Late Mesolithic and Early Neolithic Seasonal Settlement at Kõpu

Published: Environmental and Cultural History of the Eastern Baltic Region. PACT 57. Rixensart 1999, 157–172.

The westernmost part of the island of Hiiumaa, the Kõpu Peninsula, includes a complex of Stone Age sites which are well separated temporally by their positions on ancient beach ridges. The sites of the earlier settlement phase, the Late Mesolithic and Early Neolithic, are discussed in this paper. Altogether twelve Stone Age sites had been found in the Kõpu region by 1996, eight of which derive from this period, and three of them have been excavated. These can be regarded as seasonal sites of seal-hunters on the basis of the finds, especially the osteological material, mostly representing seals. Finds of quartz and flint indicate that the population came from the area of the Kunda culture.

Though several stray finds of stone axes with shaft-holes are known already from the end of the 19th century, scientific investigation into the Stone Age on Hiiumaa started only in 1981, when the archaeologist Vello Lõugas discovered the first Stone Age settlement site on the Kõpu Peninsula (Kõpu I). Later, archaeological research conducted by Aivar Kriiska on the same peninsula revealed 11 Stone Age sites, and excavations have now been carried out at five of these. Most of the discoveries belong to the Mesolithic period (pre-pottery Stone Age), but the Neolithic Narva and Combed Ware cultures are also represented.

Being the oldest part of Hiiumaa, the land emerged there at the end of the Baltic Ice Lake stage, forming a separate island until the beginning of the Limnaea Sea stage, when it merged with the central part of Hiiumaa (Sepp, 1974, p. 84; Moora and Lõugas, 1995, p. 473–475). Coastal formations representing different stages of the Baltic Sea can be observed on Kõpu, and 12 Stone Age sites have been discovered there in the course of archaeological inventories. These are located at altitudes of 19–32 m above the present-day sea level and can be related to beach ridges dating from the end of the Ancylus Lake and the Litorina Sea stages (Ratas and Raukas, 1995). The boundary between Mesolithic and Neolithic settlement traces runs at an altitude of 27–28 m, judging from the results of excavations at the Kõpu I, IV and VIII settlement sites (Table 1). On these grounds seven sites can be regarded as Mesolithic (II, IV–IX), only 1 can be positively regarded as Early Neolithic and in one case the material is still insufficient to allow definite dating (III).

The following brief survey of the archaeologically investigated Late Mesolithic and Early Neolithic sites is presented in chronological order. The finds attributed to the Kõpu IV site do not include those obtained during the excavations of 1996, as work on them had not yet been completed at the time of writing. Preliminary observations seem to confirm that they are similar to the finds of 1995 and in principle provide no new information. A few examples of quartzite in the material excavated in 1994 and 1995 have been counted among the quartz finds (separate varieties of quartz are not distinguished here).

The settlement site is located near the village of Kõpu, by the local Kiduspe–Kõpu road, close to the junction with the road to the village of Ülendi (Fig. 2). Traces of a cultural layer were found on fairly steep beach ridges at an altitude of 28–32 m above the present-day sea-level. Subfossil molluscs recovered in the immediate vicinity of the site represent a typical Ancylus Lake fauna (Tavast, 1995, p. 12).

An area of 31 m² in the lower part of the site was excavated in 1995 and an area of 21 m² in the upper part in 1996. A fire pit of diameter over 4 m was discovered in 1995. This had been dug down about 0.5 m into the ground in an uneven manner and consisted of 8 layers of stones measuring mostly 10–15 cm. The finds were located in and around this pit.

The cultural layer, containing charcoal in the pit and immediately around it and sand lightly mixed with organic material elsewhere, was more than 1 m thick in some places. The finds totalled 546, of which 313 (57 %) were fashioned from flint and 219 (40 %) from quartz, including the few quartzite ones (Table 2). There were also 7 items (1.5 %) of sandstone and 7 (1.5 %) of other rocks: biotite gneiss, leptite, granitic pegmatite etc. (identified by the geologist Tiia Rodi). The abundance of flint is outstanding relative to material from other settlements, but flint stone with secondary processing is absent. The reasons for this will need further investigation and comparison with finds from the 1996 excavation.

A major proportion of the finds consist of flint and quartz flakes, with some flint and quartz blades as well (Table 2). The 7 flint blades are 1.4–5.2 cm long and 0.7–3.2 cm wide. The small artefacts representing secondary processing, 8 scrapers and one knife, are all made of quartz. The scrapers are mostly side scrapers on flakes. Only one polished stone artefact was found: the cutting edge of a wedge made of quartz-feldspar-gneiss. Seven polishing stones and various fragments were also found (Kriiska 1996a).

Two charcoal samples taken from the fire pit were dated to 6757±51 BP (Tln-2016) and 6640±60 BP (TA-2533) (Table 1).

This site is located 400 m NNE of the IV site, near the Kõpu School, on a former beach ridge at a present altitude of 28–29 m (Fig. 2). Two trial pits with a total area of 10 m² were dug there in 1994–1995.

No constructions were found at this site, in spite of a careful search for fire pits. The finds were located in a sand layer 0.5 m in thickness. The excavated area was very rich in finds, yielding a total of 7968 artefacts, an average of 800 per m². These included 6373 (80%) fragments of quartz, or occasionally quartzite and 1519 (19 %) fragments of flint (Table 2).

Other rocks, i.e. sandstone, diabase, leptite etc. (determined by Tiia Rodi), were less well represented (1 %). The majority of the finds were quartz and flint flakes, with a few blades and cores. The blades varied in length from 0.8 to 4.7 cm, but were mostly less than 2 cm, with a width of 0.4–2.0 cm, a quarter of them being 0.5 cm wide. The flint core is cuneiform and with one platform. 83 small artefacts representing secondary processing were found (1 % of the finds): 55 side, end and side-and-end scrapers on flakes or blades (Fig. 3: 2–3, 5–6); angle and median burins, 23 of them on flakes and one on a blade (Fig. 3: 4, 7); 4 knives; and 1 scraper-burin. The larger artefacts comprised one well polished cutting edge of a wedge, made of biotite gneiss, a presumably half-finished axe fashioned from an oval gabbro pebble (Fig. 3: 1) and some undefined stones with polished surfaces, also 18 polishing stones and related fragments (Kriiska, 1996a).

2031 fragments of hazelnut shells were also gathered from excavated area of site VIII, amounting to a total weight of 64.6 g. These were radiocarbon dated to 6172±50 BP (Tln-2024) (Table 1).

Site I is located 650 m SSW of site IV, near the Kõpu graveyard, on a former coastal formation at a present altitude of 26–27 m (Fig. 2). As indicated by the brackish-water subfossil molluscs, the beach ridge had formed during the Litorina Sea transgression (Moora and Lõugas, 1995, p. 475). Vello Lõugas dug a trial pit of 7 m² at this site in 1981 and discovered a fire pit (Lõugas, 1982, p. 375–376). The excavations of 1994 were carried out close to the same place under the supervision of Kriiska. The area excavated was 34 m², and 7 fire pits were discovered (Kriiska, 1995a), all of them dug into the ground and round or oval in shape, diameter about 1–1.5 m, and with a compact cover of stones (Table 3; Fig. 4).

Altogether 502 Neolithic potsherds, 44 lumps of burnt clay and 4562 other artefacts were found at the Kõpu I site (partly published in: Kriiska, 1995a; Lõugas et al., 1996a). The latter include 3695 (81 %) items of quartz (with a small amount of quartzite), 714 (16 %) of flint (Table 2) and 153 (3 %) of other rocks: sandstone, amphibolite, Baltic red quartz-porphyry, gneiss breccia etc. (identified by the geologist Tiia Rodi and the mineralogist Juho Kirss). The majority of the finds are flakes of quartz and flint, with a few blades and cores (Fig. 5: 4). The flint blades are 1.3–3.8 cm long and 0.4–1.5 cm wide, more than a half of them being between 0.6 and 0.9 cm wide. The flint core is conical and has one platform. Small artefacts representing secondary processing numbered 57 (1 % of the finds). The overwhelming majority of these were side, end and side-and-end scrapers on flakes, less often on blades (44 items) (Fig. 5: 1–2, 5). Their cutting edges are mostly broad, and in some cases concave. The 8 burins are of the angle or median type, on flakes (Fig. 5: 3, 6). In addition to these, 2 knives, 2 quartz wedges, one bore and 4 quartz cusps without secondary processing, and 3 quartz hammerstones were found. The only large stone artefact were a well polished cutting edge of a wedge made of amphibolite gneiss, a presumably half-finished axe and what is presumed to be a weight of amphibolite, also 43 polishing stones and related fragments.

The 502 potsherds recovered (c. 0.5 kg) originate from more than 20 vessels, presumably conical in shape with tapering bottoms, and are very fragmentarily preserved (Fig. 5: 7–11). Only sherds of one or two vessels can be connected with each fire pit. Stone rubble predominates in the composition of the clay (in more than 90 % of the sherds), but it had often contained an organic admixture as well, which has burned out. A small proportion of the sherds (c. 5 %) contained an organic admixture only. The vessels were fashioned using the band technique and were often broken where the bands joined. One connecting surface is observable in 27 cases and both connecting surfaces in 21 cases. The 0.7–2.3 cm bands were connected by the U-technique, one of the connecting surfaces being convex and the other concave. The vessels were relatively thick-walled, mostly 10 mm or more, and their surfaces had been smoothed and were seldom striated. Only a few sherds carry any ornamentation (2.5 %). The main components of this ornamentation are small notches measuring a couple of mm, while some sherds are grooved and one has a pit almost piercing through it. The site also yielded 44 lumps of burnt clay, more than 70 % of them in one fire pit. The clay mostly has no admixtures and is heavily burnt.

Five samples of charcoal collected from the pits gave radiocarbon dates all confined between 5698±70 and 5330±90 BP (Table 1).

The bone material described here was collected from the Kõpu I, Kõpu IV and Kõpu VIII sites during the archaeological excavations of 1994/95/96, and preliminary results regarding the Late Neolithic Kõpu XI bone material are also included. The material is very poorly preserved at all the sites, because it was deposited in sandy and gravelly soil horizons. Also, many of the bones had been burned (24–40 %).

Where the Mesolithic Kõpu IV site contains only remains of ringed seals, Pusa hispida (Table 4), the Kõpu I site also features the grey seal, Halichoerus grypus (Table 5). The very few fish bones recovered from the sites indicate that fishing was of little importance in Mesolithic Kõpu and only the common fresh/brackish water species were available, e.g. pike. The picture changes at the point where the influence of saline water begins to be detected in the eastern Baltic (the Litorina Sea stage). The cod, Gadus morhua, and turbot, Scophthalmus maximus, remains found at Kõpu I denote immigrants from the Atlantic. Preliminary analyses show that bones of the cod greatly predominate in the refuse material at the Kõpu XI site, together with the harp seal, Pagophilus groenlandicus, from Atlantic/Arctic waters (Table 6). Kõpu XI seems to be similar in terms of its bone material to the Loona site on the island of Saaremaa, where the cod and the harp seal were the main species captured (Lõugas et al., 1996b).

The bird bones were poorly preserved, occurring in the form of small fragments. Remains of eider, some other ducks, cormorant and merganser (Table 5) provide evidence of the hunting of waterfowl. Large numbers of ducks and gulls probably nested on the island of Kõpu and hunters may well have used their eggs as well as their meat for food (Moora and Lõugas 1995, p. 478).

Cultural interpretation of the Stone Age sites on Kõpu is hindered by the absence of bone artefacts and the scarcity of polished stone implements. Reasonable comparisons can only be made in the case of the pottery from the Kõpu I site, which possesses parallels only among the material from the Kõnnu site on Saaremaa (Jaanits, 1979; 1984). In spite of some differences, such as the predominant mineral admixture, pits in the ornamentation etc., we can define the vessels at these sites as belonging to the Narva type, forming a local group within this classification. The scarcity of ornamentation, striation of the surfaces and especially the use of narrow bands with U-shaped connecting surfaces for constructiong the vessels associates the early pottery of the islanders with the continental Estonian vessels of the same period (Kriiska, 1995b) and distinguishes it from the Finnish Early Comb Ware, which is richly ornamented and built up of wide bands (Edgren, 1984, p. 30, 32), and from the comparatively abundantly decorated Western Lithuanian local group of early Narva Pottery (Rimantiene, 1979, p. 146).

The analogous settlement strategies, use of rock and similarities in artefacts are evidently attributable to a cultural continuity between the Mesolithic and Neolithic settlement periods. In all these settlements the tools were made mostly of quartz and greyish-white flint of inferior quality and of local Kõpu origin. Other rocks were used to a lesser extent. The only clear difference in terms of the latter is provided by the Baltic red quartz-porphyry, which is absent at the Mesolithic settlements and appears only in the Kõpu I material. Its proportion increases during the Neolithic period, as demonstrated by the excavations carried out at the Kõpu XI site in 1996. In view of the relatively solid opinion of researchers that the Narva culture as a whole developed from the local Kunda culture /e.g. Jaanits, 1970, p. 86; Zagorskis, 1973, p. 65; Girininkas, 1994, p. 259) and the fact that the stone axes and wedges found at the Kõnnu site on Saaremaa are peculiar to the Kunda culture, we can assume with some certainty that the early inhabitants of the ancient island of Kõpu belonged to the Kunda–Narva Culture. The basis for colonization would logically be the resources of the coastal area of Western Estonia, although nothing can be said with certainty yet, as investigations in the area is just beginning and the first pre-pottery Stone Age settlement in Western Estonia was discovered only in autumn 1996.

The beginning of settlement on Kõpu is evidently connected with changes in the hunter-gatherer economy, most notably a specialization in seal hunting. The predominance of the latter is unequivocally indicated by the osteological material described here. This links the history of the ssettlement at the site with the wider process that is known to have occurred elsewhere in the Baltic in the Late Mesolithic and Early Neolithic, e.g. the colonization of Saaremaa (Kriiska 1996b) and Åland (Nuñez 1996, p. 25).

Early settlement is generally assumed to have been seasonal (Nuñez, 1996, p. 27; Moora and Lõugas, 1995, p. 479; Kriiska, 1996b), in view of the low intensity of the cultural layer, the uniformity of the material recovered, ecological factors, e.g. the inability of small islands to support constant habitation, and presence of animal bones. All these factors are present at Kõpu as well. The most expressive argument determining the settlement pattern is the bone material, which at Kõpu represents almost exclusively seals. The period of seasonal settlement can be specified by the discovery of a bone from a ringed seal only a week old, implying that it must have been killed early in spring, when the pups are born (Moora and Lõugas 1995, p. 479). February–March was the principal and most productive seal hunting season in the coastal regions and islands of Estonia until the turn of the last century (Kalits, 1963, p. 136; Art, 1988, p. 13). This was governed by the mass migration of seals into these waters to give birth to their young, which took place at the end of February in the ringed seal and in March for the grey seal (Aul et al., 1957, p. 268–269). Though no weapons used in seal hunting have been found at the Kõpu sites, we can assume that harpoons were mostly used, and also clubs for the young seals. Both hunting methods have survived as an anachronism in the ethnographic seal-hunting tradition in Estonia (Kalits, 1963, p. 150–151). To catch a seal with a harpoon, the hunter mostly had to wait for it to come to the surface to breathe (ibid; Pettai, 1988, p. 31), although the seals resting on the ice were also hunted (Art, 1988, p. 7). There is a evidence of seal hunting with wooden clubs on the small Danish island of Hesselo in the Kattegat, which had an important sealing station in Stone Age times. More than 90 % of the Hesselö bone material is derived from grey seals, including a very high proportion from young pups. Also, six wooden clubs weree found in a bog there which might have been used to kill grey seals (Møhl, 1971). Nets were probably also used, but in autumn, when the sea was not covered by ice. This latter hunting method could have been used in the Late Neolithic (Kõpu XI) for harp seal hunting (Lõugas, 1998).

It is to be hoped that future investigations on the island of Hiiumaa will give us more information on these matters.

The authors would like to thank the geographer Toomas Kokovkin, geologists Harri Moora and Tiia Rodi, the artist Jana Ratas and all the other members of the Kõpu expedition for many years of cooperation. The investigations on Kõpu were financed by the EU/PACT project (ERBCIPACT 93–0152), the Estonian Scientific Foundation (grant no. 1022 and 2254), the Cultural Endowment of Estonia, the Estonian Academy of Sciences and the newspaper “Maaleht”.

Aivar KRIISKA
Tartu University
Lossi 3
EE-51003 TARTU, Estonia

Lembi LÕUGAS
Institute of History
Rüütli 6
EE-10130 TALLINN, Estonia

ART, E. 1988, Hülged ja hülgepüük (Seals and sealing), in E. PETTAI (ed.), Hülgepüügi meenutusi möödunud aegadest, Stockholm, p. 5–15 (in Estonian). 6c

AUL, J., LING, H. and PAAVER, K., 1957, Eesti NSV imetajad (The mammals of Estonian SSR), Tallinn, 351 p. (in Estonian).

EDGREN, T., 1984, Kivikausi (The Stone Age), in T. EDGREN, U. SALO and P.–L. LEHTOSALO–HILANDER (eds), Suomen historia (Finnish history), 1, Weilin and Göös, Espoo, p. 18–85 (in Finnish).

GIRININKAS, A., 1994, Baltu kulturos i�takos (The Origin of the Baltic Cultures), Vilnius 279 p. (in Lithuanian).

JAANITS, L., 1970, Kultuuri arengupidevusest Eestis üleminekul keskmiselt nooremale kiviajale (Über die Lückenlosigkeit der Kulturentwicklung in Estland beim Übergang von mittleren zur jüngeren Steinzeit), Studia archaeologica in memoriam Harri Moora, Tallinn, p. 81–87 (in Estonian).

JAANITS, L., 1979, Die neolithische Siedlung Kõnnu auf der Insel Saaremaa, in Eesti NSV Teaduste Akadeemia Toimetised, Ühiskonnateadused 4, p. 363–367.

JAANITS L. J., 1984, Ranneneolititsheskie kulturnyje gruppy v Estonii (Early Neolithic Cultural Groups in Estonia), in Novoje v arheologii SSSR i Finljandii. Doklady Tretego sovetsko-finljandskogo simpioziuma po vaprosam arheologii 11–15 maja 1981 g. Leningrad, p. 16–21 (in Russian).

KALITS, V., 1963, Kihnlaste elatusalad XIX sajandi keskpaigast tänapäevani (The settlement areas of Kihnu-people from the middle of 19th century to the present day), Thesis (Manuscript in the Library of University of Tartu.), 489 p. (in Estonian).

KRIISKA, A., 1995a, Archäologischen Ausgrabungen auf dem Standort der ehemaligen Steinzeitsiedlung Kõpu I (Ristipõllu), in Proceedings of the Estonian Academy of Sciences, Humanities and Social Sciences, 4, p. 410–416.

KRIISKA, A., 1995b, Narva jõe alamjooksu ala neoliitiline keraamika (The Neolithic pottery from the area of the lower reaches of the River Narva), in V. LANG (ed.), Eesti arheoloogia historiograafilisi, teoreetilisi ja kultuuriajaloolisi aspekte, Muinasaja teadus, 3, Tallinn, p. 54–115 (in Estonian with English summary).

KRIISKA, A., 1996a, Archaeological studies in the Peninsula of Kõpu, in Proceedings of the Estonian Academy of Sciences, Humanities and Social Sciences, 4, p. 398–409.

KRIISKA, A., 1996b, Viron rannikkoalueen asutus ja pyyntikulttuurin erikoistuminen kivikaudella, in Muinastutkija, 4, p. 1–6 (in Finnish).

LÕUGAS, L., 1998, Postglacial Invasions of the Harp Seal (Pagophilus groenlandicus) into the Baltic Sea, in Proceedings of the Latvian Academy of Sciences, Section B, 52, ½, p. 63–69.

LÕUGAS, L., KRIISKA, A. and MOORA, H., 1996a, Coastal Adaptation and Marine Exploitation of the Island Hiiumaa, Estonia during the Stone Age with Special Emphasis on the Kõpu I Site, in A.–M. ROBERTSSON, T. HACKENS, S. HICKS, J. RISBERG and A. ÅKERLUND (eds), Landscapes and Life, Studies in honour of Urve Miller, PACT, 50, Rixensart, p. 197–211.

LÕUGAS, L., LIDÉN, K. and NELSON, D. E., 1996b, Resource utilization along the Estonian coast during the Stone Age, in S. HICKS, V. LANG, U. MILLER and L. SAARSE, (eds), Coastal Estonia: Recent Advances in Environmental and Cultural History, PACT, 51, Rixensart, p. 399–420.

LÕUGAS, V., 1982, Beiträge zur Vorgeschichte des Westarchipels Estland, in Eesti NSV Teaduste Akadeemia Toimetised, Ühiskonnateadused, 4, p. 372–376.

MOORA, H. and LÕUGAS, L., 1995, Natural conditions at the time of primary habitation of the Hiiumaa Island, in Proceedings of the Estonian Academy of Sciences, Humanities and Social Sciences, 44/4, p. 472–481.

MØHL, U., 1971, Fangstdyrene ved de Danske strande: de zoologiske baggrund for harpunerne (Seal and whale hunting on the Danish coasts), in KUML, Årbog for Jysk Arkeologisk Selskab 1970, p. 297–239 (in Danish with English summary).

NUÑEZ, N., 1996, When the water turned salty, in Muinaistutkija, 4, p. 23–33.

PETTAI, E., 1988, Hülgeküttimisest rannakülas (Seal hunting in coastal villages), in E. PETTAI (ed.), Hülgepüügi meenutusi möödunud aegadest, Stockholm, p. 30–32 (in Estonian).

RATAS, U. and RAUKAS, A., 1995, Hiiumaa Läänemere arengu etalonina (Hiiumaa as a standard for the history of the Baltic Sea), in XVIII Eesti loodusuurijate päev, Hiiumaa loodus, Tartu–Tallinn, p. 8–11 (in Estonian).

RIMANTIENE, R., 1979, �ventoji. Narvos kulturos gyvenvietes (�ventoji 1. Ausgrabungen der Narvakultursiedelungen), Mokslas, Vilnius, 190 p. (in Lithuanian with German summary).

SEPP, U., 1974, Hiiumaa maastikuline iseloomustus (Characteristic landscapes of the Hiiumaa Island), Tallinn, Valgus, 149 p. (in Estonian).

TAVAST, E. 1995, Subfossiilsed molluskid Hiiumaa vanades rannasetetes (Subfossil molluscs in the ancient coastal sediments of Hiiumaa), in T. KUKK (ed.), XVII Eesti loodusuurijate päev, Hiiumaa loodus, Ettekanded, Teaduste Akadeemia Kirjastus, Tartu–Tallinn, p. 12–17 (in Estonian).

ZAGORSKIS, F., 1973, Agrais neolits Latvijas austrumdala (Early Neolithic in eastern Latvia), in Latvijas PSR Zinatnu Akademijas Vestis, 4, p. 56–69 (in Latvian).