KEROS ISLAND SURVEY - 2013

Keros Island Survey. C. Renfrew (Cambridge), M. Marthari (Director Emerita, ΚΑ’ ΕΠΚΑ) and E. Dellaporta (Director, 2nd EBA) report on the second and final season of intensive survey. In addition to walking a further 445 tracts, including some of the more remote and mountainous parts of the island, intensive collection was carried out at 25 areas of interest, or ‘polygons’ (complete collection was carried out in 5m² circles, with diagnostic ceramics and other special finds collected in 100m² squares). The survey overall has now defined and examined 26 polygons (Fig.), concentrated mainly in the northwest coastal region but also found in most parts of the island apart from the furthest eastern areas. These represent areas of potential intensive, diachronic use over some 6.4% of the island’s surface, with less intensive use, including agricultural use, documented by wider scatters and terracing. Most larger polygons exhibit use during different time periods, and the definition of sites within polygons during specific periods is a major aim of the forthcoming synthesis. The Early Bronze Age and Roman periods saw the widest activity, with more concentrated zones of activity in other periods. Late Bronze and Early Iron Age occupation is hardly attested, and Classical activity was strongly concentrated in areas such as Gerani and Megalo Kastro. The very low quantity of Hellenistic pottery is another notable feature of the data.

A total of 3592 chipped stone artefacts was collected: 99% are of Melian obsidian. The remaining materials include the distinctive white-spotted obsidian from Giali, local medium-quality grey-black chert, and three or four possible examples of imported cherts. The earliest material is a group of 18 Late Neolithic arrowheads (Fig..), many tanged and with good parallels from Saliagos (Antiparos), Ftelia (Mykonos) and Strofilas (Andros). These were mainly isolated finds, often in the hilly hinterland, probably where they fell during the pursuit of game. Since no other finds are so early, it seems unlikely that Keros was settled during the Late Neolithic, but was rather visited for hunting, perhaps by populations from Naxos or Amorgos where contemporary settlements existed at Zas Cave and Minoa respectively. The earliest evidence for settlement on Keros may date to the Final Neolithic period. Assemblages with percussion blades (i.e. technically comparable to the Final Neolithic material from Kephala on Kea) come from two or three areas in the interior of the island at a relatively high elevation. Most of the chipped stone material dates to the Bronze Age, and mainly postdates Early Bronze Age I on the basis of the pressure blade dominated assemblages. A number of locations along the north coast (including polygons 1, 2, 4, and 7) have evidence for production in the form of cores and knapping debris; local populations were not, therefore, reliant on obsidian workers based at Dhaskalio. The latest diagnostic finds in the chipped stone assemblage are two obsidian hollow-based points, a distinctive type associated with the Middle Helladic southern Greek mainland. Rare in a Cycladic context, these may represent a new form of prestige good or practice imported from the mainland, perhaps alongside the Grey Minyan recognised in the ceramic assemblage.

The earliest pottery so far identified corresponds to the Grotta-Pelos culture of the Cyclades, which potentially stretches into the end of the Late Neolithic (Final Neolithic elsewhere). No ceramics of canonical Late Neolithic date have been identified for consideration with the Neolithic obsidian projectiles. Pre-Dhaskalio activity or occupation on the island is tentatively suggested by four index types (cheese pot, rolled rim bowl, heavy burnished ware and organic/grog tempered fabric). Another 11 index types provide substantial evidence for occupation between late Early Cycladic I and Early Cycladic III (Fig..): several can be tied to occupation contemporary with Dhaskalio Phases A-C, particularly the radiating slashed pithoid jar handles, kerbschnitt-decorated hat-shaped vessels, and sauceboat fragments. There is also a poorly preserved frying pan fragment with concentric stamped circle decoration on one surface. Special finds include fragments of Early Bronze Age marble figurines of both the schematic and folded-arm types, and several pieces of marble bowls.

Evidence of post-Dhaskalio Bronze Age occupation on Keros is provided by a further five index types, with a significant number of tripod legs from Minoan style cooking vessels. Grey Minyan sherds, potentially from the very end of the Middle Helladic period, and a lone bridge spout from a jug or jar have also been identified. It is unclear as yet how broad a period of activity is indicated by these sherds. Potential Late Helladic or Late Minoan sherds are so few and fragmentary as to suggest that Keros was largely uninhabited during this phase.

Early Iron Age to Early Roman pottery is generally rare. (Fig.) There may be no Early Iron Age material, and the Archaic period is represented by just a handful of sherds from seventh- and mostly sixth-century decorated cups and kylikes, and a large bowl from an insular workshop. Black-glazed finewares form of the majority of the material identified across the broader period; fifth- and fourth-century Attic imports plus a few fragments of local Cycladic imitations. These are mostly plain pouring, drinking and eating vessels (skyphoi, a few plates, cups or bowls, and a kantharos) and pouring vessels such as oinochoai, plus a well-preserved fifth-century red figure bell krater – shapes suitable for the domestic table or funerary use. The few lamp fragments date to the Classical period. The sole Hellenistic piece is an intact wheelmade unpainted example of a so-called Delian lamp (c. second-century BC), a find which confirms some activity or occupation on the island during this phase.

The Middle Roman to Early Byzantine pottery gives both chronological and geographical information about the cultural and commercial relations of the island with the rest of the Mediterranean. (Fig.) It consists mainly of amphorae (LRA 1, 2, 13), red slip wares (ARS form 104, LRC forms 3 and 10) and lamps, which reveal exchange with Asia Minor, other Aegean islands, Cyprus and North Africa. Most of the identified sherd material dates to the sixth-seventh centuries AD, with few fifth-century examples.

The later Byzantine and post-Byzantine material includes late 14th-century AD protomajolica and 19th-century Çanakkale ware, Ottoman pipes of the 17th- to 19th-centuries, slipped and grooved pithoi, and various mid- to post-Byzantine amphorae. The densest concentrations of these sherds occur within polygons 1, 12 and 14, in the area of Konakia, though they also appear sparsely in the west and southeast.

The two Classical building complexes at Gerani and Megalo Kastro were planned, as were modern buildings at Konakia, tou Markou and the church of Panagia Hozoviotissa, and island-wide mapping of terracing features was completed.

Ethnoarchaeological study sought to establish a reliable toponymy, obtain information on land use, and document the material culture and architectural remains abandoned during the last century. A systematic search was made for looted Early Cycladic cemeteries, recording cavities and structures, and relating them to finds from the intensive survey. Potential graves were organized in clusters located in the area between Konakia and the hill of Gerani. Three petroglyphs were identified among the marble slabs in the area of the houses at Konakia and at tou Markou. All have engraved cavities arranged in spirals and are of the known repertoire seen in the petroglyphs now at the Museum of Apeiranthos on Naxos.