The article analyzes the materials of the site with a surface occurrence of Lotoshi artifacts (Dzungaria, Northwestern China), discovered in 2004 by the research team of the joint Chinese-Russian-American archaeological expedition. Taking into account the degree of wind corrosion of the artefacts ' surfaces, a technical and typological analysis of the monument's collections was carried out. It was found that the complex is relatively homogeneous and is characterized by a combination of Levallois and lamellar subprismatic cleavage technologies. The materials of the Lotoshi locality have analogs in the industries of the Gorny Altai, Orkhon-1 and Tolbor-4 sites (Mongolia), as well as in the Shuidongou site complexes (Ordos), which belong to the initial stage of the Upper Paleolithic. Lotoshi is the first locality discovered in Northwestern China, which belongs to the area of lamellar cultures of the early Upper Paleolithic stage. All the currently identified cultural and genetic features connect this complex with the regions of southern Siberia and northern Central Asia.

Keywords: China, the initial stage of the Upper Paleolithic, exposed complexes, stone technology.

Introduction

There are several regions in North and Central Asia where the Middle and Early Upper Paleolithic plate industries are represented. A spatiotemporal transgression is observed in the distribution of plate complexes from Gorny Altai - the westernmost region of their existence with the earliest assemblage dates-to Mongolia and Ordos - the eastern edge of their range. When reconstructing the distribution paths of Upper Paleolithic lamellar cultures, preference is clearly given to the northern route of movement of carriers of these technological traditions and / or their ideas along the mountain belt of Southern Siberia. However, the appearance of plate industries of the early Upper Paleolithic period in the center of Eurasia is not necessarily associated with the supposed migration of any population groups to such a significant distance.-

* This work was supported by the Russian Foundation for Basic Research (project N11 - 06 - 12005-ofi-m-2011).

the specified distance. Experts also support the scenario of autochthonous origin and development of this technology in several independent (possibly mutually influencing) centers. To substantiate a particular point of view on the processes of the Upper Paleolithic formation, it is necessary to attract materials from the western part of Central Asia. However, Xinjiang, which occupies a large area of Western China, has remained a huge white spot on the map of the distribution of Paleolithic sites in Central Asia for many years. At the same time, if the picture of the movement of the Silk Road realities of later times is superimposed on the Paleolithic period, then we can assume that this region was a link between the western and eastern parts of the ecumene. Despite the length of archaeological study of Xinjiang's territory, its Paleolithic past is one of the least explored stages. It is important to note that in the adjacent territories of Kazakhstan, the Russian Altai, Mongolia, and the Yellow River basin, experts have discovered numerous complexes of stone artifacts, some of which lie in stratified sites and have radiometric dates.

Interest in the archeology of Xinjiang began to form in the late 19th century, as such prominent archaeologists and orientalists as A. Stein, S. F. Oldenburg, P. K. Kozlov, S. Gedin, A. von Le Kock and others began to visit this region. Archaeological surveys began to be carried out here at the beginning of the XX century, and systematic research-only in recent decades. Geomorphologically, the territory of Xinjiang is divided into two major basins: the Tarim basin, which occupies the southern and eastern parts, and the Dzungarian Basin, which is located in the northwestern part. The Tarim basin, through which the main branch of the Silk Road passed, is more promising from the point of view of classical archaeology, it is studied much better than the Dzungarian basin, which is located away from the areas of ancient agriculture.

The first collections of Paleolithic material on the territory of Xinjiang are associated with the names of Teilhard de Chardin and Yang Zhongjian: in 1931, they discovered several large, strongly Latinized stone products of archaic appearance at three localities in the Khamian Basin (Derevyanko and Komissarov, 1997). In the 1950s, Chinese archaeologists conducted research in this region; they collected a collection of stone tools, including scrapers and axes made of large chipped pebbles. Among the artefacts found in the 1950s and 1970s at the sites of Qijiaojing and Sandaolin in Khami, Astana and Lobnor in Turfan, and others, microliths predominated. In the extreme west of Xinjiang, on the bank of the Tishkurgan River in the Pamirs, Paleolithic artifacts were discovered [Ibid.]. Only since the late 1980s have the results of these studies been published. In 1987. An Zhiming published the results of a study of materials from Neolithic sites, during which the typological sequence of traditions of making microlithic stone tools, which were mainly lifting material, was traced [Jia Weiming, Betts, Wu Xinhua, 2009, p. 169]. Based on the same materials, Wang Binghua concluded that these sites, as well as similar Late Paleolithic and Mesolithic sites of Northern China, belong to the same culture, but cannot be attributed to the period earlier than 10 thousand BC (cit. by Derevyanko and Komissarov, 1997).

After a long break, archaeological exploration in Xinjiang resumed. In the second half of the 1980s and early 1990s, joint expeditions of the Academy of Sciences of the People's Republic of China and the University of Arizona (USA) revealed several localities: sporadically occurring (with low density) Paleolithic remains were found in the Tashkent region, on the western periphery of Xinjiang [New Achievements..., 1995], as well as on the southern tip of the Tarim Basin [Ibid., 1995; Wang and Du, 1997].

In Dzungaria, research was concentrated in the valley of the Ertsisi River (Irtysh), in the Khabakhe District. Bifaces found here and scrapers found in the same valley at the Qichengzi site may be forms that represent Levallois technology [Jia Weiming, Betts, and Wu Xinhua, 2009, p. 172]. In 1993, six localities were discovered in the Ercisi Valley, two of them with rich material: 93ANE2 (660 units, including 107 guns) and 93ANE4 (133 units, including 18 guns). Among the tools, there are groups of pointed tools, end scrapers, and notched tools. The same typological set was found during a survey of the Jiaohe and Chaiwopu localities in 1995. As Zhang Chuan notes, based on these findings, it was supposed to distinguish a specific industry of "pointy-end scrapers" (cit. by Derevyanko and Komissarov, 1997, p. 45).

Paleolithic sites discovered in Xinjiang are monuments with open deposits of artifacts. The earliest stage, according to Chinese researchers, is represented by the Mousterian tradition, and the Levallois technology is probably found from the western part of the Ercisi River valley, as well as from the Kichenzhi site in the eastern part of this valley (La Wei Ming, Betts, and Wu Xinhua, 2009, p. 172). The Jiaohe Collection, according to Prof. Zhang

Senipuya is similar to the late Shuidungou artefacts (see Derevyanko and Komissarov, 1997). Thus, despite the gradually expanding source base on the Paleolithic of Xinjiang, systematic survey work on the problems of the Stone Age of the region, as well as professional publication of previously obtained materials, is relevant.

In June 2004, the research team of the joint Chinese-Russian-American archaeological expedition (headed by Academician A. P. Derevyanko of the Russian Academy of Sciences, Professor D. Olsen of the University of Arizona, and Prof. Gao Xing (Institute of Vertebrate Paleontology and Paleoanthropology of the Chinese Academy of Sciences)) an archaeological survey was conducted in the northwestern part of Xinjiang (Dzungaria). It became a kind of logical continuation of the archaeological research conducted in the mid-1980s. The Soviet-Mongolian complex archaeological expedition on the territory of the Mongolian Altai. Some routes of this expedition passed through the eastern (relative to Dzungaria) part of the Altai, in the valleys of the Barlagin-Gol and Uench rivers (Derevyanko et al., 1990). The territory of Dzungaria and the so-called Dzungarian Pass (a break in the chain of the Central Asian mountain range system located at the junction of the Russian and Kazakh Altai, the mountain systems of Mongolia and the Yellow River basin) was chosen as the object of study, taking into account its extremely important geographical location.

The Dzungarian Plain, part of Northern Xinjiang, covers an area of approximately 770 thousand ^km2. The Dzungarian Plain is included in the Altai Valley system, which is connected by "through corridors" that cross the mountain system itself (Luzin, 2007). The mountains surrounding the Dzungarian Basin from the north and west occupy an intermediate position in terms of their bioclimatic features between the mountains of Southern Siberia and the Central Asian mountain Country (Gvozdetsky and Golubchikov, 1987). In the north-west, Dzungaria is bounded by the middle-altitude Saur and Tarbagatai ranges, which are separated from the Dzungarian Alatau system consisting of several parallel high chains, in the south-southwest by the Eastern Tien Shan ranges, in the south by the Bogdo - Shan Range, and in the northeast by the Mongolian Altai Range (Selivanov, 1965).. The average height of the surrounding ridges is approx. 3 thousand meters.

The Dzungarian basin belongs to the Quaternary depression inherited from the Meso-Cenozoic basin. It accumulates thick loose deposits of gravel-pebble-boulder material and other sediments. The average elevation of the central part of the basin is 300 m above sea level and increases as it approaches the surrounding ridges. The territory belongs to arid regions: the basin receives about 50-200 mm of precipitation per year. At the same time, the evaporation potential reaches 2-3 thousand mm. According to geological or hydrological aspects, the Dzungarian plain can be divided into three zones: infiltration, oasis, and salt marshes (transitional form, semi - desert) [Chen Mengxiong, 1995]. The Dzungarian basin is 600 m higher than the neighboring Tarim Basin, so the climate here is less arid. Nevertheless, the basis of its landscape is made up of semi-deserts and steppes. Steppes develop in the foothills, and forests grow on the mountain slopes. Low mountain ranges turn into a plain, where salt marshes interspersed with oases. In the center of the Dzungarian plain are the loose sands of Dzosotyn-Elisun, which form frequent ridges. The Dzungarian plain is strongly affected by Aeolian processes, as a result of wind destruction of the inclined reservoir plain, so - called Aeolian cities appear - ridge hills and Aeolian remnants that rise to a height of 180-350 m above the bottom of the basin (Ananyev, Simonov, Spiridonov, 1992, p.136).

In the Dzungarian basin there is a poor river network with flow to the west, due to the prevailing slope of the plain surface, where the large lakes Ebi-Nur, Manas, Ulyungur, Baga-Nur are located. Lakes occupy mainly the outskirts of the plain. The largest rivers are the Irtysh, Manas, and Urungu. Most of the smaller rivers fill up only during heavy rains. In cold, low-snow winters, the rivers of Dzungaria freeze for several months. All rivers of this system originate in mountain systems and drain one or two basins.

Location and geomorphological situation of the location

Technical and typological characteristics of the artifact complex. The route of the expedition, passing through the natural borders of Dzungaria, ran through the territory of the ROC. Tachen and Altai. During the work, 32 sites of the Stone Age era were investigated. One of the most striking objects discovered during the exploration was the location of Lotoshi (Tachen district, north-west of the capital of the Xinjiang Uygur Autonomous Region, Urumqi). The expressive materials of this monument can make important additions to our understanding of the processes of formation of the Upper Paleolithic of Central Asia.

The central point of Lotoshi's location has coordinates: 46° 38 '56.0" N, 86° 02 '07.8" e. The altitude above sea level is approx. 1 thousand sq. m. Pa-

Fig. 1. Map-layout of the Lotoshi location.

mint is located on a slightly sloping foothill plain adjacent from the east to the spurs of the Dzungarian Alatau of the Saur-Tarbagatai mountain system, on both sides of the Karamai-Altai highway (Fig. 1). The plain is covered with gravelly eluvial deposits, dissected by parallel-running remnant ridges, one of which is called "Camel's Back" because of its characteristic shape (po-kit. Lotoshi). On the vast territory of the monument (total area approx. 3x5 km), outcrops of fine-grained dark siliceous material (probably silicified sandstone), shaped like blocks and tiles, were found; it served as a raw material for the manufacture of artifacts.

In the course of research in 2004, the territory of the site was divided into several research sectors, which were assigned the letter designations A, B, and C. Within these sectors, sites where artifacts were collected were additionally allocated. The sites were not the same in terms of the abundance of finds - from a few hundred to a few hundred. A sequence number was added to the alphabetic symbols of the sectors, for example, A-1, A-2, etc. The analysis of materials showed that the most numerous locality complexes differ in the ratio of percentage indicators of the main categories of stone inventory - nuclei, non-etched chips and tools. Fundamental differences in the technical and typological appearance of the inventory could not be traced. This study presents the summary results of the analysis of material from all points of the Lotoshi location.

594 artifacts were studied. The main raw material for making artifacts is a dark gray, black fine-grained siliceous material, which was initially presented, probably in the form of nodules and sub-rectangular tiles or blocks. Somewhat less frequently, pebbles were used as initial raw materials. There are also artifacts made of quartz and quartzite. The surface of the products bears traces of wind deflation. According to the degree of surface deflation, the following groups of artifacts are distinguished (534 units, chips less than 30 mm are excluded): with an undeflated ("fresh") surface-12 units (2.2 %), with a surface of a weak degree of deflation - 304 units (56.9%), medium - 195 units (36.5%), strong-23 units (4.3 %). Thus, the main part of the complex consists of artifacts with a weak and medium-deflated surface, among which products with a surface of a weak degree of deflation predominate.

Nuclidean shapes. They make up 14 % of the total number of artifacts (Table 1). When analyzing the ratio of the main morphological categories of nuclei, a large proportion of products representing the initial stage of cleavage is noteworthy; more than 20% of all nucleoid forms are preforms and test pieces of rock (Table 1). Such an abundance of preforms is explained by the fact that at the same time, the number of preforms in the on the territory of the monument there are exits of raw materials for the manufacture of artifacts.

Table 1.

Typological composition of nuclei from the Lotoshi locality

About 76 % of the nuclei on the surface have traces of a natural crust. The size of the nuclei is large: length from 68 to 267 mm (average-134 mm), width from 42 to 168 (average - 102), thickness from 22 to 180 mm (average - 52 mm). An average of 3.8 chips were removed before the nuclear impact site was stopped splitting. These data indicate that the utilization of nuclei was not intensive: most of them passed the stage of approbation or the initial stage of reduction.

Nuclei with a flat front predominate, and the proportion of subprismatic or bulk nuclei is small. This is most likely due to the characteristics of the raw materials (tiles and blocks) that determined the specifics of the technology.

Several types of nucleoid products are distinguished.

These are rocks with test chips (Table 1). These are sub-rectangular blocks or slabs of dark gray or black fine-grained sandstone or quartzite (two items). Along their natural faces, single multi-dimensional measurements were carried out without shaping the impact plane or cleavage plane. The length of objects is from 74 to 235 mm (average-148 mm), width from 58 to 146 (average - 102), thickness from 45 to 180 mm (average - 74 mm). Two items have a surface of medium and ten-a weak degree of deflation.

Preforms (Table 1). Among them, five items can be attributed to preforms of nuclei (subprismatic or planar) intended for rectangular removal in a sub-parallel direction. On one or two narrow ends of rectangular blanks, shock pads were formed. The bulge of the future workers ' front was given by parallel withdrawals. Along one of the longitudinal faces of the blank, a sinuous edge was formed by removing flakes and additional touch-up with retouching. After its removal, regular removal of workpieces could begin. The length of objects is from 114 to 235 mm (average-191 mm), width from 88 to 135 (average - 112), thickness from 47 to 80 mm (average -63 mm). One item has a medium deflation surface and five have a weak deflation surface.

Monofrontal single-site nuclei (Table 1). They are usually sub-rectangular or oval in shape. On one of the sides of the object, with the help of transverse shots, a shock pad was formed, from which plates or flakes were removed in a sub-parallel direction, sometimes traces of correction are traced on the front from laterals perpendicular to the direction of the main shots. Depending on the location of the impact pad and the degree of convexity of the front, three types of nuclei are distinguished: with a flat front (Fig. 2, 2), and subprismatic - at the narrow ends of the blanks, beveled impact pads were formed, from one of which plates were removed. They occupied from 1/2 to 3/4 of the perimeter of the convex impact pad. On the laterals, sinuous edges were formed using transverse chipping and retouching. These laterals could serve as guiding edges for future shots. This form of nuclei is typical of the early Upper Paleolithic of Transbaikalia and Mongolia (Figs. The Paleolithic complexes of Central Asia are also characterized by the third variation of Lotoshi nuclei - transverse (2 units). On the splitting front formed in the direction perpendicular to the long axis, negatives of short flakes can be traced parallel to each other. Items are 75 to 230 mm long (average-121 mm), 73 to 155 mm wide (average - 100 mm), and 55 to 87 mm thick (average - 58 mm). One core has a strong deflation surface, three - medium and eight - weak; one item is undeflated.

Monofrontal two-site nuclei (Table 1). One of the most numerous types of complex nuclei. According to the degree of convexity of the front, the direction of withdrawals, and the location of shock pads, there are several types of cores in this category. The first type consists of nuclei with sub-parallel counterexamples and a flat working edge of -12 copies. Nuclei of this type are typical for Early Upper Paleolithic industries. As a rule, they are sub-rectangular or oval in shape, planar, and in rare cases negative images are located on the laterals. The main cleavage, which was obtained from these cores, was plates or rectangular flakes. Withdrawals were made from impact pads located at opposite ends, but most often not in an alternating order, but after a series of chips from one of the pads. On average, judging by the residual negatives of the shots, three or four pieces were removed from the main impact site, two from the auxiliary one, and less often three chips. Removal from the opposite sites was carried out in order to maintain the convexity of the splitting front. Consequently, the system of cleavage of nuclei can hardly be characterized as a typical bipedal one, according to which samples from opposite sites set the shape of each other (Fig. 2, 3, 6). The nuclei are significant in size, the largest being 232x140x58 mm (Fig. 2,5). The second type is Levallois planar nuclei (3 units), the preparation of which is associated with the implementation of "preferred" chips, the shape of the latter is set by preliminary correction from laterals and opposite impact pads (Figs. 3, 1). The third type is nuclei with traces of sub - parallel counter removals and a subprismatic working front. Used as blanks-

2. Nuclei (1-3, 5, 6) and tool (4). Location of the Lotoshi.

3, 3). The cores of this type are also characteristic of the early stages of the Upper Paleolithic, but one of them has a strongly exposed surface (Fig. 3, 4). Negatives of plate removals made in the opposite direction occupy 3/4 of the perimeter of the sites and enter the laterals of the nucleus. A transverse nucleus was found (Fig. 3, 5) and three two-site monofrontal subcrosses (orthogonal) (Fig. 3, 7). The length of objects ranged from 91 to 232 mm (average -144 mm), width from 71 to 148 (average-104), thickness from 24 to 107 mm (average-51 mm). One nucleus has a strong deflation surface, 4 - medium and 17-weak.

End nuclei (Table 1). For two nuclei, the working front with the negatives of removing narrow plates is located on a narrow plane of the workpiece. The surface of products with an average degree of deflation (Fig. 3, b).

Centripetal (radial) nuclei (Table 1). Representative is a series of nuclei bearing negatives of different centripetal chips on the cleavage fronts, taken from impact pads designed on the laterals-edges. A smaller part of the objects are decorated on quartz pebbles, the basis for most of the nuclei were tiles. The nuclei that characterize the initial stage of cleavage predominate, with areas of natural crust preserved on the working fronts (75 % of the nuclei; Figs. 4, 7, 2, b). Three nuclei (Figs. 4, 3, 4) can be attributed to the classic "turtle-shaped" ones for one "preferred" removal, but this cleavage was not removed from two of them. In four subjects, the working fronts were located on both planes of the nuclei (Fig. The length of objects is from 68 to 267 mm (average-117 mm), width from 64 to 168 (average - 105), thickness from 25 to 72 mm (average -

Fig. 3. Nuclei. Location of Lotoshi.

41 mm). Three nuclei have a surface of a strong degree of deflation, 6 - medium and 16 - weak.

Bifrontal two-site nuclei (Table 1). The surface of all objects has a weak degree of deflation. On both planes of the core, negatives of parallel shots are fixed, located in accordance with the change in orientation of the longitudinal and transverse axes of the products (see Figs. 3, 2).

Chip industry (Table 2). In the collection, 31% are products with traces of natural crust. This means that a significant part of the chips were removed during the reduction process in the early stages of using the nuclei. The analysis of ways to maintain the convexity of the nucleic cleavage front helps to supplement our understanding of the features of the splitting technology at the Lotoshi site. According to the elongation index of workpieces (the ratio of the chip length (1) to its width (t)), it is possible to judge the orientation of the guide ribs and the choice of the ancient master in favor of shortened or longitudinally elongated ones

Fig. 4. Nuclei. Location of Lotoshi.

workpieces. The data obtained indicate a focus on the production of elongated blanks, even in the production of flakes. The ratio of the cleavage width (m) to the cleavage thickness (n) reflects the degree of transverse convexity of the cleavage front and shows whether the production process was aimed at producing chips that were more or less high in cross-section (obtained within the framework of a prismatic or close to it volumetric concept) or flattened (removed from nuclei with a flatter front).. According to the results of the analysis, the plates of the complex were removed by the method that assumed the maintenance of a significant convex splitting front (Table 3). The product of these removals was massive triangular and trapezoidal chips in cross-section. In terms of indicators, they are comparable to the Early Upper Paleolithic complexes of the Tolbor site in Mongolia (from 3.01 to 2.95), but differ from the same products from the Kara-Bom site, which are characterized by the concept of a more voluminous plate index (3.9) [Derevyanko et al., 2007].

Among the chips, flakes have the largest share (55.1 %). For whole flakes, the following average values are typical: length 72 mm, width 64 mm, thickness 18 mm. Flakes with parallel edges and with a sub-parallel cut characteristic of plates are very common. At the same time, the specific weight of the plates (22.7 %) is noticeably lower than that of flakes, but they remain within the parameters corresponding to the complexes of the circle of lamellar cultures of the early Upper Paleolithic period (the lower value of this indicator is 20 %). Whole plates have the following average metric values: length 97 mm, width 40 mm, thickness 15 mm. Many of the chips are chipped

Table 2.

Composition of the chip industry from the Lotoshi location

in the proximal part, a triangular negative of withdrawal is marked. Very few pointed plates with signs of bipedal cutting have been recorded. In general, the plates are very large, regular outlines. A special group consists of a small number of Levallois flakes and points: sub-rectangular and sub-triangular chips formed by counter or convergent removals, sometimes by chips of adjustment from laterals (Figs. 5, 11, 12). Attention is drawn to the extremely small number of rib-like chips and cross-section chips in the adjustment of impact pads ("tablets") The most common type of technical chipping is edge chipping with a rim.

Based on the characteristics of chips and nucleoid products of the studied sample, we can conclude that the use of planar Levallois (split-off), radial (split-off), and linear (split-off) methods is used at the site.

Fig. 5. Guns. Location of Lotoshi.

(judging by the presence of asymmetric chips with traces of convergent and centripetal faceting) and bulk flake and plate splitting. The diversity of cleavage methods is reflected in the significant variability of the dorsal faceting of cleavages (Table 4). The main (though not predominant) ways of organizing the cleavage fronts of nuclei were parallel and sub-parallel unidirectional and bipedal removal. Orthogonal (subcrest) and centripetal (radial) faceting options for dorsal surfaces at the stage of initial finishing of nuclei are very widespread.

Table 3.

Average values of the ratio of width (m) to thickness (n) and length (l) to width (m) of chips in the industry of the Lotoshi location

See Table 4.

Types of dorsal surfaces of chipped Lotoshi localities

See Table 5.

Types of Impact Pads Chipped Lotoshi Locations

Table 6.

Typological composition of tools of the Lotoshi locality complex

The share of dihedral and faceted sites is small, reaching a total of 15.3 % (Table 5). The most common types of sites are natural and smooth.

Tools (Table 6). Make up 26 % of the total number of artifacts. Since the artefacts of the Lotoshi site were located in the conditions of surface occurrence, the screening of objects with signs of pseudo-retouching was carried out with the utmost care. The tools are mainly treated with dorsal retouching (49 %), there are also tools with elements of ventral (17 %), alternating (24 %), opposite (7 %) and bifacial (3 %). Parallel retouching was used most frequently, 48% of tools have traces of it; 19% of products were processed with scale retouching, 8% with step retouching; a combination of different types of retouching was recorded on 25 % of products. Signs of weakly modifying retouching were observed on 23% of the tools, medium modifying retouching - on 40%, and strong modifying retouching - on 37%. In the Early Upper Paleolithic industries of Southern Siberia and Mongolia, products with elements of high-and medium-modulating retouching make up 60-40% of the total number of tools in the complex (Vasiliev and Rybin, 2009). Thus, the specific weight of tools subjected to intensive, modifying secondary processing is very large - 77 %. The share of tools in Lotoshi collections is noticeably higher than in the workshop industries. Probably, operations were also carried out on the territory of the monument to make simple forms of tools.

In terms of structure, the collection of chipped blanks for making tools is close to the set of non-retouched items (see Table 2). However, the share of plates in the composition of tools is slightly higher than among items without signs of secondary processing.

There are several typological groups of tools.

Crenellated tools (Table 6). They are represented by whole items (30 units) and fragments. Flakes served as blanks for 23 tools, plates-for 8, Levallois flakes - for 4 products; one tool was made from tiles. The length of items (only whole items were counted) is from 47 to 164 mm (average - 88 mm), width is from 37 to 113 mm (average - 55 mm), and thickness is from 8 to 28 mm (average - 14 mm).

Ten items have a medium deflation surface, 25 have a weak deflation surface, and one weapon has a non-deflated surface.

Toothed-notched tools are characterized by various variations in the formation of the working edge. Taking into account the specifics of the complex, this type includes items that are processed with regular retouching and have retouched recesses. The collection includes tools made in typical models of the collection.

cleavage-plates with a central negative of a triangular cleavage and flakes with a sub-parallel and radial cut of the dorsal surface (see Fig. 5. 1,3, 8, 10). On the working edge of 33 tools, elements of medium - and highly modifying parallel and scaly retouching and retouched recesses are combined, on 24 objects, signs of retouching are localized on two working edges of products or more.

Scrapers (Table 6). Tools of this type are usually made of flakes (13 units). The length of objects is from 48 to 210 mm (average-95 mm), width from 41 to 102 (average - 74), thickness from 12 to 64 mm (average -24 mm). Six items have a surface of medium and 11 - a weak degree of deflation. Multi-word forms predominate. At the same time, traces of highly modifying multi-row processing with flake or step retouching rarely occupy a significant part of the perimeter of the tool. Single-row weakly modifying parallel retouching often played an important role in the design of tools-it formed additional blades. There are several types of scrapers: longitudinal (4 units), represented mainly by variants with a single processed blade (see Fig. 5, 9); transverse (4 units), with the working edge formed in the distal part of the workpiece (see Fig. 5, 2); double angular convergent (4 units), with a single processed blade (see Fig. 5, 3; fig. 6, 1); limas (1 unit); longitudinal-transverse (2 units; see Fig. 5, 5); scrapers processed along the perimeter (2 units; see Fig. 6, 2).

Scrapers (Table 6). There are five objects with a medium deflation surface and three with a weak deflation surface. There are two main morphological variants of scrapers: angular (4 units), made on flakes, they have slightly protruding narrow scraper blades (see Fig. 6,5) and end ones (4 units) at the corners of the distal end, slightly protruding narrow scraper blades (see Fig. mm, width from 34 to 68, thickness from 18 to 22 mm, all maximum dimensions refer to a semi-ribbed plate (see Fig. 2,4)). Parallel semi-circular and vertical retouching was applied to the distal end (see Figs. 5, 7), and one product shows a correction of the impact bump and longitudinal edges using flat scaly retouching (see Figs. 6, 3).

Retouched plates (tab. 6; see fig. 6,4). Whole items (16 units) and fragments were found. The length of whole items is from 62 to 180 mm (average -108 mm), width from 33 to 76 (average - 49), thickness from 6 to 30 mm (average - 16 mm).

Two items have a strong deflation surface, 8 - medium, 14 - weak, and 1 - undeflated.

Twelve of the tools were treated with parallel retouching, 6 with scaly retouching; the remaining tools are mixed processing options, or the retouching characteristics are not reconstructed due to weathering. 19 items were processed with low - and medium-modifying retouching; the rest have traces of highly modifying retouching. In 15 guns, retouching traces are located on two longitudinal edges, but they occupy an insignificant part of the working edge. It can be noted that most of the tools bear traces of mainly light processing, which implies short-term use in labor operations.

Retouched flakes (Table 6). Include 33 whole items and fragments. Whole artifacts are 35 to 152 mm long (average - 81 mm), 25 to 141 mm wide (average - 64 mm), and 7 to 47 mm thick (average - 18 mm).

Ten items have a surface of moderate deflation, 24 - weak, 2 - undeflated.

Seventeen of the guns were treated with parallel retouching, 8 with scaly retouching, while the rest of the guns represent mixed processing options or have traces of undetectable retouching. 30 items were retouched with weak and medium modification, and 6 guns were retouched with strong modification. In 15 guns, the signs of retouching are localized on one and two longitudinal edges, but they occupy an insignificant part of the working edge. Flakes, like retouched plates, have traces of light processing, which implies short-term operation of the product.

Points (Table 6). One product has a surface of a weak degree of deflation, the rest - medium. There are several types of points. Asymmetric (2 units) are made on a chip and a large plate (dimensions 106x42x13 mm). Both longitudinal edges of the tools are treated with semi-circular parallel and scaly retouching, forming a working pointed edge deviating from the longitudinal axis of the tool (see Figures 6, 6). The complex also includes a symmetrical tip on the plate, a tip with ventral adjustment of the distal end, and a Levallois sub-triangular retouched chip. The last product has no distal end, but it can be seen that both longitudinal edges show signs of retouching and convergence (see Figs. 5, 6).

Bifaces (Table 6). The series of bifacially processed tools is small, but expressive. Three products of oval or sub-rectangular shape, biconvex in cross-section, are treated with double-sided chips on both faces and additionally corrected along the edge with retouching. They have a weak deflation surface. Their length is from 68 to 111 mm, width from 53 to 75 mm, thickness from 13 to 28 mm (Figs. 7, 1, 6). Two other products-leaf-shaped biface and partial biface-have a more elongated shape, a sub-rectangular shape-

Fig. 6. Guns. Location of Lotoshi.

in the cross-section. Their surface is of a moderate degree of deflation. Length from 88 to 182 mm, width from 34 to 71 mm, thickness from 20 to 40 mm (Fig. 7, 3).

Spiked tools (Table 6). Products of this type are typical of the Paleolithic of Central Asia; they are widely represented in the industries of many sites of the early Upper Paleolithic in Mongolia and Transbaikalia. The peculiarity of these tools is due to the formation of a perforating element - a "spike" - using various combinations of retouching and retouched recesses in the distal part of the workpieces or at the corners formed by the intersection of the longitudinal and transverse edges (Fig. 7, 5, 7). The collection includes whole items (8 units) and fragments. Flakes served as blanks for nine guns, and plates served as blanks for four. The length of items (including only whole items) is from 41 to 127 mm (average -64 mm), width is from 30 to 92 mm (average - 50 mm), and thickness is from 9 to 32 mm (average - 15 mm). Two items have a strong deflation surface, three - medium, seven - weak, and one - undeflated.

Combined tools (Table 6). A combination of various morphological elements is recorded on the workpiece: a point and a toothed tool (1 unit), a toothed tool and a multi-facet cutter (1 unit; Fig. 7, 10), a scraper and a toothed tool (4 units). All products are unfragmented. The blanks for the four guns were

Fig. 7. Guns. Location of Lotoshi.

flakes, for two-plates. Items are 42 to 142 mm long (average - 68 mm), 32 to 58 mm wide (average - 46 mm), and 8 to 16 mm thick (average - 12 mm). Three items have a surface of moderate deflation, two - weak, one - undeflated.

Designed on a chip with a cross-cut edge (dimensions 88x48x18 mm). The opposite edge is treated with a semi-circular single-row regular parallel retouch.

Choppers (Table 6). They are made of quartzite pebbles, one of the sides of which bears traces of one-sided padding forming the blade. One product has an average deflation surface (Fig. 7, 2), one - weak, and another - undeflated. The guns are 77 to 104 mm long, 75 to 110 mm wide, and 26 to 51 mm thick.

Discussion

The Lotoshi complex, as noted, includes objects with a surface of varying degrees of deflation, however,

from our point of view, there is no need to attach chronological significance to this feature and, taking it into account, distinguish cultural and chronological groups. The industry has a homogeneous appearance. Artifacts that differ in the degree of surface deflation do not form stable complexes characterized by a specific splitting technology or product typology. Some items show signs of both moderate and mild deflation, depending on the quality of the raw material or how much of the artifact was turned in the direction of the prevailing wind rose in the area. The safety of the surface of objects also depended on whether the artifacts were covered with loose deposits at some time or were located on the surface.

How can one determine the cultural and chronological affiliation of the Lotoshi complex? The composition of the cores (a significant proportion of preforms and rock testing blocks), as well as their morphological characteristics (very large average sizes of nuclei, a high specific weight of objects with traces of natural crust on the surface) suggest that the primary finishing of nucleoid forms typical for workshops was carried out on the territory of the settlement. Judging by the peculiarities of the preparation of nuclei, the splitting technology was largely determined by the initial bar-like or tile shape of nodules. The use of natural faces of workpieces made it possible not to apply the technique of removing finned plates on a large scale. The most common type of technical chipping is plates with a rim. The nuclei industry consists of one-and two-site nuclei with traces of sub-parallel, centripetal, and sub-cross (orthogonal) withdrawals. The collection includes several nuclei that can be identified as Levalloisian, as well as a few subprismatic nuclei that are similar in shape to the finds from the Early Upper Paleolithic complexes of the region. The actual Levallois detachments (flakes and spikelets) are isolated, many of them morphologically atypical. In this complex, in comparison with other industries in the region, the proportion of elongated plates with a rather high massiveness index is relatively high. The facet index is low.

The tool set is dominated by products of simple shapes that do not have signs of sophisticated secondary finishing: retouched flakes and plates, toothed-notched and awl-shaped tools. Together, they make up about 70 % of the total number of guns. The scrapers are mostly unimpressive; they, like the scrapers (among which there are some carefully designed objects), are few in number. Thus, the main part of the tools cannot be clearly attributed to a particular stage of the Paleolithic. At the same time, certain types of tools (a small series of semi-oval lenticular bifaces, points on plates, and a multi-facet cutter) have analogues in the industries of Upper Paleolithic and "transitional" monuments of neighboring territories.

The typological composition of the complex, which is characteristic of a number of stratified and reliably dated Central Asian industries, indicates that this monument belongs to the earliest stages of the Upper Paleolithic. The complex combines planar and subprismatic nuclei for removing plates, as well as radial and single Levallois nuclei; a significant proportion is made up of plates, awl-shaped tools, retouched plates, semi-oval bifaces; awl-shaped tools have a high specific weight; scrapers are not numerous.

The most geographically close analogs to the industries of the Lotoshi locality are the complexes of monuments of the southern facade of the Mongolian Altai. Artefacts of one of them - the Barlagin-Gol-2 site-are located in open deposits, they are concentrated on the slab outcrops of raw materials for the manufacture of tools. The complex consists of a significant series of planar and subprismatic nuclei for removing plates, as well as Levallois nuclei for flaking and centripetal nuclei. More than half of the chips are large plates. Levallois sub-triangular chips are presented. The tool kit includes end scrapers, awl-shaped tools, and bifacially processed tools (Derevyanko et al., 1990). The plate industry of the Mankhan-5 monument in the Mongolian Altai belongs to the early Upper Paleolithic period [Ibid.]. Materials from another site in the Mongolian Altai, Bayan - Nur-somon-13, probably belong to the same period. His collection includes many expressive bifacial tools. This complex is also characterized by a combination of subprismatic, planar, and Levallois nuclei, as well as lamellar deposits with signs of bipedal cutting.

Among the stratified sites, Mongolian sites are also the closest to Lotoshi. The industry of horizon 3 of the Orkhon-1 site (Central Mongolia), which dates from the Middle to Upper Paleolithic period, ca.40 KA BP, can be confidently compared with the materials of the Lotoshi site. The Orkhon-1 complex shows a combination of the Levallois technology, which provided for the centripetal design of the cleavage front of the so-called turtle-shaped nuclei intended for the production of oval Levallois flakes, and the Upper Paleolithic subprismatic stone splitting technology.

There is a significant proportion of plate chips in the composition of primary cleavage waste and tool blanks, and a relatively small number of Upper Paleolithic tool types, mainly represented by scrapers (Derevyanko, Kandyba, and Petrin, 2010). In Northern Mongolia, in the Selenga basin, the formation of Upper Paleolithic industries can be traced on the example of the Tolbor-4 site; its industry shows similarities with the materials of Dzungaria. On Tolbor-4 40 - 35 There was an industry represented by complexes from cultural horizons 5 and 6 (Derevyanko et al., 2007). The Levallois component of the cleavage technology here acts only as an element of the design of the nuclei. In general, the cleavage has an Upper Paleolithic character, it is reflected by planar and subprismatic nuclei with isolated laterals. These cores are very similar to samples from the Lotoshi complex, designed for removing large, massive plates. The tool kit includes a series of lenticular semi-oval bifaces, awl-shaped tools, end and corner scrapers. In terms of the composition of the industry, the materials of Tolbor-4 and Lotoshi are close. The eastern end of the range of levallu plate industries of the initial stage of the Upper Paleolithic is represented by collections of the Shuidungou site (ca. 27 Ka BP) in the prov. Ningxia, bordering Xinjiang (Shuidonggou..., 2003; Brantingham et al., 2004). Gorny Altai is the western part of the zone of distribution of plate industries related to the transition period from the Middle to Upper Paleolithic and the initial stage of the Upper Paleolithic, the earliest dates of which are 50 - 40 thousand years ago (Derevyanko and Shunkov, 2004; Derevyanko, 2010). The connecting region between Dzungaria and Gorny Altai is East Kazakhstan. Several objects of the Early Upper Paleolithic stage have been identified here. The industries of some of them have the main features of Dzungarian and Mongolian technocomplexes. For example, the industry of such a multi-layered monument as the Valikhanov site (Taimagambetov, 1990), which apparently belongs to the initial stage of the Upper Paleolithic, is characterized by disc-shaped and subprismatic nuclei, scrapers and scrapers on plates and flakes, chopping tools, and rare incisors. Shulbinka parking lot on the Irtysh and opened in the late 1990s. Stratified sites in the Irtysh basin contain plate complexes of the Early Upper Paleolithic (Petrin and Taimagambetov, 2000; Derevyanko et al., 1999). It should be noted that it is still difficult to find analogs of the Lotoshi industry in more remote areas to the south, such as the Tarim basin, Central China or Southern Kazakhstan. At the moment, cultural and genetic connections of this complex with the regions of southern Siberia or northern Central Asia and the territory of Northern China have been identified.

Conclusions

Materials of the Lotoshi locality, according to the results of studying the collections, belong to the initial stage of the Upper Paleolithic. The features of the technology used by ancient man were determined by the specifics of the stone raw materials used and the functional characteristics of the location-the workshop where the initial disposal of nuclei and the manufacture of tools were carried out.

Taking into account the wide range of analogs of the complex, it can be assumed that the location of Lotoshi was on the path of spreading the traditions of plate industries of the final Middle-early Upper Paleolithic from the Altai Mountains through the territories of Eastern Kazakhstan, Dzungaria, Western and Central Mongolia to the Yellow River bend (Ordos). On the one hand, the monument industry makes it possible to fill in the missing link in the hypothesis of the" southern route " of the Upper Paleolithic in Central Asia. On the other hand, and this point of view seems more reasonable, there are no sufficient grounds to link the Lotoshi industry with hypothetical migrations of carriers of plate technologies from Gorny Altai. If such a population movement took place, the complex under consideration would reflect the features of two bright traditions of the early Upper Paleolithic period of the Altai Mountains - Kara-bomovskaya or Ust-Karakolskaya, which have a set of clear specific features. As our comparison showed, the similarity of the Lotoshi industry with the Altai and Xinjiang technocomplexes is manifested in the most general characteristics, as well as in the proximity of the model of transition from the Middle to the Upper Paleolithic. This model is characterized by the gradual strengthening of Upper Paleolithic features based on the Levallois base of local origin. The materials of multilayered complexes indicate that the evolution of Middle Paleolithic cultural traditions in the Altai took place 70-50 thousand years AGO, and as a result of 50-40 thousand years ago, industries of the early Upper Paleolithic period were formed here. Starting from 60 thousand years AGO, in the Middle Paleolithic industry, the specific weight of Upper Paleolithic tools (scrapers, chisels, tools designed on plates) increases and, accordingly, the proportion of nuclei for plate splitting increases, end, wedge-shaped and other nuclei for removing plates appear, and the percentage of nuclei of the Levallois and radial prin decreases.-

types of cleavage. Industrial complexes of this era are characterized by features that allow us to distinguish two independent lines of development - Ust-Karakol and Kara-bomovskaya (Derevyanko, 2010). On the territory of Xinjiang and Gorny Altai, the industries of the end of the Middle-beginning of the Upper Paleolithic combine signs of the use of planar Levallois and plate-like subprismatic cleaving technology. The peculiarity of these complexes is due to the simultaneous use of typical Middle Paleolithic techniques for preparing the impact site and the chipping front using centripetal lifts and the use of techniques for maintaining the convexity of the front using chipping of rib plates and bi-longitudinal lifts organized on the subprismatic front. There are much more arguments for combining the industry of the Lotoshi site with the Mongolian complexes, as well as with the Shuidungou site industries within the framework of one local, also Levalloisian model of transition from the Middle to Upper Paleolithic.

Acknowledgements

The authors express their deep gratitude to the expedition staff who participated in field work in Xinjiang. We pay tribute to the memory of A. N. Zenin, who passed away prematurely, and also express our gratitude to V. N. Zenin, A. I. Krivoshapkin and M. V. Shunkov, who participated in the processing and analysis of the collections during their stay in Beijing in October-November 2009.

The illustrations were made by A. V. Abdulmanova and N. V. Vavilina, artists of IAET SB RAS.

List of literature

Ananyev E. S., Simonov Yu. G., Spiridonov A. I. Dinamicheskaya geomorphologiya [Dynamic Geomorphology], Moscow: Moscow State University Press, 1992, 448 p.

Gvozdetsky N. A., Golubchikov Yu. N. Mountains, Moscow: Mysl, 1987, 399 p.

Derevyanko A. P. Three scenarios of transition from the Middle to the Upper Paleolithic. Scenario one: Transition to the Upper Paleolithic on the territory of Northern Asia / / Archeology, Ethnography and Anthropology of Eurasia. - 2010. - N 3. - p. 2-32.

Derevyanko A. P., Dorzh D., Vasilevsky R. S., Larichev V. E., Petrin V. T., Devyatkin E. V., Malaeva E. M. The Stone Age of Mongolia: Paleolithic and Neolithic of the Mongolian Altai. Novosibirsk: Nauka Publ., 1990, 646 p. (in Russian)

Derevyanko, A. P., Zenin, A. N., Rybin, E. P., Gladyshev, S. A., Tsybankov, A. A., Olsen, D., Tseveendorzh, D., and Gunchinsuren, B., Technology of stone splitting at the early Stage of the Upper Paleolithic of Northern Mongolia, in Archeology, Ethnography, and Anthropology of Eurasia. - 2007. - N1. - p. 16-38.

Derevyanko A. P., Kandyba A.V., Petrin V. T. Paleolith of Orkhon. Novosibirsk: IAET SB RAS Publ., 2010, 384 p. (in Russian)

Derevyanko A. P., Komissarov S. A. K probleme sin'tszyanskogo paleolita [On the problem of the Xinjiang Paleolithic] / / Problemy arkheologii, etnografii, antropologii Sibiri i sopredel'nykh territorii: mat-ly V Godovoi utogovoi sessii IAET SB RAS. - Novosibirsk: Izd. IAET SB RAS, 1997. - Vol. III. - pp. 43-47.

Derevyanko, A. P., Petrin, V. T., Taimagambetov, Zh.K., Rybin, E. P., Kozlovsky, M. K., Larionov, V. A., Larionova, P. P., and Sitnikova, L. G., Paleolith of the Leninogorsk Basin in the Western Altai, Problemy arkheologii, etnografii, antropologii Sibiri i sopredel'nykh territorii: mat-ly VII Godovoi sessii IAET SB RAS. Novosibirsk: Izd-vo IAET SB RAS, 1999, vol. V., pp. 84-88.

Derevyanko A. P., Shunkov M. V. Stanovlenie verkhnepaleoliticheskikh traditsii na Altae [Formation of Upper Paleolithic traditions in the Altai]. - 2004. - N 3. - p. 12-40.

Luzin B. P. Bolshoy Altay kak klimaticheskiy barrier [Big Altai as a climatic barrier]. Izv. Alt.gosudarstvennogo unita. - 2007. - N 3. - p. 39-46.

Petrin V. T., Taimagambetov Zh. K. Complexes of the Shulbinka Paleolithic site from the Upper Irtysh region. Almaty: Institute of History. research of the Kazakh state national Al-Farabi University; IAET SB RAS, 2000. - 165 p.

Selivanov E. I. Geomorphology of Dzungaria, Moscow: Nedra Publ., 1965, 154 p.

Taimagambetov Zh. K. Paleolithic site named after Ch. Valikhanov. Alma-Ata: Nauka KazSSR Publ., 1990, 125 p. (in Russian)

Shuidonggou. Report of the 1980 excavation (Shuidonggou. 1980 nian fajue baogao). Beijing: Kexue chubanype Publ., 2003, 233 p. (in Chinese).

Brantingham P. J., Gao X., Madsen D.B., Bettinger R.L., Elston R.G. The Initial Upper Paleolithic at Shuidonggou, Northwestern China // The Early Upper Paleolithic beyond Western Europe. - Berkeley; Eos Angeles; L.: Univ. of California Press, 2004. - P. 223 - 241.

Chen Mengxiong. Impacts of Human Activities on the Hydrological Regime and Ecosystems in an Arid Area of Northwest China // Man's Influence on Freshwater Ecosystems and Water Use (Proceeding of Boulder Symposium, July 1995). -Beijing: IAHS Publ., 1995. - N230. -P. 131 - 139.

Jia Wei Ming, Betts A., Wu Xinhua. Prehistoric Archaeology in the Zhunge'er (Junggar) Basin, Xinjiang, China // Eurasion Prehistory. - 2009. - Vol. 6, N 1/2. - P. 167 - 198.

New Achievements in Xinjiang Cultural Relics and Archaeology (1979 - 1989) // Xinjiang Institute of Cultural Relics and Archaeology. - Urumchi: Xinjiang People's Press, 1995. -P. 1 - 8.

Wang B., Du G. New Achievements in Xinjiang Cultural Relics and Archaeology (1990 - 1996) // Xinjiang Institute of Cultural Relics and Archaeology. - Urumchi: Xinjiang Fine Arts Photography Press, 1997. -P. 29 - 70.

The article was submitted to the editorial Board on 05.04.12, in the final version-on 10.04.12.

Permanent link to this publication: