Bryn Mawr Classical Review

Bryn Mawr Classical Review 2007.03.13

Donald C. Haggis, Kavousi I: The Archaeological Survey of the Kavousi Region. Prehistory Monographs 16.   Philadelphia, PA:  INSTAP Academic Press, 2005.  Pp. 165; figs. 69, pls. 39.  ISBN 1-931534-18-7.  $80.00.  

Contributors: J.T. Ammons, P.M. Day, J.E. Foss, L. Joyner, E. Kiriatzi, M.S. Mook, M.W. Morris, M. Relaki, M.E. Timpson. Edited by G.C. Gesell and L.P. Day

Reviewed by Saro Wallace, University of Reading (
Word count: 2917 words

An impressive publication from INSTAP Academic Press, the first of the Kavousi series, gives us much to look forward to. Haggis' (hereafter H.) dissertation was completed in 1992, but the intervening years have lent polish to the text and expanded the illustrations. Important information is still emerging from the area, not least from the author's excavations at Azoria,1 and has been referred to here, bringing the study up to date.

The Kavousi excavation focused on the sites of Kastro and Vronda in east Crete, selected because of their Early Iron Age remains. The survey around the sites covered a strategic part of the island -- the north part of the narrow Ierapetra isthmus linking the main arable area of central Crete to the far east: the coasts of this area form the south part of the large, sheltered Mirabello bay. The boundaries were drawn to include multiple environmental zones, including mountains c. 1000m asl, water-rich foothills and both inland and coastal plains. After short prefaces by the editors of the series (directors of the Kavousi project) and by H. himself, the book comprises an overview of the study's background, a description of the physical landscape, details of the survey's method and findings, and a detailed commentary on the pottery, followed by a synthesizing history of settlement. Overarching conclusions form a small section at the end.

Though the survey is presented in the general preface as a background to the excavations, it seems always to have looked well beyond this remit. It presents one of the best selections of surface ceramics so far published from Crete, used to date and evaluate 101 sites of different periods from the Neolithic to the Roman. The advance this represents cannot be overstated. Considering Crete's dense survey coverage, little surface material has been published, restricting comparison with excavated assemblages. The analysis of surface ceramics here draws on a good understanding of regional pottery styles and production technologies obtained from dense excavation coverage in the wider region -- not only at the Kavousi sites but others like Gournia, Myrtos, Kalo Chorio and Mochlos, where the author has first-hand experience with excavated material. Close integration of the data allows relatively precise chronological categories to be used, though broad by comparison with excavation sequence phases -- e.g. LM IIIA-B, MMIII-LMIB, EM I-II. Some still seem too narrow for the realities of this survey: for example H. admits in several places that LM I material is difficult to separate from LM IIIA-B in surface contexts, while still divorcing the phases in synthetic maps.

The study is progressive in pursuing the systematic use of fabric analysis on coarse wares, which dominate survey assemblages, to help date pottery and analyze exchange patterns. The synthetic pottery discussion correlates fabrics to shapes by period and is supported by a macroscopic fabric typology by Mook (all periods) and a petrographic study by Day and Kiriatzi (Neolithic-EM periods only). To date, many Cretan excavators have chosen to minimize attention to their coarse wares, or focused on contexts where they are absent such as tombs (particularly for the EIA): this study shows their great value in writing history and must surely convince excavators to publish and discuss much more coarse material.

The findings have strong parallels in other Cretan surveys, but local differences show up as well. Sites, often of limited size (the largest Bronze Age examples do not seem to exceed 3ha) tend to cluster closely together, making the political independence of any one site difficult to ascertain, and suggesting that basic political units often subsumed several settlements, especially in periods when settlement densities are high, as in Proto- and Neopalatial. The clusters focus on the most workable and moisture-retaining phyllite soils, and many have coastal/plain topography. Yet the foothill zone to the east of the survey area (also with phyllite soils and ranging up to 400m asl) periodically comes into use for settlement, thanks to its springs, attractive phyllite soils, and naturally defended locations.

First settlement evidence comes from the final Neolithic period, paralleled in other parts of Crete, where an apparently significant increase and spread of population took place. Six sites were retrieved by the survey or revealed by excavations within its boundaries. The fact that FN material so often comes from subsoil contexts suggests to Haggis heavy alluviation in parts of the region soon after this (perhaps connected to extended cultivation from the EM period) and leaves open the possibility of earlier Neolithic habitation. Though most of these predominantly coastal sites (including some, like Chrysokamino, clearly connected to off-island trade) continue in use into EM, Azoria shows that a focus on inland defensible hilltop sites for settlement in FN alone also applies here. H. stresses the ceramic grounds for dividing FN clearly from EM, rather than assuming a delayed advent of EM styles or influences to the east from central Crete producing a long transition (the idea of delay or isolation is echoed in treatments of LM II and EIA pottery). In fact, there is more and more evidence that the FN period (though the characteristics and absolute chronology of FN pottery phase are disputed) marks a major historical horizon across Crete.

EM I-II sees more extensive use of coastal valleys and plains for settlement, with a larger number of small village-sized sites. By EM III a major unit seems to have developed near the coast in the Tholos plain, with smaller villages scattered in other good arable areas as around Chomatas/Chordakia. In Protopalatial we sees the emergence of a number of notably large sites, including inland examples, within a more complex hierarchy of smaller rural units. Some built on existing settlements, as at Tholos, but others seem new foundations. It is worth noting that many of these sites continue into Neopalatial, and thus their size during their earliest use cannot be assessed. But the explosion in numbers of sites, and their variety of types, definitely suggest a larger population in newly complex relationships. There is much greater spread of settlement into the foothill zone on the east: some of the sites here. such as Chondrvolakes 68, seem to resemble the fortified farm- or guard-houses seen elsewhere in central and eastern Crete, perhaps marking borders for emerging palatial territories. A cluster of sites (9-14) above sheer cliffs on the coast at Mt. Chomatas may show concern with defence for larger settlement groups: Monastiraki Katalimata (site 100) is another important case of this type.

Interestingly, this complex system emerges in a region outside the direct ambit of any known palatial centre, though the area of modern Ierapetra may have housed a palace, as H. notes, and Gournia probably acted as a centre for a large region. The lack of any similarly large settlement in the Kavousi area suggests even the largest Protopalatial sites functioned as subcenters within a territory centered elsewhere, perhaps ending somewhere in the wall of hills to the east. The petrographic studies, however, do show that trade and other connections between the Gournia/Mirabello area and the east Cretan peninsula operated through or passed around the region during the period.

The Neopalatial period sees continuation of the major settlement nodes, apparently with some degree of nucleation. The most important cases appear at Ayios Antonios, Tholos and Chordakia, on or at the fringes of the lowland plains. The foothill zone is not widely used: the lack of concern with defensibility is paralleled in other parts of the island through most of this period. H. notes a startling degree of contraction to the main settlement nodes in LM IIIA-B. While this may reflect real concentration of power at fewer centers in a changing administrative and political environment, I doubt there were as few small farmhouse-type sites in LM III A-B as he suggests: those which are noted often have long term use from at least Protopalatial, indicating a basic stability of infrastructure and highlighting potential problems in dating material. One reason for doubting real abandonment of the area at this time is its fairly rich population in both LM I and LM IIIC. The gap seems too short and the demands for landscape infrastructure from the redistributive systems which continued to operate in Crete through LM IIIA-B too great for rural abandonment to have occurred. Again nearby Gournia, as a major LM IIIA-B nucleation, may distort the picture.

For LM IIIC, while H. contributes a detailed discussion of the move to Kastro and Vronda, an equally important contribution is the documentation of small sites in the Avgo valley and at Panagia Skali, which enlarges our view of settlement at this period: the partly defensive site pattern in the foothill/mountain zones was supplemented by fieldhouses and small hamlets in inland and protected positions. By the end of the EIA significant nucleation had occurred at Azoria, and Haggis assesses the relationships of that settlement with the rest of the Ierapetra isthmus, where several other large polities existed. The Vrokastro survey suggests a return to residence in the coastal zones occurred at least by early Archaic, c. 700 BC, as security risks lessened and coastal trade picked up. In the Kavousi region, however, neither coastal nor small-scale rural settlement seems very important in the region during the period that Azoria flourished. This contrasts starkly with the record around other polis sites such as Phaistos, perhaps due to differences in each polity's scale of operation as a regional center and/or in the relative intensity of agriculture in each region.

A process of conflict and further nucleation, perhaps targeted on the polis site of Kato Chorio Profitis Elias, south of the survey region, seems likely to have led to Azoria's abandonment by the fifth century. By Roman times, when Hierapytna controlled the whole region. the coast was certainly in use at an industrial scale for grain and olive oil exports: H. identifies the presence of warehouses at the long-rooted coastal site of Tholos, and a number of small settlements reappear on the phyllite soils during the first century BC.

H. weaves pottery, as an essential dating and interpretative tool, closely into the fabric of his argument throughout the text in a way alien to some earlier Aegean surveys: we never lose touch with the primary data or find that it has been 'farmed out' by period to specialists, with disjointed results. The familiarity shown with ceramics of all periods studied and a wide range of ceramic products should be emulated by all serious students of archaeological landscapes. Notwithstanding, specialist studies enrich the findings greatly. Fabric studies support H.'s model of clustered settlements being linked by kin, economic and political bonds in several periods, including the Protopalatial and EIA, showing closely shared fabric recipes. Petrography shows the close interaction from Early Minoan onwards between the Kavousi area and the west Mirabello as well as the Myrtos/Malia zone. Neopalatial fabrics show considerable standardisation in tempers used and harder and more consistent firing, reflecting a more centralised context of production and distribution.

Geomorphological and soils studies by Morris, Timpson, Ammons and Foss contribute vitally to the discussion of subsistence/settlement relations. They show that major colluvial deposition in the kampos occurred at several points including pre-LBA (reflecting land disturbance by expanding cultivation) and the post-EIA period (when it may reflect some abandonment of cultivation as settlement moved away from foothill sites); the extension of agriculture under Roman rule and its subsequent retraction seems to have had the same effect. All the site data were collected by a single field walker dividing the landscape into the most practicable and meaningful 'fields', then walking linear 'passes' c. 20-25 metres apart, recording visual observations and sherd counts for each pass. Sites were then revisited for systematic collection of material in measured units, using a set of passes again, rather than a grid. Within larger and more complex sites individual loci are numbered and plotted, rather than identifying these areas as separate sites, which has often led to interpretative confusion. The study demonstrates how small-team or single-person projects can offer a deep, seamless, organic understanding of a region which is often superior to the more mechanistic, disjointed coverage by large surveys.

H.'s discussion of historical and ethnographic evidence adds a comparative dimension helpful in understanding settlement dynamics, particularly their relationship to subsistence practice. For example, understanding of the settlement clustering phenomenon is developed through study of farming settlements traditionally clustered in the best arable areas. Data on the historically low productivity of the Kavousi plain until the advent of modern deep wells, traditionally deterring extensive settlement in favor of better-watered upland valleys, helps in understanding long-term choices in settlement patterns. Yet the focus of most prehistoric settlement does seem to be in the plain itself, and the exceptional outreach into the hills in Protopalatial and LM IIIC needs politically-centered explanation: in general H. pays little attention when describing use of this zone to individual site types and characteristics, rather than use of the zone as a whole. His understanding of recent land-use equips H. well to observe small pieces of information on ancient landscape infrastructure - large walls in the Tholos plain are suggested to have been Bronze Age river dams; a MM date is suggested for terraces at Chondrovolakes based on sherds found there, and he notes EIA tombs at Aloni are also built on terraces, both suggesting Bronze Age investment in this mountainous part of the landscape.

The sites are presented in a 60-page gazetteer and a series of maps and plans, some at large scales for densely settled areas. These show the approximate size of each site as a shaded polygon. Size is not, however, always able to be accurately estimated by period, making the maps potentially misleading if used in isolation. The use of toponyms alongside numbers in the site descriptions makes site checking easy; the book's index usefully includes all toponyms. 29 of the 69 full-page figures are dedicated to closely-spaced (but absolutely clear) artefact drawings, allowing a very large amount of sherd material to be published. The plates include landscape photographs, vital to understanding the sites in their setting, photomicrographs of pottery thin sections and (even more useful than the latter for the general archaeological reader), simple magnified views of coarse ware fabric surfaces to illustrate a typology of fabrics. These greatly aid the comparison and dating by fabric of sherds lacking other diagnostic features by scholars working throughout the region. If the means had been found to reproduce these in color, as the recent selection published by Moody et al.,2 they would have formed an even better reference.

Reference to the artefacts illustrated from each site is facilitated by the use of site numbers as the main identifiers for each artefact. H. also provides a concordance table linking these numbers (and the original site numbers used in the fieldwork) with the sherds' catalog numbers. In an attempt to reduce monotony, the typology of 18 major pottery fabric types described at length in Mook's text is used for easy reference throughout the pottery discussion and catalogue. This system can be awkward to use at first and raises interpretative questions about truly identical fabric manufactures over long periods or between sites: the use of this system means we are not enabled as readers to pick up on small variations in fabric between sites. Nonetheless, the categorization has clearly proved well-grounded, matching other diagnostic dating features. Attempts at 'easy-reference' classification are also rather awkward when topography is discussed. Descriptions of site locations in the gazetteer are rather sterile -- I think a more vivid, characterful word-picture is needed to allow the reader to really visualize these landscapes.

The environmentally determinist assumptions informing some 1970s-80s Aegean survey are absent here. The book also shows that regional studies offer complex layers of historical insight in their own right, rather than functioning just as 'microcosms' within generalizing models. H. constantly highlights the potential of the natural landscape to condition settlement location while showing, through a sensitive documentation of changing settlement balance over time, the role of agent choice and decision-making under different historical conditions. Among other issues, we are provoked by the study to consider how suddenly complexity can emerge, how differently political impetus and subsistence considerations can drive the use of mountainous environments; how maritime proximity competes with arable land access over time in settlement considerations; how parallel pottery technologies and common sourcing of ceramics across a region can reinforce political and social connections and thereby drive regional history; how cult sites, defense sites and farming satellites relate to large settlement nuclei but also adapt to use particular natural landforms or soils. This history of the particular expands and changes many generalist models of Bronze Age Crete and regions beyond.

The authoritative manner in which a wide range of evidence types is treated here at primary level encourages us to treat the author's interpretations with respect. But the latter are never thrust very dynamically at us, even in the sections dedicated to synthetic comment. Instead, arguments are often constructed obliquely, with a selective piling-up of observations behind one general point. This unemphatic presentation can produce a bland effect, making important conclusions difficult to pick out. Despite H.s clear ability to look beyond his region to wider patterns, the text has only a modest amount of comparative discussion, which can again lead to a bland or flat impression. This reticence seems deliberate: the author's main aim is to offer a clear primary level of analysis of the data, leaving further discussion for future papers by the author as well as other scholars.3


1.   D. C. Haggis, M. S. Mook, C. M. Scarry, L. M. Snyder, and W. C. West, 2004, "Report on the Azoria excavations 2002-3." Hesperia 73 (2004): 339-400; B. Hayden (with contributions by H. Dierckx, G. W. M. Harrison, O. Rackham, J. Moody, G. Postma, A. B. Stallsmith), Reports on the Vrokastro area, Eastern Crete. Volume 2. The settlement history of the Vrokastro area and related studies (Philadelphia: University of Pennsylvania Museum of Archaeology and Anthropology, 2004); D. Haggis (with contributions by H. Dierckx, M. Hahn, G. W. M. Harrison, O. Rackham, J. Moody, O. Postma, M. K. Risser, A. B. Stallsmith), Reports on the Vrokastro area, eastern Crete. Volume 3: The Vrokastro Regional Survey Project: sites and pottery (Philadelphia: University of Pennsylvania Museum of Archaeology and Anthropology, 2005); L. V. Watrous, L. V., et al, forthcoming, The Gournia Survey .
2.   J. A. Moody, H. L. Robinson, J. Francis, L. Nixon and L. Wilson, "Ceramic fabric analysis and survey archaeology: the Sfakia survey, BSA 98 (2000): 37-105.
3.   Small point noted: I could not find Kamara tou Tholou (site 26) on the map of LM IIIA-B sites.

Read Latest
Index for 2007
Change Greek Display
Books Available for Review

HTML generated at 13:31:42, Friday, 03 April 2009