Kaletepe Deresi 3 (Turkey): Archaeological evidence for early human settlement in Central Anatolia

Available online at www.sciencedirect.com

Journal of Human Evolution 54 (2008) 99e111

Kaletepe Deresi 3 (Turkey): Archaeological evidencefor early human settlement in Central Anatolia

Ludovic Slimak a,*, Steven L. Kuhn b, Helene Roche c, Damase Mouralis d,Hijlke Buitenhuis e, Nur Balkan-Atlı f, Didier Binder g,

Catherine Kuzucuoglu c,h, Herve Guillou i

a Economies Societes et Environnements Prehistoriques, UMR 6636, 5 rue du Chateau de l’Horloge, BP 647, 13094, Aix-en-Provence cedex 2, Franceb Department of Anthropology, Unioversity of Arizona, Tucson, AZ 85721-0030, USA

c Prehistoire et technologie, UMR 7055, Nanterre, Franced Laboratoire de Geographie Physique, UMR 8591, 1 place Aristide Briand, 92195 Meudon, France

e Groningen University, Archaeological Research and Consultancy, Hollandf Istanbul University, Department of Prehistory, Turkey

g UMR 6130, 250 Avenue Einstein, 06560 Valbonne, Franceh Institut Francais d’Etudes Anatoliennes, MAE, Nuru Ziya Sok. 22, PK 54, Beyoglu, 80072, Istanbul, Turkey

i LSCE, Bat. 12, avenue de la terrasse 91198 Gif-sur-Yvette, France

Received 9 May 2006; accepted 3 July 2007

Abstract

Located in the Central Anatolian Volcanic Province, Kaletepe Deresi 3 was discovered in the summer of 2000 and has been under investi-gation since that time. Volcanic activity in the region generated a number of obsidian intrusions that have attracted humans to the area throughoutprehistory. The stratigraphic sequence at Kaletepe Deresi 3, more than 7 m in depth, presents a series of archaeological horizons representing theLower and Middle Paleolithic. The site contains the longest open-air Paleolithic sequence excavated in Turkey, as well as the first in situ Acheu-lean industry documented in Anatolia. Tephras in the upper Middle Paleolithic horizons and the rhyolithic bedrock bracket the timespanrepresented at Kaletepe Deresi 3. The lithic industry at the site illustrates a wide range of technological behaviors and documents changesin raw-material exploitation and artifact manufacture through the Lower and Middle Paleolithic.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Lithic technology; Lower Paleolithic; Middle Paleolithic; Obsidian; Raw material exploitation; Technological evolution; Tephrochronology; Turkey

Introduction

The Anatolian peninsula is the geographic and biogeo-graphic junction between the Near East, central Asia, andEurope. This region is potentially of great relevance to under-standing cultural and demographic interactions within Eurasiaand between Eurasia and Africa during the Pleistocene. Forexample, the origins of Acheulean technologies in Europeand western Asia and their possible links with population

* Corresponding author.

E-mail address: [email protected] (L. Slimak).

0047-2484/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jhevol.2007.07.004

dispersals out of Africa remain open questions subject to con-siderable debate (Carbonell et al., 1999; Goren-Inbar et al.,2000; Anton and Swisher, 2004). Resolving such questionswill demand data from areas such as Turkey that lie betweenbetter-studied regions such as Western Europe and the Levant.Due mainly to a lack of primary field research, little is cur-rently known about the early human presence in the regionduring the Pleistocene. Only a handful of Paleolithic siteshave been excavated in Turkey to date, and almost all ofthem are situated on or near the Mediterranean coast.

This paper presents a synthesis of results from the excava-tions at Kaletepe Deresi 3 (KD3), a Paleolithic site located in

100 L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

a volcanic region within Central Anatolia (Turkey) (Fig. 1).Research at KD3 has been conducted by a joint French andTurkish team since 2000 (Slimak, 2004a; Slimak et al.,2004). The site contains the first stratified, multicomponentPaleolithic archaeological sequence documented in CentralAnatolia. The KD3 stratigraphy contains multiple Middleand Lower Paleolithic archaeological layers, including theonly documented in situ Acheulean assemblages in Turkey.The archaeological sequence embodies a number of long-term changes in human behavior related to artifact productionand raw-material exploitation across the Lower and MiddlePaleolithic. The site can provide data relevant to questionsabout hypothetical Acheulean dispersal events and the abilitiesof Lower Paleolithic hominins to survive in cold, temperateenvironments.

Geology, sedimentology, and tephrochronology

The KD3 site is located within the Gollu Dag volcaniccomplex, which contains a number of chemically and macro-scopically distinct obsidian sources (Wagner and Weiner,1987; Bigazzi et al., 1993, 1998; Chataigner et al., 1998; Poi-devin, 1998). Obsidians from Gollu Dag were exploited fortool-making throughout prehistory and were widely traded inthe eastern Mediterranean region during the Neolithic (Cau-vin, 1996; Cauvin and Balkan-Atlı, 1996; Chataigner et al.,1998; Binder and Balkan-Atlı, 2001). The KD3 site is locatedseveral kilometers east of the volcano’s caldera, close to theKomurcu obsidian source, which is one of the largest and cul-turally most important obsidian sources in the region, withoutcrops extending over more than 3 km of the mountain’slower flanks. The Paleolithic occurrences were discovered(by the senior author) during fieldwork at the early Neolithicworkshops on the Komurcu obsidian source.

Fig. 1. Location of Kaletepe Deresi 3 and other sites discussed in text: (1) Kalet

Latamne.

The KD3 site is situated on the south bank of a seasonalstream bed (dere), near the southeastern edge of the main Ko-murcu obsidian area. Given the topography of the bank, it wasnecessary to expose the deposits through two separate steppedexcavations (called ‘‘locus Amont’’ and ‘‘locus Aval’’) locatedclose to one another. Each excavation locus is approximately15 m2 in size. Stratigraphic sections revealed a series of allu-vial and colluvial layers more than 7 m in depth containingboth archaeological horizons and volcanic tephras. This allowsfor presentation of Paleolithic remains in the framework ofPleistocene volcanic events in the area.

The KD3 site is situated within the Central Anatolian Vol-canic Province (CAVP). Volcanic activity in the area beganduring the Miocene, if not earlier, and continued during theQuaternary with the creation of stratovolcanoes (Hasan Dagand Erciyes Dag), monogenic vents, and major rhyolitic com-plexes (Acıgol and Gollu Dag). The Gollu Dag complex,within which the KD3 site is situated, is approximately12 km in diameter. Its genesis began with the eruption of largepyroclastic flows and the collapse of a caldera, followed by theextrusion of nine rhyolitic domes. These extrusions present anoriginal model associating phreatomagmatic and phreatopli-nian activity, pelean activity, the intrusion of annular dykesof obsidian and perlite, and the extrusion of rhyolitic domes(Mouralis et al., 2001, 2002). Activity has been dated to theearly and middle Pleistocene: fission-track ages of obsidiandykes range between 1.48� 0.09 and 0.98� 0.06 Myrs (Big-azzi et al., 1993, 1998), and K/Ar ages between 1.1� 0.02 and0.44� 0.07 Myrs (Mouralis, 2003). This obsidian was usedextensively for tool-making during the Paleolithic, Neolithic,and Chalcolithic periods. The creation of the Gollu Dag com-plex before 1.0 Mya places a maximum age limit on the ar-chaeological deposits at KD3.

The archaeological excavations are situated on the rightbank of a ravine on the southeastern slope of Kabak Tepe,

epe Deresi 3, (2) Karain Cave, (3) Dursunlu, (4) Yarımburgaz Cave, and (5)

101L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

one of the domes extruded inside the Gollu Dag rhyolitic com-plex. An intrusive obsidian dyke associated with this extrusionhas been K/Ar dated to 1.1� 0.02 Ma (Mouralis, 2003). Sec-tions present alluvial and colluvial layers interstratified withanthropogenic formations and tephras, allowing us to proposea tephrochronological framework for some of the recent volcanicevents in the area. Basic sedimentological units are describedbelow.

Sedimentological features (Fig. 2)

The stratigraphic sequence at KD3 can be divided into threemajor cycles of deposition. There are slight differences be-tween the two excavation loci, but the basic sequences aresimilar. For simplicity, the following description refers to lo-cus Aval, which had the deeper and more complete sequenceat the end of the 2005 campaign.

The earliest sedimentary cycle, cycle 1, consists of collu-vial deposits containing reworked rhyolitic, perlitic, and obsid-ian blocks in a silt-and-sand matrix. The basal sedimentsare clast-supported, with a maximum grain size (Tmax) of39 cm. The clasts are angular and composed exclusively ofgray perlitic rhyolite. This formation, which appears to havebeen deposited by gravitational processes, covers a paleoslopethat is still exposed in parts of the right bank of the dere. Theamount of fine-grained sediment decreases with depth withincycle 1, so that higher up in the cycle, sediments are matrix-supported, with smaller blocks (Tmax between 10 and16 cm), suggesting a range of slope processes including soli-fluction. Archaeological levels III, IV, V, Vam, and VIeXIIare contained within the colluvial sediments of cycle 1.

Fig. 2. Stratigraphy of Kaletepe Deresi 3. Archaeological levels are indicated

with roman numerals. Note that not all levels appear in this section.

Cycle 2 consists of much finer, silt-to-silty-clay sediments,with occasional larger ryholite and andesite blocks. It appearsto have been deposited by low-energy alluvial processes, in-cluding slope wash. In parts of the section, the contact be-tween cycles 1 and 2 has been destroyed by the incision ofa paleochannel 120 to 50 cm deep. This paleochannel is filledwith coarse sand and gravel and shows irregular laminationsand cross-bedded structures, indicating an alluvial origin.Archaeological levels II and II0 are contained within the intactsediments of cycle 2.

Sedimentary cycle 3 consists of middle-to-fine-grainedsands, small gravels, and open-work lenses with better sorting.The coarser sands are composed of pumice, rhyolite, and ob-sidian. The top part of the whole section presents a massive,light-brown, silty-to-sandy formation containing pumicelikecoarse sands and gravels. These sediments are exposed overthe modern slopes in the area around the site. Archaeologicallayers I and I0 are contained within sedimentary cycle 3.

Cycle 3 includes six air-fall tephra layers covering a basaldark-brown paleosol. Downstream, where the tephras havebeen reworked (by alteration and bioturbation), they formclay and silty-sand lenses, but upstream these layers are insitu. Tephra R1 is a whitish lapilli pumice fall, 15 cm thick.Its upper part seems to have been reworked. The R2 and R4tephras are gray ash falls, 3 and 4 cm thick, respectively.The R3 and R5 tephras are white lapilli pumice falls contain-ing rhyolite and obsidian xenoliths; they are 25 and 20 cmthick, respectively. The top of R4 has been reworked by bio-turbation and/or pedogenetic processes. The R6 tephra isa black-colored ash fall with vesiculated clasts that is gener-ally intensely bioturbated. Despite the reworking features,the homogeneity of the material in each layer indicates thein situ character of the tephra falls; stratigraphy also showsthat the timespans between each fall (eruption) were verybrief. No archaeological materials have been recovered fromthe tephra layers.

Chemical characterization of the tephra deposits

Geochemical analyses of single glass shards from the sixtephra layers at KD3 interstratified between archaeologicallevels I0 and II were performed using an electronic microprobe(EPMA) in the ‘‘Magmas et Volcans’’ Laboratory in Cler-mont-Ferrand (France). The mean SiO2 content of glass ele-ments (glass shards and pumices) from the R1eR5 tephrasranges from 76.5% to 75.8%, while the value for the R6 tephrais 63.9% (Table 1). Figure 3, the ‘‘total alkali silica’’ (TAS) di-agram (after Le Bas et al., 1986), shows that the R1eR5tephras correspond to the rhyolite field, while R6 correspondsto the dacite and trachydacite fields. Although this classifica-tion is generally applied to whole-rock analyses, the low min-eral content of the tephras justifies it in this instance; use of themethod also facilitates comparisons with other publishedresults. Low standard errors confirm the lack of reworking,except in the cases of R1-r and R3. The higher standard errorexhibited by the R6 tephra is typical of less differentiatedglass.

Table 1

Geochemical analyses of tephra from Kaletepe Deresi 3

Tephra Number of analyses Na2O MgO K2O CaO TiO2 FeO MnO SiO2 Al2O3 Total

R1-r* 9 4.000 0.032 4.562 0.678 0.101 0.779 0.066 73.471 12.737 96.425

0.230 0.026 0.140 0.081 0.067 0.189 0.060 2.306 0.474 2.734

R1 8 3.745 0.062 4.440 0.729 0.068 0.960 0.053 72.451 12.447 94.955

0.132 0.014 0.141 0.052 0.039 0.070 0.043 0.433 0.135 0.415

R2 7 3.676 0.063 4.539 0.717 0.052 0.916 0.051 72.753 12.379 95.146

0.143 0.033 0.136 0.080 0.031 0.110 0.051 0.489 0.114 0.392

R3 12 3.869 0.057 4.780 0.639 0.069 0.810 0.028 74.425 12.545 97.222

0.247 0.031 0.466 0.139 0.044 0.187 0.036 2.149 0.298 2.518

R4 10 3.712 0.084 4.491 0.719 0.078 0.924 0.042 72.139 12.449 94.638

0.207 0.042 0.173 0.055 0.040 0.088 0.026 0.330 0.127 0.469

R5 6 3.780 0.071 4.539 0.732 0.113 0.973 0.046 71.375 12.514 94.143

0.138 0.018 0.265 0.112 0.039 0.052 0.039 0.514 0.165 0.862

R6 14 4.561 1.468 3.130 3.602 1.784 6.349 0.121 62.188 14.039 97.240

0.573 0.312 0.394 0.490 0.211 1.049 0.035 2.355 0.406 1.808

Analytical details: operator: D. Mouralis, Microprobe Cameca SX 100; ‘‘Magmas et Volcans’’ Laboratory, Clermont-Ferrand; beam defocalized to 10 mm; tension

15 kv; intensity 6nA; count time 10 s.* R1 reworked.

102 L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

Origins and ages of the R1eR6 tephras

Four volcanoes in the CAVP have seen rhyolitic activityduring the Quaternary: Hasan Dag, Erciyes Dag, Acıgol, andGollu Dag. Erciyes Dag and Gollu Dag are unlikely to be or-igins for the R1eR5 tephras: the first is located about 100 kmaway, and the second was active mainly during the early Pleis-tocene (Bigazzi et al., 1993; Mouralis, 2003). The artifacttechnology identified in archaeological levels I and II, locatedbelow the ash falls, is Middle Paleolithic in character and iscertainly more recent. Moreover, the grain size and thicknessof the R1eR5 tephras point to a distal position, thus excludingan origin within the Gollu Dag complex itself.

In order to locate the origin of the R1eR5 tephras, their geo-chemical profiles were compared with data from analyses ofmore than 100 tephras collected from the main Quaternary vol-canoes in Cappadocia (Mouralis et al., 2002; Mouralis, 2003).The geochemistry of the major elements in the R1eR5 tephrasshows strong similarities with that of falls produced during the

Fig. 3. Glass geochemistry of tephras from the Kaletepe Deresi 3 section. TAS

diagram from Le Bas et al. (1986). Normalized to 100%. For analytical details,

see Table 1 caption.

syn-caldera activity in the Acıgol area (Fig. 4; Table 1). Thehigher SiO2 content in the R3 tephra probably reflects lowerreliability of results, as demonstrated by a higher standard error.

The Acıgol tuff was known to have erupted between 180 ka(fission-track age of the obsidian reworked during the erup-tions) and 75 ka (fission-track age of obsidian younger thanthe tuff) (Bigazzi et al., 1993; Druitt et al., 1995). New field-work allowed one of us (DM) to propose a more detailedtephrochronological framework and a minimum age of160 ka (Table 2) for the Acıgol tuff. Indeed, inside the Acıgolvolcanic complex, the syn-caldera tuff overlaps an intrusiveobsidian K/Ar dated to 160� 3 ka. Moreover, the oldest scoriacone inside the Acıgol caldera is K/Ar dated to 154� 4 ka(Mouralis, 2003); this age postdates the collapse of the caldera

Fig. 4. Geochemistry of Kaletepe Deresi 3 rhyolitic tephras and comparison

with some products from Cappadocian Quaternary volcanoes. Normalized to

100%. For analytical details, see Table 1 caption.

Table 2

K/Ar age from Acıgol rhyolitic complex

Sample K* (wt%) Weight molten (g) 40Ar* (%) 40Ar*

(10e12 moles/g)

Age (ka� 2s) Age

(mean value)

Material

DAM e 124 3.683� 0.037 1.61181 5.537 1.034 162� 4 Obsidian under

DAM e 124 3.683� 0.037 6.253 1.011 158� 4 160� 3 Acıgol Tuff

CKUZ/DM-02 1.468� 0.015 1.59906 3.672 0.392 154� 5

CKUZ/DM-02 1.468� 0.015 2.50569 4.005 0.390 153� 4 154� 4 Scoria cones inside

Acıgol Caldera

CKUZ/DM-03 1.656� 0.017 1.59906 4.035 0.373 130� 4

CKUZ/DM-03 1.656� 0.017 2.67446 3.617 0.398 138� 4 134� 3

Age calculations are based on the decay and abundance constants from Steiger and Jager (1977).

103L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

and the emission of the syn-caldera tuff. These ages areroughly consistent with the Mousterian artifact assemblagesrecovered from archaeological levels I and II, above theR1eR5 tephras (see below). The origin of the R6 tephra isunknown. Because of its trachytic-to-dacitic geochemistry, itshould be attributable to one of the 30 monogenic vents lo-cated immediately north of Gollu Dag rhyolitic complex. Ac-cording to Ercan et al. (1987, 1991) and Olanca (1994), mostof these volcanoes have a basaltic-to-andesitic geochemistry,though some have more differentiated products. Publisheddata concerning some of these monogenic volcanoes indicatemiddle Pleistocene ages. For example, K/Ar ages of twolava flows from Koytepesi and Catal Tepe dated by Olanca(1994) are respectively 181.8� 9.8 and 72.2� 7.2 ka. Kuzu-cuoglu et al. (1998) obtained a 92� 4 ka age for a lava flownorthwest of the Nenezi Dag. These data are roughly consis-tent with the age of the R1eR5 tephras.

Archaeological remains and technological trends

The archaeological finds from KD3 consist almost exclu-sively of lithic artifacts. Bone is very scarce in the site; the

Table 3

Content of lithic artifact assemblages from Kaletepe Deresi 3

Level Flakes Tools Cores Fragments Bifaces

n % n % n % n % n

I 46 75.41 4 6.56 11 18.03

I0 9 100.00

II 387 63.34 40 6.55 14 2.29 169 27.66

II0 107 72.30 10 6.76 5 3.38 26 17.57

III 207 63.69 24 7.38 21 6.46 70 21.54

III/IV 46 43.40 7 6.60 16 15.09 32 30.19

IV 676 44.04 85 5.54 169 11.01 509 33.16 3

V 209 54.01 19 4.91 25 6.46 88 22.74 1

Vam 30 32.97 8 8.79 9 9.89 29 31.87 1

V0 4 6.90 2 3.45 50 86.21

VI 3 4.76 1 1.59 57 90.48 1

VI0 31 49.21 1 1.59 26 41.27 1

VII 17 23.29 1 1.37 1 1.37 38 52.05 1

VIII 81 53.29 3 1.97 51 33.55 2

IX 30 28.85 67 64.42

X 14 46.67 14 46.67

XI 40 50.63 2 2.53 24 30.38

XII 32 69.57 11 23.91

Total 1969 49.96 199 5.05 268 6.80 1272 32.28 9

faunal sample is currently limited to a single equid mandibleand isolated teeth from level II (described below) and a fewfragments of bone from lower levels. This is probably theresult of high acidity and the abrasive nature of the sediments.However, we cannot rule out the possibility that the scarcity ofbone is related to the kinds of activities that took place at KD3prehistorically.

The locations of all objects observed during excavationswere recorded in three dimensions (x, y, z), facilitating illustra-tion of the spatial distribution of artifacts. In order to examinethe vertical distributions of finds and to check distinctionsbetween archaeological levels defined based on field observa-tions, all of the lithic artifacts from the Amont locus were pro-jected onto the xz and yz axes (not shown). The resultingprojections show a very good definition of levels along the xaxis (perpendicular to the stream). Due to the natural slopeof the deposits, projections along the y axis are less clear.

Taking all archaeological levels together, nearly 4000 lithicartifacts have been recovered from the two localities at KD3 asof the end of the 2005 excavation season (Table 3). The ar-chaeological assemblages include a variety of volcanic rocks,including obsidian and rhyolite, andesite, and basalt. Most of

and cleavers Choppers and chopping tools Polyhedrons Total

% n % n %

61

9

1 0.16 611

148

3 0.92 325

3 2.83 2 1.89 106

0.20 56 3.65 37 2.41 1535

0.26 36 9.30 9 2.33 387

1.1 9 9.89 6 6.59 92

2 3.45 58

1 1.59 63

1.59 4 6.35 63

1.37 15 20.55 73

1.32 2 1.32 13 8.55 152

1 0.96 6 5.77 104

2 6.67 30

13 16.46 79

3 6.52 46

0.23 109 2.77 115 2.92 3941

104 L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

these raw materials are available within a few hundred metersor less of the site, but a few allochthonous obsidian specimenshave been recognized based on macroscopic characteristics.Raw-material selection varied over the long sequence atKD3. Figure 5 shows the proportions of obsidian andcoarse-grained stones in the assemblage from the Amont lo-cus; the assemblages from the Aval locus contain more obsid-ian overall, but the trends are the same. In the most recentarchaeological layers (I, I0, II and II0), obsidian is the dominantraw material, although individual assemblages show differentmodes for working the obsidian, including Levallois and dis-coid debitage. In the oldest occupationsdlevels IV, V, andVIeXIIdhominin toolmakers used local rhyolites, andesites,and basalts much more extensively. These course-grainedraw materials were most often used for the production ofheavy choppers and other core tools and for the manufactureof large flakes. The obsidian present in the earliest assem-blages resulted mainly from handaxe production. The individ-ual assemblages are described in more detail below, startingwith the most recent layers.

Level I and I0, the most recent archaeological componentspreserved at KD3, cover tephras R1eR5. Thus, they mustpostdate the deposition of these tephras at around 160 ka. Asof the end of the 2005 seasons, these levels have yieldedonly 61 and 9 artifacts, respectively. There is evidence for pro-duction of both Levallois and Kombewa flakes, and manyblanks exhibit facetted platforms. Retouched tools are limitedto some fragments of sidescrapers. Both the technologicalcharacteristics and the retouched tools suggest that these as-semblages are attributable to the Middle Palaeolithic, sensulato.

Level II is covered by the R1eR5 tephra layers and is thusolder than 160 ka, dating to at least oxygen-isotope stage 6.This level has provided one of the largest artifact assemblages

Fig. 5. Proportions of obsidian and coarse-grained volcanic rocks (rhyolite,

andesite, and basalt) in levels IIeV (Amont locus only, n¼ 1472).

in the sequence at KD3 (611 specimens as of the end of the2005 season). Only a few artifacts from level II are made onrhyolite and basaltdthe great majority are of obsidian. Afew obsidian specimens have been attributed to the outcropsat Sırca Deresi, located 5 km from Kaletepe, based on distinc-tive visual characteristics. Scarce as it is, the presence of thiskind of obsidian is surprising given the profusion of raw ma-terial of equal or better quality in the immediate surroundingsof the site. Thus far, only level II has yielded raw materialsthat are demonstrably not local in origin.

The archaeological assemblage from level II providesevidence for a variety of Levallois production strategies. Thecommon element to all Levallois technology is the systematicpreparation of the core in the initial stages of exploitation inorder to obtain large flakes of more-or-less predeterminedshape. In level II, two main Levallois production systems re-sulted in manufacture of different kinds of end product: bladesfrom unipolar cores and flakes from centripetally preparedcores (Fig. 6.3). In addition, attributes of some of the bladessuggest that they were produced from unipolar, non-Levalloiscores. Retouched tools from level II (9% of total artifacts) areexclusively represented by Mousterian scrapers and points(Fig. 6.1e2).

Two maxillary molars from a relatively young equid indi-vidual were recovered from level II in locus Aval. A once-complete equid mandible was recovered from the same levelin locus Amont; in fact, it was the presence of this specimenin the stream cut that originally alerted archaeologists to thepresence of Pleistocene deposits in the Kaletepe Deresi local-ity. Because of the scarcity of bone in the site, the presence of

Fig. 6. Kaletepe Deresi 3, artifacts from level II: (1) point, (2) scraper with

ventral thinning, (3) centripetal Levallois core. Drawings by M. Grenet.

Fig. 8. Kaletepe Deresi 3, artifacts from levels II0 and III: (1) large flake, level

II0; (2) flake with use-wear or postdepositional modification, level II0; (3) irreg-

ular scraper or denticulate, level III; (4) core with large removal, level III.

Drawings by L. Slimak, M. Grenet.

105L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

these elements is exceptional. The mandible was closely sur-rounded by a number of stones, and fine aeolian sand wasfound between the stones and around the mandible. We spec-ulate that this less acidic sand protected the bone fromdisintegration.

The mandibular specimen originally included both left andright portions, but the left portion was exposed and has almostcompletely deteriorated. Although part of the ramus and alve-olar walls were broken, most of the right mandible is intactand all dental elements are present (Fig. 7). The surface ofthe bone and the isolated teeth are quite weathered and showchemical erosion, and thus no cut marks could be observed.The jaw is quite robust, indicating a large animal, and thewear pattern on the teeth suggests a mature individual, perhaps16e20 years of age (using modern horses as referents). Adetailed analysis of the dental remains shows this to be a rela-tively primitive form of equid, most closely related to modernasses and hemiones. An isolated molar (which could be fromthe same individual) is also more primitive than late Pleisto-cene cabaline equids from Umm-el-Tlel, Syria (Eisenmannet al., 2002). Unfortunately, there is virtually no comparativePleistocene equid material from Anatolia. The KD3 specimensdo appear somewhat more primitive than modern equids, andthus their presence is consistent with the apparent late middlePleistocene date for level II.

Levels II0 and III are represented by 148 and 325 artifacts,respectively. Obsidian is the dominant raw material in II0,whereas coarse-grained rocks were more extensively used inIII. The lithic industry of level II0 (Fig. 8.1e2) includesa few typical Levallois flakes, but products and by-productsof discoid debitage (Slimak, 1998e1999, 2003, 2004b) arepredominant. Retouched tools include a few irregular scrapers.Level III is characterized by the production of heavy blanks,with thick unprepared platforms and ‘‘open’’ interior platformangles (>> 90 degrees). These flakes were sometimes trans-formed into denticulate and notched tools (Fig. 8.3). Thefew technological elements from layer III indicate unipolarflake production using cores with plain, unprepared platforms.This method of production exhibits some affinities with Leval-lois debitage in that the broadest surface of the core wasexploited (Fig. 8.4), but it lacks the extensive preparation ofplatform and core face that typifies Levallois. The resultingartifacts show minimal preparation and a lack of predetermi-nation in their shapes. Level III contains neither indicationsof true Levallois production nor well-made Mousterian side-scrapers. However, it has yielded a few obsidian flakes typicalof soft-hammer biface production.

Levels IV, V, and VIeXII show major technical contrastswith the more recent Paleolithic levels at Kaletepe. Obsidian

Fig. 7. Right equid mandibular dental row from Kaletepe Deresi 3.

was much less extensively utilized and was limited to certaintechnological procedures. In some levels, flakes and largetools were produced using local rhyolites, basalts, and andes-ites almost exclusively. Knapping experiments show that thelocal versions of these raw materials are anisotropic, withhighly variable working properties, due both to their poor silic-ification and to the presence of many internal fractures andother irregularities. However, the presence of some obsidianartifacts and natural obsidian nodules in the sediments indi-cates that this raw material was available locally during thetime levels IV, V, and VIeXII were deposited.

The archaeological materials from layer IVoccur within the‘‘chaos’’ of natural rhyolite and andesite blocks, apparently re-sulting from the collapse of a small cliff or rock face; this isclearly visible within the upper and middle parts of cycle 3in the sedimentary profile (Fig. 2). The artifact assemblagefrom level IV contains many large cores and heavy core tools(Fig. 9). Blank production is very simple, with almost no prep-aration of core faces and platforms. Nonetheless, there seemsto have been a systematic attempt to detach a large flake fromthe broadest exploitable face of the core (Fig. 10). Dependingof the original shapes of the blocks used and their internalflaws and fractures, this simple strategy resulted in a variety

Fig. 9. Kaletepe Deresi 3, rhyolite chopper from level IV. Drawing by

M. Grenet.

Fig. 10. Kaletepe Deresi 3, artifacts from level IV: (1) obsidian core, (2) rhy-

olite core. Drawings by M. Grenet.

Table 4

Flake, tool, and core ratios for Kaletepe Deresi 3

Level (Flakesþ fragments)/core Flake tools/core

I, I0 d dII, II0 689/19 (36.3/1) 50/19 (2.6/1)

III 277/21 (13.2/1) 24/21 (1.1/1)

IV 1185/169 (7/1) 85/169 (0.50/1)

V, Vam, V0 410/36 (11.4/1) 27/36 (0.75/1)

VIeXII 536/7 (76.6/1) 2/7 (0.29/1)

Levels I and I0 do not contain cores.

106 L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

of residual core forms, including large discoidal and polyhe-dral cores. Materials of different flaking quality, includingobsidian and rhyolite/andesite, seem to have been worked inbasically the same way (Fig. 10). This suggests that the modal-ity of blank production was not simply the result of exploitingpoor-quality raw materials, but a consistent behavioral ten-dency. Rhyolite and andesite blocks flaked as cores weresometimes obtained by breaking apart much larger massesof raw material (Tmax ca. 1 m), either conchoidally or alongnatural fracture planes.

The layer IV assemblage contains comparatively few flakesand flake tools compared to the number of cores and core frag-ments; there are only seven flakes and fragments per core,fewer if choppers and chopping tools are counted as cores(Table 4). In part, this may be due to the tendency of thecoarse-grained raw materials to fracture along internal planes,or it may be an indication that the larger flakes were habituallymoved to another location. Winnowing of the assemblage bywater or gravitational action may also have contributed tothis imbalance. Some of the small flakes collected do show

intentional retouch, mainly forming large notches. The greatmajority of core tools from layer IV are large choppers andchopping tools. As is typically the case, there is a gradationbetween simple flake cores, choppers/chopping tools, and dis-coids, and some artifacts can fit into more than one category(e.g., Collina-Girard, 1975; Toth, 1985). However, some ofthese choppers show evidence for production of a sharp work-ing edge. A few fragmentary, atypical bifaces have been re-covered from level IV as well.

Level V underlies the ‘‘chaos’’ of natural rock, which hasserved to fossilize the archaeological layer and protect it

Fig. 12. Kaletepe Deresi 3, cleavers from level V: (1) obsidian cleaver, (2) an-

desite cleaver. Drawings by M. Grenet.

107L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

from erosion by the stream. The lithic industry is in an excel-lent state of preservation. This level had only been exposedover a surface of excavation of about 6 m2 at the end of the2005 season. To this point, there is little evidence for true flakeproduction in level V. Instead, manufacture seems to havebeen aimed at the production of large core tools. The assem-blage is characterized by the exploitation of different formsof coarse-grained volcanic rock, including basalt, to obtainvery thick and heavy tools (some more than 3 kg), dominatedby chopping tools. Polyhedral artifacts of andesite are alsopresent. Obsidian was employed mainly in the production ofhandaxes using both hard- and soft-hammer percussion(Fig. 11). This level has also yielded two cleavers on flakes,one of obsidian (Fig. 12.1) and the other of andesite(Fig. 12.2). The first artifact was produced using a knappingtechnique that resembles the Tabelbella Tachengit Acheuleancleaver technology of North Africa (Tixier, 1956), and itsdistal edge was resharpened by a large coup de tranchet.The association of handaxes and cleavers in layer V at KD3is broadly diagnostic of the Acheulean in southwest Asia.

Levels VI through XII occur in similar sediments to level V,a colluvial accumulation formed of small fragments of decom-posing rhyolitic bedrock and obsidian. To date, these levelshave been exposed over an area of only 3 m2, in locus Avalonly, within the confines of a small natural hollow or cuvettein the bedrock. Levels VIeXII are much thinner than manyof the overlying levels and appear to represent successive ep-isodes of occupation/deposition separated by brief intervals ofnatural sedimentation.

Although the excavated area is small, levels VIeXII arevery rich. As of 2004, more than 600 artifacts have beenrecovered from these levels. The materials from all six levelsare sufficiently similar to be described as a single unit. Likelevel V, the assemblages from levels VIeXII can be character-ized as Acheulean. They include several handaxes, includingsmall asymmetrical (Fig. 13) and planoconvex forms, aswell as many typical waste flakes from bifacial reduction. Asingle obsidian cleaver-flake along with technological debrisresulting from cleaver production has also been recovered.

Fig. 11. Kaletepe Deresi 3, obsidian handaxe from level V. Drawing by

M. Grenet.

Bifacial technology was executed exclusively in obsidian.Other local raw materials were employed to different techno-logical ends. The most abundant class of shaped artifact inlevels VIeXII consists of polyhedrons (n¼ 53, 9.2% of the as-semblage), which are made almost exclusively of andesite.The polyhedrons vary in diameter from 5 to 15 cm. The small-est were produced from thin tabular pieces of andesite usinghammer-on-anvil percussion. Pebble tools (choppers andchopping tools) of basalt, andesite, and rhyolite are also pres-ent in levels VIeXII.

Discussion and conclusions

The stratigraphic sequence at Kaletepe Deresi 3 presentsone of the most complete, and possibly the oldest, Paleolithicsequence in the Anatolian peninsula. As more absolute datesbecome available, the site promises to shed new light on themovement of technological innovations and perhaps of homi-nin populations between the Near East and Europe during thePleistocene. The archaeological sequence can be divided intothree chronocultural phases (presented in reverse chronologi-cal order).

The most recent archaeological phase at KD3 includeslevels I0, I, and II. These three layers are characterized byLevallois debitage and the presence of well-made sidescrapers,and they are attributable to a Middle Paleolithic or Mousterianoccupation. The dominant raw material is obsidian. Levels I0

and I are separated from level II by tephras R1/R5, whichdate to roughly 160 ka.

The middle phase of occupation comprises levels II0 and III.The assemblages are characterized by non-Levallois reduction,both discoid and heavy unipolar flake production. Retouched

Fig. 13. Kaletepe Deresi 3, obsidian bifaces from level VI. Drawings by

L. Slimak.

108 L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

pieces include irregular scrapers and notched and denticulatedtools. In level III, these tools were manufactured on thickflakes with broad platforms and ‘‘open’’ platform angles,whereas in level II0, blanks were produced mainly using thediscoid method. There is also evidence for manufacture of bi-facial artifacts in level III. Obsidian was the main raw materialexploited in level II0, whereas coarse-grained volcanic stoneswere more extensively exploited in level III, at least withinlocus Amont.

The earliest phase of occupation at KD3, consisting of ar-chaeological levels IV, V, and VIeXII, shows a variety oftechnological and economic behaviors associated with theexploitation of both obsidian and course-grained volcanicraw materials. Level IV is marked by multiple methods ofproducing large flakes and by the manufacture of heavychoppers and chopping tools. Levels V and VIeXII also con-tain evidence for production of large flakes and core tools fromcoarse-grained volcanic rocks, along with bifacial technologyon obsidian and cleavers manufactured from both andesite andobsidian. The andesite polyhedrons, which are most abundantin levels VIeXII, may represent another, independent techno-logical trajectory. The assemblages from levels V and VIeXIIcan be considered a local expression of the Acheulean.

It is rather difficult to identify appropriate comparisons forthe artifact assemblages from Kaletepe Dersi 3 within Turkey.

First, relatively few Lower and Middle Paleolithic sites inAnatolia have been excavated. Second, the Kaletepe assem-blages are heavily dominated by the by-products of artifactproduction. While other activities may well have taken placeat this locality, the lithic assemblages derived mainly fromthe exploitation of abundant local raw material. In contrast,other documented assemblages come from sites located fartherfrom raw material sources, and their assemblages are domi-nated by other phases in the chaınes operatoires.

The closest comparators for the industries of levels I, I0, andII, and indeed the only well-dated early Middle Paleolithic as-semblage currently known in Turkey, come from Karain caveE. The Karain cave complex is located on the Mediterraneancoast near Antalya, nearly 500 km from KD3 (Yalcinkayaet al., 1992; Otte et al., 1998). The Karain E stratigraphycovers a period of approximately 500,000 years and includesLower, Middle, and Upper Paleolithic layers. The collectionfrom layer F at Karain E is characterized by Levallois blankproduction associated with the manufacture of Mousterianscrapers. Current evidence indicates that assemblage F fromKarain E dates to around 200e250 ka (Otte et al., 1998),which would be roughly consistent with the likely age of levelII at KD3.

With respect to an earlier, Lower Paleolithic presence inTurkey, three sites, Karain E, Yarımburgaz Cave, and Dur-sunlu, can be compared with the earlier (pre-Middle Paleo-lithic) layers at KD3. Dursunlu, the oldest of the three andthe site closest to Kaletepe, is located in Central Anatolia,northwest of the city of Konya. Deposits at the Dursunlu local-ity bear evidence of a hominin presence antedating theBrunhes/Matuyama boundary and postdating the Jaramillosubchron (Gulec et al., 1999). The small lithic assemblagerecovered to date provides evidence of flake production frombipolar (hammer-and-anvil) and polyhedral cores in quartzand other raw materials. No evidence of bifacial technologyhas been found in Dursunlu, but the assemblage is relativelysmall. Overall, Dursunlu bears little obvious resemblance toany of the assemblages from KD3, though radical differencesin the kinds of raw materials used in the two sites hindercomparisons.

The oldest artifact assemblage in the Karain caves, assem-blage A, shows a somewhat closer resemblance to the materialfrom levels II0 and III at KD3. The Karain assemblage exhibitsso-called ‘‘Clactonoid’’ features, with large flakes, ‘‘open’’platform angles, and little or no shaping of the core (Yalcin-kaya et al., 1992; Otte et al., 1998). However, there is currentlyno evidence of biface production from the early layers at Kar-ain. The earliest layers at Karain are currently undated, but arethought to be middle Pleistocene in age.

Yarımburgaz Cave, in Aegean Thrace, is a multichamberedkarstic cavern containing evidence for Lower Paleolithic occu-pations dating to the latter half of the middle Pleistocene(Blackwell et al., 1990; Kuhn et al., 1996; Stiner et al.,1996). Like assemblage A from Karain E, the lithic materialfrom Yarımburgaz Cave is characterized by thick flakes andsimple methods of flake production. The majority of the toolassemblage consists of irregular scrapers and denticulated

109L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

and notched tools, but some heavy chopping tools are presentas well. Blanks were produced from simple discoid cores (inflint) and by the hammer-on-anvil bipolar technique (inquartz). There is no evidence of bifacial technology at Yarım-burgaz Cave. Unlike Dusunlu, the excavated area and the lithicassemblage are both large, so the absence of bifaces is nota function of sample size. Level II0 at KD3 resembles theYarımburgaz assemblage in that it shows an association be-tween simple discoid-method blank production and toolswith irregular edges.

Until the discovery of KD3, no in situ archaeological de-posits in Turkey provided evidence for bifacial technology ina Lower Paleolithic context. Levels VeXII at KD3 cast newlight on this issue, clearly demonstrating the use of bifacialhandaxe and cleaver-flake technologies, associated with chop-per/chopping tools and polyhedrons, in Pleistocene-aged sed-iments. The assemblages from layers VeXII could presentsome technical similarities with sites in the Levant, such as‘Ubeidiya (Stekelis, 1966; Bar-Yosef, 1989; Bar-Yosef andGoren-Inbar, 1993; Shea, 1999; Guerin et al., 2003) and par-ticularly with Gesher Benot Ya’aqov (Goren-Inbar et al.,2000) in the Jordan Valley and Latamne, Syria (Clark, 1966,1967, 1968). These latter localities are among the best-docu-mented Acheulean sites and some of the oldest human settle-ments known in the eastern Mediterranean region.

The simple presence of a Lower Paleolithic industry withhandaxes and cleavers in level V at KD3 has potentially signif-icant implications with respect to the distribution of Acheu-lean/Mode II technologies in western Eurasia. Prior to thediscovery and documentation of this site, it was an open ques-tion as to whether the Acheulean was present on the CentralAnatolian plateau (see Kuhn, 2002). Moreover, based on theirstratigraphic position immediately overlying bedrock dated to>1.0 Mya, levels V and VIeXII at KD3 may provide evidencefor a comparatively early Acheulean presence in the region.As such, findings from levels V and VIeXII may be usefulin testing hypotheses about the origins and dispersal ofMode II technologies (e.g., Carbonell et al., 1999; Goren-Inbaret al., 2000; Anton and Swisher, 2004). Much depends on theabsolute ages of the basal archaeological levels at KD3, and inparticular on the amount of time that elapsed between the for-mation of the basal rhyolites and the earliest occupation at thesite. Establishing a better-controlled geochronological frame-work is one of the main priorities for ongoing research atthe site.

Establishing the ages of the earliest layers at KD3 will alsoshed light on the capacities of early hominins to cope withcold-temperate environments and prolonged cold seasons.Sites in western Eurasia securely dating to before 700 ka aresituated at low elevations and/or low latitudes (Bar-Yosefand Goren-Inbar, 1993; Swisher et al., 1994; Gibert et al.,1998; Goren-Inbar et al., 2000; Anton and Swisher, 2004) orin areas with mild local microclimates such as the southernBlack Sea coast of Georgia (Gabunia et al., 2001; but seeZhu et al., 2004). This has led some researchers to concludethat cold-temperate environments were a major obstacle to ho-minin colonization (Dennell and Roebroeks, 1996; Roebroeks,

2001; Hoffecker, 2002). The KD3 site is situated at an eleva-tion of 1600 meters on the Central Anatolian plateau. Todaythe area is characterized by a harsh continental climate withlong, cold, windy winters, and conditions would have beenmore severe during the colder parts of the Pleistocene. Under-standing the ages of the various levels at KD3 and their rela-tions to global climate cycles would tell us whether theearly occupations were confined to warm periods or whetherthey date to cold glacial events.

Continued work at KD3 also has potential to cast light onother little-studied aspects of behavior during the Lower andMiddle Paleolithic. The KD3 site is the only Lower or MiddlePaleolithic artifact production/raw-material extractionsite cur-rently known in Anatolia, and one of the very few known inthe eastern Mediterranean (but see Vermeersch et al., 1990;Barkai et al., 2002). While this fact makes it difficult tocompare it with sites located more distally to points of raw-material procurement, it also means that KD3 provides a per-spective on technological behavior complementary to otherknown localities. The shifts over time between coarse-grainedrocks and obsidian as the preferred raw materials are espe-cially interesting (Fig. 5). By many standards, obsidian wouldseem to be the superior raw material. Yet, the first hominids atKD3 (represented by levels VeXII) used obsidian mainly forproducing handaxes and cleavers, preferring coarse-grainedmaterials for other technological applications. Obsidian wasalmost completely ignored in level IV in favor of andesite,rhyolite, and basalt, which are tougher than obsidian but whichpossess much poorer flaking properties. Only in the later partof the sequence did toolmakers show a strong preference forthe ‘‘best’’ raw material (Fig. 5). These shifts in raw-materialselection indicate that major changes in strategies for the man-ufacture and use of stone tools occurred with or before the ap-pearance of the Middle Paleolithic in this region.

Acknowledgements

Research described in this paper was performed under theauspices of the Mission for Anatolian Prehistory (French Min-istry of Foreign Affairs), and in partnership with the Univer-sity of Istanbul, Department of Prehistory (project 115/15052003 and 235/29042004). Support was also provided bythe French Centre National de la Recherche Scientifique andthe Institut Francais d’Etudes Anatoliennes Georges Dumezil.

References

Anton, S., Swisher, C.C., 2004. Early dispersals of Homo from Africa. Annu.

Rev. Anthropol. 33, 271e296.

Barkai, R., Gopher, A., laPorta, P.C., 2002. Palaeolithic landscape of extrac-

tion: flint surface quarries and workshops at Mt. Pua, Israel. Antiquity

76, 672e680.

Bar-Yosef, O., 1989. The excavations at ‘Ubeidiya in retrospect: an eclectic

view. In: Bar-Yosef, O., Vandermeersch, B. (Eds.), Investigations in South

Levantine Prehistory. Prehistoire du Sud-Levant. BAR International Series,

Oxford, pp. 101e111.

110 L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

Bar-Yosef, O., Goren-Inbar, N., 1993. The Lithic Assemblages of ‘Ubeidiya,

a Lower Paleolithic Site in the Jordan Valley. Qedem 34. Monographs of

the Institute of Archaeology, Hebrew University, Jerusalem.

Bigazzi, G., Yegıngil, Z., Ercan, T., Oddone, M., Ozdogan, M., 1993. Fission

track dating obsidians in Central and Northern Anatolia. Bull. Volcanol.

55, 588e595.

Bigazzi, G., Poupeau, G., Yegıngil, Z., Bellot-Gurlet, L., 1998. Provenance

studies of obsidian artefacts in Anatolia using the fission-track dating

method: an overview. In: Cauvin, M.C., Gourgaud, A., Gratuze, B.,

Arnaud, N., Poupeau, G., Poidevin, J.-L., Chataigner, C. (Eds.), L’obsidi-

enne au Proche et Moyen Orient. Du Volcan a l’Outil. BAR International

Series, Oxford, pp. 69e89.

Binder, D., Balkan-Atlı, N., 2001. Obsidian exploitation and blade technology

at Komurcu-Kaletepe, Central Anatolia. In: Caneva, I. (Ed.), Beyond

Tools: Reconsidering Definitions, Counting and Interpretation of Lithic

Assemblages. Proceedings of the Workshop on PPN Chipped Lithic Indus-

tries (November 1998), Venezia.

Blackwell, B., Schwarcz, H.P., Porat, N., Howell, F.C., Arsebuk, G., 1990.

Electron spin resonance (ESR) dating of Ursus teeth from Yarımburgaz

cave, Turkey. Geol. Soc. Am. Abstr. 22A, 120e121.

Carbonell, E., Mosquera, M., Rodriguez, X.P., Sala, R., Van der Made, J.,

1999. Out of Africa: the dispersal of the earliest technical systems recon-

sidered. J. Anthropol. Archaeol. 18, 119e136.

Cauvin, M.-C., 1996. L’obsidienne dans le Proche-Orient prehistorique: Etat

des recherches en 1996. Anatolica 22, 1e31.

Cauvin, M.-C., Balkan-Atlı, N., 1996. Rapport sur les recherches sur l’obsidi-

enne en Cappadoce, 1993e1995. Anatolia Antiqua 4, 249e271.

Chataigner, C., Poidevin, J.L., Arnaud, N.O., 1998. Turkish occurrences of ob-

sidian and use by prehistoric peoples in Near East from 14000 to 6000 BP.

J. Volcanol. Geotherm. Res. 85, 517e537.

Clark, J.D., 1966. Acheulean occupation sites in the Middle East and Africa:

a study in cultural variability. Am. Anthropol. 68, 202e229.

Clark, J.D., 1967. The middle Acheulian occupation site at Latamne, northern

Syria, I. Quaternaria 9, 1e68.

Clark, J.D., 1968. The middle Acheulian occupation site at Latamne, northern

Syria, II: Further excavations (1965), general results, definition and inter-

pretation. Quaternaria 10, 1e71.

Collina-Girard, J., 1975. Les industries archa€ıques sur galets des terrasses qua-

ternaires de la plaine du Roussillon (Pyrenees-Orientales, France). These

de doctorat, Universite de Provence, Aix-Marseille I, 416 p.

Dennell, R., Roebroeks, W., 1996. The earliest colonization of Europe: the

short chronology revisited. Antiquity 70, 535e542.

Druitt, T.H., Brenchley, P.J., Gokten, Y.E., Francaviglia, V., 1995. Late Quater-

nary rhyolitic eruption from the Acıgol Complex, central Turkey. J. Geol.

Soc. 152, 655e667.

Eisenmann, V., Helmer, D., Sania Segui, M., 2002. The big Equus from the

Geometric Kebaran of Umm El Tlel, Syria: Equus valeriani, Equus capen-

sis or Equus caballus? In: Buitenhuis, H., Choyke, A., Mashkour, M.

(Eds.), Archaeozoology in the Near East V. Archaeological Research

and Consultancy, Groningen, pp. 62e73.

Ercan, T., Akbasxalı, A., Yıldırım, T., Fısxekci, A., Selvi, Y., Olmez, M., Can, B.,

1991. Acıgol (Nevsxehir) yoresindeki senozoyik yasxlı volkanik kayacların

petrolojisi. Maden Tetkik ve Arama Dergisi 113, 31e44.

Ercan, T., Yıldırım, T., Akbasxalı, A., 1987. Gelveri (Nigde) e Kızılcın

(Nevsxehir) arasındaki volkanizmanın ozellikleri. Jeomorfoloji Dergisi 15,

27e36.

Gabunia, L., Anton, S., Lordkipanidze, D., Vekua, A., Justus, A., Swisher, C.,

2001. Dmanisi and dispersal. Evol. Anthropol. 10, 158e170.

Gibert, L., Gibert, L., Iglesias, A., Maestro, E., 1998. The ‘‘Oldowan’’ assem-

blages in the Orce region, southeast Spain. Antiquity 72, 17e25.

Goren-Inbar, N., Feibel, C., Verosub, K.L., Melamed, Y., Kislev, M.E.,

Tchernov, E., Saragusti, I., 2000. Pleistocene milestones on the out-

of-Africa corridor at Gesher Benot Ya’aqov, Israel. Science 289,

944e947.

Guerin, C., Bar-Yosef, O., Debard, E., Faure, M., Shea, J., Tchernov, E., 2003.

Oubeidiyeh, carrefour biogeographique entre l’Afrique et l’Eurasie au

Paleolithique ancien. In: Vandermeersch, B. (Ed.), Echanges et Diffusion

Dans la Prehistoire Mediterraneenne. Comite des Travaux Historiques Et

scientifiques, Paris, pp. 131e145.

Gulec, E., Howell, F.C., White, T., 1999. Dursunluda new lower Pleistocene

artifact-bearing locality in southern Anatolia. In: Ullrich, H. (Ed.), Homi-

nid Evolution: Lifestyles and Survival Strategies. Edition Archae, Berlin,

pp. 349e364.

Hoffecker, J.F., 2002. Desolate Landscapes: Ice-Age Settlement in Eastern

Europe. Rutgers University Press, New Brunswick.

Kuhn, S., 2002. Paleolithic archaeology in Turkey. Evol. Anthropol. 11,

198e210.

Kuhn, S., Arsebuk, G., Howell, F.C., 1996. The middle Pleistocene

lithic assemblage from Yarımburgaz Cave, Turkey. Paleorient 22,

31e49.

Kuzucuoglu, C., Pastre, J.F., Black, S., Ercan, T., Fontugne, M., Guillou, H.,

Hatte, C., Karabıyıkoglu, M., Orth, P., Turkecan, A., 1998. Identification

and dating of tephra layers from Quaternary sedimentary sequences of In-

ner Anatolia, Turkey. J. Volcanol. Geotherm. Res. 85, 153e172.

Le Bas, M.J., Le Maitre, R.W., Streckeisen, A., Zanattin, B., 1986. A chemical

classification of volcanic rocks based on the total alcali silica diagram.

J. Petrology 27, 745e750.

Mouralis, D., 2003. Les complexes volcaniques quaternaires sur les paysages

de Cappadoce (Golludag et Acıgol e Turquie): evolutions morphodynami-

ques et implications environnementales. These de doctorat, Universite

Paris 12, 303 p.

Mouralis, D., Kuzucuoglu, C., Pastre, J.F., Turkecan, A., 2001. Le volcanisme

recent de Cappadoce: manifestations volcaniques du Pleistocene superieur

et de l’Holocene enregistrees dans le maar d’Eski Acıgol e Resultats pre-

liminaires. 8eme Congres Francais de Sedimentologie, Association des

Sedimentologistes Francais, Orleans (12-14 november 2001), Livre des re-

sumes, pp. 251e252.

Mouralis, D., Pastre, J.F., Kuzucuoglu, C., Turkecan, A., Atici, Y.,

Slimak, L., Guillou, H., Kunesch, S., 2002. Les complexes volcaniques

rhyolitiques quaternaires d’Anatolie centrale (Gollu Dag et Acıgol, Tur-

quie): genese, instabilite, contraintes environnementales. Quaternaire 13,

219e228.

Olanca, K., 1994. Geochimie des laves quaternaires de Cappadoce (Turquie).

Les appareils monogeniques, Ph.D. Dissertation, Blaise Pascal University,

Cermont-Ferand.

Otte, M., Yalcinkaya, I., Kozlowski, J., Bar-Yosef, O., Lopez Bayon, I.,

Tasxkıran, H., 1998. Long-term technical evolution and human remains in

the Anatolian Palaeolithic. J. Hum. Evol. 34, 413e431.

Poidevin, J.L., 1998. Les gisements d’obsidienne de Turquie et de Transcaucasie.

Geologie, geochimie et chronometrie. In: Cauvin, M.C., Gourgaud, A.,

Gratuze, B., Arnaud, N., Poupeau, G., Poidevin, J.-L., Chataigner, C. (Eds.),

L’obsidienne au Proche et Moyen Orient. Du Volcan a l’Outil. BAR

International Series, Oxford, pp. 106e203.

Roebroeks, W., 2001. Hominid behaviour and the earliest occupation of

Europe: an exploration. J. Hum. Evol. 41, 437e461.

Shea, J.J., 1999. Artifact abrasion, fluvial processes, and ‘‘living floors’’ from

the early Paleolithic site of ‘Ubeidiya (Jordan Valley, Israel). Geoarchaeol-

ogy 14, 191e207.

Slimak, L., 1998e1999. La variabilite des debitages disco€ıdes au Paleolithi-

que moyen: diversite des methodes et unite d’un concept. L’exemple

des gisements de la Baume Neron (Soyons, Ardeche) et du Champ

Grand (Saint-Maurice-sur-Loire, Loire). Prehist. Anthropol. Med 7-8,

75e88.

Slimak, L., 2003. Les debitages disco€ıdes mousteriens, evaluation d’un con-

cept de debitage. In: Peresanim, M. (Ed.), Disco€ıd Lithic Technology.

BAR International Series, Oxford, pp. 33e65.

Slimak, L., 2004a. Implantations humaines et exploitations des obsidiennes en

Anatolie Centrale durant le Pleistocene. Paleorient 30, 7e20.

Slimak, L., 2004b. Les dernieres expressions du mousterien entre Loire et

Rhone. These de doctorat, Universite de Provence, Aix-Marseille I.

Slimak, L., Roche, H., Mouralis, D., Buitenhuis, H., Balkan-Atlı, N., Binder, D.,

Kuzucuoglu, C., Grenet, M., 2004. Kaletepe Deresi 3 (Turquie), aspects

archeologiques, chronologiques et paleontologiques d’une sequence pleisto-

cene en Anatolie Centrale. Palevol 3, 411e420.

111L. Slimak et al. / Journal of Human Evolution 54 (2008) 99e111

Steiger, R.H., Jager, E., 1977. Subcommission on geochronology: convention

on the use of decay constants in geo- and cosmochronology. Earth Planet.

Sci. Lett. 36, 359e362.

Stekelis, M., 1966. Archaeological Excavations at ‘Ubeidiya, 1960e1963: the

Lower Pleistocene of the Central Jordan Valley. Israel Academy of Sci-

ences and Humanities, Jerusalem.

Stiner, M.C., Arsebuk, G., Howell, F.C., 1996. Cave bears and Palaeolithic ar-

tefacts in Yarımburgaz Cave, Turkey: dissecting a palimpsest. Geoarcheol-

ogy 11, 279e327.

Tixier, J., 1956. Le hachereau dans l’Acheuleen nord Africaine. Notes typolo-

giques. Bull. Soc. Prehist. Fr. 55, 628e644.

Toth, N., 1985. The Oldowan reassessed: a close look at early stone artifacts.

J. Archaeol. Sci. 12, 101e120.

Vermeersch, P., Paulissen, E., VanPeer, P., 1990. Paleolithic chert exploitation in the

limestone stretch of the Egyptian Nile Valley. Afr. Archaeol. Rev. 8, 77e102.

Wagner, G.A., Weiner, K.L., 1987. Deutches Archaologisches Institu

Demircihuyuk e die Ergebnisse der Ausgrabungen 1975e1978. In:

Herausgegeben von Manfred Korfan. Band II: Naturwissenschaftliche

Untersuchungen. Verlag Philipp von Zabern, Mainz am Rhein, pp. 26e29.

Yalcinkaya, I., Otte, M., Bar-Yosef, O., Kozlowski, J., Leotard, J.M.,

Tasxkıran, H., 1992. Karain 1991, Recherches paleolithique en Turquie du

sud: rapport provisoire. Paleorient 18, 109e122.

Zhu, R.X., Potts, R., Xie, F., Hoffman, K.A., Deng, C.L., Shi, C.D., Pan, Y.X.,

Wang, H.Q., Shi, R.P., Wang, Y.C., Shi, G.H., Wu, N.Q., 2004. New evidence

on the earliest human presence at high northern latitudes in northeast Asia.

Nature 431, 559e562.