of palace-centered states emerged from numerous, small polities. Because
the exotic objects and pictorial art of the shaft graves show close ties with Minoan
Crete, at that time at the apex of its Neopalatial prowess, it has often been suggested
that small groups of nascent elites on the mainland cultivated a ‘‘special
relationship’’ with one or more Minoan palaces to gain access to exotic materials
and artisans. The Minoan influence is certainly real and even profound in some
areas, including iconography, ceramic forms and styles, metalworking, and to some
extent funerary architecture, but the graves and their furnishings betray many other
influences and craft traditions, from Anatolia and Egypt, but more prominently of
local or other mainland origin. In recent years, attention has been drawn to Kolonna
on the island of Aegina, the most prominent site in the Aegean before the rise of the
expansive state at Mycenae. Kolonna possesses a continuous sequence of massive
fortifications built and modified over 500 years and a shaft grave remarkably similar
in form and content to those at Mycenae (Kilian-Dirlmeier 1998), only earlier.
Kolonna’s connections to Minoan Crete and the Cycladic Islands are obvious in the
shaft grave and the Middle Bronze Age settlement, and some argue that Kolonna
may have acted as mediator of goods and ideas to nascent elites at Mycenae
(Niemeier 1995; Rutter 2001, pp. 126–130), and possibly as direct competitor
(Pullen and Tartaron in press). Regrettably, in spite of decades of excavations at
Kolonna since the late 19th century, relatively little information has been published,
but renewed Austrian excavations may soon allow us to fully appreciate Kolonna’s
role in these crucial centuries (Gauss 2006, in press; Lindblom 2001; Walter 2001).
In recent years, archaeologists have endeavored to move beyond simply debating
the provenience or inspiration of the objects in the shaft graves, or invoking vague
concepts of ‘‘secondary state formation,’’ as if these answers were in themselves
causal explanations. Instead, analyses that embrace entire regions and full ranges of
material remains have led to new questions about the mechanisms by which social
inequality emerged and the specific processes by which leaders arose and marked
their status in material ways. A sort of narrative can be distilled that depicts this
period (c. 1700–1400 B.C.) as one of transformation in which chiefdoms arose at
many sites in southern and central mainland Greece, and a few of these managed to
consolidate power as small palace-centered states by c. 1400 (though there are some
who see Mycenae as a special case that qualifies as a palace state already during the
use of the shaft graves in Grave Circle A, c. 1600: French 2002; Kilian 1987,1988a;
but cf. Laffineur 1995). Incipient imbalances in wealth and population stimulated
the emergence of chiefdoms in an environment of competition and emulation,
involving conspicuous consumption and display, and possibly warfare. Later,
certain principalities were able to form palatial states through the suppression of
regional competition (Voutsaki 1995,1998; Wright 1995,2003).
The principal debate about this narrative involves the way that political
hegemony within a region was achieved; generally, existing models emphasize
either conflict or consensus. A major obstacle in moving from generalized to more
nuanced reconstructions is that this era is known primarily from burials; because
J Archaeol Res (2008) 16:83–161 101
there is little information on settlements, architecture, and other aspects of the world
of the living, most arguments are based on funerary evidence. Voutsaki (1995,1998,
1999,2001) has carefully analyzed the burial evidence, tracking changes in tomb
form, shifts from intramural burial to extramural cemeteries and individual to
multiple burial, and trends toward more complex burial rituals and increasing
amounts of wealth deposited as grave goods (see also Boyd 2002). She concludes
that funerary behavior was an active agent in social change, not a passive reflection
of wider social tensions; thus, differentiation, display, and competition in mortuary
practices were deliberate strategies on the part of nascent elites to create and
maintain social distinctions. For Voutsaki, the catalyst for all of this activity was
Minoan expansion; specifically, the influx of Minoan and Cycladic goods disrupted
the egalitarian social structures of Middle Helladic Greece with novel ideas and
ways to distinguish oneself through the creation and expression of prestige. These
new objects and styles were put to work as political capital through conspicuous
ritual deposition in tombs. A comparative analysis of mortuary change in two
prominent regions, Messenia and the Argolid, reveals contrasts in the pace and
strategies of political consolidation by two polities (Pylos and Mycenae, respec-
tively) that would emerge as palace centers (Voutsaki 1998). We may expect that
each region has a unique historical trajectory, and it will be especially important to
discover why in some regions palace centers never developed at all, in certain cases
despite obvious ecological and geographic advantages (Cherry and Davis 2001;
Pullen and Tartaron in press; Wright 2004a).
In Voutsaki’s analyses, and indeed in most reconstructions, palace centers
became possible because their political elite were able to eliminate rivals in a
regional competition for hegemony (Bennet and Davis 1999). An alternative
interpretation focusing on consensus is offered by Wolpert (2004). Drawing on
work by Pauketat and others (Fotiadis 1999; Pauketat 1994; Pauketat and Emerson
1999), Wolpert argues that notions of hegemony through antagonistic, open
competition rely on misunderstandings of classic ethnographic phenomena such as
the potlatch, and suggests instead a process of regional consensus through
negotiation of legitimate practice within a kinship structure. From this perspective,
consensus establishes hegemony more effectively than violence and subjugation,
and legitimacy is established when rule by a particular lineage appears as natural or
organic, not oppressive. Wolpert specifically rejects the notion that objects and
influences from the Minoan world could stimulate the ‘dissolution of earlier
frames of reference: the more localized identities, the kinship order, the segmentary
alliances’’ (Voutsaki 1998, p. 47). He seems to believe that instead of emerging
victorious after a period of intense political and military struggle, palace centers like
Mycenae and Pylos incorporated their regions into a community of cult centered on
ancestor veneration and vertical kinship. This process was played out in the reuse
and lavish consumption of prestige goods in shaft and cist graves over many
generations (the other main types, tholos and chamber tombs, are typically too
heavily disturbed to reconstruct separate burial events). The consolidation of
consensus involved rival lineages renegotiating social networks and establishing the
meaning of unfamiliar symbolic codes appropriated from external sources.
102 J Archaeol Res (2008) 16:83–161
Consensus models have their weaknesses. For example, what are we to make of
the proliferation of bronze weapons and martial iconography at that time (Voutsaki
2001, fig. 3)? Wolpert’s is a mainly theoretical argument, with little guidance about
specific methods or material indicators that might establish its validity. Upon close
examination, Wolpert and Voutsaki agree in most respects, differing chiefly on the
mechanisms of social change reflected or effected by mortuary behavior. There is in
fact a middle way, that after a period of competition rival chiefs recognized the
futility of cyclical conflict and chose to cooperate in the creation of a larger, more
productive political and economic entity (Wright 1995, p. 73). This is a useful
debate because it focuses attention on potential biases in the ways that we assign
causality to material patterns in the archaeological record. But it is also important to
keep in mind that in ascribing such a pivotal role to conspicuous mortuary behavior,
we are only dimly aware of the realm of the living, where other symbols and
activities must have been implicated in these momentous changes.
The meager information we possess regarding settlements has come from
regional surface surveys, small-scale excavations, and architectural studies, which
have added details about the expansion of nascent centers and the transition to
palatial political systems (Bennet 1999; Cherry and Davis 2001; Cosmopoulos
2006; Davis 1998; Davis et al. 1997,1999; Maran 1995; Nelson 2001; Wright 2003,
2004a). In Messenia, intensive survey has documented the growth of settlement at
Pylos from the beginning of the Shaft Grave era to the formation and expansion of
the palace (Bennet 1999), as well as an apparent centralization of population around
the palace after 1400 as formerly active settlements in the hinterland diminish in
importance (Bennet and Shelmerdine 2001). But the survey also added nuance by
tracking diachronic histories of several small settlements, demonstrating their
changing fortunes and functions over time, first within a competitive environment
and later as part of the Pylian state.
There has been no surface survey at Mycenae and its immediate environs, but in
the early 1990s a systematic survey of monuments, old excavation trenches, and
roads outside the citadel walls was undertaken (French et al. 2003; Jansen 1997,
2002). The remains of the MH III–LH II settlement within the citadel at Mycenae
are deeply buried beneath those of later periods, but the proliferation of richly
appointed tombs (shaft graves, tholos tombs, and chamber tombs) certainly sets
Mycenae apart from other developing centers of the Argolid (Voutsaki 2001). The
expansion of Mycenae is instead best seen in the data from intensive surveys in
adjacent regions to the east (the Berbati Valley: Wells and Runnels 1996) and
northwest (Phlius and the Nemea Valley: Casselmann et al. 2004; Cherry and Davis
2001; Wright 2004a; Wright et al. 1990). These surveys consistently show near or
total abandonment of much of the Argolid and southern Corinthia for four centuries
from the end of the Early Bronze Age to the last phase of the Middle Bronze Age,
until the time of the shaft graves at Mycenae. A burgeoning center like Mycenae
might logically seek to fuel its growth by expanding into these rich and nearly
uninhabited agricultural landscapes, but the emerging picture is complex and
variable. By LH II, the Nemea Valley appears to have been incorporated into
Mycenae’s sphere (Cherry and Davis 2001), but the initial resettlement of sites in
the valley (e.g., Tsoungiza) in MH III was probably narrowly focused on local
J Archaeol Res (2008) 16:83–161 103
arable and pastoral lands (Wright 2004a, p. 126). In fact, before Mycenae annexed
the Nemea Valley into its sphere, the inhabitants of Tsoungiza may have been
attached to the Phliasian Plain to the west, where an early settlement and rich
cemetery are known at Aidonia (Casselmann et al. 2004; Wright 2004a, p. 125).
The Berbati Valley, adjacent and readily accessible to the east, was not fully
exploited by Mycenae until well into palatial times (LH IIIA2), perhaps because the
small settlement at Mastos managed to maintain its independence (Schallin 1996,
pp. 170–173). Furthermore, the dynamics of expansion among the developing
centers in the Argolid, including the exploitation of the richest agricultural land in
the extensive Argive plain to Mycenae’s south, are little known since no systematic
survey has been performed there.
Thus, in the current state of research we are merely beginning to assemble the
details of the transformation from small, egalitarian farming and herding
communities to palace-centered states. A key contribution of survey has been to
introduce variability in time and space to a process that, peering back from the
palaces, looked linear and uniform. A recently formed project focusing on the
Argolid during the Middle Helladic period and the transition to the Late Helladic (c.
2000–1500 B.C.) should place the indigenous and exogenous contributions to
emerging complexity in proper perspective (Voutsaki 2004). By examining and
reanalyzing decades of funerary, settlement, skeletal, and iconographic data, the
investigators hope to identify the seeds of later developments in the Middle Helladic
background and to sort out intra- and interregional dynamics. In the future, this kind
of research design can be extended to less-studied regions, which may allow us to
move beyond simplistic, totalizing formulations such as secondary state formation
under Minoan influence toward frameworks that better accommodate variability
across space and time, such as dual-processual analysis (see above) or dynamic
models that track the different stages in the life cycles of states (Marcus 1998;
Parkinson and Galaty 2007; but for critique see Haggis in press).
Organization and political economy of the Mycenaean palaces
In the past decade many important works have appeared on the organization and
operation of the Mycenaean palaces (Galaty and Parkinson 1999a; Shelmerdine
1998,1999,2001a, b; Voutsaki and Killen 2001). One of the most significant trends
in recent scholarship has been a reformulation of the debate on how the political
economy of the palaces actually operated on a daily basis, and particularly a new
emphasis on the roles of individuals and their interactions with the state. There has
long been a tendency to portray the Mycenaean palaces as impersonal structures
whose managerial control was ‘‘pervasive, monolithic and monopolistic’’ (Bennet
2001, p. 25), based partly on comparison with obsolete notions of an ‘‘Asiatic’
palatial economy in the Near East (Cherry and Davis 1999, pp. 94–95). In such a
world, individuals are faceless props lurking behind systemic titles and roles, their
actions a passive response to structures beyond their control. Several scholars have
sought to break the monopoly of the palaces in the economic sphere by identifying
palatial and nonpalatial sectors, the latter referring to certain areas of agriculture and
104 J Archaeol Res (2008) 16:83–161
craft production in which the palaces may have shown little interest or exerted little
control (Galaty 1999a, b; Halstead 1992a, b,1999a, 2001; Parkinson 1999;
Whitelaw 2001b). Nonpalatial sectors have usually been identified by a virtual
absence of mention in the Linear B archives, particularly those involving nonluxury
goods, or from evidence that a particular activity was decentralized at a remove
from the palaces. Palatial interest was instead focused on ‘‘the creation and control
of easily transported products with pronounced ritual, aesthetic, or commercial
value’’ (Galaty 1999a, p. 57), the mark of a wealth-financed economy (D’Altroy and
Earle 1985; Galaty and Parkinson 1999b; Parkinson 1999). While the two-sector
approach has spawned a productive debate, many critics find it problematic
(Nakassis 2006, pp. 11–16). The assumption that silence in the texts or distance of
an activity from the center should imply a lack of palatial involvement is
unwarranted: workshops that are archaeologically attested at the palaces, as well as
craft specialists including scribes, receive no mention. At the same time, the tablets
do mention vast herds of sheep controlled by the palace at Mycenaean Knossos that
grazed in distant pasture and taxes extracted from distant communities in the Pylian
polity (Cherry and Davis 1999, pp 96–97; Halstead 1999b, 2001; Killen 1999,
p. 89). Further, in more extreme formulations this model tends to create and reify a
false dichotomy, as if two parallel economies existed as separate, nonintersecting
entities in daily life. To be sure, there were economic activities over which the
palace exerted greater or lesser control and had more or less interest, but it is at the
intersections and overlaps of official and nonofficial action that we may glean the
true penetration of the state in the lives of nonelite individuals.
A good example of this debate is the question of palatial interest in pottery
production. In the palace at Pylos, more than 10,000 vessels were found in the final
destruction deposit, many of these stacked neatly in pantries waiting to be deployed
for state-sponsored feasts. This assemblage is significant because it seems to
represent a nearly complete record of the pottery in use at the moment of the
palace’s destruction (Whitelaw 2001b, p. 77). Were these produced by palatial
potters, commissioned from outside artisans, or obtained from nonpalatial markets?
The relative scarcity of references to potters in the Linear B archives has been noted
(e.g., Palaima 1997; Whitelaw 2001b), but Hruby (2006, p. 198) contends that their
number is consistent with the scale of production needed to supply the palace in a
given year. Bulk acquisitions of pottery are recorded and four potters are mentioned,
including one landholding potter at Pylos accorded the epithet ‘‘royal.’’ This
‘royal’’ potter has been variously interpreted as an elite, attached craftsperson who
supplied the pottery found in the destruction deposit (Wiener in press), an
independent producer on whom a ‘‘royal seal of approval’’ was bestowed (in the
sense of ‘‘potter to the king’’) (Knappett 2001, p. 94), or possibly a procurer of
pottery from local producers for the palace (Nakassis 2006, p. 16). Nevertheless,
calculations by Whitelaw (2001b) based on ethnographic data on labor and pottery
breakage indicate that all of the palace’s annual needs for pottery could have been
met by a single workshop; the total corpus at the palace accounts for only a tiny
percentage of total ceramic production within the Pylian state. Hruby (2006,
pp. 199–209) takes this argument much further in an exhaustive analysis of nearly
6,700 fineware vessels from pantries 18–22 at the palace. Based on ethnographic
J Archaeol Res (2008) 16:83–161 105
data and observations on fabric uniformity, consistent motor habits, metrical
standardization, and even fingerprints and palmprints preserved in clay, she makes a
persuasive case that a single potter produced all the vessels in the pantry rooms, if
not in the entire palace. Interestingly, the potter was inexperienced and the work
emphasized mass production rather than quality.
Although Whitelaw concludes that the palace exercised no meaningful control
over the production of ceramics, being content to allow departments of the state to
acquire pottery from a small number of provincial workshops through direct
taxation or obligations, more direct palatial involvement cannot be excluded.
Because the tablets at Pylos represent temporary records that may preserve only the
final six or seven months of the existence of the palace (Palaima 1997), it is
possible, as Voutsa (2001, p. 160) has observed, that a large shipment in summer (a
logical seasonal schedule for pottery production) simply predated the administrative
cycle of the tablets preserved in the destruction, which is thought to have occurred
in the spring. Some see interest on the part of the palace in specific kinds of
ceramics only (Knappett 2001). Recent chemical and petrographic analyses (Galaty
1999a, b) raise the possibility that a single workshop monopolized high-quality
sources of kaolinite clay for the manufacture of fine, wheelmade vessels such as the
kylix (pl. kylikes), a drinking goblet found in the thousands in the palace pantries.
Although Galaty (1999a, p. 59) concludes that ‘‘an independent and parallel local
economyonly superficially intersected with the palace economy,’’ he finds it
plausible that the palace tried, in mature palatial times, to control a few large potting
establishments to lower costs and ensure supplies of the finest vessels. This may
depend on the debated point of whether value or prestige was ever attached to
kylikes, used in funerary ritual and feasting but also an everyday vessel and at Pylos
mostly undecorated (Galaty 1999a; Knappett 2001, p. 94). If so, Voutsaki’s (2001)
observation on the distributional pattern of prestige goods over time is relevant: in
the early Mycenaean period, there was intense competition for prestige goods,
followed in the palatial period by a progressive restriction of access, culminating in
late palatial times (LH IIIB). It is unclear whether this strategy signifies confident
prosperity or fear of the impending crisis of c. 1200 B.C. What is clear, however, is
that dichotomous categories such as ‘‘attached’’ and ‘‘independent’’ craft specialists
fail to capture the complexity of relationships between official and nonofficial
actors, or producers and consumers in the Mycenaean political economy. The
potters mentioned in the texts may have been both or neither (Hruby 2006, p. 226;
Knappett 2001, p. 95; see also Costin and Wright 1998).
Because numerous individuals are referred to by name and/or title, and their
interactions with or on behalf of the state are described, the Linear B archives
present an unusual opportunity to apply agency theory (Bennet 2001; Manning
1998; Nakassis 2006). The aim of recent work has not been merely to show that we
can identify individuals who did things, or to imagine individuals operating free of
structural constraints. Rather, following the now universally cited concepts of
‘structuration’’ (Giddens 1979,1984) and ‘‘habitus’’ (Bourdieu 1977,1990), an
agency approach to the Linear B texts emphasizes the dynamic interplay of humans
and the structures and practical knowledge they inherit, a continuous process of
106 J Archaeol Res (2008) 16:83–161
creating and mutual shaping in which individual action may have discernible
consequences. This perspective holds the promise of a humanized Mycenaean state,
reproduced but also animated and transformed by myriad individual acts and
decisions. Nakassis (2006) presents a detailed prosopographical study of named
individuals, who often appear in multiple texts with multiple roles and areas of
responsibility, to argue that these Mycenaeans were knowledgeable agents capable
of manipulating social contexts for their own advantage. Examining the records of
bronzesmiths, shepherds, and others, Nakassis proposes that many such persons
were heavily invested in the palatial economy, because in return for managing the
resources of the state, they were granted access to otherwise unattainable wealth and
status. Bennet (2001) associates the palace scribes with the names of elite
supervisors of various activities, including the provisioning of feasts at the center
and in the hinterland or distributing raw materials to craftsmen.
There are, of course, limits to what we can learn about agency from the tablets
alone. The individual actions and decisions of a vast hinterland of common farmers
and herders cannot be extracted from the tablets, though some of their interactions
with the palace are known. A view from the hinterland is sorely needed to
counterbalance a strong palace-centered bias (Cosmopoulos 2006, pp. 207–208). To
some extent, archaeological surface survey has succeeded in placing lower-order
settlements and activity areas on the map. In Messenia, the Minnesota Messenia
Expedition (see above) in the 1950s and 1960s (McDonald and Rapp 1972), the
Pylos Regional Archaeological Project (PRAP) in the 1990s (Davis et al. 1997), and
the ongoing Iklaina Archaeological Project (IKAP; Cosmopoulos 2006) have
investigated the region at three very different scales. IKAP has narrowed its focus to
a single district administered by the palace at Pylos, using an intensive, total
coverage strategy to accord particular attention to the lower tiers of settlement.
Nevertheless, published excavations of small, provincial Mycenaean settlements
have been rare, and this has curtailed much discussion of individuals in the
hinterland. Addressing this imbalance has long been expressed as a priority, but
reaction has been slow, in part due to the constraints on fieldwork explained above.
The few new excavations at small Mycenaean sites are thus almost inordinately
important and deserve brief mention. In 2006, IKAP began a second phase
involving excavation at Traghanes, a site identified with the regional town a-pu
the Linear B records, and at other sites discovered by the survey. IKAP has a
tremendous advantage in that it is operating under the auspices of the Athens
Archaeological Society, a body outside the purview of the Greek Archaeological
Service, with potentially fewer restrictions on the project’s scope and duration. At
Mitrou, a small island just off the eastern coast of central Greece, excavations begun
in 2004 have already yielded structures of the poorly represented early Mycenaean
period (LH I) and more abundantly of the postpalatial phases of LH IIIC, which
transition to the Early Iron Age (Van de Moortel and Zahou 2005). These
excavations afford our best hope of illuminating topics that can scarcely be
broached at present, e.g., archaeologies of households and communities, domestic
production and consumption, and evidence of the presence of the state in daily life.
Defining Mycenaean political and economic structure at close range is essential,
but there are still questions about the way that Mycenaean polities related to one
J Archaeol Res (2008) 16:83–161 107
another. In spite of separate and distinct origins for each palace center, a striking
uniformity evolved among them in crafts, administrative systems, and unifying
institutions such as kingship (Shear 2004; Wright 1995, p. 64), typically explained
as the result of some form of peer polity interaction. Yet at a recent conference,
Near Eastern specialist Nicholas Postgate noted the remarkable homogeneity of the
Linear B script and the administrative system it served at the palaces, and
challenged experts to explain why there could not have been a single Mycenaean
political authority (Voutsaki and Killen 2001, p. 13). The fact that there was no
ready, conclusive answer underscores our imperfect knowledge of the political
structure of the Mycenaean world and the need for empirical fieldwork and more
robust theory. Is it unthinkable to imagine a politically unified Mycenaean world?
The idea that the Greek mainland was ruled from a single palace (i.e., Mycenae) was
discarded long ago, but an old proposal has recently resurfaced that there may have
been a pan-Mycenaean elite of related families ruling the palace centers. This notion
turns on the recurrence of identical names in the Linear B tablets among a group of
prominent officials known as ‘‘collectors’’ over a period of six to seven generations
between the earliest archives at Knossos and the latest at Pylos and Thebes (Killen
1979; Olivier 2001). Among Linear B experts there is disagreement about whether
these are family names unique to a restricted elite or simply popular names
reflecting a cultural koine
´(Rougemont 2001, p. 138).
In any case, the peer polity concept falls apart once one ventures beyond the
confines of the core area of southern and central Greece, or even into the intervening
territories between palace states. In the last decade there has been increasing interest
in defining and exploring ‘‘cores’’ and ‘‘peripheries’’ in the Mycenaean world, using
a variety of approaches (Dakoronia 1999; Kyparissi-Apostolika and Papakonstan-
tinou 2003). The Mycenaeans were active participants in eastern Mediterranean
trade networks, from which the palaces obtained essential raw materials, partic-
ularly metals including copper, tin, silver, and gold. As the palace economies
expanded in search of reliable sources of supply, certain peripheral locations
witnessed Mycenaean presence, ranging from sporadic visits to full-blown colonies.
Broadly speaking, to the east the old civilizations in Egypt and Syro-Palestine were
more powerful and politically complex than the Mycenaeans, and there was little
prospect of Mycenaean colonization or significant cultural impact. Kardulias
(1999a, b) advocates a world systems approach, envisioning the Mycenaean world
in a ‘‘core-core’’ relationship with Egypt and the Levant through intensive trade
relations. Yet we still do not know the frequency with which Mycenaean merchants
traveled to distant places in Egypt and the Levant, as opposed to trading through
middlemen, at Ugarit or Cyprus, for example, who then transshipped Mycenaean
products farther on (Bell 2005; Cline in press; Pulak 2005; Whittaker 1997, pp.
104–115). The Linear B archives are virtually silent on exchange within the
Mycenaean world and without, a situation that has not changed in the decade since
Bennet and Galaty wrote, in spite of the recovery of many new tablets, mainly at
Thebes (Bennet, personal communication 2006; see Aravantinos et al. 2001).
To the north and west and on the Aegean Islands and coasts, however,
Mycenaeans encountered many societies at lower levels of complexity than
themselves. In the past, these interactions have been interpreted primarily in terms
108 J Archaeol Res (2008) 16:83–161
of the interest of the palaces in establishing and securing access to desired trade
goods. Relations were presented as asymmetrical between a dominant (Mycenaean)
core and a passive periphery; that is, there was no agency in the periphery (Stein
1999, pp. 10–26; Tartaron 2005). A salutary trend of recent years has been to shed
this Myceno-centric point of view and instead explore these encounters as unique
and historically contingent outcomes of interaction between Mycenaeans and
indigenous populations, which were diverse in terms of social organization and
motivations for accommodation or resistance (Melas 1991). The general question of
Bronze Age Aegean emporia in the eastern and central Mediterranean has been
taken up in the weighty proceedings of a recent conference in the Aegaeum series
(Laffineur and Greco 2005), and many authors consider specifically the form of
Mycenaean presence: do the material remains indicate simple episodes of trade at
multicultural emporia, or was there deeper cultural penetration in the form of
emulation, merchants’ enclaves, or even colonies of Mycenaean immigrants?
Notable work that emphasizes this interaction as a dynamic and negotiated process
has been accomplished in Macedonia (Andreou and Kotsakis 1999; Buxeda I
´s et al. 2003; Kiriatzi et al. 1997), Epirus (Soueref 1999; Tartaron 2001,
2004,2005), Thessaly (Adrimi-Sismani in press; Feuer 1994,1999,2003), the
Aegean Islands (Karantzali 2005; Momigliano 2005; Privitera 2005), the Ionian
Islands (Soyoudzoglou-Haywood 1999), Anatolia (Mu
¨ller Celka 2005; Niemeier
1998,2005), Cyprus (Cadogan 2005), and the central Mediterranean (Alberti and
Bettelli 2005; Buxeda I Garrigo
´s et al. 2003; D’Agata 2000; Jones and Vagnetti
1991; Militello 2005). In spite of increasingly sophisticated theoretical approaches
and careful reassessment of relevant assemblages, the evidence, frequently
consisting mainly or solely of portable goods such as pottery and lacking useful
information on such areas as funerary or religious behavior, is often not up to the
task of distinguishing among the various options.
Even when there is good reason to suspect a colony, Mycenaean impact in
peripheral lands was primarily coastal, superficial, and discontinuous (Tartaron
2005). Certain landfalls may be understood as ‘‘ports of trade’’ or ‘‘gateway
communities,’’ but there was little direct penetration of Mycenaean culture into the
interior save for portable objects that were probably conveyed by indigenous
traders. In the region of Thessaly, which borders the Mycenaean core area by land
and sea, decades of survey and excavation have made it possible to trace the
attenuation of Mycenaean influence as one moves inland from the Aegean coast and
northward along a land frontier (Eder 2003,2006; Feuer 1983,1994,1999). Yet
ongoing excavations at the remarkable Mycenaean-era settlement at Dimini on
Thessaly’s Aegean coast (Adrimi-Sismani 1994,1996,1999–2000) are revealing a
high-order Mycenaean settlement, possibly a palace center, which underscores the
maritime orientation of Mycenaean expansion and demands a reevaluation of the
geographical meaning of the ‘‘core area’’ (Adrimi-Sismani in press).
A further interesting development is an effort to break down the Aegean region
into ‘‘small worlds’ (Broodbank 2000; Horden and Purcell 2000, pp. 51–172;
Sherratt and Sherratt 1998), referring to the small-scale, intensive networks of
interaction among communities of the Aegean Islands and coasts (compare Chase-
Dunn and Mann 1998 on the Wintu of northern California). These interactions
J Archaeol Res (2008) 16:83–161 109
sustained essential ties among small communities living with limited subsistence
and human resources. Bronze Age small worlds have been proposed for the
Cycladic Islands (Broodbank 2000), the coast and offshore islands of southwestern
Anatolia (Momigliano 2005), and the Saronic Gulf (Pullen and Tartaron in press).
Momigliano finds evidence that Iasos on the Anatolian coast was part of a small-
scale exchange network with the Cycladic and Dodecanesian Islands of the Aegean,
which in turn was nested within larger-scale exchange at major emporia such as
Miletos (on coastal Anatolia) and Trianda (on the island of Rhodes: Karantzali
2003). Another small-scale network is proposed for small Mycenaean settlements
ringing the Saronic Gulf, centered (geographically and culturally) on the major site
of Kolonna on the island of Aegina (Pullen and Tartaron in press). The Saronic Gulf
has become one of the most important new research areas in the Mycenaean world,
thanks to the discovery of a number of major and minor Mycenaean sites on the
islands and coasts (Siennecka 2002), notably on Salamis Island (Lolos 1996,2001,
2002) and along the previously poorly studied western shores, where an important
settlement at Galatas (Konsolaki-Yannopoulou 1999,2003b), a Mycenaean
sanctuary at Ayios Konstantinos on the Methana peninsula (see below; Kon-
solaki-Yannopoulou 2001,2002,2003a), and a Mycenaean harbor at Korphos
(Rothaus et al. 2003; Tartaron et al. 2003) have all been located recently. The
Saronic Gulf is ideally positioned for an investigation of emergence, consolidation,
and resistance in the rise of Mycenaean power in the northeastern Peloponnese.
Ritual and religion in the Minoan and Mycenaean worlds
Although the beliefs underlying Greek Bronze Age religion(s) will always remain to
some extent elusive, considerable progress has been made in the last decade on
illuminating the range of ritual expressions and cult places associated with them,
with some new, albeit tentative steps toward belief. These advances have been made
possible by the careful analysis and comparison of contextual assemblages, by an
unprecedented integration of artifactual, iconographic, glyptic, and textual data, and
by an increasing use of anthropological models and interpretive frameworks.
Minoan ritual and religion
Minoan religion is more easily approached than Mycenaean, owing to a tightly
associated suite of iconographic, artifactual, and architectural elements of widespread
ritual practices. For example, recurrent themes of female goddesses and worshipers,
symbols such as double axes and ‘‘horns of consecration,’’ and specific kinds of clay
figurines and pouring and drinking vessels often co-occur in painted fresco scenes,
inscribed sealstones and signet rings, and artifactual assemblages. Based on contextual
associations, several kinds of ritual spaces have been identified, including cult rooms
within palatial and nonpalatial structures; large, open courts attached to palaces;
isolated shrines located on conspicuous mountain peaks or hillsides, known as peak
sanctuaries; and caves. The iconographic images provide clues to certain ritual acts
110 J Archaeol Res (2008) 16:83–161
and equipment, as well as the roles of human and divine participants, but the identities
of those depicted and the meanings of the ritual acts and accompanying symbols are
generally problematic; moreover, these images represent a narrow, elite slice of
Minoan religious life (Cain 2001; Fitton 2002, pp. 172–178; Peatfield 1992; Rehak and
Younger 2001, pp. 437–439). Recent excavations at a number of sites in Crete have
shed new light on the changing nature of religious practice in shrines—public and
private, and in palatial and nonpalatial settings (Rehak and Younger 2001, pp. 433–
440; Watrous 2001, pp. 193–196, 220–221).
Some of the earliest evidence for Minoan ritual comes from large, communal
built tombs that supply good evidence for an ancestor cult that endured for a
millennium or more in the eastern half of Crete in the EM and MM periods. In a
study of the stone tholos tombs (circular in plan with corbel-vaulted superstructures;
pl. tholoi) of the Mesara plain, Murphy (1998) makes a comprehensive case for
ancestor cult and shows that long-term funerary practices reflect broader develop-
ments in social and political organization. The tombs were built close to the
settlements that used them, typically situated in prominent places with commanding
views of the surrounding landscape, but mainly on land unsuitable for cultivation.
This locational pattern is ripe for a landscape archaeology interpretation (e.g.,
Buikstra and Charles 1999). The tombs were built for permanence using better
materials than were used for the settlements themselves. Many had pavements
attached to them, suitable for public rituals. Some were used for up to a thousand
years, though not necessarily continuously, as there were many demographic shifts,
and surely some of the social units (families, clans, etc.) that used individual tombs
became extinct from time to time. Such tombs could be reoccupied by newly arrived
groups to press claims to land and resources through fictive ancestral lineages. The
treatment of the dead is consistent with ancestor cult: earlier burials were moved
aside, but skulls and some long bones were carefully stacked in tombs or in
specially constructed antechambers that served as ossuaries. Obsidian blades found
with bones bearing cut marks were probably used to deflesh corpses, and signs of
burning may indicate periodic fumigation (Branigan 1987). Drinking, pouring, and
ritual vessels such as rhyta are common, and food remains, while rare, indicate that
feasting may sometimes have accompanied drinking ceremonies. Other stone
features are interpreted as altars and libation stones.
Citing ethnographic studies of death and society, Murphy (1998) interprets the
tholoi as territorial markers manifesting the claims of a living community to land
and resources through explicit links of descent from ancestors who occupied them in
the past (Saxe 1970). She explains the contents of the tombs in terms of a three-
phase funerary process of conversion of a living individual to an ancestor: burial,
with rites of separation; transition, a liminal phase in which the body exists but is
decomposing; and incorporation into the community of ancestors after the flesh has
disappeared (Van Gennep 1909). Each of these phases was accompanied by
communal rituals emphasizing continuity and stability, remembering and forgetting
(Hamilakis 1998).
Changes in the use of the tombs over time correspond to significant changes in
Minoan society. After an initial phase in EM I in which there is little sign of social
differentiation in the contents or uses of the tholoi, EM II witnessed a gradual
J Archaeol Res (2008) 16:83–161 111
increase in the placement of prestige goods and symbols of authority, including
seals, daggers, and gold and other precious metals, with certain of the deceased.
This process accelerated in EM III/MM IA in the run-up to the first palaces;
cemeteries then contained conspicuously larger tombs, and grave offerings included
exotic goods from foreign lands. Some bodies were placed individually in large
ceramic containers (larnakes), along with personal items such as seals and work
tools. Annex chambers were built onto the tombs as small cult rooms or ossuaries.
The ritual nature of some of these chambers is suggested by benches, stone bowls,
and stacks of the mass-produced conical cups that in later centuries became
essential components of Minoan ritual (Wiener 1986). Because of their small size
and difficulty of access, these chambers seem intended to exclude. Murphy (1998,p.
36) concludes that Late Prepalatial chiefs sought to ‘‘assert their control over local
resources and also to legitimize growing social disparities by manipulating the
rituals carried out at the tombs and by controlling access to the ancestors of the
Soles (2001) takes these conclusions one step further by linking the history of
funerary practice with patterns of political economy through the end the Bronze Age
and beyond. He characterizes the Minoans as ‘‘a very old, ancestor-worshipping
culture’’ (Soles 2001, p. 233), adding to the archaeological evidence cited above the
suggestion that the settlement pattern of small farms, towns, and country estates that
persisted through the Neopalatial period implies the existence of a large middle
class of free, land-owning families. Ancestor worship is characteristic of societies in
which the distribution and ownership of land are widespread, since land and
resources belonging to the ancestors must be preserved and passed down through
the generations. This factor perhaps had the effect of preserving ancestor worship,
even after the emergence of the palaces and the introduction of elite, inaccessible
rock-cut chamber tombs in the Neopalatial period. Soles attributes the demise of
ancestor worship on Crete instead to the imposition by the Mycenaeans of a feudal
society as recorded in the Linear B archives, in which ancestor cult lost its purpose
since landless peasants had little stake in the land and its resources. Soles’
impression of an egalitarian, ancestor-worshipping (almost utopian) Minoan society
brought to heel by an oppressive Mycenaean regime seems simplistic, but the
changes he documents are real and call for explanation.
Cult at peak sanctuaries and caves emerged at about the same time as the first
palaces, and there may be a connection between the establishment of powerful
central authorities and the activity at these new ritual spaces. The rural peak
sanctuaries offer an interesting contrast to the representations of deities and
worshippers in frescoes, sealstones, and rings. Typical finds at a peak sanctuary
include large numbers of clay anthropomorphic and zoomorphic figurines, clay and
stone offering tables, pottery of varying form and quality, and ash layers without
bones (Kyriakidis 2005). Despite the fact that obvious representations of deities are
absent among finds at Minoan peak sanctuaries, Minoan scholars have generally
tried to identify a deity or deities to which these shrines were dedicated. Peatfield
criticizes this ‘‘theistic premise,’’ noting that ‘If you assume that religion is
primarily about gods, then you are forced to go looking for them’’ (Peatfield 2001,
p. 54). Instead, he focuses on the unusual poses of the human figurines—with hand
112 J Archaeol Res (2008) 16:83–161
to head or chest, and apparently swaying bodies—to argue for an emphasis not on
gods but on bodily experience and the ecstatic performance of visionary epiphany,
divination, healing, or altered states of consciousness (see also German 2005;
Morris 2001; Morris and Peatfield 2002). Like the frescoes, these figurines depict
and memorialize enacted rituals, but of a very different type. A similar suggestion
has been made for Minoan cult places found deep in caves, based on comparable
figurines and other ritual objects, as well as the susceptibility to altered states that
may result from profound darkness and isolation (Tyree 2001). This emphasis on the
performative aspects of Minoan religion is informed by psychological studies and
ethnographic accounts of trances induced by narcotics, dancing, chanting, and
rhythmic sounds, as well as shamanic traditions of healing and divination (Goodman
1990; Lewis 2003; Price 2001). A general emphasis on theater and performance in
Neopalatial religion also has been inferred from the paucity of identifiable cult
buildings, the construction of open areas where communal gatherings could be held,
and the portable nature of cult equipment at that time (Rehak and Younger 2001,p.
In Neopalatial times, the palaces seem to have exercised unprecedented control
over many spheres of religious activity, including the peak and cave sanctuaries, as
part of a political strategy in which elites also may be religious officials (Adams
2004; Rehak and Younger 2001, pp. 439–440). Still, some scholars detect a
distinction between ‘‘official’’ and ‘‘popular’’ religion at the Neopalatial palace
centers (Gesell 2004). At Knossos, a cult area in the palace’s west wing yielded
faience ‘‘snake goddess’’ figurines with associated ritual equipment in luxury
materials, notably faience but also bone, ivory, rock crystal, and gold and silver foil.
These assemblages have been interpreted as the remains of elite worship of a
fertility goddess with a chthonic dimension (Jones 2001; Marinatos 1993), open
only to those with access to the palace’s innermost sanctum. Signs of coexistence
with popular religion can be seen at the palace centers of Phaistos and Malia, where
shrines in peripheral locations within the palace may represent a link with the wider
community. These shrines contain no objects in luxury materials and no
representations of the Minoan goddess. Instead, the cult objects, including clay
female figurines probably in attitudes of worship, stone libation tables and altars,
seashells, and pottery, are made from readily available and inexpensive materials
(Gesell 2004, pp. 132–133). This contrast underscores the restrictive nature of the
religious practices of palace elites.
In nonpalatial towns of the Neopalatial period, recurring ‘‘cult assemblages’
have been identified at a few sites, including Pseira (Betancourt 2001) and Kommos
(Shaw 2004). Ritual objects in these assemblages may include some or all of the
following: bull-shaped figurines, triton shells, offering tables, double axes, stone
chalices, and groups of rhyta (sing. rhyton, a perforated pouring vessel that was
widely used in Minoan cults to transfer liquids and pour libations: Koehl 2006). At
the small Minoan town of Pseira, where 60 buildings have been excavated,
Betancourt (2001) has identified three containing cult assemblages, which he terms
‘‘ rhyton hoards.’’ In one of these buildings, the ritual equipment was found in
storage, suggesting periodic ceremonies. In another, the House of the Rhyta, the cult
objects came from an upper floor with a carefully plastered and painted room, while
J Archaeol Res (2008) 16:83–161 113
the ground floor was occupied by a kitchen and a number of large storage jars. The
excavator’s inference of feasting associated with ritual ceremonies is strengthened
by a common spatial association of storage and food preparation with presumed cult
locations on Crete (Gesell 1985).
A fundamental reorientation in Cretan religious practice after the widespread
destruction of Neopalatial palaces, towns, and country houses at the end of LM IB
(c. 1450 B.C.) is one manifestation of a sharp break with the social and political
order of the Neopalatial period that affected almost all areas of art and culture
(Rehak and Younger 2001, pp. 441–464). Only the palace at Knossos survived
relatively unscathed into LM II, then possibly controlled by Mycenaeans as
indicated by the new Linear B-based administration and other changes. While this
would appear to suggest Mycenaean presence, the assumption of an invasion and
takeover from the mainland has been challenged (Preston 1999), and this has
provided scope for an interesting debate about ethnicity and identity in Final Palatial
and Postpalatial Crete (Brogan et al. 2002, p. 89). From a mortuary perspective,
Preston (2004) rejects an ethnic distinction between Mycenaean rulers and Minoan
subjects, arguing instead for conspicuous display of mixed cultural symbols, not for
the purpose of asserting ethnicity but rather as part of intraisland competition among
elites and a gradual ‘‘Mycenaeanization’’ that implies economic and cultural
influence but not political domination. An alternative interpretation is offered in
Burke’s (2005) reading of the famous Ayia Triada painted sarcophagus of the early
Postpalatial years (c. 1370–1360 B.C.), decorated on four sides with a complex cult
scene combining traditional Minoan images and symbols with contemporary
Mycenaean elements. For Burke, the sarcophagus exemplifies the strategic
appropriation of Minoan symbols by a Mycenaean elite ‘‘who were asserting
political, ideological, and economic dominance by means of art and architecture in
religious settings’’ (2005, p. 405). The excavators of Mochlos on the northern coast
of east Crete have found evidence for a foreign reoccupation early in the Final
Palatial period with close ties to Knossos and an apparent interest in reemerging
maritime trade (Brogan et al. 2002). The LM IB/LM II transition provides a
plausible case for reading dramatic changes in cultural identity and will continue to
be debated as new evidence emerges almost continuously.
The Shrine of the Double Axes at Knossos, founded near the end of the Final
Palatial (LM IIIIA2) just before the final destruction of the palace, shows the
transition between the old palace-based cult and a new, popular religion no longer
under palatial control (Gesell 2004). Among the features that prefigure Postpalatial
religion is a new kind of female terracotta figure, known as the Minoan Goddess
with Upraised Arms (MGUA), found displayed on a bench along with two terracotta
female figurines with hands on breasts, one terracotta male figurine holding a bird,
two pairs of horns of consecration, and a miniature double axe. This group, found
in situ, appears to depict a ritual ceremony in which votaries make offerings to a
goddess in front of traditional Minoan sacred symbols (Gesell 2004, p. 134). The
MGUA is thought to be the successor of the Minoan snake goddess, appropriated
and adopted as the standard cult image of the popular religion that spread across
Crete in the subsequent Postpalatial period (LM IIIB–IIIC).
114 J Archaeol Res (2008) 16:83–161
Recent discoveries of in situ or moderately disturbed shrines and ritual
assemblages have clarified the variability of Postpalatial religious practice, allowing
us to reconstruct the appearance and display of cult objects and furniture and the
locations of cult settings within settlements and houses (Eliopoulos 2004; Gesell
2004; Klein 2004;Shaw2004). Small public shrines have been found in at least
eight Postpalatial towns, typically near the edge of the habitation area but easily
accessible (Gesell 2004, pp. 135–143). Most have outdoor areas suitable for public
ritual, and the number of rooms in the cult complex varies from one to eight,
incorporating functions of display and ceremonial space, storage, and food
preparation. A distinctive ritual assemblage of ceramic objects (no luxury materials
were used) includes MGUAs, offering bowls on long tubular stands, and ceramic
plaques, some depicting ritual scenes, that were pierced with holes to hang on the
wall behind benches on which the cult objects were displayed. These objects recur
in sets and have been found together frequently enough that a typical ceremonial
room with benches can now be reconstructed (Gesell 2004, p. 143, fig. 7.14). Yet an
interesting variability in popular religion is suggested by the discovery of domestic
ritual spaces at Kommos, a harbor town on Crete’s southern coast (Shaw 2004). One
small shrine occupying the corner of a room was recovered largely in situ (Shaw
2004, fig. 10.6).Stacks of small vessels rested on either side of a small table made
of stone slabs. On the table were placed miniature spouted jugs and a small cup
containing pebbles and seashells. On the floor in front of the table, a triton shell and
more pebbles rested on a burned slab, and a pair of bowls full of ash were tucked
under the table. Braziers, ceramic containers possibly used to carry coals and burn
incense, were found in all of the shrine’s phases from LM II to LM III. In the
absence of characteristic ritual objects, the excavator speculates that the focus of
ritual may have been on sea, earth, and sky, the essential elements of nature.
Whether right or wrong, this interpretation highlights the ambiguity of identifying
cult in the archaeological record: in this case, ordinary objects take on a ritual
meaning because of their contextual associations. By and large, however, there
seems to be no rigorous or consistent methodology meant to test and potentially
falsify such claims. Such a method has been proposed by Kyriakidis (2005) for the
peak sanctuaries, building on work by Renfrew (1985) and others, which queries
first the ritual nature and then the religious content of the context or assemblage in
question. This framework could profitably be applied across the board to Aegean
Bronze Age religion.
Mycenaean ritual and religion
Mycenaean religion is harder to penetrate because there are fewer archaeological
contexts and artifacts that can be assigned unambiguously a religious function, and
because the Mycenaeans adopted Minoan religious iconography and cult objects but
apparently without harboring the same underlying meanings or beliefs (Ha
¨gg 1985,
1996; Soles 2001). The Linear B archives inform us of deities and religious
personnel, their lands and obligations, and appropriations for certain kinds of feasts
and religious activities, but the archaeological visibility of these aspects of religion
J Archaeol Res (2008) 16:83–161 115
is quite low. On the other hand, the places (e.g., bench shrines) and artifacts (e.g.,
figures, figurines, frescoes) to which we usually attach a religious function are not
directly attested in the texts. The one class of object that is found universally in
potentially cultic settings, the ceramic female figurine, appears in so many and such
diverse contexts (citadel cult centers, tombs, domestic contexts, dumps, fill) that its
status as a ritual indicator has been questioned (French 1972; Tzonou-Herbst 2002,
pp. 264–265); an alternative explanation is simply the universality of Mycenaean
ritual practice (Wright 1994). For example, at Tiryns, such figurines have been
found preferentially around hearths and doorways, suggesting a protective function
in everyday Mycenaean beliefs (Kilian 1988b). Even so, there have been important
strides made in studies of Mycenaean ritual and religion, through a combination of
new discoveries, innovative reanalysis of older materials, stronger theoretical and
comparative perspectives, and the kind of close integration of archaeology with the
Linear B texts that had been called for many years ago (Bennet 1988; see Lupack
1999; Shelmerdine 2001a, pp. 369–372, 380–381). Most of this new knowledge
involves the ritual equipment and expressions of a belief system about which we
still have little understanding (Ha
¨gg 1996, p. 600; Whittaker 1997, pp. 160–162),
but there has been some new work on the content of Mycenaean religious beliefs in
connection with a cult of the dead and belief in an afterlife (Gallou 2005).
The most salient current discussions on Mycenaean religion focus on the
variability of cult activity over time, within polities and across regions, and in the
range of cult places, ceremonies, and participants (Whittaker 2001). It is perhaps
easiest to give a general sense of the development of Mycenaean cult over time.
¨gg (1996, p. 611–612) proposes three phases that reflect variable Minoan
influence on an essentially mainland religion. The first phase, in the 16th century,
involved the importation or imitation of Minoan luxury and cult objects, yet with no
adoption of Minoan beliefs as cult objects were not used in the same way. At
Kynortion hill near Epidauros, we see a mainland cult merely embellished with
Minoan cult objects such as the double axe. The second phase, the 15th century, is
poorly known archaeologically, but Minoan influence on Mycenaean iconography
and cult material was strongest. This was a time of close relations between elites on
the mainland and their counterparts on Crete, and some Mycenaean elites may have
shared Minoan religious beliefs as part of aristocratic cults that had little to do with
the rest of the populace. The third phase is the Mycenaean Palatial period, when
Minoan elements gradually disappeared, except for the continued imitation of
Minoan cult symbols in iconography and the survival of certain Minoan objects
such as rhyta for pouring libations. In spite of the continued popularity of Minoan
religious symbols, Kontorli-Papadopoulou (1996, pp. 101–102) shows the devel-
opment of idiosyncratic mainland iconographic features, including warrior
goddesses, processions of females bearing gifts, lions and griffins arranged
antithetically or in repetitive lines, and a more explicit depiction of cult actions.
Thus, Mycenaean religion throughout the Late Bronze Age was Helladic, with a
superficial borrowing of Minoan and, to a lesser extent, Syrian and Egyptian
elements, showing influences from beyond the Aegean.
Several publications have appeared on various aspects of Mycenaean sanctuaries
and cult buildings (e.g., Albers 1994,2001; Moore and Taylour 1999; Whittaker
116 J Archaeol Res (2008) 16:83–161
1997; Wright 1994). Based on textual and archaeological evidence, each community
of any size established one or more sanctuaries, but few of these have been
recognized or investigated. The list of securely identified sanctuaries and cult places
is rather short, including the palaces and other intramural cult locations at Pylos,
Mycenae, and Tiryns; the small town sanctuaries at Phylakopi on the island of
Melos and Ayios Konstantinos on the Methana Peninsula; and the rural hill
sanctuary at Mt. Kynortion near Epidauros.
The most significant recent discovery is the Mycenaean sanctuary at Ayios
Konstantinos, overlooking the Saronic Gulf (Hamilakis, 2003; Hamilakis and
Konsolaki, 2004; Konsolaki-Yannopoulou, 1999,2001,2002,2003a). This
sanctuary is important for numerous reasons: its inconspicuous position within a
small, peripheral village; the in situ condition of the remains, which permits
chronology and ritual performance to be reconstituted; and the distinctiveness of the
cult objects, which show regional variability that cannot be characterized as a
chronological effect. The cult centered on the small Room A (4.3 92.6 m), whose
furnishings consisted of a floor of mixed earth and pebbles, a stepped bench in the
northwest corner opposite the entrance, a low platform along the south wall, a
podium in the center of the room, and a hearth in the southeast corner. The finds
date the use of the room to LH IIIA–LH IIIB (early 14th to late 13th century), i.e.,
squarely in the Palatial period. On and around the bench, excavators found more
than 150 terracotta figurines, tripod altar tables, pottery, and a triton shell similar to
those found in Minoan shrines. The corpus of figurines is unusual in that it consists
mainly of bovids (cattle and oxen) and horses, with several rare groups including
horses with helmeted riders, horses with chariot groups, and ridden and yoked oxen.
The standard Mycenaean female figurines that are so abundant elsewhere are
virtually absent. Other aspects of the sanctuary are well attested elsewhere,
however. Like most Mycenaean cult places outside the palaces, this sanctuary lacks
monumental construction or decorative elaboration. The pottery includes kylikes,
bowls, alabastra, and rhyta, all common ritual shapes. Certain structural features, a
stepped bench on which figurines were displayed, and platforms on the wall
opposite the bench and in the center of the room, probably served as attention-
focusing devices in the rituals and connect this sanctuary with others such as the
Temple in the Cult Centre at Mycenae. Of utmost significance is the hearth, which
was filled with ash and animal bones as well as scattered sherds from tripod cooking
pots. Analysis of the faunal remains revealed a predominance of burnt, juvenile pig
bones, with lesser representation of sheep and goat (Hamilakis 2003; Hamilakis and
Konsolaki 2004). The presence of all body parts suggests that these animals were
burnt offerings to the deity rather than meals roasted for human consumption. An
important distinction should be made between sacrifice, the ritual killing of an
animal followed by consumption of the meat, and burnt offering, where the focus is
on the destruction of the animal body, perhaps understood symbolically as having
been consumed by the deity (Hamilakis and Konsolaki 2004, p. 145). This is the first
evidence found in a primary use context for burnt animal offerings in Mycenaean
Greece, although we can infer animal sacrifice at Pylos (see below) and elsewhere
and the practice must have been widespread (Hamilakis and Konsolaki 2004,
p. 144).
J Archaeol Res (2008) 16:83–161 117
The features at Ayios Konstantinos that appear anomalous are difficult to assess,
since we possess few Mycenaean sanctuaries and thus do not know the true range of
variation. We do not know whether the sanctuary was autonomous, serving the
needs of a small rural community, or tethered to a regional center, such as
the recently discovered site at Galatas (Konsolaki-Yannopoulou 1999,2003b) or the
still poorly known Mycenaean settlement at Kolonna on Aegina. Ayios Konstan-
tinos may have been like one of the outlying communities to which the palaces sent
animals for sacrifices and feasting (Bennet 2001, p. 33; Dabney et al. 2004).
Hamilakis and Konsolaki construct a narrative for the performance of ritual at Ayios
Konstantinos involving the embodied sensory experience of food, drink, music (the
triton shell may have been used as a horn), and symbolic communication with
deities and ancestors through the sights and smells of burnt offerings. These
experiences, shared by a few members of society, might be translated to power and
authority in wider social arenas (Hamilakis and Konsolaki 2004, pp. 146–147). This
account, while appealing, shows speculation running well ahead of what we know
archaeologically about the site, the region, and Mycenaean sanctuaries generally.
Still less certain is the way that Mycenaean religion was organized and practiced
(for a good overview, see Shelmerdine 2001a, pp. 362–372). A distinction
commonly drawn between ‘‘official’’ and ‘‘popular’’ cult (Ha
¨gg 1996) has been
challenged as a false dichotomy (Albers 2001; Wright 1994), but the question of
where meaningful distinctions can be discerned in the range of cult practice
remains. Scholars have focused on other dichotomies such as rural/urban (Wright
1994, p. 60), public/private (Albers 2001, p. 132, n. 6), presence/absence of palatial
control, restricted access/open access (Wardle 2003, p. 317), or simply evolution of
cult over time (Wright 1994, p. 60) as alternative ways to think about this
variability. For example, Albers (1994, pp. 9–10) identifies five kinds of Mycenaean
cult settings: (1) the megaron and court of the palace; (2) ‘‘public communal’
sanctuaries in peripheral locations at palace centers and outside them; (3)
commoners’ houses and workplaces; (4) gate sanctuaries; and (5) house sanctuaries.
On the other hand, Wardle (2003, p. 317) perceives three groups on the basis of
access and location: (1) isolated sanctuaries with relatively unrestricted access; (2)
those associated with settlements but showing no sign of restricted access; and (3)
those within settlements to which access may have been restricted to an elite
The debate over the status of cult spaces within the palace centers highlights
problems of definition and interpretation. It is generally agreed that some kind of
cult centered on the core of the palace itself, architecturally defined as the megaron,
a highly elaborated form of the basic Mycenaean domestic unit: a long, linear
structure consisting of a porch, vestibule, and main room in which the throne and a
large hearth were installed. Following ideas first proposed by Kilian (1988a),
Wright (1994) developed the concept of a ‘‘hearth-wanax cult,’’ in which the king
(wanax) presided over a state cult that emphasized the symbolism of the hearth as
the center of the domestic sphere and the primacy of the ruler as father and chief.
Strong evidence for cult in the megaron comes from the palace at Pylos, where a
thematic fresco program shows a bull being led in a procession in the vestibule,
followed by scenes of banqueting and ritual toasting in the throne room. Next to the
118 J Archaeol Res (2008) 16:83–161
central hearth, excavators found an altar table and miniature kylikes thought to have
been used for libations. A plastered depression next to the throne also has been
interpreted as a receptacle for libations.
Other cult spaces within the palace centers have provoked more disagreement.
Wright (1994, pp. 60–61) draws a sharp contrast between the hearth-wanax cult and
the kind of peripheral sanctuaries or cult complexes he calls ‘‘citadel cult centers.’
The best-known example of the latter is the Cult Centre at Mycenae (French 2002,
pp. 84–92; Moore and Taylour 1999; Wardle 2003), an irregular agglomeration of
buildings arranged around a small courtyard housing cult rooms with altars,
platforms, benches, and frescoes as well as storage rooms and workshops. Wright
(1994, p. 61) argues that the Cult Centre at Mycenae (and other citadel cult centers
by extension) was of ‘‘lesser importance’’ than the palace cult, noting the vernacular
architecture, the lack of elaborate decoration or monumentality, the peripheral
location against the fortification wall at a remove from the palace, and the late
appearance in mature palatial times, suggesting that the rituals enacted there were
not essential to the early palace. Further, because the citadel cult centers survived
the demise of the palaces, they appear not to have involved the wanax or state-
sponsored religion. Albers (2001) responds that it is misleading to compare these
two manifestations of cult in this way, because they served quite different purposes
for the state. She calls the citadel cult centers ‘‘public communal sanctuaries,’’ and
the distinction she draws is between the megaron as cult place of the human ruler
and the peripheral sanctuaries as abodes of the deity, where they resided and were
accessible for human contact. Accordingly, both Albers (2001) and Whittaker
(1997,2001) define them as temples. For Albers, public communal cult was official,
organized by the palace administration and executed by priestly functionaries
according to a fixed annual schedule of religious festivals—an arena provided by the
king for public cult but restricted to a small, elite group who carried out ritual on
behalf of the entire community. This image of a highly restricted Cult Centre at
Mycenae may require reassessment, however, in light of new information that it
may have operated for several decades entirely accessible from the outer town
before the extension of the fortification wall cut off access to the complex from the
south after 1250 B.C. (LH IIIB2) (Wardle 2003). This may mean that public,
popular ritual took place in the courtyard, while access to the interiors of the small
cult buildings remained restricted.
The recognition of a ceremonial feasting deposit from faunal and artifactual
evidence at the small Mycenaean town of Tsoungiza northwest of Mycenae may
provide the rural counterpart to the public communal sanctuary (Dabney et al.
2004). This deposit contained a dump of head and foot bones from butchered cattle,
pig, and sheep/goat; ceramics dominated by plain vessels used for serving food and
drink; a fragmentary terracotta female figure; and a number of female and animal
figurines. The excavators interpret it as the deliberately discarded remains of a
regional feast intended to maintain political and economic alliances among elites
from several towns and villages. Such feasts may even have been provisioned by the
palace at Mycenae (Bennet 2001, p. 33). The presence of a terracotta female figure
and smaller female and animal figurines connects this assemblage with sanctuaries
such as the Cult Centre at Mycenae.
J Archaeol Res (2008) 16:83–161 119
Feasting in the Aegean Bronze Age has been the theme of recent archaeological
work that conspicuously integrates archaeological, environmental, and textual data.
Two important edited volumes appeared in 2004, The Mycenaean Feast (Wright
2004c) and Food, Cuisine and Society in Prehistoric Greece (Halstead and Barrett
2004). In these volumes, excellent survey chapters draw together the evidence from
ceramic and metal drinking sets and other feasting equipment; iconography of
feasting on frescoes, seals, and vessels; zooarchaeology; and Linear B archives (Day
and Wilson 2004; Halstead and Isaakidou 2004; Killen 2004; Palaima 2004; Rutter
2004; Wright 2004d, e). Some of this new scholarship considers the social
implications of palace-sponsored feasting: Bendall (2004) analyzes the distribution
of banqueting vessels at the palace at Pylos, concluding that although persons of
high and low status participated in and contributed to feasts, there was a strict
hierarchy of banqueting that reproduced and perpetuated social inequalities.
Perhaps the most remarkable example of integrative research on feasting
concerns the reconstruction of specific feasts involving animal sacrifice at Pylos
(Halstead and Isaakidou 2004, pp. 143–150; Isaakidou et al. 2002; Stocker and
Davis 2004). Archaeologists and faunal experts reexamined six sets of deliberately
deposited cattle bones from discrete locations around the palace dating to LH IIIB.
The bones, which consisted almost entirely of mandibles and leg joints, were burned
and showed signs of dismembering or filleting. One of these deposits representing at
least 10 head of cattle was found in Room 7, an archives office, along with 22
miniature kylikes, a spearhead and sword, and fragments of a large storage jar with a
number of Linear B tablets underneath them. Carl Blegen, the original excavator,
recognized these as the remains of sacrificial and votive offerings but was mystified
by their presence in an archive room. A recent analysis of the distribution and
content of the texts indicates that Room 7 was an office where an archivist revised
texts and monitored the flow of tablets that would later be archived in adjacent
Room 8 (Pluta 1996–1997). Several tablets from Room 7 concern provisioning of
animals for sacrifices and associated feasts. One of these, Un 718, describes the
offering of a bull to be made to Poseidon, possibly by the wanax himself along with
a military commander and others. The Ta tablet series, found elsewhere in the
palace, records banqueting equipment and sacrificial animals for a feast marking the
appointment of a new officeholder (Killen 1998). This particular feast was allotted
22 seats at 11 tables, matching exactly the number of miniature kylikes found in the
Room 7 deposit (Palaima 2000). Although this may be a coincidence, a plausible
scenario begins to emerge in which the miniature kylikes and other equipment were
brought to Room 7 to be processed before returning to storage, along with the cattle
bones as proof of completion of the ritual feast. This interpretation accords well
with studies of the flow of information among the various offices and workshops at
the palace, as well as the meticulous recording of activities in areas of palatial
interest (e.g., Shelmerdine 1998,1999). The contextual information suggests that
the Room 7 deposit was still being processed when the palace was destroyed and
that sacrifices were made to Poseidon to the very end (Palaima 1995). The case
of Pylos Room 7 shows how we may recognize the archaeological consequences of
particular feasts (Dietler and Hayden 2001, pp. 8–9), and illustrates the potential of
true interdisciplinary research.
120 J Archaeol Res (2008) 16:83–161
The study of feasting leads to the recognition of differences in Minoan and
Mycenaean ritual practice that may cautiously be generalized to contrasts in religion
and in social and political organization (Borgna 2004). Minoan feasting, whether in
connection to funerary rites, religious ceremonies, or other rituals, focused more on
community unity and identity than on individual aggrandizement. The Mycenaeans,
on the other hand, feasted in smaller, restricted areas where interelite generosity and
hospitality were emphasized in reciprocal transactions among individuals. Minoan
feasts presented an ideology of community solidarity, but elites displayed their
status through privileged roles in the ritual and in separate, exclusive celebrations.
The Mycenaean palaces organized many feasts, presenting an ideology of equality
but always separating elites from the lower strata of society through unequal
displays of conspicuous consumption. The provision of Mycenaean feasts contained
a patronizing element: as Bendall (2004) proposes, one’s social status was clearly
marked at a feast by the location, equipment, and companions one was assigned.
Though nonelite individuals were included in regional banqueting, the intent seems
to have been to underscore the superiority of the palace through lavish contributions
of food and drink that were beyond the means of small communities and common
people. For Bendall (2004, p. 128), to participate in a Mycenaean banquet was to
accept and perpetuate the status quo of hierarchical inequality, a grim ‘‘bread of
servitude.’’ Few would take such a bleak view, and many would characterize
Bendall’s interpretation as a misunderstanding of gift exchange theory (Nakassis,
personal communication 2006), but this is a good example of a locus of interaction
where the state is present in a material and symbolic way in the lives of common
people in the hinterland. Borgna (2004, pp. 146–147) invokes dual-processual
theory to assert contrasting pathways to power—the Minoans pursuing corporate
strategies to structure and constrain social action, and the Mycenaeans effecting
social exclusion through network strategies—though she acknowledges that the
reality is far more complex.
Archaeological science
Aegean prehistorians have had a long and fruitful relationship with the natural and
physical sciences, starting with Heinrich Schliemann’s first excavations at Troy in
the 1870s, where he pioneered a multidisciplinary approach by incorporating
geography, geology, cartography, meteorology, ethnology, anthropology, botany,
photography, and technical analyses of metal (Runnels 2002, p. 8). That said, the
application of rigorous science in Greek Bronze Age archaeology has been uneven
at best, but there have been significant advances and greater consistency on all
fronts in the last decade. Much of the impetus for collaboration with nonsocial
scientists has come first from regional archaeological projects and their emphases on
environmental resources and surface geomorphology, and second from big
questions that require scientific, often archaeometric, input, such as the sourcing
of metals and ceramics to investigate trade networks and technologies. The typical
regional-scale project now includes geology and geomorphology, climate studies,
remote sensing and geophysics, and GIS as complements to traditional methods of
J Archaeol Res (2008) 16:83–161 121
excavation, survey, ethnoarchaeology, and artifact analysis. Many Aegean regional
projects are represented in the five-volume publication of the European POPULUS
project, The Archaeology of Mediterranean Landscapes (Bintliff and Sbonias 1999;
Francovich et al. 2000; Gillings et al. 2000; Leveau et al. 1999; Pasquinucci and
´ment 2000), which presents case studies of methods in landscape archaeology
with the general aim of establishing a set of ‘‘best practices’’—although many
would regard this as an unrealistic or even undesirable goal (Blanton 2001). Among
the topics discussed are demography, environmental reconstruction, GIS, remote
sensing and geophysics, geochemistry, geomorphology, and the interpretation of
surface artifact material. A good overview of current Aegean projects with a strong
scientific component may be found in Metron: Measuring the Aegean Bronze Age in
the Aegaeum series (Foster and Laffineur 2003).
Environmental studies
Environmental studies have been fundamental to contextualizing human societies,
and as practiced in Aegean regional studies typically involve geomorphology,
paleoclimate, zooarchaeology, and archaeobotany (e.g., Davis et al. 1997; Given
and Knapp 2003; Wiseman and Zachos 2003; Zangger et al. 1997). A sampling of
current environmental research, with a focus on soils and botanical evidence, may
be found in Landscape and Land Use in Postglacial Greece (Halstead and Frederick
2000). In addition to case studies describing methods and results of paleoenviron-
mental reconstruction, this volume takes on some thorny problems: the difficulty of
correlating the effects of climate and human activity in chronological and causal
terms, and the extent to which humans have been responsible for episodes of
environmental degradation such as landscape destabilization resulting in erosion
and catastrophic soil loss. With palynological and geoarchaeological evidence that
often can be only roughly dated, it is difficult enough to establish contemporaneity
with specific, well-dated human occupations, let alone causality (Halstead 2000, pp.
118–121). For this reason, debate continues on the prevalence of human agency in
episodes of destabilization that are often observed in the paleoenvironmental record.
In the Argolid, regional surveys of the 1970s–1990s generated a large body of
geological and archaeological data that seemed to indicate human agency in certain
episodes of massive Holocene soil erosion (Runnels 1995,2000; van Andel et al.
1986,1990; Zangger 1994). Two of these were attributed to careless slope clearance
by farmers and another to widespread grazing and the collapse of agricultural
terraces. This interpretation has been criticized for what some see as poorly dated
sequences that leave ambiguous the causal relationship between humans and
episodes of landscape destabilization, and for the extrapolation from a few samples
to broad, regional patterns (Bintliff 1992; Butzer 2005; Endfield 1997; Moody 1997,
2000; Whitelaw 2000). Others, citing ethnographic and ethnohistorical evidence,
argue that traditionally, Greek farmers and pastoralists who cannot afford to
overexploit or otherwise endanger their resources put in place informal, self-
regulating taboos and constraints (Forbes 2000; Koster 1997). This debate is not
easily resolved, but there is a clear need to build local, well-dated, and correlated
122 J Archaeol Res (2008) 16:83–161
sequences (e.g., Krahtopoulou 2000) as a prelude to making wider inferences about
agency in regional land–human relationships.
In spite of these problems, enormous resources are expended in flotation for the
recovery of botanical and microfaunal material (Megaloudi 2006). An excellent
recent example is the use of recovered animal, fish, plant, and stone remains to
reconstruct the contents and use of a kitchen at pre-eruption Akrotiri (Birtacha et al.
in press). Organic residue analysis by gas chromatography-mass spectrometry
figures prominently in investigations of diet and trade. Recently, a major exhibition
and accompanying book, Mycenaeans and Minoans: Flavours of Their Time
(Tzedakis and Martlew 1999), brought together evidence from residue analysis of
pottery; plant, faunal, and molluscan remains from excavations; iconographic and
written sources; and isotopic and other studies of human remains to describe Aegean
Bronze Age cuisine. At Neopalatial Mochlos, residues from large vessels in the Vat
Room of Building C.7 are part of a persuasive argument for an industry in perfumed
oils and unguents (Koh 2006).
Geomorphology has figured prominently in reconstructions of Bronze Age
coastlines and harbors (Besonen et al. 2003; Jing and Rapp 2003; Rothaus et al.
2003; Zangger et al. 1997). In Messenia, Zangger and colleagues identified an
artificial rectangular basin that they believe served as a protected port for the palace
at Pylos (Shelmerdine 2001a, p. 339; Zangger et al. 1997, pp. 619–623). To keep
the port free of sediment, Mycenaean engineers excavated a small lake and a
channel by which the Selas River could be partially diverted to the harbor, while the
sediment carried by the river was trapped in the artificial lake. This reconstruction
has met with some skepticism, but a project of this scale and sophistication was well
within the impressive engineering skills of the Mycenaeans, who built monumental
tombs, fortifications, roads, bridges, and dams and drained the vast Lake Kopais in
central Greece to reclaim agricultural land (Knauss 2001; Loader 1998). Another
coastline reconstruction project employed sedimentological analysis of dozens of
geological cores to contextualize a possible Mycenaean colony on the Ionian
seacoast in Epirus (Besonen et al. 2003). The reconstruction restores a broad bay
that extended inland some 6 km or more in the Bronze Age, placing the Mycenaean
settlement in a strategic position overlooking a sheltered harbor.
Geomorphologists have become close partners in regional survey archaeology.
Some surveys have begun to attach a geomorphologist to each survey team for real-
time interdisciplinary consultation (Given et al. 2002; Tartaron et al. 2006a, pp.
468–470). Other creative applications of geomorphology have originated from the
need to study surface sites when there is no recourse to excavation. The Laconia
Rural Sites Project investigated 20 small, rural surface sites by combining
controlled artifact collections, geophysical prospection, and soil chemistry of 20-cm
auger cores (Cavanagh et al. 1996,2004; Mee and James 2000). The soil from the
cores was analyzed for a range of properties, including color, texture, consistency,
artifact and ecofact content, organic carbon, mineral magnetic properties, and
J Archaeol Res (2008) 16:83–161 123
elemental content of phosphorus, copper, lead, calcium, and potassium. Among the
many interesting results was a strong positive correlation between artifact
distribution/density and most of these soil properties. This correspondence suggests
that these markers of human activity have survived together for thousands of years,
with little evidence for widespread erosion, even at sites on sloping terrain. In
another case, geomorphologists and archaeologists collaborated to provide evidence
for dating partially collapsed limestone architecture (Tartaron et al. 2006b).
Geomorphologists constructed a relative chronology based on the progressive
development of karstic dissolution features in the stone, with two broad phases in
antiquity. Controlled artifact collections demonstrated a clear association of Early
Helladic (EH) II artifacts with the older phase architecture, allowing the plan of a
fortified coastal site of that period to be recognized.
Geophysical prospection is widely practiced on both excavations and surveys, as
improvements in instrumentation, electronics, and data processing have increased
coverage per unit time, with greatly enhanced reliability, precision, and visualiza-
tion (Kvamme 2003). In a thorough review of techniques and the history of
geophysical prospection in the Mediterranean, Sarris and Jones (2000) emphasize
both the research value of geophysical techniques and their potential as rapid,
nondestructive means to gather information on the nature and structure of
archaeological sites and features endangered by modern development. A dramatic
increase in interest began around 1990, and now much of the geophysical work is
carried out by specialized Greek research units such as that directed by Sarris at the
Institute for Mediterranean Studies at Rethymnon, Crete, which publishes its work
annually in Archaio-telepiskopika Nea (in Greek and English). These publications
show the power of integrated geophysical survey using several different instruments
(Vafidis et al. 2005) and illustrate many superb examples of GIS imaging of
integrated geophysical, archaeological, and environmental data. To mention just one
project relevant to the Greek Bronze Age, at the spectacular Mycenaean center at
Dimini, geophysical mapping of 29,000 m
using magnetic, electromagnetic, and
soil resistance methods revealed the two main megaron complexes and numerous
other structures, which have been confirmed by ongoing excavations (Sarris 2002).
Geographic information systems
As everywhere, there has been an explosion in the use of GIS to organize and
process the massive amount of spatial information that any project generates.
Whereas until recently GIS had been used mainly to analyze data retroactively
(Gillings 2000), most new surveys now come to the field armed with complex
‘archaeological knowledge systems,’’ including multiple GIS layers of environ-
mental and cultural data linked via databases to the new data generated in the field
(for links to online examples, Gates et al. 2004). The proliferation of GIS in Aegean
124 J Archaeol Res (2008) 16:83–161
prehistory has been facilitated by the increasingly user-friendly interface of off-the-
shelf GIS packages such as ESRI’s ArcGIS, as well as the wider availability of
inexpensive satellite imagery and digital terrain (or elevation