The Kulbulak Bladelet Tradition in The Upper Paleolithic of Central Asia

ARCHAEOLOGY,

ETHNOLOGY

& ANTHROPOLOGY

OF EURASIA

Archaeology Ethnology & Anthropology of Eurasia 41/2 (2013) 2–25

E-mail: [email protected]

© 2013, Siberian Branch of Russian Academy of Sciences, Institute of Archaeology and Ethnography of the Siberian Branch of the Russian Academy of Sciences. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.aeae.2013.11.002

2

Introduction

Knowledge of the Upper Paleolithic in western Central Asia is rather patchy which hampers comprehensive interpretation of cultural developments in the Asian part of Eurasia during the terminal part of the Upper

K.A. Kolobova1, D. Flas2, A.P. Derevianko1, K.K. Pavlenok1, U.I. Islamov3, and A.I. Krivoshapkin1

1Institute of Archaeology and Ethnography, Siberian Branch, Russian Academy of Sciences,Pr. Akademika Lavrentieva 17, Novosibirsk, 630090, Russia

E-mail: [email protected]; [email protected]; [email protected] Fund for Scienti c Research (FRS-FNRS), University of Liège,

Place du XX Août 7, Bat. A1 4000 Liège, BelgiumE-mail: damien [email protected]

3Institute of Archaeology, Uzbekistan Academy of Sciences,Akademika Abdulaeva 3, Samarkand, 140051, Republic of Uzbekistan

E-mail: [email protected]

THE KULBULAK BLADELET TRADITION IN THE UPPER PALEOLITHIC OF CENTRAL ASIA*

The article describes lithic industries of the Upper Paleolithic levels of Kulbulak, Uzbekistan, a key site in southwestern Central Asia based on materials from excavations at several sites in the western and northwestern Pamir–Tien Shan region (Kulbulak, Kyzyl-Alma-2, Dodekatym-2, and Shugnou). A new cultural and technological tradition is introduced, for which the authors suggest the name Kulbulakian. Its distinctive features are the bladelet technique and a microlithic set including backed pieces and triangular microliths. Stages in the evolution of the Kulbulakian tradition are reconstructed: origins, development, peak, and disappearance of the carinated technology. Industries belonging to this tradition have shaped the general appearance of the Upper Paleolithic in the area in question.

Keywords: Upper Paleolithic, bladelet production techniques, carinated technology, western Central Asia.

Pleistocene. As a result, all researchers studying the Upper Paleolithic in Uzbekistan and western Central Asia at large mention its heterogeneity, whereby virtually each new site is regarded as a special phenomenon (Vishnyatsky, 1999). Moreover, the almost complete absence of absolute dates even for a few stratified sites impedes the identi cation of the chronological and cultural variability of the Upper Paleolithic in this region. Cultural and chronological schemes of the Upper Paleolithic suggested by some researchers (Ranov, 1972; Tashkenbaev, Suleimanov, 1980; Davis, Ranov,

*Supported by the Russian Foundation for Basic Research (Projects Nos. 12-06-33041 mol-a-ved and 12-06-31235 mol-a) and the Russian Foundation for the Humanities (Project No. 12-31-01322).

PALEOENVIRONMENT. THE STONE AGE

K.A. Kolobova et al. / Archaeology, Ethnology and Anthropology of Eurasia 41/2 (2013) 2–25 3

1999) outline general trends in the development of the Upper Paleolithic material culture but do not give a detailed assessment of the origins of cultures and their interactions. That so little knowledge of the area exists is odd compared to the many Mesolithic sites that have been discovered and studied in western Central Asia. Several researchers believe that this disproportion may be accounted for by the depopulation of the region in the interval between 40–25 ka BP caused by abrupt aridization and decline of animal and vegetable resources (Ranov, 1972; Davis, Ranov, 1999). Other hypotheses propose that most Upper Paleolithic sites may not yet have been discovered because they could be buried at a considerable depth (Abramova, 1984) or have been destroyed by seismic activity and mud ows (Ranov, 1988).

It is widely believed that the principal features of the local Upper Paleolithic are spatial patchiness precluding the separation of cultures, general archaism of industries, and absence of the blade technique (Ranov, 1964; Ranov, Nesmeyanov, 1973; Abramova, 1984; Korobkova, Dzhurakulov, 2000; Ranov, Karimova, 2005). In this junction, discovery of new sites within this chronological period as well as the reexamination of lithic collections from previously located sites will assist in clarifying these unresolved issues.

The efforts of the International Archaeological Expedition of the Institute of Archaeology and Ethnography SB RAS working in Uzbekistan since 1998 have resulted in the discovery of new Upper Paleolithic sites in the western Tien Shan: Dodekatym-2 and Kyzyl-Alma-2 (Kolobova et al., 2010; Kolobova, Krivoshapkin, Derevianko, Islamov, 2011). The expedition’s efforts were mainly focused on resuming excavations at Kulbulak, a key Paleolithic site in the region. This article presents key data obtained during the examination of Kulbulak Upper Paleolithic horizons conducted in 2007–2011.

Site location and previous excavations

Kulbulak represents a strati ed open-air site (41 00 31 N; 70 00 22 E) located on the southeastern face of the Chatkal Range in the Tashkent Province, Republic of Uzbekistan. O.M. Rostovtsev discovered the site in 1962 on a long promontory on the right bank of the Dzharsai River where it joins the Kyzyl-Alma, the right tributary of the Akhangaron (Fig. 1).

Systematic excavations at the site were conducted in several stages. M.P. Kasymov, the first principal excavator of the site, worked from 1963 to 1985 with interruption. The maximum depth of excavations of

Quaternary deposits in a small area reached 19 m; the size of the area excavated exceeded 600 sq. m. According to Kasymov’s interpretation, the excavated sediments can be attributed to the Lower (22 layers), Middle (24 layers), and Upper (3 layers) Paleolithic. The Upper Paleolithic strata contained artifacts that, in Kasymov’s opinion, illustrate the continuation of the Mousterian tradition, although typical Upper Paleolithic cores and tools are also present in the industry (Kasymov, 1990).

In 1994–1995, the Joint Russian-Uzbekistan Expedition headed by N.K. Anisyutkin excavated the upper portion of the sediments and identi ed that layers 1–4 contained lithic artifacts. These researchers attributed assemblages from layers 1 and 2 to the Upper Paleolithic and correlated these strata with Upper Paleolithic layers 2 and 3 identi ed by M.P. Kasymov. It was noted that the assemblages consisted of side- and endscrapers, burins, chisel-like tools, borers, notch-denticulate tools, blades, bladelets, akes, chips, and cores (Novye issledovaniya…, 1995).

In 2007–2011, excavations at Kulbulak were resumed by the International Expedition organized by the Institute of Archaeology and Ethnography SB RAS (Novosibirsk, Russia), the Institute of Archaeology of the Uzbekistan Academy of Sciences (Samarkand, Uzbekistan), and the Royal Museums of Art and History (Brussels, Belgium). The study was aimed at collecting lithic artifacts within a strati ed context, obtaining new data on the site’s stratigraphy, and taking samples for absolute dating. Works was carried out in three areas, two focusing on the Middle Paleolithic layers and one on the Upper Paleolithic layer.

Fig. 1. Map showing the location of Upper Paleolithic sites in the Pamir–Tien Shan region.

4 K.A. Kolobova et al. / Archaeology, Ethnology and Anthropology of Eurasia 41/2 (2013) 2–25

0 1 m

Stratigraphic and planigraphic contexts of the Upper Paleolithic strata

The Upper Paleolithic horizons of Kulbulak excavated in 2007–2011 were incorporated into lithological layer 2 (the overlying sediments were excavated during previous stages of site study). Layer 2 is composed of greenish-gray loam (sandy loam) containing sand, gravel, grit, and rare inclusions of rock debris. The layer shows some evidence of bioturbations. It contains pellets of carbonate concretions and dense clay; spots of ferrous oxide are visible in some places (mostly at the bottom of the layer). Despite the proluvial origin of the sediments, archaeological materials were only insigni cantly moved within the layer. This conclusion is evidenced by a vast number of chips and small shatter pieces as well as by the mostly horizontal position of artifacts forming accumulations separated by empty spaces (planigraphic evidence of “activity zones”). In addition, several broken artifacts were found with their re ttable parts located close to each other both horizontally and vertically. Excavations have shown that in lithological layer 2, Upper Paleolithic artifacts were deposited in two quantitatively unequal culture-bearing horizons that may represent two episodes of human occupation of various duration separated by a short period of time. The upper horizon containing numerous artifacts (cultural horizon 2.1) suggests long-term human habitation, while the lower horizon (cultural horizon 2.2) indicates short-term human habitation (Fig. 2).

Lithic artifacts from the Upper Paleolithic layers

of Kulbulak

In the analysis of core reduction techniques, fragments, shatter, chips and small akes not exceeding dimensions

of a maximum length of 2 cm were included in the debitage category. These by-products were not taken into account in calculating percentages of various groups of artifacts within each layer. In the metric analysis of small laminar blanks, small bladelets and microblades were included into the category of “bladelets,” i.e., akes that are at least twice as long as they are wide,

provided their width does not exceed 12 mm. However, the term ‘microblade’ is used in the description of some blanks in order to emphasize the diminutiveness of some artifacts (Kolobova, Krivoshapkin, Derevianko, Islamov, 2011).

Lithic industry from cultural horizon 2.2

The assemblage comprises 11,851 artifacts (Table 1) including 10,024 specimens (84.6 %) classi ed as by-products. Core-like pieces number 94 including 72 distinct nuclei (Table 2) that were reduced by at (Fig. 3, 2, 5, 6), narrow-face (Fig. 3, 7, 9, 13), and prismatic (Fig. 3, 8, 10) aking. Levallois cores for pointed blanks (Fig. 3, 12) and akes (Fig. 3, 11) can be distinguished among the flat flaking cores. Only two nuclei were classi ed as carinated (Fig. 3, 3, 4) within the group of prismatic cores. One core illustrates the combination of prismatic and at aking techniques (Fig. 3, 1).

The assemblage from this cultural horizon consists of 90 core trimming elements (hereafter, CTE) (Table 3). Lateral akes prevail (45 specimens); “tablets” are also numerous (22 specimens). Then follow crested and semi-crested akes (12 specimens) and akes resulted from trimming of the at core’s hinge (7 specimens). The assemblage contains one flake made from the rejuvenated lateral margin of the carinated core as well as three burin spalls.

A hammerstone made on a rolled volcanic pebble was noted in the assemblage.

Fig. 2. Kulbulak northern wall pro le (upper part).


Table 1. Composition of the Kulbulak lithic industries

Table 2. Typology of cores

Primary reduction categoryHorizon 2.1 Horizon 2.2

Number % Number %

Core-like artifacts* 472 7.7 94 5.2

Core trimming elements* 290 4.7 90 4.9

Points* 13 0.2 11 0.6

Flakes* 2392 39.2 884 48.4

Blades* 870 14.3 293 16.0

Bladelets* 1907 31.3 408 22.3

Blade akes* 158 2.6 47 2.6

Total without waste** 6102 13.9 1827 15.4

Waste (shatter, scales, chips)** 37,751 86.1 10,024 84.6

Total 43,853 100 11,851 100

*Percent of total number of artifacts in the horizon (waste disregarded).**Percent of total number of artifacts in the horizon.

TypeHorizon 2.1 Horizon 2.2

Number % Number %

1 2 3 4 5

Prismatic 134 39.9 12 16.7

Single platform for laminar blanks 42 … 7 …

Single platform convergent for laminar blanks 2 … 1 …

Single platform for akes 10 … 0 …

Double platform for laminar blanks 13 … 2 …

Combination 1 … 0 …

Double-platform, cone-shaped with two aking surfaces 2 … 0 …

Carinated 64 … 2 …

on transverse blanks 19 … 1 …

on longitudinal blanks 7 … 1 …

on nodule 38 … 0 …

Flat aking 129 38.4 48 66.7

Single platform parallel with one aking surface:

for laminar blanks 47 … 9 …

for bladelets on blanks 4 … 4 …

for akes 33 … 7 …

for akes on blanks 0 … 3 …

Double platform convergent for akes and points with one aking surface 0 … 1 …

Double platform parallel for akes with two aking surfaces 0 … 1 …

Double platform parallel for laminar blanks with two aking surfaces 1 … 0 …

Double platform bidirectional for laminar blanks with one aking surface 23 … 8 …


Table 2 (end)

Table 3. Typology of core trimming elements

1 2 3 4 5

Double platform bidirectional for akes with one aking surface 5 … 3 …

Orthogonal 6 … 5 …

Disc-shaped 2 … 0 …

Radial 4 … 3 …

Cube-shaped 4 … 1 …

Levallois 0 … 3 …

for akes 0 … 1 …

for points 0 … 2 …

Narrow-face 72 21.4 11 15.3

Double platform on blank with one aking surface 1 … 0 …

Single platform bladelet cores on blanks 24 … 0 …

Single platform for laminar blanks 24 … 2 …

Double platforms bladelet cores with one aking surface 2 … 0 …

Wedge-shaped bladelet 20 … 8 …

CoresMulti-platform bladelet core with multiple aking surfaces 1 … 0 …

Double platform blade core with one aking surface 0 … 1 …

Combination 1 0.3 1 1.4

Total 336 100 72 100


Number % Number %

Striking platform rejuvenation akes from prismatic cores:

“tablets” 58 19.9 22 24.4

“half-tablets” 1 0.3 0 0.0

Striking platform rejuvenation akes from carinoid cores – “tablets” 2 0.6 0 0.0

Core’s hinge rejuvenation akes 27 9.2 7 7.8

Lateral shortened akes 93 32.9 29 32.2

Lateral blades 42 14.4 16 17.8

Crested blades 9 3.0 3 3.3

Crested shortened akes 3 1.0 0 0.0

Crested bladelets 3 1.0 0 0.0

Semi-crested blades 17 5.8 3 3.3

Semi-crested shortened akes 15 5.1 6 6.7

Plunging akes 3 1.0 0 0.0

Carinated lateral akes 3 1.0 1 1.1

Burin spalls 15 5.1 3 3.3

Total 290 100 90 100


Fig. 3. Cores from cultural horizon 2.2.

0 3 m

1

2

3

4

5 67

8

9

10

11

12 13


Blanks (1643 specimens) are represented by akes, bladelets, blades, blade akes, and points (Table 1).

The toolkit is comprised of 75 specimens (Table 4). Grattoirs of varying morphology represent the most numerous category (Fig. 4, 1, 2, 10–12). Sidescrapers can be subdivided into convergent (Fig. 4, 15, 16), single straight longitudinal, and backed types (Fig. 4, 18). The toolkit also contains 16 blades with traces of retouch of various morphologies (Fig. 4, 14) including one backed blade. Notch-denticulate tools, retouched akes, chisel-like implements (Fig. 4, 6–9), spurs, retouched points (Fig. 4, 17), backed knives (Fig. 4, 13), and tools with signs of trimming are less numerous.

Several tools (5 specimens) due to their miniature size were included in a separate group designated as microindustry. This group includes micrograttoirs (Fig. 4, 3, 4), Dufour bladelets (Fig. 4, 5), and a retouched bladelet.

Lithic industry from cultural horizon 2.1

The assemblage consists of 43,853 artifacts (Table 1) including 37,751 (86 %) debitage pieces (chips, shatter, and small akes). The category of core-like artifacts includes 472 specimens, of which 336 are typologically distinct cores (Table 2) that were reduced using at surface (Fig. 5, 17, 18; 6, 11–17), narrow-face (Fig. 6, 1–10), and prismatic (Fig. 11–16) aking techniques. One core demonstrates a combination of prismatic and narrow-face reduction strategies that were applied at various stages of core utilization.

The category of CTE includes 290 specimens (Table 3). Lateral akes (46.5 %), “tablets” (20.7 %), and crested and semi-crested akes (16.2 %) dominate. Some akes resulted from the rejuvenation of carinated cores (Fig. 7, 1–5). Flakes produced by trimming

Fig. 4. Stone tools from cultural horizon 2.2.

0 3 m

12

3 4

5

6 7

8

9

10 11

1213

1415

1617

18


Table 4. Typology of tools


Number % Number %

Grattoirs 90 23.6 14 18.6

endscrapers 58 … 8 …

retouched over ¾ of perimeter 6 … … …

angular 11 … 3 …

lateral 21 … 3 …

Microtools 60 15.7 5 6.6

micrograttoirs 6 … 2 …

Dufour bladelets 4 … 2 …

triangular microlith 1 … 0 …

backed bladelets 5 … 0 …

retouched bladelets 25 … 1 …

microchisel tools 19 … 0 …

Chisel-like tools 56 14.7 4 5.3

with one working edge 41 … 2 …

with two working edges 15 … 2 …

Retouched points 8 2.1 3 4

Retouched blades 36 9.4 16 21.3

Sidescrapers 20 5.2 10 13.3

double longitudinal-transverse 1 … 0 …

longitudinal convex 1 … 1 …

double longitudinal straight 1 … 0 …

double longitudinal alternate convex-concave 1 … 0 …

angle 1 … 0 …

longitudinal straight 15 … 4 …

convergent 0 … 5 …

Notched tools 10 2.6 5 6.6

Denticulate tools 9 2.4 3 4

Burins 12 3.1 0 0

at multifaceted 1 … 0 …

side multifaceted 6 … 0 …

side single-faceted 5 … 0 …

Spur-like tools 10 2.6 4 5.3

Borers 8 2.1 0 0

Knives 6 1.6 3 4.1

Unifacial tool 1 0.3 0 0

Chopper 1 0.3 0 0

Retouched akes 53 13.9 5 6.6

Trimmed tools 0 0 2 2.6

Truncated blade 1 0.3 0 0.0

Backed blade 1 0.3 1 1.3

Total 382 100 73 100



0 3 m

1

2

3

4

5

6

7

89

1011

12

13

14

15

16 17 18



0 3 m

12

34

5

6

78

9

10

11

12

13

14

15 1617


Fig. 7. Lithic artifacts from cultural horizon 2.1.

0 3 m

12 3

4

5

6

78

9

10 1112

13

14

15

16

17

18

19

20

21

22 23


the at core’s hinge, “plunging” akes produced by rejuvenating the base of prismatic single- and double platform cores, and burin spalls are also represented in the assemblage.

Two hammerstones fashioned on rounded pebbles were found, one fragmented in the past (Fig. 7, 23).

Blanks (5340 specimens) are represented flakes, bladelets, blades, blade akes, and points (Table 1).

The toolkit comprises 382 specimens (Table 4). Grattoirs of various morphological types (Fig. 7, 6–18, 21) form the most numerous tool category. Then follow chisel-like tools with one (Fig. 8, 31–34) and two working edges (Fig. 8, 35–41) and retouched blades (Fig. 7, 19, 20, 22). Sidescrapers can be subdivided into single and double. The former are represented by straight and convex varieties (Fig. 8, 47). The series of double scrapers includes straight longitudinal, longitudinal alternative convex-concave, longitudinal-transverse, and angle forms. Notched and denticulate tools are rather numerous. The category of perforators includes spurs and borers. Burins are represented by multifaceted at forms (Fig. 8, 46) as well as by multifaceted (Fig. 8, 44, 45) and monofaceted varieties of side forms. Retouched points are few in number (Fig. 8, 42, 43). Knives can be classi ed into implements with natural and prepared backs. The assemblage includes a unifacial tool, chopper, truncated blade, and a backed blade (Fig. 8, 17). Retouched flakes form numerous though typologically insigni cant group.

Miniature tools were included into a separate group defined as a microindustry. This microindustry is primarily represented by retouched bladelets including 16 small laminar blanks bearing facets of non-modifying utilization retouch over various parts of long edges (Fig. 8, 8, 9, 25–27) and 9 blanks with traces of intentional retouching (Fig. 8, 10–16). The second important category of microtools includes chisel-like implements not exceeding 20 mm in length and 5 mm in thickness (Fig. 8, 18–21, 29, 30). Micrograttoirs are typologically distinct and attributable to end (Fig. 8, 22–24) and thumbnail varieties. Backed bladelets (Fig. 8, 6, 7, 28) were noted. The industry also comprises Dufour bladelets including one atypical (Fig. 8, 4) and others fashioned by alternate retouch over both long edges (Fig. 8, 1–3). A triangular microlith is a unique nd in this assemblage (Fig. 8, 5).

Comparison of lithic industries from cultural horizons 2.1 and 2.2

Because the principal artifact types and primary reduction products are well represented in assemblages

from both horizons, comparison is possible despite the signi cant difference in artifact numbers. In general, the similarity observed in the technological and techno-typological features of these two industries makes it possible to de ne the Upper Paleolithic technocomplex of the site as homogenous. However, certain differences can be traced both in the choice and use of raw material as well as in the production of akes and their modi cation into tools.

Sources of raw material are located in the immediate vicinity of the site. Pebbles of volcanic rocks and int were transported from the nearest banks of the Kyzyl-Alma and Dzharsai. Flint nodules were brought from the outcrops of organogenic limestone that situated at a distance not exceeding 1.5 km from the site. The petrographic analysis has shown that most artifacts were made of primarily light gray to white and yellowish, sometimes brown and dark gray int of low quality. The int has cryptocrystalline and ne-grained structures;

in many cases, grains of unrolled quartz not exceeding 0.5 mm in size are present. The ints are heterogeneous; many represent a form transitional to transparent light chalcedony with a concentric zonal structure. Some artifacts show thin veins and isometric stocks of ne quartz crystals. A lesser quantity of rocks reveals the relict uidal texture. In addition, a considerable portion of artifacts were made on uidal, acid volcanic rocks of varying degrees of silici cation. The volcanic rocks are slightly porphyritic; feldspar and quartz are present in idiomorphic inclusions (N.A. Kulik, personal communication of 2007). Chalcedony and quartzite are rare at Kulbulak. Flint and volcanic rocks are the main raw materials in both industries, but int predominates (Tables 5, 6). Among the blades, the amount of int used is even higher, being maximal among microblades. The smaller the blade, the larger the proportion of those made of int in both assemblages.

Both technocomplexes are based on at, prismatic, and narrow-face reduction strategies. Parallel aking was also used; convergent, orthogonal, and radial flaking patterns are less common. Flat flaking was the dominant technique in both complexes. Double platform cores indicate parallel alternate knapping. The percentage of double platform cores is notably lower than that of single platform cores. The negative scars of nal removals show that the cores of at aking were used for blanks of various types: akes, blade akes, blades, and bladelets. The cores show signs of

insigni cant rejuvenation as well as crudely prepared aking surfaces and striking platforms.

The proportion of at aking cores in horizon 2.1 is considerably lower than in horizon 2.2 due to the increased number of prismatic cores. Prismatic cores


Fig. 8. Stone tools from cultural horizon 2.1.

0 3 m

0 2 m

12

34 5 6 7

8

9

1011 12

13 14 1516

17

18 19 20 21

22 2324 25 26

27 28

2930

31

3233 34

35

36 37 38

39

40

41

42

43

44

4546 47


were aked using parallel and convergent reduction techniques. The bidirectional prismatic reduction technique is the same in both assemblages; dorsal surfaces of blanks preserve mostly unidirectional aking scars. Prismatic cores from horizon 2.1 are closer to standards than those from horizon 2.2: they demonstrate regular detachment of laminar blanks; their flaking surfaces are longer; some cores are close to pyramidal nuclei. Both collections contain prismatic cores of similar types, although their numbers are different. This is mostly the case in carinated cores: only two were found in horizon 2.2, while in horizon 2.1 they number 64. This is the largest series of carinated cores from one site currently known in western Central Asia. The main stages of core utilization have been reconstructed based on the attributive analysis of these artifacts (Kolobova, Krivoshapkin, Flas et al., 2011).

Carinated artifacts are morphologically similar to certain types of narrow-face cores. The portion of narrow-face cores in horizon 2.1 is slightly greater than in horizon 2.2. Such cores were used in the production of bladelets and blades with a predominantly straight pro le. The technocomplex of horizon 2.1 was more oriented towards the production of small laminar blanks than that of horizon 2.2.

The portions of CTE in the collections are roughly equal. Lateral flakes including those with blade

proportions are most numerous in both assemblages. Semi-crested flakes of various modifications are also well represented. It should be noted that in the assemblage from horizon 2.2, the amount of prismatic core rejuvenation akes exceeds the number of prismatic cores. This horizon also contained a considerable number of “tablets,” which corresponds with the number of prismatic cores in this assemblage.

The percentages of blades are approximately equal in both collections. Judging by the lengths of the complete artifacts, most blades from horizon 2.1 are somewhat shorter than those from horizon 2.2. The width values of all blanks also point to diminution of blade akes (Fig. 9). Generally, small blades were preferred in both industries. Blanks with a straight pro le dominate both assemblages, although blades with a curved pro le are more numerous in horizon 2.1 (Table 7). Blade akes with a triangular cross-section prevail in horizon 2.2 (Table 8). Horizon 2.1 contains a slightly greater amount of trapezium-shaped and polygonal blades. The dorsal surfaces of such blanks bear the negative scars of parallel and unidirectional removals (Tables 9, 10). The number of spalls bearing scars from convergent aking is relatively large. The artifacts in both industries have primarily plain residual platforms. In horizon 2.1, their percentage increases due to dihedral and polyhedral forms rather numerous in horizon 2.2

Table 5. Distribution of artifacts from horizon 2.1 in terms of raw material

Table 6. Distribution of artifacts from horizon 2.2 in terms of raw material

Raw materialFlakes Blades Bladelets Microblades Total

Number % Number % Number % Number % Number %

Flint 2028 84.8 765 87.9 1050 96.0 796 97.9 4639 89.7

Volcanic rock 354 14.8 100 11.5 40 3.7 16 2.0 510 9.9

Quartzite 10 0.4 5 0.6 3 0.3 1 0.1 19 0.4

Chalcedony 0 0.0 0 0.0 1 0.1 0 0.0 1 0.0

Total 2392 100 870 100 1094 100 813 100 5169 100

Raw materialFlakes Blades Bladelets Microblades Total

Number % Number % Number % Number % Number %

Flint 751 85.0 240 81.9 243 91.4 133 93.7 1367 86.2

Volcanic rock 132 14.9 53 18.1 22 8.3 7 4.9 214 13.5

Quartzite 1 0.1 0 0.0 1 0.4 0 0.0 2 0.1

Chalcedony 0 0.0 0 0.0 0 0.0 2 1.4 2 0.1

Total 884 100 293 100 266 100 142 100 1585 100


Fig. 9. Distribution of blade preforms in terms of width. Cultural horizons 2.1 and 2.2.

Table 7. Distribution of blades in terms of pro le shape

Table 8. Distribution of blades in terms of cross-section shape

Pro leHorizon 2.1 Horizon 2.2

Number % Number %

Straight 432 65.3 150 71.1

Curved 126 19.0 27 12.8

Twisted 104 15.7 34 16.1

Total 662 100 211 100

ShapeHorizon 2.1 Horizon 2.2

Number % Number %

Triangular 371 42.9 146 50.3

Trapeziform 373 43.2 105 36.2

Polygonal 71 8.2 21 7.2

Segment 14 1.6 6 2.1

Angular 35 4.1 12 4.1

Total 864 100 290 100

(Tables 11, 12). Thus, up the pro le, the size of blade akes diminishes; the number of blades with a curved

pro le and trapezium-shaped cross-section is greater; the quantity of plain, punctiform, and linear striking

platforms becomes larger; and the portion of products from their reduction slightly increases.

The assemblages are characterized by a growing number of small blades and prismatic blade cores. It


Table 9. Distribution of artifacts in terms of dorsal scar pattern, horizon 2.1

PatternBlades Bladelets Blade akes Flakes Points Total

Number % Number % Number % Number % Number % Number %

Parallel unidirectional 527 60.6 1299 68.1 121 76.6 958 40.1 1 7.7 2906 54.4

Parallel opposite 57 6.6 68 3.6 10 6.3 132 5.5 0 0.0 267 5.0

Convergent 167 19.2 391 20.5 3 1.9 205 8.6 12 92.3 778 14.6

Radial 0 0 0 0 0 0 60 2.5 0 0 60 1.1

Plain 6 0.7 49 2.6 2 1.3 232 9.7 0 0 289 5.4

Irregular 0 0 0 0 0 0 1 0 0 0 1 0

Orthogonal 14 1.6 9 0.5 1 0.6 50 2.1 0 0 74 1.4

Secondary 47 5.4 50 2.6 18 11.4 532 22.2 0 0 647 12.1

Natural 26 3.0 18 0.9 3 1.9 138 5.8 0 0 185 3.5

Transverse 0 0 0 0 0 0.0 74 3.1 0 0 74 1.4

Divergent 1 0.1 0 0.0 0 0.0 0 0.0 0 0 1 0

Unidenti able 25 2.9 23 1.2 0 0.0 10 0.4 0 0 58 1.1

Total 870 100 1907 100 158 100 2392 100 13 100 5340 100

Table 10. Distribution of artifacts in terms of dorsal scar pattern, horizon 2.2

PatternBlades Bladelets Blade akes Flakes Points Total


Parallel unidirectional 171 58.4 275 67.4 27 57.4 487 55.1 0 0.0 933 58.9

Parallel opposite 19 6.5 8 2.0 5 10.6 45 5.1 0 0.0 72 4.5

Convergent 61 20.8 84 20.6 1 2.1 83 9.4 11 100.0 228 14.4

Radial 0 0.0 0 0.0 0 0.0 20 2.3 0 0.0 20 1.3

Plain 10 3.4 11 2.7 5 10.6 52 5.9 0 0.0 73 4.6

Irregular 0 0.0 0 0.0 0 0.0 1 1.0 0 0.0 1 0.1

Orthogonal 4 1.4 3 0.7 1 2.1 13 1.5 0 0.0 20 1.3

Secondary 22 7.5 11 2.7 8 17.0 115 13.0 0 0.0 148 9.3

Natural 5 1.7 6 1.5 0 0.0 40 4.5 0 0.0 51 3.2

Transverse 0 0.0 0 0.0 0 0.0 25 2.8 0 0.0 25 1.6

Unidenti able 1 0.3 10 2.5 0 0.0 3 0.3 0 0.0 14 0.9

Total 293 100 408 100 47 100 884 100 11 100 1585 100

should be noted that this increase mostly concerns the category of small blade akes less than 6 mm in width. The width values of all blanks of this type suggest that the assemblage of horizon 2.1 is dominated by

narrower bladelets (Fig. 9). Small blade akes with a mostly straight pro le (Table 13) were produced in both industries. The upper cultural horizon contains a greater number of bladelets with a curved pro le,


Table 11. Distribution of artifacts in terms of residual striking platforms types, horizon 2.1

Table 12. Distribution of artifacts in terms of residual striking platforms types, horizon 2.2

Table 13. Distribution of bladelets in terms of pro le shape

Residual striking

platforms

Blades Bladelets Blade akes Flakes Points Total


Plain 251 54.2 316 36.6 63 58.9 804 63.3 6 66.7 1 440 53.1

Dihedral/polyhedral 76 16.4 62 7.2 5 4.7 52 4.1 1 11.1 196 7.2

Faceted 4 0.9 2 0.2 1 0.9 32 2.5 1 11.1 40 1.5

Linear 58 12.5 290 33.6 8 7.5 17 1.3 0 0.0 373 13.7

Punctiform 17 3.7 124 14.4 27 25.2 235 18.5 0 0.0 403 14.9

Cortical 27 5.8 45 5.2 3 2.8 97 7.6 1 11.1 173 6.4

Unidenti able 30 6.5 25 2.9 0 0.0 33 2.6 0 0.0 88 3.2

Total 463 100 864 100 107 100 1 270 100 9 100 2 713 100

Residual striking

platforms

Blades Bladelets Blade akes Flakes Points Total


Plain 67 45.6 81 46.0 24 68.6 322 62.8 2 40.0 496 56.6

Dihedral 53 36.1 19 10.8 2 5.7 38 7.4 1 20.0 113 12.9

Faceted 2 1.4 0 0.0 2 5.7 40 7.8 2 40.0 46 5.3

Linear 11 7.5 52 29.5 0 0.0 4 0.8 0 0.0 67 7.6

Punctiform 1 0.7 14 8.0 7 20.0 88 17.2 0 0.0 110 12.6

Cortical 3 2.0 10 5.7 0 0.0 21 4.1 0 0.0 34 3.9

Unidenti able 10 6.8 0 0.0 0 0.0 0 0.0 0 0.0 10 1.1

Total 147 100 176 100 35 100 513 100 5 100 876 100

Pro leHorizon 2.1 Horizon 2.2

Number % Number %

Straight 817 60.4 180 69.0

Curved 235 17.4 22 8.4

Twisted 300 22.2 59 22.6

Total 1352 100 261 100

while the number of bladelets with a twisted pro le remains unchanged. Analysis of bladelets belonging to various size categories has shown that smaller bladelets have a more curved pro le. Small blade akes with a triangular cross-section prevail in both assemblages (Table 14). This is suggestive of the aking technique

with one ridge. Judging by the patterns of negative scars left on the dorsal surfaces of the bladelets parallel unidirectional reduction was the leading technique in both industries. The portion of blanks with convergent aking pattern is also signi cant (Tables 9, 10). Plain

residual striking platforms dominate the assemblage


from horizon 2.2. Punctiform and linear platforms are also numerous (Tables 11, 12); their number increases considerably in horizon 2.1.

The category of akes evidently illustrates processes of core decortication, because many of them (37.3 % in horizon 2.1 and 29.6 % in horizon 2.2) have some cortex retained on their surface. Many blanks can be de ned as primary decortication akes.

Analysis of the metric parameters of all elongated akes (Fig. 9) has shown that many blades and bladelets

were produced through common reduction strategies and their size depended only on the degree of core utilization. All elongate blanks tend to decrease in size in the upper layers.

The most marked changes traced in the upper horizon compared to the lower horizon concern the primary reduction technique: various carinated cores used to detach blades with curved profiles become considerably more frequent. The portions of small blade akes with twisted pro les are pretty much the same in

both complexes, while the proportion of akes with a curved pro le increases (within the limits of 10 %) in horizon 2.1. Thus, the morphological parameters of the akes re ect changes that occurred in primary reduction

processes. It can be inferred that carinated cores were mostly used in the production of akes with curved pro les, while akes with twisted pro les resulted from the nal stages of reduction of various core types.

Against the background of a predominance of blades and bladelets with a triangular cross-section, the upper horizon revealed an increase in the proportion of blades and bladelets with trapeziform and polyhedral cross-sections. The category of small blade flakes demonstrates a different tendency. The portion of bladelets with a trapeziform cross-section increased, while bladelets with a triangular cross-section remain dominant and this proportion increases in groups with a smaller width and vice versa, the larger the preform, the higher the probability of two ridges having been used

to detach it and, accordingly, the wider (and longer) it is. This corresponds to observations made in the technological analysis of assemblages from horizons 2.1 and 2.2 collected in 2007–2009 (Pavlenok, 2011).

None of the artifacts in either technocomplex show signs of frequent rejuvenation of the striking platform, yet the rate of platform rejuvenation is slightly higher in the industry of the upper cultural horizon. Flakes display the lowest rate of platform rejuvenation, which is higher in blade akes and highest in small specimens: 28.2 % of bladelets and 14 % of blades from horizon 2.1 show signs of platform rejuvenation. The technocomplex of horizon 2.2 shows similar characteristics. In the process of producing blade akes, the striking platforms were rejuvenated through overhang trimming, while in the production of bladelets this was achieved through inverse reduction trimming

The number of bladelets with a punctiform and linear striking platform is considerably larger in horizon 2.1 than in horizon 2.2. The proportion of bladelets with a reduced platform also increases (traces of inverse reduction trimming and overhang trimming co-occur with punctiform or linear striking platforms in 73.7 % and 75.0 % of cases).

Tools were fashioned on blades, bladelets, akes, points, pebbles, and cores. The proportion of tools on akes is considerably higher in horizon 2.1 (54.1 %

versus 41.0 % in horizon 2.2). This does not correspond to the decrease in percentages of shortened blanks in the category of non-prepared blanks. It was probably linked to the increased number of chisel-like tools and scrapers that were mostly fashioned on blanks of this type. For this reason, in the industry of the upper horizon, the greater portion of akes was chosen for tool production. In addition, the toolkit of horizon 2.1 demonstrates a considerable decrease in the proportion of blades (23.0 % versus 38.3 % in horizon 2.2). This can be explained partially by the fact that blades were replaced by tools on bladelets, the percentage of which is notably

Table 14. Distribution of bladelets in terms of cross-section shape

ShapeHorizon 2.1 Horizon 2.2

Number % Number %

Triangular 1165 61.3 276 69.2

Trapeziform 578 30.4 99 24.8

Polygonal 83 4.4 11 2.8

Segment 38 2.0 5 1.3

Angular 36 1.9 8 2.0

Total 1900 100 399 100


increased (4.1 % in horizon 2.2 and 11.5 % in horizon 2.1). The percentage of pieces made from small blades increases as a function of the frequency of carinated aking and, accordingly, of the number of small blade

preforms with curved pro le.It should be noted that the toolkits from these

two technocomplexes show considerable similarity. However, as noted above, the proportion of scrapers and chisel-like tools as well as microtools including both implements on bladelets and microblades, and exceptionally small scrapers and chisel-like tools, increases in horizon 2.1. The presence of small tools constitutes a speci c feature of the lithic industry.

Results of the comparative analysis suggest that the lithic industries of horizons 2.1 and 2.2 were similar in function. A few minor differences should be mentioned though. The excavated area of horizon 2.2 appears to have been a workshop camp displaying a complete cycle of stone reduction, from transportation to the manufacture of artifacts, some of which may have been carried away. Finds from horizon 2.1 also demonstrate a complete production cycle. Certain tools (mostly chisels and scrapers) remained at the camp and were used in operations involving wood, bone, and skin processing. This supposition is supported by the greater portion of artifacts in the assemblage and by the narrow specialization of retouch evident in most tools.

Human fossil

In 2009, the bottom of horizon 2.1 yielded a tooth, which, according to B. Viola’s personal communication (2009), is a third lower premolar of Homo sapiens. The tooth was found in an undisturbed stratigraphic context and is well preserved. So far this nd represents the rst human fossil in western Central Asia associated with a distinct species of Homo.

Discussion

Results of technological and typological analyses suggest that the industries of horizons 2.1 and 2.2 at Kulbulak represent a single Upper Paleolithic tradition. Differences may be both chronological and functional.

The nearest site is Kyzyl-Alma-2 discovered in 2007, located 1200 m north–northwest of Kulbulak in close vicinity to int outcrops. At Kyzyl-Alma-2, four lithological units were recorded containing Upper Paleolithic artifacts and disturbed by slope processes. The Kyzyl-Alma artifacts resemble nds from Kulbulak horizon 2.2, suggesting that both sites were linked

within a single ecological system: blanks of cores and tools were transported from the workshop near the outcrops of rock (Kyzyl-Alma-2) to the principal site (Kulbulak) (Kolobova, Pavlenok, Flas et al., 2010).

In 2005, site Dodekatym-2 was discovered in the middle course of the Paltau River, the right tributary of the Chatkal (Republic of Uzbekistan). Four cultural horizons (5–2) attributed to the advanced Upper Paleolithic were recorded at Dodekatym-2 in a relatively intact state. These horizons contained lithic industries representing a single Upper Paleolithic cultural tradition, whose bearers inhabited the site ca 23–21 ka BP (uncalibrated dates). Technocomplexes from Dodekatym-2 (horizons 5–2) and Kulbulak (horizons 2.1 and 2.2) share many technological and techno-typological traits. Owing to its highly diagnostic characteristics, the assemblage from lower horizon 5 of Dodekatym-2, despite its small size, is reliably comparable with that of Kulbulak horizon 2.1. The most speci c feature of the Dodekatym lithic industry is the predominance of carinated cores among the primary reduction elements. The high percentage of cores correlates with their abundance in the lithic industry from Kulbulak layer 2.1. The carinated cores at Dodekatym posse the same technological and techno-typological characteristics as the carinated forms from Kulbulak. In terms of basic technological and typological features both industries represent the same cultural tradition. Dodekatym-2 horizon 5 is either contemporaneous to or somewhat later than Kulbulak.

Archaeological materials from the upper culture-bearing layers of Dodekatym-2 undoubtedly demonstrate the development of the industry from horizon 5 and, accordingly belong to the same cultural tradition as the assemblages from Kulbulak layer 2, although document later evolutionary stages. The main trend is the gradual decline in the practice of using carinated cores for bladelets with a curved profile and shift to single platform prismatic cores for bladelets with a straight pro le. At the same time, the proportion and variety of microtools increases. First of all, the percentage of backed bladelets rises. Starting from horizon 4, triangular microliths appear in the assemblages and become the principal and most standard tool type in the assemblage of Dodekatym-2 horizon 2.

Lithic industries from Dodekatym-2 horizons 2–4 share strong affinities with the assemblage from Kulbulak horizon 2.1. They include typologically identical, specific tool types: orthogonal chisel-like tools with double edges, scrapers with ventrally retouched laterals, spurred scrapers, micro-endscrapers and miniature chisel-like implements. The Kulbulak assemblage also comprises several backed bladelets.


One of the diagnostic tools linking the two industries is a triangular microlith from Kulbulak horizon 2.1. Both typologically and morphologically it is identical to oblique triangles from Dodekatym-2.

Another key Upper Paleolithic locality in the region is the strati ed site of Shugnou discovered in 1968. Five culture-bearing layers have been identi ed at Shugnou and attributed by V.A. Ranov to the Upper Paleolithic (layers 4–1) and Mezolithic (layer 0) (Ranov, 1973; Ranov, Nikonov, Pakhomov, 1976; Ranov, Karimova, 2005). Techno-typological analysis with elements of the attribute approach performed in 2010–2011 speci ed the characteristics of the technocomplexes. Based on new ndings and comparisons, the industries of all the cultural horizons at Shugnou were assigned to a single Upper Paleolithic cultural tradition. To all appearances, the camp was inhabited intermittently between 30 ka BP and 23–21 ka BP (Ranov, Kolobova, Krivoshapkin, 2012). Judging by the character of primary reduction and secondary treatment, the industries of the lower layers (4–2) at Shugnou illustrate the closest correlation with the assemblage from Kulbulak horizon 2.2. The predominance of at cores and the presence of prismatic cores indicate similarity with the industry of horizon 4, whereas carinated and narrow-face cores are similar to those from horizons 3 and 2. It should be noted that carinated cores from both sites are identical and were utilized within a framework of a common technological scheme (Kolobova, Krivoshapkin, Flas et al., 2011). Both technocomplexes contain a small number of distinct microtools. However, the proportion of carinated implements in Kulbulak horizon 2.2 is signi cantly lower than in the Shugnou assemblage. This observation may point to the intermediate chronological position of Kulbulak horizon 2.2 (between cultural layers 4 and 3/2 of Shugnou).

Parameters of the lithic industry from Shugnou layer 1 show the closest correlation with the assemblage from Kulbulak horizon 2.1. Both assemblages contain a signi cant number of carinated cores and small blade flakes with a curved profile. Similar tool types are represented in the assemblages; primarily these are retouched and backed bladelets that are also available in the collections from Shugnou layers 3 and 2. Triangular microliths found at both sites (one specimen in each assemblage) are noteworthy. However, the industry from Shugnou layer 1 seems to be more developed. It contains a greater amount of carinated cores and several pyramidal blade cores. In this assemblage, more residual striking platforms bear signs of reduction and the proportion of small tools is also higher. All this seems to suggest that Shugnou layer 1 is younger than Kulbulak horizon 2.1. Alternatively, the difference may be due to

functional specialization, whereby the industry of the Kulbulak workshop camp appears somewhat older.

In the authors’ opinion the industries of Shugnou and Kulbulak (horizon 2), represent different stages in the evolution of the same cultural and technological tradition. Their features are as follows: blades detached from at and prismatic cores (Shugnou layer 4); the appearance of carinated tools and a greater number of bladelets (Kulbulak horizon 2.2 and Shugnou layers 3 and 2); the predominance of carinated techniques and a greater number of microliths (Kulbulak horizon 2.1, Shugnou layer 1); and eventually the abandonment of the practice of detaching blanks from carinated cores (Shugnou layer 0).

Assemblages from two layers of Kharkush, a site located in the offshoots of the Gissar Range, also share some similarities with both Shugnou and Kulbulak industries (Filimonova, 2007; Ranov, Karimova, 2005). In the context of cultural and technical correlations, the Samarkandskaya site, a key Upper Paleolithic locality in western Central Asia, should certainly be mentioned. It comprises three cultural horizons, which, according to some researchers, are roughly contemporaneous and can be attributed to the advanced Upper Paleolithic. An impressive series of “carinated endscrapers” resembling carinated cores from Kulbulak horizon 2.1 is particularly noteworthy (Korobkova, Dzhurakulov, 2000). Mention should also be made of the Valikhanov site (southern Kazakhstan). Assemblages from this site can also be correlated with Kulbulak industries. The site comprises six cultural horizons. The date 24,800 ± 1100 BP (uncalibrated) was obtained for the first horizon. The Valikhanov assemblage is interpreted as being a speci c early Upper Paleolithic industry with a considerable number of archaic features. In addition, researchers have remarked that the presence of carinated endscrapers links the Valikhanov assemblage with the Samarkandskaya, Shugnou, Khodja-Gora, and Kara Kamar lithic industries (Taimagambetov, Ozhereliev, 2009). The recently discovered site of Maibulak in southeastern Kazakhstan comprises three cultural horizons with lithic artifacts few in number. Horizon 3 dated to 34,970 ± 665 BP contained a distinct microblade series. AMS dates are available for horizon 1 (30–28 ka BP) and horizon 1 (24,330 ± 190 BP). The lithic industries associated with these layers contain scrapers of various modifications including carinated forms (Ibid.).

The Kara Kamar site (northern Afghanistan) outside western Central Asia, which contained carinated implements is also worth mentioning. C. Coon, the site’s principal researcher, identi ed four archaeological horizons. The third horizon comprised an Upper


Paleolithic industry of the “Aurignacian type.” Ten radiocarbon dates falling within the interval 32–20 and >32 ka BP were generated for this layer. C. Coon believed that the date 30 ka BP was the most plausible for this horizon (Coon, Ralph, 1955). The main typological group in this industry is carinated scrapers according to V.A. Ranov and A.V. Vinogradov (Vinogradov, 2004; Ranov, Karimova, 2005) or keeled scrapers/bladelet cores according to R. Davis (2004).

In sum, in the Pamir–Tien-Shan region, one and the same cultural and technological tradition existed at Kulbulak, Kyzyl-Alma-2, Dodekatym-2, and Shugnou. The Upper Paleolithic industries of adjacent regions (Samarkandskaya, Kharkush, Maibulak, Valikhanov, and Kara Kamar) are marked by predominantly carinated tools, linking them with these sites.

In a broader context, the Upper Paleolithic industries of the Pamir–Tien Shan region are paralleled by those of the Baradostian in Zagros, specifically by the best known ones – those of Shanidar, Warwasi, and Yafteh. Their characteristic features are a developed carinated bladelet production technology and a large number of carinated burins. The predominant tools are Dufour blades and Arjeneh points (Olshewski, 1993a; Olszewski, Dibble, 1994; Olszewski, 1999; Otte et al., 2011). A series of uncalibrated dates for Yafteh falls within the 35–31 ka BP interval (Otte et al., 2011). Recently a new culture was described in western Iran, named Rostamian. A series of uncalibrated dates for its key site, Ghar-e Boof, is within the 37–31 ka BP period (Conard, Ghasidian, 2011). The Ghar-e Boof industry appears to be very close to the Baradostian. In the Zagros Mountains, the Baradostian was succeeded by the Zarzian, whose best known sites are Warwasi, Shanidar, Zarzi, and Palegawra. Zarzian primary reduction is characterized by single platform prismatic bladelet cores and carinated cores with wide aking surface. Microtools dominate the toolkit and include triangular inequilateral microliths (morphologically close to the Dodekatym triangular microliths), backed bladelets, and retouched bladelets. The Zarzian culture is broadly dated within the range of 20–12 ka BP (Olshewski, 1993b; Wahida, 1999). Final Upper Paleolithic assemblages resembling the Zarzian industry have been reported in the Levant (Ohalo II, Fazael X, Ein Gev I, Fazael IIIA and IIIB) (Nadel, 2003).

Being the easternmost manifestation of the Levantine Aurignacian (Olszewski, Dibble, 1994, 2006), the Baradostian links Kulbulak and Shugnou with Near Eastern Aurignacian sites such as Ksar Akil, Hayonim, Kebara, Yabrud II, etc. The Levantine Aurignacian assemblages date to 32–26 ka BP. They contain carinated implements of various forms,

bladelets with curved pro le, Dufour bladelets, Arjeneh points, and a considerable number of bone and horn implements including points with a split base and personal adornments, primarily pendants of various sorts (Mellars, Tixier, 1989; Belfer-Cohen, Goring-Morris, 2007).

Given the geographical remoteness of the Levant and Zagros from western Central Asia and the different volume of archaeological information available on them, direct cultural links between the Upper Paleolithic industries of these regions can hardly be postulated. Yet numerous technological and typological parallels along with the fact that those industries are nearly contemporaneous may indicate at least similar evolutionary trajectories if not direct or indirect cultural diffusion. The reasons underlying these similarities have yet to be established.

Conclusions

New ndings from the 2007–2010 excavation seasons at Kulbulak prompt us not only to revise the materials from earlier excavations (Kasymov, 1990) but also to revisit the interpretation of the Upper Paleolithic of that region as a whole. The new study of the Kulbulak, Kyzyl-Alma-2, Dodekatym-2, and Shugnou Upper Paleolithic industries (Kolobova et al., 2010; Kolobova, Krivoshapkin, Derevianko, Islamov, 2011; Ranov, Kolobova, Krivoshapkin, 2012) suggests that they represent one and the same Upper Paleolithic tradition, which the present authors named Kulbulakian (Kolobova, Krivoshapkin, Derevianko, Islamov, 2011; Ranov, Kolobova, Krivoshapkin, 2012). This tradition originated in the Upper Paleolithic of the western and southwestern Tien Shan, and its features are bladelet technology with a peculiar microlithic toolkit which includes backed tools and triangles. At present, the origin, maturation, peak, and decline of that tradition, and its eventual transformation into the Mesolithic culture can be traced. Distinct stages in its evolution have been described based on the stratigraphy of the sites, technological and typological characteristics of the industries, and absolute dates.

The early stage (Kyzyl-Alma-2, Kulbulak horizon 2.2, and Shugnou layers 4–2) is characterized by prevalence of uni- and bidirectional parallel aking of flat cores. However, in these technocomplexes, prismatic and narrow-face cores for blades and bladelets are present and increase in number. Carinated cores for detaching curved blades are either absent (Shugnou layer 4) or an exception (Kyzyl-Alma-2, Kulbulak horizon 2.2, Shugnou layers 3 and 2). The proportion of


bladelets among the blanks is insigni cant. The toolkits are dominated by scrapers of various sorts (primarily endscrapers). Ventral or alternate variants of scrapers are present. The assemblages also comprises mostly longitudinal sidescrapers, retouched points, chisel-like implements, and some microtools (retouched and Dufour bladelets). The assemblage from Shugnou layer 3 represents the terminal stage of the Kulbulakian tradition. It includes a considerable number of carinated artifacts, bladelets with a curved pro le, and distinct microtools. This stage can be tentatively dated to 30 (35?)–25 ka BP.

The developed Kulbulakian tradition (Kulbulak horizon 2.1, Shugnou layer 1, Dodekatym-2 horizon 5) is characterized by predominance of prismatic aking for laminar and small laminar blanks. Carinated cores for bladelets with curved pro le are most numerous in the category of prismatic nuclei. The carinated cores were reduced within a single technological scheme. Bladelets including those with curved profile form a fairly high percentage of all blanks. Narrow-face cores were also mostly used for bladelet production. The prismatic double platform blade cores are also available. However, the toolkits show growing numbers of flake blanks that were used in the production of formal tools. Scrapers of various forms dominate the assemblages; ventral varieties of the end- and angular scrapers are present. Some assemblages comprise a signi cant number of chisel-like tools. The proportion of microtools increases; these are mostly retouched and Dufour bladelets. Several backed bladelets and solitary inequilateral triangular microliths were recorded. Absolute dates suggest that this stage is older than 21–23 ka, likely 23–25 ka.

At the nal stage (Dodekatym horizons 4–2), the Kulbulakian assemblages gradually evolved into an industry characterized by prismatic unidirectional aking and a growing proportion of backed tools and

triangular microliths. Carinated cores for bladelets with a curved profile were replaced by prismatic, single platform cores for bladelets with a straight profile; these nuclei became the dominant core type. This was possibly linked with the increased necessity for triangular microliths fashioned on such blanks. Scrapers and chisel-like implements, including the orthogonal varieties, continue to be prevailing tool types. However, microtools primarily represented by backed bladelets and triangular microliths became more signi cant. This stage is later than 20 ka.

The origins of the Kulbulakian tradition could be related to the gradual evolution of local nal Middle Paleolithic and transitional blade industries such as Khudji in Tadjikistan, Obi-Rakhmat in Uzbekistan, and

Kulbulak (layer 23, 2007–2010 excavations) (Ranov, Amosova, 1984; Derevianko et al., 2001; Krivoshapkin et al., 2010). These industries were aimed at the production of blades and points, detached from at and subprismatic cores. Blades detached from narrow-face and subprismatic cores are numerous as are bladelets resulting from the reduction of core-burins, narrow-face wedge-shaped, truncated-faceted and subprismatic cores; some bladelets are retouched. Several carinated pieces were found in Kulbulak layer 23 and Obi-Rakhmat layer 21. Those industries, therefore, include virtually all elements that might have evolved into the early Kulbulakian tradition: developed blade and point technique, volumetric core reduction, and numerous constituents of the bladelet technique including narrow-face wedge-shaped cores and carinated pieces.

Given the dates of the upper horizons of Obi-Rakhmat (Derevianko et al., 2001) and Khudji (Ranov, Amosova, 1984), direct parallels between Kulbulak layer 23, Obi-Rakhmat, and Khudji, on the one hand, and the early stage of Kulbulakian Upper Paleolithic tradition, on the other (Shugnou layer 4, Kyzyl-Alma-2, Kulbulak horizon 2.2) are hardly warranted. Even if the lower chronological boundary of that stage is taken, a long gap remains, which requires additional analysis. However, the idea that the Kulbulakian industry was introduced from without cannot be supported by any factual data, so the idea of in situ origin is more plausible and should be accepted as the working hypothesis.

The subsequent evolution of the Kulbulakian tradition can apparently be traced in Mesolithic cultures of western Central Asia (Islamov, 1980; Ranov, Karimova, 2005). Because no absolute dates of Mesolithic sites excavated in the late 20th century are available, and several sites were attributed to that stage solely on the basis of the microblade and microlithic technique, the discovery of developed microliths in the Upper Paleolithic (Dodekatym-2) requires a revision of earlier interpretations of “Mesolithic” sites. In the rst place, absolute dates are needed. When they have been obtained, Upper Paleolithic (speci cally Kulbulakian) sources of local Mesolithic cultures will hopefully become more apparent.

Acknowledgements

Drawings of the artifacts were made by N.V. Vavilina, Institute of Archaeology and Ethnography SB RAS. The authors are grateful to their colleagues from the Institute of Archaeology and Ethnography SB RAS and from the Institute of Archaeology, Uzbekistan Academy of Sciences, for critical comments and discussions in the course of eld work and preparation of the paper.


References

Abramova Z.A. 1984Pozdnii paleolit Aziatskoi chasti SSSR. In Paleolit SSSR.

Moscow: Nauka, pp. 304–308.Belfer-Cohen A., Goring-Morris A.N. 2007From the beginning: Levantine Upper Palaeolithic cultural

change and continuity. In Rethinking the Human Evolution. Cambridge: Univ. of Cambridge, pp. 199–205. (McDonald Inst. for Archaeological Research Monographs).

Conard N., Ghasidian E. 2011The Rostamian cultural group and the taxonomy of Iranian

Upper Paleolithic. In Between Sand and Sea: The Archaeology and Human Ecology of Southwestern Asia. Tubingem: Kerns Verlag, pp. 33–53.

Coon C.S., Ralph E.K. 1955Radiocarbon dates for Kara Kamar, Afghanistan. Science,

vol. 122, No. 3176: 921–923.Davis R.S. 2004Kara Kamar in Northern Afghanistan: Aurignacian,

Aurignacoid, or just plain upper Paleolithic? In Archaeology and Paleoecology of Eurasia. Novosibirsk: Izd. IAE SO RAN, pp. 211–217.

Davis R.S, Ranov V.A. 1999Recent work on the Paleolithic of Central Asia. Evolutionary

Anthropology, vol. 8: 186–193.Derevianko A.P., Krivoshapkin A.I., Anoikin A.A., Islamov U.I., Petrin V.T., Saifullaev B.K., Suleimanov R.H. 2001The initial Upper Paleolithic of Uzbekistan: The lithic

industry of Obi-Rakhmat Grotto (on the basis of materials recovered from strata 2–14). Archaeology, Ethnology and Anthropology of Eurasia, No. 4: 42–63.

Filimonova T.G. 2007Verkhnii paleolit i mezolit Afgano-Tajikskoi depressii.

Cand. Sc. (History) Dissertation. St. Petersburg.Islamov U.I. 1980Obishirskaya kultura. Tashkent: Fan.Kasymov M.R. 1990Problemy paleolita Srednei Azii i Yuzhnogo Kazakh-

stana (po materialam mnogosloinoi paleoliticheskoi stoian-ki Kulbulak). D.Sc. (History) Dissertation. Novosibirsk.

Kolobova K.A., Krivoshapkin A.I., Derevianko A.P., Islamov U.I. 2011The Upper Paleolithic site of Dodekatym-2 in Uzbekistan.

Archaeology, Ethnology and Anthropology of Eurasia, vol. 39, No. (4): 2–21.

Kolobova K.A., Krivoshapkin A.I., Flas D., Pavlenok K.K., Islamov U.I. 2011Karenoidnye izdelia paleoliticheskoi stoianki Kulbulak:

opyt tehniko-tipologicheskoi klassi katsii. Vestnik NGU. Ser.: Istoriya, lologiya, vol. 10, iss. 7: Arkheologia i etnogra ya: 87–99.

Kolobova K.A., Pavlenok K.K., Flas D., Krivoshapkin A.I. 2010Stoyanka Kyzyl-Alma-2 – novyi pamyatnik epokhi verkh-

nego paleolita Zapadnogo Tian-Shania. Vestnik NGU. Ser.: Istoriya, lologiya, vol. 9, iss. 5: Arkheologiya i etnogra ya: 111–123.

Korobkova G.F., Dzhurakulov M.D. 2000Samarkandskaya paleoliticheskaya stoyanka kak etalon

verkhnego paleolita Srednei Azii: (Spetsifika tekhniki rasschepleniya i khozyaistvenno-proizvodstvennoi deyatelnosti). Stratum Plus, No. 1: 385–462.

Krivoshapkin A.I., Kolobova K.A., Flas D., Pavlenok K.K., Islamov U.I., Lukiyanova G.D. 2010Industriya sloya 23 stoyanki Kulbulak po materialam

raskopok 2010 goda. In Problemy arkheologii, etnografii i antropologii Sibiri i sopredelnykh territorii: Materialy Godovoi sessii IAE SO RAN 2010 g., vol. XVI. Novosibirsk: Izd. IAE SO RAN, pp. 105–110.

Mellars P., Tixier J. 1989Radiocarbon-accelerator dating of Ksar Akil (Lebanon) and

the chronology of the Upper Palaeolithic sequence in the Middle East. Antiquity, vol. 63: 761–768.

Nadel D. 2003The Ohalo II int assemblage and the beginning of the

Epipalaeolithic in the Jordan valley. In More than Meets the Eye: Studies on Upper Paleolithic Diversity in the Near East. Oxford: The Short Run Press, pp. 216–230.

Novye issledovaniya paleolita v Akhangarone (Uzbekistan). 1995N.K. Anisyutkin, U.I . Islamov, K.A. Krakhmal,

B. Saifulaev, N.O. Khushvakov. St. Petersburg: IIMK RAN. (Arkheologicheskiye izyskaniya; iss. 28).

Olshewski D. 1993aThe Late Baradostian Occupation at Warwasi Rockshelter,

Iran. In The Paleolithic Prehistory of the Zagros-Taurus. Philadelphia: The Univ. Museum, Univ. of Pennsylvania, pp. 187–206.

Olshewski D. 1993bThe Zarzian occupation at Warwasi Rockshelter, Iran. In The

Paleolithic Prehistory of the Zagros-Taurus. Philadelphia: The Univ. Museum, Univ. of Pennsylvania, pp. 207–336.

Olszewski D. 1999The Early Upper Paleolithic in the Zagros Mountains.

In Dorothy Garrod and the Progress of the Paleolithic Studies in the Prehistoric Archaeology of the Near East and Europe. Oxford: Oxbow Books, pp. 167–180.

Olszewski D., Dibble H. 1994The Zagros Aurignacian. Current Anthropology, vol. 35:

68–75.Olszewski D., Dibble H. 2006To be or not to be Aurignacian: The Zagros Upper

Paleolithic. In Towards a De nition of the Aurignacian. Lisboa: Instituto Portugés de Arqueologia, pp. 355–373.

Otte M., Shidrang S., Zwyns N., Flas D. 2011New radiocarbon dates for the Zagros Aurignacian from

Yafteh Cave, Iran. Journal of Human Evolution, vol. 61: 340–346.

Pavlenok K.K. 2011Tekhnologii obrabotki kamnya v verkhnem paleolite

Zapadnogo Tian-Shanya (po materialam stoyanki Kulbulak). Cand. Sc. (History) Dissertation. Novosibirsk.

Ranov V.A. 1964O sviaziah kultur paleolita Srednei Azii i nekotoryh stran

zarubezhnogo Vostoka. Moscow: Nauka.


Ranov V.A. 1972K probleme vydeleniya lokalnykh kultur v Srednei Azii.

In Kamennyi vek Srednei Azii i Kazakhstana. Tashkent: Fan, pp. 34–47.

Ranov V.A. 1973Shugnou – mnogosloinaya paleoliticheskaya stoyanka

v verhoviah r. Yakhsu (raskopki 1969–1970 gg.). Arkheo-logicheskiye raboty v Tajikistane, iss. 10 (1970 g.): 42–61.

Ranov V.A. 1988Kamennyi vek Yuzhnogo Tajikistana i Pamira. D.Sc.

(History) Dissertation. Novosibirsk.Ranov V.A., Amosova A.G. 1984Raskopki musterskoi stoyanki Hudji v 1978 godu. Ar-

kheologicheskiye raboty v Tajikistane, iss. 18: 11–58.Ranov V.A., Karimova G.R. 2005Kamennyi vek Afgano-Tajikskoi depressii. Dushanbe:

Devashtich.Ranov V.A., Kolobova K.A., Krivoshapkin A.I. 2012The Upper Paleolithic assemblages of Shugnou, Tajikistan.

Archaeology, Ethnology and Anthropology of Eurasia, vol. 40, No. 2: 2–24.

Ranov V.A., Nesmeyanov S.A. 1973Paleolit i stratigra ya antropogena Srednei Azii. Dushanbe:

Donish.Ranov V.A., Nikonov A.A., Pakhomov M.M. 1976Lyudi kamennogo veka na podstupakh k Pamiru (paleo-

liticheskaya stoyanka Shugnou i ee mesto sredi okruzhaiu-

schikh pamyatnikov). Acta Archaeologica Garpatica, vol. XVI: 5–18.

Taimagambetov Zh.K., Ozhereliev D.V. 2009Pozdnepaleoliticheskie pamyatniki Kazakhstana. Almaty:

Kazak universitet .Tashkenbaev N.H., Suleimanov R.H. 1980Kultura drevnekamennogo veka doliny Zarafshana.

Tashkent: Fan.Vinogradov A.V. 2004Zagadochnyi Kara-Kamar. In Arkheologiya i paleoekologiya

Evrazii. Novosibirsk: Izd. IAE SO RAN, pp. 58–79.Vishnyatsky L. 1999The Paleolithic of Central Asia. Journal of World Prehistory,

vol. 13, No. (1): 69 122.Wahida G. 1999The Zarzian industry of the Zagros Mountains. In Dorothy

Garrod and the Progress of the Paleolithic: Studies in the Prehistoric Archaeology of the Near East and Europe. Oxford: Oxbow Books, pp. 181–208.

Received August 7, 2012.

The Kulbulak Bladelet Tradition in The Upper Paleolithic of Central Asia

Documents