Précédent Suivant

The Neolithic landscape and settlement of the Island of Gökçeada (Imbros, Turkey)

p. 89-94

Résumé

The recent archaeological discoveries on the island of Gökçeada (Imbros) shed new light on the early Prehistory of the North Aegean Islands. The earliest finds date from the Middle Palaeolithic period, and possible Mesolithic/ Epi-palaeolithic chipped stone tools were discovered in the eastern part of the island. Stratigraphic excavations at the site of Uğurlu have clarified the spatial extent of the settlement from Preceramic or Initial Neolithic occupation onwards. Uğurlu Phase VI is dated to 6700-6500 cal. BC. A site near the Salt Lake is probably contemporary with Uğurlu Phase VI. The Neolithic Phase V (6500-6000 cal. BC) at Uğurlu has signs of continuity, and the permanent settlers were agriculturalists who introduced domestic sheep, goats, cattle and pigs to the island. The distribution of Melian and Central Anatolian obsidian suggests long-distance exchange mechanisms during the Neolithic. The sea level and shoreline in the Aegean were different during prehistoric times compared to the present day. During the low sea level of the Last Glacial Maximum the island of Gökçeada, together with all of the North Aegean Islands, was connected to the mainland. Gökçeada, together with Lemnos, became an island probably just after the Younger Dryas, and they were connected by an isthmus. Around 7000-6500 cal. BC, sea level was 20 m lower than today and the separate island of Gökçeada lay close to the Gelibolu Peninsula.


Texte intégral

Introduction

1The field of island archaeology provides an opportunity to address the following questions. Why did people settle islands? How did they manage in an insular setting? What type of relations existed between island societies or between islanders and mainlanders (Broodbank 2000; Knapp, 2008; Dawson, 2014)? These key research themes were also central to our research on Neolithic societies at the island of Gökçeada. The island of Gökçeada (Imbros) is about 17 km from the Gallipoli Peninsula and covers an area of 289.5 km² (Figure 1). The Uğurlu Archaeological Project was initiated in 2009 with the aim of exploring the role of the Northern Aegean Islands in the Neolithization of South East Europe. The project was also focused on understanding the Neolithic environment and with stratigraphical analysis of Prehistoric sites on the island.

2The earliest finds on the island date from the Middle Palaeolithic period. Middle Palaeolithic activity is documented around the chert outcrops in the eastern part of the island (Özbek and Erdoğu, 2014; Erdoğu, in press). Levallois flakes and points are dominant (Figure 3 B). Possible Mesolithic/ Epi-palaeolithic finds, consisting mainly of micro f lake tools, were also found in the eastern part of the island, around the Eksino Stream. Uğurlu is the only early Neolithic settlement on the western part of the island (Figure 2). It is a low mound covering an area of approximately 250 × 200 m on a gentle slope at the eastern foot of Mount Isa (Doğanlı). The main Uğurlu-Dereköy road cuts through the site. The Pilon stream lies at the eastern part of the site, and there is also a nearby spring. During the six years of excavation six main cultural phases, designated I-VI (counting from top to bottom), and at least 12 layers of occupation, have been revealed so far. The earliest Phase VI is dated to 6700-6500 cal. BC and Phase V is dated to 6500-6000 cal. BC (Table 1). Another Early Neolithic site was discovered near the Salt Lake in the southeastern part of the island.

Image

Figure 1: Location of Uğurlu in the Northeastern Aegean.

Image

Figure 2: The Neolithic site of Uğurlu and its environment.

Image

Table 1: Radiocarbon dates for Uğurlu.

Geological background and sea level change

3Gökçeada is a mountainous island and Mount Doruk (Elias), with an altitude of 673 m, is the highest point. Narrow valleys, such as Büyükdere in the north and Ballidere in the west, and the northern part of the Salt Lake, provide the only agricultural land. The solid geology is composed mainly of volcanic rocks. Volcanic rocks on the island occur as vents, intrusion centres or as volcanoclastics interlayered with the sedimentary strata. They usually constitute topographic highs and have an andesitic composition (Kurtulus et al., 2010). The commonest minerals present in andesitic rocks are: plagioclase feldspar, more specifically andesines; amphiboles, usually hornblende; pyroxenes, usually augite; spinel, such as titanomagnetite, ilmenite and magnetite; quartz; apatite; and biotite. At least 5 chert outcrops in the eastern part of the island were discovered. Black-coloured fine- grained chert of Kuzu Limanı was mainly used in the Neolithic.

4The sea level and the shoreline of the Aegean were different during prehistoric times compared to the present day and our project was focused on the following questions: When did Gökçeada begin to separate from the mainland and become an island? How far was it across the sea to the mainland in the Early Neolithic?

5Starting at the time of the Last Glacial Maximum (around 21 000 years ago), sea level was about 120 m lower than today (Lamback and Purcell, 2005). The island of Gökçeada, along with all the North Aegean Islands, was connected to the mainland. Sea level was 55-60 m lower during the Younger Dryas between 10800 and 9600 cal. BC. Gökçeada, together with Lemnos, became an island probably just after the Younger Dryas, and they were connected by an isthmus (Perissoratis and Conispoliatis, 2003; Efstratiou, 2014). Subsequently the rise in the level of the Aegean Sea continued at a slower pace, reaching a height approximately 20 m lower than today during the Neolithic period, around 7000-6500 cal. BC (for discussion see Özbek and Erdoğu, 2014), and the island of Gökçeada lay close to the Gelibolu Peninsula, ca. 10 km. In addition, geomorphological investigations indicate that deep bays were formed at the river estuaries in the island (Öner, 2000). Today, the seashore is about 2 km from the site of Uğurlu, but the site was located far from the shoreline during the Neolithic period. Archaeobotanical research demonstrates that the first settlement was located in an area where small lakes and swamps occur, and the Neolithic inhabitants of Uğurlu lived near a spring in this well-watered fertile area.

6The Uğurlu project was also intended to improve our knowledge of the palaeoenvironment. A coring programme was conducted to reconstruct the Neolithic environment. A single core, to a depth of ca. 3.24 m, was taken from the marsh some 2 km south of Uğurlu, but no Early Neolithic levels were detected.

The Early Neolithic Occupation

7The earliest occupation at Uğurlu is located in the eastern part of the settlement, close to the Pilon stream. Phase VI is represented in sounding trenches only. No architectural structures, with the exception of scattered stones in clusters, and a hearth, were found in this Phase. Three AMS radiocarbon dates range from ca. 6682-6570 cal. BC (1 α). No ceramics or any other clay objects were recorded. Several disc- shaped shell or stone beads and bone awls were found. The chipped stones of Phase VI are characterised by thin blades made using the pressure technique. Only 8 obsidians (6 blades) were found in this phase, and local f lint/chert was used much more frequently. A flotation sample of 100 litres was taken from Phase VI and only one possible cereal grain was identified.

8The site near the Salt Lake (Figure 1) is probably contemporary with Uğurlu Phase VI. The site lies on the eastern bank of the Salt Lake and most of its surface is flooded by saltwater all year round, except in summer (Figure 3). However, it is unclear whether or not the Salt Lake existed during the Neolithic period. Only chipped stone implements were found at this site, and no pottery or ground stones were discovered. Raw materials are similar to the raw materials used at Uğurlu. Only two pieces of obsidian were found on the surface (Figure 3A). Flakes are dominant and there are also flake cores. Blades are quite rare and most of them were made using the pressure technique. Some of the blades may have been used as sickle inserts and tools are relatively abundant. A foliate point is noteworthy.

9So far two possible occupational layers of Phase V have been recorded. The early layer of Phase V is represented in sounding trenches and consists of earthen floors with an extremely dense concentration of animal bones, f lint and obsidian tools and bone spatulas, and a fish hook was also found (Figure 4A). A single-room, earthen-floored building about 5 × 4 m, has been excavated in the late layer of Phase V (Figure 4). It is characterized by small room size (9.2 m²), thick walls (1.00 m) and massive exterior buttresses. The northern wall stands to a height of about 1 m, and a fireplace is set inside the wall. Following its abandonment the fireplace seems to have been filled with earth and stones, and a stone axe was deliberately left within it. The northern wall stretches about 3-5 m to the east with a parallel wall ca. 1 m wide, creating a courtyard. A sherd with a human motif in relief and a head from animal bone from an Acrolithic figurine were found in the courtyard. The nose of the figurine head was shown in relief while the eyes were highlighted with red paint (Figure 4 B). Several broken bone tools, a small stone axe made of serpentine, and a small broken malachite bead, were found in situ in the building. The malachite bead is one of the earliest malachite finds in Southeast Europe. There are two malachite veins running between the villages of Dereköy and Tepeköy, close to the site. A single AMS radiocarbon date from the early layer of Phase V yielded an age of 6478-6436 cal. BC (1α), and an additional AMS radiocarbon date from the late layer of Phase V yielded an age of 6085- 6010 cal. BC (1 α).

Image

Figure 3: The Salt Lake. Figure 3A: An obsidian blade from the Salt Lake site; Figure 3B: Middle Palaeolithic finds.

10A large number of soil samples in Phase V were processed by flotation. Domestic cereals, including einkorn wheat (Triticum monococcum), six-rowed barley (Hordeum vulgare) and pea (Pisum sativum L.), were present throughout the sampled deposits. Einkorn wheat was the most abundant of the cultivated species. The faunal data show that domestic sheep and goat are dominant in a very early phase V. Domestic pig, cattle and dog also occur. The presence of wild boar, red deer, hare and fox indicate hunting.

11Most of the chipped stone production made use of raw materials found in the island: different coloured flint/cherts, jasper and quartz. Obsidian was also used as a raw material. Melian obsidian was widely used, but Central Anatolian East Göllü Dağ obsidian was also available (Miliç, 2014). Flakes are more common than blades. Tools are very rare and scrapers are the most common type. Polished stone axes and adzes are abundant, especially in the late layer of Phase V. The raw material used in their manufacture probably comes from local sources. On the other hand, stone axes made from nephrite were imported to the island from outcrops at the foot of Ganos Mountain in the Gallipoli Peninsula (Özbek and Erol, 2001).

12Pottery is the most common artefact found in Phase V. The vast majority of the pottery is red slipped and burnished. Black burnished sherds were found in small quantities. All pottery is handmade and thin walled. Deep bowls with an “S” profile, hole-mouth vessels and straight-sided shallow dishes are common shapes. Bases are either flat or have a low pedestal. Vertically-placed tube-like and knob-like perforated tubular lugs, as well as small crescent shaped lugs, are characteristic. A human-faced sherd and a sherd with a human motif in relief are unique. A comparison of the results of the thin section analysis and the chemical composition of the sherds and clay samples indicates that the Neolithic pottery of Uğurlu is composed of local clays (Erdoğu, 2014). Only one type of clay could be distinguished and it is non-calcareous of andesitic/dacitic composition. The minerals consist of quartz, alkali and plagioclase feldspars, amphiboles and titaniferous minerals, especially magnetite and titaniferous magnetite. Organic temper is almost absent. Only a very small number of organic inclusions were identified and they were probably not deliberate additions. A very small number of bone fragments were identified in a few samples.

Image

Figure 4: Neolithic Building at Uğurlu Phase V; Figure 4A: Bone fish hook; Figure 4B: Bone figurine head.

Conclusion

133 AMS radiocarbon dates confirm that phase VI at Uğurlu began rather early, at least as early as the Preceramic or Initial Neolithic layers of Ulucak VI and Çukuriçi XIII in Western Anatolia (Horejs et al., 2015; Çevik and Abay, in press). The radio-carbon dates from Knossos X and Franchthi “Initial Neolithic” indicate the existence of the Preceramic phase in the Aegean between 7000 to 6600 cal. BC. (Perlès et al., 2013; Reingruber, 2015). Possible Mesolithic/Epi Palaeolithic finds indicate an indigenous population on the island. Future studies may reveal whether the indigenous populations played a role in the process of Neolithization. The Preceramic site near the Salt Lake is also important. The absence of pottery and the presence of obsidian and especially pressure blades are noteworthy. Phase V at Uğurlu demonstrates that permanent settlement on the island was achieved by people with an agricultural economic base, including cultivated plants and domesticated animals. Settlers were agriculturalists and they introduced domestic sheep, goats, cattle and pigs to the island. Strong parallels to Uğurlu V pottery were found in Western Anatolian Neolithic sites as well as at the Marmara region sites and at Hoca Çeşme IV-III in Turkish Thrace (Bertram et Karul, 2005; Figure 1 and 3). Neolithic Gökçeada (Imbros) is characterized by the striking evidence of long-distance communications. The distribution of Melian and Central Anatolian obsidian suggests long-distance exchange mechanisms, and travel must have involved seafaring for at least part of the journeys.

Acknowledgements

14The Uğurlu Archaeological project is supported by the Turkish Ministry of Culture and Tourism and the University of Trakya, Edirne (BAP 2014-25). I thank Levent Atıcı, Soultana Valamoti and Denis Guilbeau for sharing their data.

Bibliographie

Des DOI sont automatiquement ajoutés aux références bibliographiques par Bilbo, l’outil d’annotation bibliographique d’OpenEdition. Ces références bibliographiques peuvent être téléchargées dans les formats APA, Chicago et MLA.

References

BERTRAM J.-K., KARUL N., «From Anatolia to Europe: The ceramic sequence of Hoca Çeşme in Turkish Thrace», in Lichter C. (ed.), Byzas 2 – How Did Farming Reach Europe? Anatolian-European Relations from the Second Half of the 7th Through the First Half of the 6th Millennium Cal BC, Istanbul, 2005, p. 117-130.

BROODBANK C., An Island Archaeology of the Early Cyclades, Cambridge, Cambridge University Press, 2000.

ÇEVIK Ö., ABAY, E., «Neolithisation in Aegean Turkey: Towards a More Realistic Reading», in Yalçın Ü. (ed), Anatolien und seine Nachbarn vor 10.000 Jahren (in press).

EFSTRATIOU N., «The last forager camp of Ouriakos on the island of Lemnos: Human groups on the move at the turn of the Holocene in the Northern Aegean», Eurasian Prehistory, 11, (1-2), 2014, p. 75-96.

10.4324/9781315424774 :

DAWSON H., Mediterranean Voyages. The Archaeology of Island Colonisation and Abandonment, Walnut Creek, CA, Left Coast Press, 2014.

ERDOĞU B., «Gökçeada Uğurlu Archaeological Project: A Preliminary Report from the 2011-2013 Field Seasons», Anatolica, XL, 2014, p. 157-179.

10.4324/9781315230603 :

ERDOĞU B., «Whither The Aegean Neolithic», in Tsirtsoni Z., Reingruber A., Nedelcheva P. (eds.), Going West? The spread of farming between the Bosporus and the Lower Danube Region (in press).

10.1007/s10963-015-9090-8 :

HOREJS B., MILIÇ M., OSTMANN F., THANHEISER U., WENINGER B., GALIK A., «The Aegean in the early 7th Millennium BC. Maritime Networks and Colonization», Journal of World Prehistory, 28, 2015, p. 289-330.

10.1093/acprof:oso/9780199237371.001.0001 :

KNAPP B.A., Prehistoric and Protohistoric Cyprus. Identity, Insularity, and Connectivity, Oxford, Oxford University Press, 2008.

10.1007/s10064-010-0270-6 :

KURTULUS C., IRMAK T.S., SERTCELIK I., «Physical and mechanical properties of Gökçeada: Imbros (NE Aegean Sea) island andesites», Bulletin of Engineering Geology and the Environment, 69 (2), 2010, p. 321-324.

10.1016/j.quascirev.2004.06.025 :

LAMBECK K., PURCELL A., «Sea-level change in the Mediterranean Sea since the LGM: Model predictions for tectonically stable areas», Quaternary Science Review, 24, 2005, p. 1969-1988.

10.1016/j.jas.2013.08.002 :

MILIÇ M., «PXRF characterisation of obsidian from central Anatolia, the Aegean and central Europe», Journal of Archaeological Science, 41, 2014, p. 285-296.

ÖNER E., «Yeni Bademli Höyük Çevresinde Jeoarkeolojik Arastirmalar (Gökçeada-Imroz)», 15, Arkeometri Sonuclari Toplantisi, 2000, p. 19-32.

ÖZBEK O., EROL K., «Étude pétrographique des haches polies du Hamaylitarla et Fenerkaradutlar (Turquie)», Anatolia Antiqua, 9, 2001, p. 1-7.

ÖZBEK, O., ERDOĞU, B., «Initial occupation of the Gelibolu Peninsula and the Island of Gökçeada (Imbroz) in the Pre-Neolithic and Early Neolithic», Eurasian Prehistory, 11, (1-2), 2014, p. 97-128.

10.1016/S0025-3227(03)00047-1 :

PERISSORATIS C., CONISPOLIATIS N., «The impacts of sea-level changes during latest Pleistocene and Holocene times on the morphology of the Ionian and Aegean seas (SE Alpine Europe)», Marine Geology, 196, 2003, p. 145-156.

10.1017/S0003598X00049826 :

PERLÈS C., QUILES A., VALLADAS H., «Early seventh-millennium AMS dates from domestic seeds in the Initial Neolithic at Franchthi Cave (Argolid, Greece)», Antiquity, 87(338), 2013, p. 1001-1015.

10.4312/dp.42.9 :

REINGRUBER A., «Preceramic, Aceramic or Early Ceramic: The radiocarbon dated beginning of the Neolithic in the Aegean», Documenta Praehistorica XLII, 2015, p.147-158.

Précédent Suivant

Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.