The earliest Neolithic complex in Siberia > the Ust-Karenga 12 site and its significance for the Neolithisation process in Eurasia

The definition of the term ‘Neolithic’ in Siberia and the northern and eastern parts of Asia implies first of all the presence of pottery (e.g. Oshibkina 1996a; Barnes 1999; Kuzmin 2003, 2006; Kuzmin and Orlova 2000). In this case, pottery is determined as containers made of fired clay (e.g. Darvill 2002. 337–338). Therefore, the concept of Neolithisation for Siberia as well as East Asia means the emergence of pottery-making. In this paper, we present a systematic description of the earliest pottery assemblage from Siberia known so far, Ust-Karenga. Previously, it was published only in brief (e.g. Vetrov 1985; Kuzmin 2002; Kuzmin and Orlova 2000.361).


Introduction
The definition of the term 'Neolithic' in Siberia and the northern and eastern parts of Asia implies first of all the presence of pottery (e.g.Oshibkina 1996a; Barnes 1999;Kuzmin 2003Kuzmin , 2006;;Kuzmin and Orlova 2000).In this case, pottery is determined as containers made of fired clay (e.g.Darvill 2002. 337-338).Therefore, the concept of Neolithisation for Siberia as well as East Asia means the emergence of pottery-making.In this paper, we present a systematic description of the earliest pottery assemblage from Siberia known so far, Ust-Karenga.Previously, it was published only in brief (e.g.Vetrov 1985;Kuzmin 2002;Kuzmin and Orlova 2000.361).
The cluster of 16 prehistoric sites in the Karenga River mouth, located on the boundary between the upper and middle courses of the Vitim River in northern Transbaikal, Siberia (Figs. 1-2), was discovered in the second half of the 1970s when a systematic survey was conducted in the Vitim River basin by researchers from Irkutsk State University (Aksenov and Vetrov 1977;Vetrov et al. 1978).The geographical coordinates of the Ust-Karenga cluster are: 54°28' northern latitude and 116°31' eastern longitude, as determined with the aid of a U.S. Operational Navigation Chart, scale 1:1 000 000 (sheet ONC E-8).The Ust-Karenga sites lie in the Vitim River valley; the water level elevation at the confluence of the Vitim and Karenga rivers is about 600 metres above sea level (asl).The Vitim River cuts through the low mountain system of the Vitim Tableland with heights of about 800-1200 m asl (End-rikhinsky 1974), and the highest points in the vicinity of Ust-Karenga are about 1200-1700 m asl.The climate of the Vitim Tableland is of ultra-continental type, with hot summers and cold winters (e.g.Suslov 1961).The mean January temperature is -30°to -33 °C; and average July temperature is up to +20 °C.The annual amount of precipitation is about 350-400 mm (Gvozdetsky and Mikhailov 1978.350).The area is covered with dense conifer forests (taiga), consisting mainly of Dahurian larch [Larix dahurica, in some sources Larix gmelinii (e.g.Shahgedanova et al. 2002)].

Materials and Methods
The Ust-Karenga 12 site, which is the most representative for our study, was discovered in 1976.The finds in cultural layer 7 included pottery fragments, along with stone artefacts of typical final Upper Palaeolithic appearance (wedge-shaped cores, Araya type transversal burins, bifaces, and scrapers).It was separated from the underlying and overlying cultural layers by about 1 metre of sterile sediments both above and below (Figs.3-4).Excavations of Ust-Karenga 12 were conducted in a series of periodical campaigns, from 1976 until recently.The total excavated square at Ust-Karenga 12 for layer 7 is 214 m 2 .As for geoarchaeological studies, palynological data were obtained for layer 7 (Vetrov and Kuzmin 2005), and a series of radiocarbon (hereafter -14 C) dates was generated.The first 14 C dates were released in the mid-1990s (Vetrov 1995a), and new results were produced and published in the late 1990s and the 2000s (e.g.Vetrov and Kuzmin 2005;Vetrov et al. 2006).

Results
The cluster of archaeological sites at the confluence of the Vitim and Karenga rivers is located on the alluvial terrace of the Karenga River, at a height of 20-25 m above the water level (Fig. 2).The general stratigraphy of the Ust-Karenga cluster, mainly derived from the Ust-Karenga 12, 14, and 16 sites, is as follows (Fig. 3):
Cultural layer 7 as the earliest component of the Ust-Karenga Neolithic culture is the main focus in this report.It was excavated at several sites; the most representative locale is Ust-Karenga 12 (Fig. 2), for which a major part of archaeological and palaeo-environmental information was obtained.The thickness of layer 7 is from 2 to 10 cm (Figs.3-5).It con-tains several well-preserved hearths and artefact concentrations around them; these spots are up to 6 m in diameter.
The predominant raw material is flint obtained from pebbles collected in the channels of the Vitim and Karenga rivers.
In terms of raw materials used for the manufacture of the stone tools, it is important to note the pre- sence in cultural layer 4 of the Ust-Karenga 16 site of some artefacts made on rocks 'exotic' for the middle stream of the Vitim River, hyalodacite and graphitite (Vetrov et al. 2000).The sources of these raw materials are located downstream from the Ust-Karenga cluster, at a distance of up to 400 km.This fact demonstrates that the inhabitants of the middle course of the Vitim River had active contacts with the neighbouring territories of Eastern Siberia, with distances of 400 km and possibly up to 600 km (Vetrov et al. 2000).
The pottery from cultural layer 7 is unique in all Siberia.Numerous potsherds were excavated, including large fragments, and this allows the reconstruction of the size and shape of vessels.The vessels are of parabolic type, from 17-20 to 35 cm high, and from 12 to 20 cm in diameter.The sharp-based bottom looks mammiformed.The design is mainly combpattern (Fig. 8), and also zigzag, herringbone, and cogged stamped (Figs.9-10) (Vetrov 1985;Kuzmin and Orlova 2000.361).Both external and internal sides have traces of grooves made with grass fibre or comb trail to smooth the surface of the clay during the pottery-making process .Ornamentation was made mainly by cog-wheel (Vetrov 2006).The distinctive feature of the Ust-Karenga pottery is that it is plant fibre-tempered.The number of vessels used at the Ust-Karenga 12 site may be estimated as about ten.For the whole Ust-Karenga cluster, about 16-18 vessels can be reconstructed.
14 C dating of cultural layer 7 was conducted using two kinds of datable material -charcoal from hearths and the cultural layer in general, and pottery temper (Tab.1).The extraction of carbon from organic-tempered pottery was performed by low temperature combustion with oxygen (O' Malley et al. 1999;Derevianko et al. 2004;Vetrov et al. 2006).The carbon yield of three pottery samples was about 0.8-1.0%, which makes the 14 C dates on pottery temper quite reliable in terms of the origin of carbon.We assume that the 14 C-dated carbon comes predominantly from short-lived plant fibre temper, and not from clay carbon itself, which may be much older than the time of vessel manufacture.Calibration was done with the aid of Calib Rev. 5.0.1 software (available online: www.radiocarbon.org).
The results of 14 C dating are presented in Table 1.Charcoal from cultural layer 7 at a depth of 1.00 m below the surface, found in small depressions in direct association with pottery, was dated to c.  Holocene transition in Eastern Siberia (e.g.Krivonogov et al. 2004).

Discussion
In the light of Late Glacial pottery in Transbaikal, we should examine the adjacent regions of Siberia to see if there are any other Neolithic complexes known with ages similar to the Ust-Karenga culture.In Transbaikal, two other sites may contain pottery of the final Pleistocene age.At the Ust-Kyakhta site, in the southernmost part of the region near the border with Mongolia, the 1978 excavation campaign of the first cultural layer revealed stone artefacts, wedgeshaped cores and scrapers, ostrich eggshell beads, and about 10 small pieces of pottery, including two rim fragments about 2 cm long (Aseev 2003.35-37;Medvedev 1995).The pottery is tempered with mineral particles and crushed ostrich eggshells.The diameter of the vessel was up to 10 cm.A 14 C date on animal bone from cultural layer 1 is 11 505 ± 100 BP (SOAN-1552).At the Studenoe 1 site in the Chikoi River basin, southern Transbaikal, the earliest pottery was found in cultural layers 9 and 8 (Khlobystin and Konstantinov 1996.306).It is represented by fragments of a sharp-based vessel with thin walls and string impressions.This thin-walled (0.2-0.3 cm) pottery was made using the paddle and anvil technique (Tseitlin and Aseev 1982.110).The overlying cultural layers 7 and 6 have similar pottery.The 14 C dates associated with this pottery are: 10 450 ± 300 BP (GIN-5493) for cultural layer 7b; 9620 ± 250 BP (GIN-5492) for layer 7; and 10 780 ± 150 BP (GIN-4577) for layer 6 (e.g.Konstantinov 1994.85;Kuzmin and Orlova 2000.359).However, Konstantinov (1994.85)rejected these 14 C values; he also stated that the reason for such an old age of the Transbaikal Neolithic remains unclear, and determined the age of the Early Neolithic in Transbaikal as c. 6500-5500 BP (Konstantinov 1994.153-155).Therefore, the situation with final Pleistocene 14 C dates in possible association with pottery at the Studenoe 1 is still obscure.
String and cord impressed pottery became common in Siberia after c. 7000-6000 BP (e.g.Kuzmin and Orlova 2000).Nowadays, in the light of the very early 14 C age of the Ust-Karenga complex pottery, the question 'How old is comb-patterned pottery in Siberia?' becomes an important issue related to the Neolithisation of the region.The earliest sites with pottery decorated with a comb-pattern and incised ornamentation, besides the Ust-Karenga complex,

Fig. 6. Cores from the cultural layer 7 of Ust-Karenga 12 site (on Figures 6-11, each bar unit is 1 cm long).
Material dated 14 C age, BP are known now in the central West Siberian Plain.They are located in the upper reaches of the Konda River, within the larger Ob River basin, about 3000 km west of the Ust-Karenga sites (Fig. 1).A cluster of Neolithic sites was found in the 1960s on the shore of the Satyiginsky Tuman Lake, in the Sumpanya River mouth area (geographical coordinates: 59°48' N, 64°49' E).Pottery with both incised and comb ornamentation on the surface of sharp-bottomed vessels was determined as the 'Sumpanya' type (Kovaleva et al. 1984;Krizhevskaya and Gadzhieva 1991).At the Sumpanya IV site, a series of charcoal 14 C dates were obtained: 6850 ± 60 BP (LE-1440) from a hearth; 6520 ± 70 BP (LE-1813) from a burnt tree log; and 6590 ± 70 BP (LE-1814) from the dwelling floor.At the Sumpanya II site, charcoal collected in association with Sumpanya-type pottery was dated to 6530 ± 70 BP (LE-1818) (Kovaleva et al. 1984.38).At the Sumpanya VI site, three 14 C dates on charcoal from the cultural layer with Sumpanya pottery were generated: 6100 ± 70 BP (LE-2540); 9130 ± 80 BP (LE-2554); and 9920 ± 80 BP (LE-2772) (Krizhevskaya and Gadzhieva 1991.85).Kosarev (1996.262)and Timofeev and Zaitseva (1996.344)accepted the 14 C dates from these sites in the range of c. 6850-6100 BP.However, they did not include the 14 C values of c. 9130-9920 BP from the Sumpanya VI site in their databases (Timofeev and Zaitseva 1996;Timofeev et al. 2004).Furthermore, 14 C dates for Sumpanya IV sites in excess of c. 10 000 BP, released after the original publication of the site's materials, i.e., 10 100 ± 100 BP (LE; No. is not given); 10 910 ± 100 BP (LE-1817); and 11 970 ± 120 BP (LE-1812) (Krizhevskaya and Gadzhieva 1991.85),were not taken into account.Indeed, it is hard to explain such a large variation in a date series from the same site, especially in the case of Sumpanya IV.This was noted by Krizhevskaya and Gadzhieva (1991) due to the absence of earlier cultural complexes at the Sumpanya IV site.
New archaeological and chronological data were recently gained from other sites in central Western Siberia with the Sumpanya type of pottery.At a cluster of sites on the shore of Lake Andreevskoe near the city of Tumen (geographical coordinates: 57°01' N, 65°51' E), four pottery types were determined at locality VIII (Usacheva 2001).The earliest pottery of Sumpanya appearance with incised and comb ornamentation from dwelling 7 is associated with a 14 C date of 9140 ± 60 BP (LE-2296).
Therefore, it is possible to correlate tentatively the Sumpanya pottery type from Western Siberia with 14 C dates of c. 9900-6100 BP; more research is needed to explain the older values of c. 10 100-12 000 BP. Currently, it is safer to accept the 'conservative' opinion on the Holocene age of the Sumpanya pottery (e.g.Kosarev 1996;Usacheva 2001).As was recently highlighted, the discrepancy between the 14 C and archeological ages in West Siberian prehistoric complexes is most commonly connected with uncertain taphonomic situations, when carbon material which could not be related to human occupation was 14 C-dated (Kosintsev et al. 2004.21).
Another important issue is the search for the 'roots' of the Ust-Karenga culture.Based on the most recent results, final Pleistocene pottery is known from East Asia, including the southern part of China, the Japanese Archipelago, and the Russian Far East (Amur River basin) (e.g.Derevianko and Medvedev 1995;Barnes 1999;Lapshina 1999;Keally et al. 2004;Kuzmin 2006;Nesterov et al. 2006).Pottery seems to appear almost simultaneously in these three different regions of East Asia, at c. 13 700-13 300 BP, and in each case pottery-making technology was most probably invented independently (e.g.Kuzmin 2006.368-369).There are some similarities and differences between the pottery from the Ust-Karenga complex, the Incipient Jomon of Japan, and the Initial Neolithic of the Amur River basin.For example, plant fibre tempering is common in the Ini-tial Neolithic complexes of Osipovka and Gromatukha in the Amur River basin (e.g.Kuzmin 2006;Derevianko and Medvedev 2006.130),although some plant-tempered pottery is known in the Incipient Jomon (e.g. Jomon Jidai Sosoki 1996.46, 63;Keally et al. 2003.5).On the other hand, pottery from the Amur River basin is flat-based, while most of the Incipient Jomon vessels are sharp-based.Therefore, the possible source of pottery origins for the Ust-Karenga complex may be provisionally suggested in the Amur River basin.This does not exclude the possibility of the independent invention of pottery-making in northern Transbaikal at the end of the Pleistocene, c. 12 200-10 700 BP.At the modern stage of research, the final answer to the question 'What is the origin of the Ust-Karenga pottery?' remains open to discussion.
As for the implications of the Ust-Karenga pottery to the broader Eurasian aspect of the emergence of the Neolithic (sensu Chard 1974;Barnes 1999;Kuzmin 2006.362), it is important to keep in mind the very early emergence of pottery-making in remote northern Transbaikal, far from traditional 'centres' of the origin and spread of prehistoric technological innovations such as East Asia and the Near East (e.g.Sherratt 1980).Based on the results of archaeological studies in East Asia, Siberia, and Europe in the last few decades, it becomes clear that the process of Neolithisation was very 'unlinear' (e.g.Budja 2005;2006), and there is no direct correlation between environmental conditions and the appearance of pot- On the other hand, some authors (Dolukhanov 2004. 231-235;Dolukhanov et al. 2005Dolukhanov et al. .1456Dolukhanov et al. -1457) ) have accepted early 14 C dates from the pottery sites in East Asia and Siberia, and suggested the spread of pottery-making from the east to the west, reaching the southeastern periphery of Eastern Europe at c. 7000 calBC, which roughly corresponds to c. 8000 BP (Reimer et al. 2004(Reimer et al. .1054)).This conclusion re-mains quite debatable, and more research is needed in order to understand the spatial-temporal patterns of the Neolithisation of Eurasia.
It is feasible to see two main trajectories of the Neolithisation process in Eurasia: the 'agricultural' route from the Levant towards Europe (e.g.Mellaart 1994), and the 'hunter-gatherer' route from East Asia towards Siberia and Europe (e.g.van Berg 1997).
They represent two fundamentally different processes: the emergence of food production and the appearance of food containers, and should be treated separately in terms of the meaning of the term 'Neolithisation'.In this case, extreme caution should be taken when one is trying to model the spread of the Neolithic in Eurasia.A recent attempt by Davison et al. (2006) (see also Timofeev et al. 2004.36, 63, 70-72) seems to mix 'apples and oranges', by determining the Neolithic as an agricultural phenomenon which emerged in the Near East.However, their model (Davison et al. 2006.648)shows the spread of the 'Neolithic' from the Levant, where it is dated to c. 10 300 BP at Jericho (e.g.c. 7000 BP, while agriculture was unknown in these regions until at least the beginning of the Bronze Age, c. 4500-4000 BP (e.g.Merpert 1994;Oshibkina 1996b).This is due to combining two different phenomena, the Levantine-derived 'agricultural' Neolithic and the pottery complexes of 'hunter-gatherer' type originating somewhere in East Asia.

Conclusion
The discovery and excavations of the Ust-Karenga cluster in northern Transbaikal brought to light new data on the emergence of the Neolithic in Siberia.It is evident that cultural layer 7 at the Ust-Karenga 12 site contains the oldest pottery west of the Amur River basin, and it is also one of the earliest ceramic complexes in northern Eurasia, dated to c. 12 200-10 700 BP (c. 12 200-10 200 calBC).The modelling of the Neolithisation process in Eurasia should be conducted with a more complete understanding of the nature of this phenomenon.In East Asia and Siberia, the origin of the Neolithic is related to the appearance of pottery vessels for storing and processing food in hunter-gatherer communities long before the invention or adoption of agriculture and/or animal husbandry.

Fig. 1 .
Fig. 1.General position of the Ust-Karenga cluster of prehistoric sites in Northern Asia.
12 180-12 170 BP (or c. 12 200-11 900 calBC).The hearth charcoal gave slightly younger ages, c. 11 240-10 750 BP (or c. 11 300-10 700 calBC).Three pottery temper 14 C dates, c. 11 070-10 600 BP (or c., are similar to those on charcoal.Therefore, it is safe to say that the age of cultural layer 7 at the Ust-Karenga 12 site is about 12 200-10 600 BP (or c. 12 200-10 200 calBC; 14 150-12 150 calBP), and this makes the pottery from cultural component 7 the oldest in Siberia.The quite 'advanced' appearance of the Ust-Karenga pottery may mean that it originated even earlier, if we take into account that an area of only 25 m 2 of cultural layer 8 has been excavated so far.Thus, we should not exclude the possibility that pottery at the Ust-Karenga 12 site may be found in the earlier component 8, dated to c. 12 700-13 600 BP.The palaeo-environmental reconstruction of cultural later 7 is based on the results of palynological analysis.An environment of cold grass steppe and open pine-larch forest, with dwarf birch, alder, and coldadapted lycopodium moss (Selaginella sibirica) existed at the time of site activity at c. 12 200-10 600 BP.This kind of vegetation is typical of the Pleistocene-
We are pleased to acknowledge Professor M.Budja (University of Ljubljana, Slovenia)for his hospitality and a chance to share knowledge about the Ust-Karenga 12 site with participants at the 13 th Neolithic Seminar'The Mesolithic and Neolithic Cultural  and Population Trajectories in Eurasia', in Ljubljana  on November 10-12, 2006.This research was supported by the Russian Foundation for Basic Research (RFFI), grants06-06-80258, 06-06-80108, and 05- 06-97208;and by the Russian Foundation for the Humanities (RGNF), grant 06-01-00466A.The Department of Archaeology, University of Ljubljana, kindly provided accommodation for Y. V. Kuzmin during the 14 th Neolithic Seminar.
Kuijt and Bar-Yosef 1994), to the southern and central parts of Eastern Europe about 3000 years after its emergence, i.e., at