THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF * (KERMANSHAH, IRAN) J. JAUBERT 1, F. BIGLARI 2, V. MOURRE 3, L. BRUXELLES 4, J.-G. BORDES 1, S. SHIDRANG 2, R. NADERI 7, M. MASHKOUR 5, B. MAUREILLE 1, J.-B. MALLYE 1, Y. QUINIF 6, W. RENDU 1 & V. LAROULANDIE 1 1

UMR 5199-PACEA, University Bordeaux 1, Avenue des Facultés, F-33405 Talence, France. [email protected], [email protected], [email protected], [email protected], [email protected], [email protected] 2 Center for Paleolithic Research, National Museum of Iran, 30 Tir st., Emam Khomaini Ave. P.O.Box 11365/4364, Tehran, Iran. [email protected], [email protected], [email protected] 3 UMR 5608-UTAH, Toulouse, France. [email protected] 4 INRAP, Centre de Recherche en Anthropologie, Toulouse, France. [email protected] 5 UMR 5197, Muséum national d’histoire naturelle/ CNRS “Archéozoologie, Histoire des Sociétés Humaines et des Peuplements Animaux”, 55, rue Buffon 75005 Paris, France. [email protected] 6 CERAK, University of Mons, 9 rue de Houdain, B-7000 Mons, Belgium. [email protected] 7 Student of Archaeology, University of Abhar, Abhar, Iran Résumé: La première campagne de terrain de la mission franco-iranienne portant sur le Paléolithique a concerné une cavité perchée du massif de Bisotoun, région de Kermanshah (sud du Zagros, Iran) découverte en 1986. Les sept premiers cosignataires ont participé à la phase terrain (2004) qui a consisté en une étude globale et interdisciplinaire de la petite grotte de Mar Tarik dont le remplissage avait été malheureusement affecté par des travaux clandestins qui ont détruit l’essentiel du gisement à l’exception de lambeaux de sédiments indurés et conservés dans la salle du fond. Outre des témoignages de fréquentation ou d’occupations d’âge historique, protohistorique (Chalcolithique) qui ne sont pas décrits ici, les vestiges paléolithiques révèlent une homogénéité évidente, concernant uniquement (ou presque) le Moustérien du Zagros. Une datation U/Th d’un spéléothème antérieur au niveau moustérien a été rapporté au Dernier interglaciaire (OIS 5e). Il est possible de distinguer la faune post-paléolithique des vestiges pléistocènes par leur état de surface mais sans contexte stratigraphique et avec un faible échantillonnage, l’histoire des différents stocks de faune, qui est complexe, n’est pas aisée à établir. De même pour les restes humains attribuables à Homo sapiens sp. Un essai de datation 14C SMA sur un fragment d’os humain insuffisamment riche en collagène n’a pas donné d’estimation chronologique. L’industrie lithique montre une chaîne opératoire fractionnée typique d’un site de consommation secondaire avec faible activité de taille et importation d’outils déjà préparés, les outils retouchés représentant 44 % du total de l’industrie. La production est dominée par la méthode Levallois, difficile à identifier en cas de ravivages multiples. Les outils retouchés sont typiques du Moustérien du Zagros: racloirs, racloirs déjetés, convergents, pointes, outils à dos amincis. Une pierre en calcaire fin est gravée de lignes ou quadrillages géométriques, mais, sans provenance stratigraphique claire, son appartenance au Moustérien, bien que possible, n’est pas assurée.

INTRODUCTION*

THE SITE OF MAR TARIK

The excavation in the Mar Tarik Cave takes place in a French-Iranian project begun in 2004 to update data about Palaeolithic period in Iran, especially about the end of Middle Palaeolithic and the beginnings of Upper Palaeolithic (50 – 30 kyr BP) (Jaubert et al., 2005). The study of the Iranian Middle Palaeolithic is resumed on the basis of a double objective: 1, to better define the lithic techno-complexes and the chronological position of the classical Zagros Mousterian; 2, to invest other geographical units than Zagros, starting with the Central Plateau (for example Niyasar) and Isfahan region. The preliminary results presented hereafter concern the first objective of our joint program.

In Central Western Zagros Mountains, the chain of Kermanshah, of which Bisotun massif constitutes the south-eastern end (Fig. 1.1), dominates an interior sedimentary plain by walls of 1.000 to 1.400 meters height. This mountain range, constituted by a very thick calcareous frame (Jurassic, Middle and Upper Cretaceous), corresponds to a great overlapping accident and forms a narrow scaled internal band Zagros crushed zone. In this sector, three great morphological sets can be distinguished: the high calcareous relieves, the detrital piedmont and the alluvial plain. Mar Tarik is located at the contact of the two first sets (Fig. 1.2). With Mar Dodar and the neighbouring Mar Aftab, Mar Tarik (dark cave in Kurdish) is one of the three highest cavities, with a rather long access (about 250 m above the Chamchamal Plain). Like Mar Aftab and Mar Dodar, it is

*

The research has been accomplished with financial support provided by the French Foreign Affairs Ministry (Division des Sciences sociales et de l’Archéologie, mission n°891 Iran) and Iranian Cultural Heritage & Tourism Organization (ICHTO).

7

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

Fig. 1.1. General map of Iran with Mar Tarik location

Fig. 1.2. Local map of Kermanshah area and Bisotun massif with location of Middle Palaeolithic and raw material sites

8

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

calcareous stratum (Fig. 1.3). From Hunter’s Cave, which opens close to the Darius relief and the spring, it takes at least half an hour to reach that point. There is a narrow, flat terrace between the cave and the talus slope. The cave entrance is a 3 m scramble up from this terrace, with hand and footholds in the rock at full stretch for a tall man. The climbing rout is worn smooth with use, indicating prolonged commute. For easy access to the cave we used speleological ladder. The entrance, which is approximately 1,50 m of width, faces southeast, so the first few meters of the cave benefit from an almost continuous sunlight during the day. About 4 m behind the entrance, the cave turns sharply toward the northeast, and as a result the main corridor and bottom chamber plunging in darkness. The long and narrow chamber is 20 m deep by approximately 2 m of mean width and at the highest point in the ceiling is about 1.80 m high (Fig. 1.4).

Fig. 1.3. Mar Tarik and Bisotoun plain. The entrance is closed the white protection (drift stand) (photo J. Jaubert)

located between the two caves of Bisotun massif already known in the past (Hunter’s Cave and Ghar-e Khar: Coon 1951, Young and Smith 1966). This and two other neighboring cave sites were discovered and published by one of us (Biglari 2000).

The whole of the cavity is fixed on this horizontal accident. Its overall morphology corresponds to a flattened pressure pipe, comparable in any point with a gallery out of joint of layer. Its width varies from two to four meters, for a height ranging between one and two meters. In order to facilitate its description, the cavity can be divided into several sections: the entrance area, the central part and the terminal chamber with detrital deposits.

Its entrance, located at 1.540 m of altitude, is hardly guessed while approaching the foot of the impressive calcareous cliff which forms the Bisotun massif, with the favour of a long fault or joint which clearly divides the

Fig. 1.4. Mar Tarik. Plan and longitudinal section (A-A’) of the cave (topography and drawing L. Bruxelles, V. Mourre)

9

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

Fig. 1.5. Mar Tarik, lithic industry discovered by. F. Biglari & coll. 1-5, 7-9, 16 Convergent, déjetés scrapers; 6, foliated point; 10, Scraper; 11, Truncated-facetted; 12-13, Levallois blanks; 14-15 Double scrapers (drawing F. Biglari: Biglari 2000)

of classical argillaceous silt more or less charged with stones (Biglari 2000). On more than around fifty pieces collected in the 1986 survey, the proportion of tools was high (37 out of 55), which is not very surprising for this kind of site: scrapers, double scrapers, convergent scrapers, Mousterian points, déjetés scrapers, generally on

THE DISCOVERY The first reconnaissance in 1986 allowed the collect of an interesting series of lithic implements attributed to Zagros Mousterian on the surface or the sediments disturbed by looters in the bottom chamber of the cave, which made up 10

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Fig. 1.6. Mar Tarik, lithic industry discovered by. F. Biglari & coll. 1, 3, Burin; 2, Levallois core; 4-6, Notches (drawing F. Biglari: Biglari 2000)

Levallois flake and a small double-side foliated piece (or foliated scraper), exceptional for Zagros Mousterian (Fig. 1.5-1.6).

pieces, only a few scattered vestiges, ascribable to Chalcolithic (ceramic), the historical periods, antique, medieval or modern, with or without microfauna, obviously of Holocene age. Concerning this frequentation of the cavity, we are dealing with simple visits or animal and/or human specific passages and here, there is no clearly identifiable occupation by a homogeneous level of accumulation.

At the time of our visit on spring 2003 to prepare this excavation, we collected there a handle of additional pieces including a beautiful thick scraper with “scalariforme” retouch evoking certain tools of European Mousterian of Quina type.

According to the surface discoveries, we established the principal zone of excavation in the final part of Mar Tarik. Initially, the blocks of the coarse fraction were pushed back or evacuated and the mixed sediments, which piled up in the centre were sifted in order to regularize the ground. In the second time, after setting up of a squaring, we started with the excavation of 2 m2 in the centre, quickly prolonged on both sides in order to carry out a trench, perpendicular with the axis of the corridor (‘trench’ Fig. 1.4). The excavation was carried out methodically with recovery of the vestiges per m2 until meeting the top of sterile clay or scales of calcareous bedrock going up in the centre of the trench. The substratum was rather quickly reached and cleaned. A

The 2004 campaign The first work consisted in cleaning the sectors of the porch and the entrance, where we excavated or disencumbered calcareous bedrock of accumulations of dusty and locally more or less indurate sediments. Cleaning concerned several points of the gours or anfractuosities of the limestone that had trapped and retained lenses of sediment. All of the removed sediment of course was screened with various meshes. The “excavated” zones in this former part of the cave delivered, in addition to some isolated Mousterian lithic 11

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

Fig. 1.7. Mar Tarik: stratigraphy of the end chamber (L. Bruxelles). Calcareous. Irregular and vacuole calcite. Massif stalagmite. Breccia with indurate blocks, artefact and bones. Cailloutis in indurate silt with artefact and bones. Silt loam indurate

quite astonishing observation was that, in our impression, even in plunging anfractuosities of limestone, the sediments were mixed, jumbled up by former work probably by looters. The excavated sediment looks brown dark, rather rich in stones and including a little fauna, some shards of ceramics (Middle and Late Chalcolithic, Iron Age) and Mousterian lithic industry, the latter homogeneous.

the left wall, at the end of the corridor and supposed to seal intact scraps of filling (Fig. 1.7).

GEOLOGY The karstologic study and the history of its sedimentary dynamics could be drawn (L. Bruxelles) and will be published in detail in a future monographic paper.

Lastly, the third and last stage was the establishment of a test trench at the base of the stratified levels located on

The sequence observed in the terminal chamber of the 12

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Tab. 1.1. Detailed results of U/Th date on a stalagmite fragment from Mar Tarik sequence (Y. Quinif) Sample

[U]ppm

[7447]

0,496 ± 0,004

234

U/238U

0,984 ± 0,008

230

Th/234U

230

Th/232Th

0,678 ± 0,009

15,8 ± 0,4

[234U/238U]t=0

Age (kyr)

0,978

123,6 [+3,4/-3,2]

cave (Fig. 1.7) shows the following succession, from base to top:

clearly distinct from that observed in the stalagmite of the lower level.

− at the base, a large stalagmitic concretion constitutes the oldest identified element in this section. It is a massive stalagmite, of around 50 cm height, shaped by a resumption of flow. Calcite, of a rather dark yellowish colour, is well crystallized and translucent. For the dating, the most compact central part of the stalagmite has been taken. The result is reliable: 123,6 [+3,4/-3,2] kyr BP has been obtained by Y. Quinif (CERAK, University of Mons, Belgium) on a stalagmite fragment (tabl. 1.1). The well-crystallized concretions materialize fossilization or a very slowed down functioning of the cavity. They are also related to a bioclimatic context favourable to their deposit, certainly a hot and wet period of the Pleistocene, isotopic substage 5e as U-Th date indicates.

As supposed, we found some lithic implements, in particular a typical Mousterian retouched point on Levallois blank (fig. 1.13 n°32) in the most indurate part of the sediment. Some small remains of fauna were also collected and they will be of course in priority submitted to dating test. Unfortunately, the collagen preserved was not sufficient to obtain AMS date.1 The excavation of this scrap of intact filling is important because this last, which thus has escaped with the devastator activity of the clandestine diggers and other former rehandlings, proves that Mar Tarik contained well a homogeneous level of Zagros Mousterian whose age is necessarily former to the last phase of concretion formation of the final room.

− US 5: The stalagmitic formation is fossilized by an accumulation of argillaceous silts, more or less hardened, light brown slightly orange. This formation is clearly stratified and can be cut up in small plateaus from 1 to 2 cm thick. They are bent in two points, phenomenon which is perhaps ascribable with an old racking. The recording of tectonic movements is not possible here because none of the morphologies of the initial digging of the cavity is fractured or shifted. The limono-argillaceous filling illustrates a reloading of the cavity. This argillaceous detrital sedimentation, rated, remembers the varves observed in the mountain’s karsts.

THE MATERIAL ASCRIBABLE TO THE MIDDLE PALAEOLITHIC From the excavation, collected in the scrap of breccia excavated in the end of the final chamber, collected by square in the trench whose filling was mixed, collected on the surface during the cleaning of the gallery or discovered in the sifted sediments from the excavation, the archaeological material ascribable to the Middle Palaeolithic is subdivided in three categories of vestiges: − patinated remains of fauna, of orange colour, which are distinguished rather easily from recent Holocene fauna (Chalcolithic or Protohistoric and Historic times?).

− US 4: An accumulation of rubble from 40 to 50 cm thick contains archaeological material (Mousterian) and fragments of patinated bones, of orange colour. It fills a secondary hole in the clays. The higher part of this level is very hardened by calcite and constitutes a true breccia. The level of gravel (US 4), in gullying position on clays (US 5), forecasts of a radically different type of sedimentation. These limestone fragments associated with archaeological material and faunal remains show the opening of the cavity and its proximity with surface. The gravel, which appears for the first time, can be due to the bursting by the freezing of the benches most sensitive to the cryoclasty, logically the beginning of Last Glacial.

− The lithic industry, which except for one or two elements for which a hesitation can still exist, is ascribable in its entirety to the Middle Palaeolithic. Those among us who know well the productions of the Middle Palaeolithic as well as by Upper Palaeolithic specialists share this analysis and we hardly have more doubt as for the homogeneous character of the series. − Lastly, and here we will regret for a long time the mixed up character of the discovery context, a slab of fine-grained limestone engraved on its two faces, which we would associate to the Middle Palaeolithic. The presence of Chalcolithic2 vestiges mixed within the same stratigraphic unit prevents us from being categorical.

− US 3: Between the top of the breccia and the vault, a series of small squat columns, developed on the edges of the gallery, on the level of the principal joint. The circumvented and vacuolar concretion formation is

1 2

13

Beta analytic, Miami. We wait about a complementary study from K. Abdi.

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

FAUNA

Taphonomy, Zooarchaeology

Preliminary analysis

We analysed faunal remains to establish the faunal composition of the assemblage and to characterise its different origins. All remains have been recorded in a taxonomical and taphonomical way. Moreover bone surface modifications have been studied to determine the agents of the accumulation. The observation of bone surfaces has been done under a binocular microscope (x40).

The identification of the fauna was first realized in Iran in the National Museum (M.M.). The preliminary study of this assemblage suggests that the Mar Tarik Faunal assemblage has particularly suffered of taphonomic processes. They are in majority badly preserved. Bones are eroded and teeth are split. Two different wears one brownish and the other one much clearer indicate different taphonomic episodes and specially a chronological mixture. The identifiable bones belong in majority to Caprids (Sheep and Goat). This taxonomic group is present in all the stratigraphic units, which yielded animal bones. Apart from Caprids, another small herbivore is also present which can be allocated to Gazella sp. Tooth remains of a small Canid (Fox / Chacal (Vulpes sp. / Canis aureus) are present in the fauna. Also teeth remains belonging to Hare (Lepus) could be identified in the remains. A single fragment of tooth (root) belongs to larger mammals, the single indication for the presence of a large mammalian species in the cave.

Taxonomical determination has been limited because of the lack of comparison material in the university of Bordeaux 1. For this reason, we used taxonomical size categories. For example, the avifauna has been separated in three categories: − Small sized birds (Sparrow) − Medium sized birds (Blackbird) − Large sized birds (Duke) Ungulates remains have been quoted by size groups (Brain 1981).

The excavated sediments were sieved and thus a large number of micro-vertebrate remains was also recovered during excavation. These remains belong to fish, amphibian / reptiles, birds and mammals. Among the Amphibian / Reptile group the presence of snake and land turtles was observed. Within the fish remains, pharyngeal bones of sweet water fish were identified. The major part of the micro-vertebrate remains found in the Mar Tarik fauna belongs to birds and micro-mammals (bats, rodents and insectivores). These remains are generally well preserved and identifiable. Like the fish and amphibian/ Reptile remains these remains were transferred to France for a deeper analysis.3

The assemblage is composed of 492 faunal remains (tabl. 1.2). The microfauna (Reptiles, Amphibians and Micromammals) constitutes most of the assemblage (54%). Avifaunal remains (Anseriformes, Galliformes, Pigeon, Passeriformes) contribute to 22% of the faunal collection. Ungulates are found in 9% of the material. Leporids and carnivores represent only 4% of the assemblage. Only three remains of turtle and fish have been recorded. Most of the material is well preserved; only 32% of the bones reach the third Behrensmeyer’s stage (Behrensmeyer 1978). Twenty percent of the remains exhibit a chemical attack. Its origin cannot be clearly established (biological versus weathering). Digestion marks are present on few bird bones belonging to Pigeon and little Passeriformes. Tooth marks are present on some of the Anseriformes bones.

The first conclusions based on these preliminary observations are that the major food resources were herbivores of medium size (Caprids and Gazelles). Note worthy the presence of the young individuals within this group. Is this resulting of hunting activities of other carnivores or mirroring hunting strategies of man? These remains are not represented in a significant number to allow the construction of mortality profiles. The presence of young herbivores how ever poses the question of a possible seasonality of the occupation, if the anthropic origin of the remains is admitted.

Cut marks and burnt bones are poorly represented. Only 6% of the assemblage is affected by such marks. These last are observed on all taxonomical groups. The occurrence of cut marks on bird bones and carnivores remains has to be underlined. The well preservation of bone surfaces cannot explain the poverty of human evidences.

The presence of the micro vertebrates and their diversity could indicate a non-anthropic origin of those accumulations and could be related to the accumulation of raptor pellets non simultaneous to the occupation of the site. This could be examined through the taphonomic profile of the remains once the specific analysis of each group has been done.

To conclude, several agents contributed to the faunal assemblage at Mar Tarik. Unfortunately, we were not able to quantify the contribution of each of them. Digested bones and teeth marks attest of the participation of carnivores. Some of the bones (Pigeon, Passeriformes) were introduced in the cave by Owl’s pellets. Cut marks and burnt bones are the evidence of human accumulation. Natural accumulation cannot be excluded for a part of the remains. The lack of stratigraphical data makes unable to

3

Excepted the rodent remains, which will be studied at the Rodent research center of the Mahhad University in Iran.

14

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Tab. 1.2. Faunal spectrum and number of remains of Mar Tarik Taxa

NR

Total Carnivores

22

Felids

1

Canids

5

Middle sized canids

6

Badger

3

Stoat

2

Small Carnivores

4

Medium sized carnivores

1

Leporids

19

Total Ungulates

44

Equids

2

Ovicaprins

17

Ungulate size 1

10

Ungulate size 1/2

4

Ungulate size 2

8

Ungulate size 2/3

1

Ungulate size 3

2

Mammals

13

Total avifauna

106

Anseriformes

10

Pigeon

11

Galliformes

2

Small sized birds

27

Medium sized birds

44

Large sized birds

11

Birds

1

Mesofauna

19

Turtle

3

Fish

3

Microfauna

263

TOTAL

492

the World. It is not useless to remind that with more than 1,6 millions of km2, a large variety of environments, Iran has immense potentialities of discoveries of new intact or with little reshaped by intense activities agricultural, sites. For sure, the main advances in the knowledge of the Pleistocene human settlement, and particularly the distribution of anatomically modern Humans, their biologic and cultural relationships with archaic populations of the Near and Middle East, will result, in the future, from researches realized in this country. Let us remind that except still confidential discoveries, regrettably not always confirmed, and rare vestiges known by the international scientific community, such as those of Belt (Mazanderan), Hotu (Behsharh) reported to the beginning of Holocene, Tamtama (Rezaiyeh) whose chronological age is always unknown4 and Bisitun (Kermanshah) which are supposed to be Upper Pleistocene, our knowledge on the Prehistoric human fossils of Iran is very limited (Coon 1975). Of course we have also to underline the existence of the famous site of Shanidar (Iraq) near of the western border of the North of Iran, which has delivered an exceptional set of Mousterian human remains, considered as Middle East Neanderthals (Solecki 1971, Trinkaus 1983). Our works realized recently in the cavity of Mar-Tarik favoured the discovery of two collections of human vestiges. The most numerous set is reported to the stratigraphic unit 0 (US 0). Two other pieces (among which one could be fauna) come from the US 1. The inventory of these remains is the following one: 9 teeth and fragments of teeth, 7 cranial fragments, 7 elements of the infra-cranial skeleton (phalanx, metatarsal, hipbone…) and 2 pieces which are not necessarily human (fragments of a long bone diaphysis and an infra-cranial bone). With such a set of human remains (US 0 + US 1), the conclusions are limited. Nevertheless, we note the presence of: − 1, a series of essentially fragmentary permanent teeth and the absence of temporary tooth. The NMI (by exclusion) can be estimated to two, − 2, a series of fragments of mature cranial vault bones of one or different individuals. Again, we note the absence of clearly immature vestige, − 3, some phalanx of feet and hands and, again, the absence of immature piece, − 4, and finally, one fragment of a right mature hipbone.

assign the exploitation of bird and carnivores to Middle Palaeolithic humans. We hope that further excavations will provide more data on Middle Palaeolithic subsistence in Iran.

Again, according to our study and because of the inherent difficulties of very fragmentary, badly preserved and vague chronological position of such a collection of human vestiges, the results are necessarily limited. But we can underline that no piece shows any important pathology or evidence of voluntary human damage

Human remains The study of the prehistoric populating of Iran is one of the problems among the most fascinating of this region of

4

A re-study of the femur from Tamtama by N. Minugh-Purvis and E. Trinkaus showed that it belongs to deer!

15

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

(except a fragment of long bone (but for which we are not sure that it is a human one).

The sources to which the craftsmen of Mar Tarik supplied themselves remain difficult to determine. The importance of deteriorations of tectonic origin affecting materials in primary position would encourage us to seek a source in secondary position, in order to profit from a first selection related to transport in fluvial conditions. However, the natural surfaces observed on the vestiges leads us to moderate this assumption. Indeed, the presence of fluvial ‘neocortex’ (= pebble cortex) is relatively rare. On the other hand, more are the pieces presenting fresh cortex or natural surfaces, rather indicating procurement in primary position.

Considering the taxonomical status of these human remains, none of these vestiges present some characteristics, which would exclude them from the variability of extant Homo sapiens sapiens and we must also note that they don’t present any Neanderthal apomorphy. But, if we accept the hypothesis that some of these pieces are reported to the Upper Pleistocene, (even if they don’t present any peculiar robusticity which is very frequent in most of the Upper Pleistocene populations) they have a great scientific interest, as we don’t know so much about any Upper Pleistocene population of Iran. Remember that the temporal bone of Darra-i-Kur (Afghanistan), if really Upper Pleistocene and Mousterian, could represent a specimen belonging to such a population (Angel 1972). The conclusion of the preliminary study of this specimen is that it is an anatomically modern temporal except some isolated traits which were difficult to interpret in terms of Neanderthal versus AMH variability. Then, it could strengthen the interest of the anthropological study of the Mar Tarik’s human pieces to try to get direct absolute dates of some of them.

Even if it appears difficult to determine their exact origin, we tried to describe and quantify the various groups of employed raw materials. It arises very clearly that the redbrown group of cherts is most common by far and more diversified. The group of the clear cherts, orange yellow, then grey cherts follows it. About fifteen pieces not belonging to these groups correspond to more rare materials, including particular cherts and some pieces in chalcedony-like rocks with very fine grain. The origin of those latter was not localized yet. All the materials seem to be flaked using the same methods and no true economy of the raw materials is perceptible on the scale of the series. It is worthy however to announce here that certain rare materials are represented only by flakes or fragments of flakes, for the majority retouched and Levallois.

LITHIC INDUSTRY Mar Tarik delivered 360 lithic remains, among which 286 were found during 2004 field season and 74 during previous surface collection. Most of the remains are coming from the main chamber located at the bottom of the cave (312); partial cleaning of the entrance and of the corridor delivered respectively 32 and 16 remains only.

Surface conditions Surface conditions are extremely diversified, a single piece showing sometimes two or three of them. As a whole, the pieces are very frequently smoothed and with edge damages. Differences, rather foreseeable, are perceptible between the three sectors: the rounded pieces are more frequent in the entrance and especially in the corridor; the patinated pieces are more frequent in the bottom room, where they were prone to important chemical attacks in particular in the form of gypseous concretions.

Raw materials A geological and geomorphological investigation made it possible to draw solids regional bases relating to these fields and to supplement previous work by F. Biglari and S. Heydari (Heydari 2004; Biglari 2004). The raw materials employed in Mar Tarik are mainly radiolarian cherts, presenting very important variations of facies in terms of texture, of silicification degree and of colour. These variations can also take the form of resilicified diaclase plans, from which the colour is then different from that of their matrix. The side variations can be observed at the level of an outcrop, or even on a block.

General characteristics From a technological point of view, no significant difference is perceptible between the series coming from the various parts of the site. It will be however question here only of the material coming from the main room, which one can regard a priori as more homogeneous. Indeed, the elements likely to be intrusive are extremely rare there: only a distal fragment of unretouched bladelet with trapezoidal section and regular axis (from mixed sediments, US 0), a blade and a very regular blade fragment (both from surface collection) could not correspond to the Mousterian occupation of the site. The other blades and blade fragments fit well in the series: they are generally uni- or bipolar Levallois products. Only a few supports with triangular section could concern a

This kind of raw material outcrops in the so-called Radiolarit Belt of Kermanshah, about 10 to 15 kilometres south of the site, on an area including Gakia and Harsin localities (Biglari 2004). Radiolarian chert in primary position is also present in the Bisotun massif itself, particularly in the form of secondary silicifications of breccias, but the Gakia-Harsin sources seem to offer more homogeneous, better-silicified and especially more accessible materials. 16

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Fig. 1.8. Mar Tarik, lithic industry: 1, Denticulate on Levallois flake; 2, Levallois flake (bipolar); 3, medial fragment of retouched Levallois blade; 4, proximal fragment of Levallois blade (bipolar) (Drawing J. Jaubert – S. Shidrang)

distinct volumetric structure, without however being different from the operative schemes known for the Middle Palaeolithic of the Middle East (see below).

These first observations obviously result in regarding the series as the expression of a particular economic and functional facies, characterized by the extreme scarcity of the operations of flaking on the spot, or at least inside the cave, and the importation of already flaked supports, even retouched. Indeed, in spite of a systematic sifting, only two flakes of retouch could be identified. In addition, four characteristic flakes of secondary sharpening were recognized within the series (Fig. 1.11, n° 20-21), including one turned into a scraper.

The series is largely dominated by the flaking products in a broad sense, complete or broken (> 87%). The cores are very rare (N = 3 + 1 fragment), just as the various negative bases or the flaking by-products (Fig. 1.14 n°34). The specificity of the series is confirmed by the abundance of the retouched pieces, in particular with regard to the flakes and flakes fragments: respectively 54,7% and 43,7% of the latter are retouched, against 44,4% for the whole of the series (Table 1.2). This proportion is very important for an industry of this age but is quite common in the regional context (e.g. Hole and Flannery 1967, Dibble and Holdaway 1993, Baumler and Speth 1993).

Production schemes This type of economic and functional facies rather badly lends itself to the precise reconstitution of the utilized production schemes, insofar as many diagnostic parts are missing (cores, core shaping flakes, flaking accidents,

17

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

Fig. 1.9. Mar Tarik, lithic industry: 5, Mousterian point showing a burin blow; 6-8 and 11, Mousterian points or convergent scrapers (Levallois); 9, atypical limace; 10, straight-convex double scraper (Levallois) (Drawing J. Jaubert – S. Shidrang)

etc). It is however possible to propose a general reconstitution of these schemes analyzing the flaking products themselves and the rare unearthed cores.

particular organization of removals: they are primarily Levallois blanks, resulting generally from unipolar (Fig. 1.9 n°5, 6, 8; Fig. 1.10 n°14, 16-17; Fig. 1.14 n°35), bipolar (Fig. 1.8 n° 2-4) or centripetal methods (Fig. 1.10 n°15; Fig. 1.11 n°18). The supports described as “débordants” or “débordants unipolar” are compatible

Certain products can be directly attached to a specific production scheme, even to a volumetric structure or a

18

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Tab. 1.3. Lithic industry of Mar Tarik: proportions of retouched pieces according to the various technological categories Unretouched

Retouched

Use wear

Total

N

%

N

%

N

%

Total

Flake

50

39,1

70

54,7

8

6,3

128

Flake fragment

85

53,8

69

43,7

4

2,5

158

Blade

1

20

2

40

2

40

5

Blade fragment

11

52,4

9

42,9

1

4,8

21

Bladelet fragment

1

100

–

0

1

Core

1

50

1

50

2

Core fragment

1

100

–

0

1

Core on flake

1

100

–

0

1

Undeterm. negative base

1

33,3

2

66,7

3

Pebble fragment

–

0

1

100

1

Debris

29

96,7

1

3,3

30

Chunk

4

44,4

5

55,6

9

Total

185

51,4

160

44,4

15

4,2

360

Fig. 1.10. Mar Tarik, lithic industry: 12, proximal end-scraper (Levallois); 13, convex scraper (Levallois); 14, partial scraper (Levallois); 15, truncated flake (Levallois); 16-17, convex simple scrapers (Levallois)(Drawing J. Jaubert – S. Shidrang)

19

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

Fig. 1.11. Mar Tarik, lithic industry: 18, Déjeté scraper (Levallois); 19, transverse or déjeté scraper (demi-Quina); 20, flake from secondary sharpening of scraper; 21, Scraper made on scraper secondary sharpening flake (Drawing J. Jaubert – S. Shidrang)

besides with the unipolar and convergent methods of Levallois flaking (Beyries and Boëda 1983, Boëda 1995).

confirmed only by the study of series comprising the complete production scheme.

Of three cores of the series, two can be described as bipolar Levallois (Fig. 1.14 n°34); the third is a core on flake showing centripetal removals on its two faces, and it is quite impossible to relate it to a known volumetric structure (Fig. 1.14 n° 33).If the Levallois production scheme is dominating within the series, it is also advisable to point out the presence, among the differentiated products, of some supports maybe related to other volumetric structures: they are supports described as “unipolar”, “thick bipolar”, “convergent laminar” and as “under crested blade”. These supports are generally elongated and show robust sections, often triangular (Fig. 1.12 n°23, 28 and Fig. 1.14 n°35). They could concern a flaking concept characterized by a more “volumetric” exploitation of the core. As for the rest of the series, the hard hammer is the only technique employed. If this kind of flaking has already been described in the Levant Middle Palaeolithic (e.g. Meignen 1994), its presence in the Middle Palaeolithic of Zagros could be fully

Retouched tools As evoked previously, the retouched pieces are particularly abundant in Mar Tarik. These pieces do not correspond however always to traditional tool types of the Middle Palaeolithic, far from it: that is related to the fact that the retouch is very often irregular and/or marginal, at such point for example that the real scrapers are rather rare. One will note the relative frequency of the pieces with convergent bilateral retouch (15%): Mousterian points, limace, convergent scrapers (Fig. 1.9 and 1.12). The tools of Upper Palaeolithic type are rare but typical: they include four burin (Fig. 1.9 n°5) and five endscrapers (Fig. 1.10 n°12). Only one “truncated-faceted piece” was found, although this kind of thinning (similar to Kostienki or Nahr-Ibrahim) is quite common in other regional series (Dibble 1993). As previously mentioned, an uncommon foliated bifacial piece was discovered during first surface collection. A general typological

20

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Fig. 1.12. Mar Tarik, lithic industry: 22-28, tools with convergent retouched edges: 22-24 proximal fragments of Mousterian points; 25, déjeté scraper; 26, convergent scraper (Levallois); 27, distal fragment of thick point; 28 distal fragment of point (Levallois) (Drawing J. Jaubert – S. Shidrang)

breakdown inspired of the Bordes list-type (Bordes 1961) can be proposed (Table 1.4).

The initial phases of these schemes are very poorly represented there, not to say absent. The series comprises practically only already flaked supports, of which approximately one out of two is retouched. The cores are extremely rare. Some indices of retouch and secondary sharpening of tools were detected. The production of the supports obviously did not take place on the spot and only

Synthesis In a synthetic way, the industry of Mar Tarik corresponds to the final phases of one or several operative schemes.

21

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

Fig. 1.13. Mar Tarik, lithic industry: 29-30, Double scrappers (29: dos aminci); 31, scraper on cortical flake; 32 Mousterian point discovered in situ (Drawing J. Jaubert – S. Shidrang)

the finished products, rough or even – at least for some of them – already retouched, were imported. This industry thus represents a particular economic and functional facies, related most probably on the localization and the operating mode of the site itself. Unfortunately, this one will hardly be precisely reconstructed, due to the indigence of the faunal material.

very regular, cordiform, asymmetrical in section (triangular) and weighing 346 g. It presents a kind of cortical back opposed to a sharper edge. The lower edge doubly concave is reworked by two removals. The upper edge is convex and interrupted by the hollow of an old negative scar. The brittleness and the thinness of the edges caused some shocks, which notch them more or less.

THE ENGRAVED SLAB

In waiting of a macroscopic examination and a drawing after photograph, we can describe its broad outline:

On the border between squares E5 and D5, alas in the mixed levels (US 0), was found a slab of fine-grained limestone (grainstone), engraved on its two faces (Fig. 1.15). The stone is rather large (15,5 x 11,7 cm), flat (greater thickness: 1,5 cm), smooth, of dark grey colour,

Face A One distinguishes clearly, with a minimum of oblique light, a clumsy, broad and vaguely orthogonal squaring (measured angle: 105°). In the largest length, there is a

22

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Fig. 1.14. Mar Tarik. lithic industry: 33, core on flake or retouched tool; 34, Bipolar Levallois core; 35, Retouched Levallois blade (broken point?); 36, fragment of retouched convergent tool. (Drawing J. Jaubert – S. Shidrang)

series of three parallel strips, themselves made up of recoveries gearing down the layout with the manner of a repeated scraping rather than the passage of a notched tool. Certain lines stop or are recut, but the three strips recut the whole length, that is to say nearly on 15 cm long

for the largest. The one on the left starts very close to the edge (less than one centimetre), the two following ones being spaced on average of 2 cm. A recent scale recuts the left strip. The number of features for each strip varies from 2 to 8 or 9. In the width, there are only two strips; 23

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

Tab. 1.4. Mar Tarik, breakdown of the retouched pieces MAR TARIK

Nb

%

Mousterian point

3

Elongated Mousterian point

3

Limace (atypical)

1

Simple straight scraper

1

Simple convex scraper

15

Simple convex scraper 1/2 Quina

1

Simple concave scraper

1

Simple concave scraper 1/2 Quina

1

Simple concave convex scraper

1

Double straight convex scraper

1

Double biconvex scraper

3

Double irregular scraper

1

Convergent scraper

1

Concave convex convergent scraper

3

Convex convergent scraper

4

Straight convex convergent scraper

5

Irregular convex convergent scraper

1

Déjetéscraper

4

Transversal convex scraper

2

1,2

Reverse scraper

1

0,6

Bifacial scraper

1

0,6

End-scraper

5

3,2

Burin

4

2,5

Notch

1

0,6

Truncated-faceted piece

1

0,6

Sundry (other slightly retouched pieces)

92

58

Foliated bifacial point

1

0,6

158

100

TOTAL

and 4,7 cm; two engraved lines in the width, spaced approximately by 5 cm. A set of very fine cracks, probably natural, comes to complicate its reading. Only the first horizontal line is strongly stressed and is clearly duplicated.

3,9 0,6

Contrary to the bone supports for which various marks always can be left by involuntary actions (meat removal) or by carnivores, the regularity of engravings and the rate of the features on this type of limestone slab exclude the result from a natural action or domestic activities (support of work, block...).

13

The problem of the dating It is regrettable to have discovered this limestone engraved slab in the mixed sediments of Mar Tarik because, contrary to lithic industry, one cannot suppose de facto his belonging to the Middle Palaeolithic occupation. However, several arguments could plead for this:

3,2

− the engraved features were carried out with a stone tool; − the bottom of engravings is rather patinated and engravings are not fresh;

11

− the pattern fits rather well with the few available comparative elements for the end of the Middle Palaeolithic: parallel lines of engravings (Temnata, Bulgaria), groups of incisions (Lartet shelter, France), crossed lines (Fréchet, France; La Quina, France), striated lines (Chez-Pourré, Champlost, France), encased arcs of circle (Quneitra, Golan plateau), cross (Tata, Hungary), chevrons, zigzags (Bacho-Kiro, Bulgaria), etc; − of various anthropic passages in Mar Tarik, only two periods can be suspected for this engraved stone, namely Mousterian and Chalcolithic, and no comparison element is known for local Chalcolithic. If the presumption of a Mousterian age is confirmed, this discovery would be important. Mar Tarik would then become the first site of Zagros having delivered this kind of graphic expression, thus joining the short list of the engraved lithic supports so far listed for the Middle Palaeolithic of Eurasia (Lorblanchet 1999, Jaubert 1999).

the main one at the higher third of the slab could follow a natural line (crack) and presents about the same number of juxtaposed features as the vertical bands. The lower strip undulates a little more, made up of only two engraved lines meeting or re-cutting themselves like waves.

MAR TARIK AND ZAGROS MOUSTERIAN Several lessons can be drawn from this preliminary study of the Mar Tarik lithic industry:

Face B • Initially, except for a few pieces, nothing allows us to separate the lithic series in several sets: on the contrary, it is of an obvious homogeneity. Even according to the split character of the chaîne opératoire, all indicates that it results from the stay of a group whose knappers controlled the same schemes for flake production and tools preparation.

On the opposed face, engravings are more limited: in fact, the squaring pattern is comparable, but the features are single and are different from the banding scraped of face A. They are simple engraved lines, quite as irregular and clumsy according to the same rate: three somewhat parallel lines laid out in the largest length, separated by 2

24

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Fig. 1.15. Mar Tarik, engraved limestone slab (photo J. Jaubert) • Then, the production schemes, dominated by Levallois flaking, even if they were underestimated in previously published works for lack of diagnosis criteria based on resolutely technological approaches, are those usually described for Zagros Mousterian: Shanidar (Solecki and Solecki 1993), Warwasi (Dibble and Holdaway 1993), Bisotun (Dibble 1984), Kunji (Baumler and Speth 1993) to quote only the principal published studies including inventories. The examination of the plates is for this purpose unambiguous.

1.16), like “Tabun B”, from now on called Late Levantine Mousterian (LLM, 70-45 ky) (Bar-Yosef, Meignen 2004). The latter are associated with Neanderthal remains in Kebara (≈ 60 ky), Amud or Dederiyeh. In Quneitra (Golan), the LLM is dated by ESR to 53,9 ± 5,9 (LU) or 39,2 ± 4,2 (EU). The main difference between Zagros Mousterian and LLM holds only in the smallest proportion of retouched pieces, since unretouched Levallois points generally dominate these techno-complexes.

• The only real difference would hold in the presence of a proportion of tools higher than the figures announced for example for Shanidar, although it is necessary to examine the criteria of calculation more closely. This is not surprising anyway for an altitude site.

• The relationship is on the other hand even stronger with Taurus Mousterian, for example the higher levels of the Middle Palaeolithic sequence of Karain on the SouthAnatolian coast (Turkey) where, as in Zagros, the proportion of retouched tools is equivalent (e.g. Yalçinkaya 1998, Yalçinkaya et al. 1993). It is quite the same with the attested facies (and for those which are dated) of the Caucasus: Armenia, Georgia, northern slope of the Caucasus, Russia (Golovanova, Gorbanichev 2003), etc. (Fig. 1.16).

• An additional and new element for Zagros Mousterian would be the presence of a hard hammer laminar production scheme which it will be appropriate, of course, to check on a more consequent series. • Without prejudging result of possible dating, everything leads us to place Mar Tarik among the typical series of Zagros Mousterian, namely a Recent Middle Palaeolithic, with Levallois flaking, important retouched tools dominated by the convergent retouched tools (15%): points, small elongated points, atypical limace, convergent or déjeté scrapers.

References ANGEL, J.L., 1972. A Middle Palaeolithic temporal bone from Darra-i-Kur, Afghanistan. In: Dupree L. (Eds), Prehistoric research in Afghanistan (1959-1966). Transactions of the American Philosophical Society, vol. 62, pt. 4: 54-56.

• These industries probably correspond to the most recent phase of Near and Middle East techno-complexes (Fig.

BAUMLER, M.F. and SPETH, J.D. 1993. A Middle Paleolithic assemblage from Kunji Cave, Iran, in D.I.

25

Fig. 1.16. Simplified synoptic table of the lithic techno-complexes dated from the recent Middle Palaeolithic in Near East, Middle East, Anatolia, Caucasus and Zagros

IRAN PALAEOLITHIC / LE PALEOLITHIQUE D’IRAN

26

J. JAUBERT ET AL.: THE MIDDLE PALAEOLITHIC OCCUPATION OF MAR-TARIK, A NEW ZAGROS MOUSTERIAN SITE IN BISOTUN MASSIF

Olszewski et H.L. Dibble dir., The Paleolithic Prehistory of the Zagros-Taurus, The University Museum, University of Pennsylvania, University Monograph 83, University Museum Symposium Series volume V, p. 1-73.

HEYDARI, S. 2004. Stone raw material sources in Iran: some case studies. Persiens Antike Pracht, Bergbau – Handwerk – Archäologie, Deutsches BergbauMuseum Bochum, Katalogbeiträge in englischer Sprache / Catalogue , p. 124-129.

BEHRENSMEYER, A.K. 1978. Taphonomic and ecologic information from bone weathering. Paleobiology, 4, p. 150-162.

JAUBERT, J. 1999. Chasseurs et artisans du Moustérien. Paris, La Maison des roches, 157 p. JAUBERT, J., BIGLARI, F., BORDES, J.-G., BRUXELLES, L., MOURRE, V., SHIDRANG, S., NADERI, R. and ALIPOUR, S., 2005. New Research on Paleolithic of Iran: Preliminary Report of 2004 Iranian-French Joint Mission. Archaeological Reports, ICHTO, Tehran (4), p. 18-26.

BEYRIES, S. et BOËDA, É. 1983. Étude technologique et traces d’utilisation des “éclats débordants” de Corbehem (Pas-de-Calais), Bulletin de la Société Préhistorique Française, t. 80, 9, p. 275-279. BIGLARI, F. 2000. Recent finds of Paleolithic period from Bisotun, Central Western Zagros Mountains, Iranian Journal of Archaeology and History, Serial n° 28, Vol. 14, n° 2, p. 4-5 (English abstract) & pp. 50-60 (in Farsi).

LORBLANCHET, M. 1999 La naissance de l’Art. Genèse de l’art préhistorique. Paris, Éd. Errance, 304 p. MEIGNEN, L. 1994. Le Paléolithique moyen au ProcheOrient: le phénomène laminaire. In: S. Révillion et A. Tuffreau (Dir.), Les industries laminaires du Paléolithique moyen, CNRS, Dossiers de Documentation Archéologique n° 18, p. 125-159.

BIGLARI, F. 2004. The preliminary observations on Middle Palaeolithic raw material procurement and usage in the Kermanshah plain, the case of DoAshkaft cave. In: T. Stollner, R. Slotta et A. Vatandoust (Dir.), Persiens Antike Pracht, [Persian Antiques Splendor, mining crafts and archaeology in ancient Iran, Volume: 1, Catalogue], Bergbau – Handwerk – Archäologie, Deustches BergbauMuseum Bochum, p. 130-138.

MEIGNEN, L. and BAR-YOSEF, O. 2004. Réflexions sur la fin du Paléolithique moyen et les débuts du Paléolithique supérieur au Proche-Orient. In: P. Van Peer, D. Bonjean et P. Semal (Dir.), Session 5 : Paléolithique moyen, BAR S1239 – Actes du XIVe Congrès de l’UISPP, Liège, 2-8 sept. 2001, p. 235246.

BOËDA, É. 1995. Levallois: a volumetric construction, methods, a technique. In: H.L. Dibble & O. Bar-Yosef (Dir.), The definition and interpretation of Levallois technology, Philadelphia, Monographs in World Archaeology n° 23, Prehistory Press, p. 41-68.

SOLECKI, R.S. 1971. The first flower people. New York: A.A. Knop, 290 p. SOLECKI, R.S. and SOLECKI, R.L. 1993. The pointed tools from the Mousterian occupations of Shanidar, Northern Iraq. In: D.I. Olszewski et H.L. Dibble (Dir.), The Paleolithic Prehistory of the ZagrosTaurus, The University Museum, University of Pennsylvania, University Monograph 83, University Museum Symposium Series volume V, pp. 119-146.

BORDES, F. 1961. Typologie du Paléolithique ancien et moyen, Bordeaux, Delmas Éd., Publications de l’Institut de Préhistoire de l’Université de Bordeaux, Mémoire n° 1, 111 p. BRAIN, C.K. 1981. The Hunter or the Hunted? An Introduction to African cave Taphonomy. University of Chicago Press, Chicago, 365 p.

TRINKAUS, E. 1983. The Shanidar Neandertals. New York: Academic press, 502 p.

COON, C.S. 1951. Cave Explorations in Iran 1949, Philadelphia, The University Museum, University of Pennsylvania, 125 p.

YALÇINKAYA, I. 1998. La grotte Karain: généralités dans le contexte anatolien. In: M. Otte (Dir.), Préhistoire d’Anatolie. Genèse de deux mondes (The Paleolithic of Eastern and Southeastern Europe), Liége, ERAUL 85.

COON, C.S. 1975. Iran. In: Oakley K.P., Campbell B. G., Molleson T.I. (Eds), Catalogue of fossil hominids, Part III: Americas, Asia, Australia. London: Trustees of the Bristish Museum (Natural History), p. 118-120.

YALÇINKAYA, I., OTTE, M., BAR-YOSEF, O., KOZŁOWSKI, J.K., LÉOTARD, J.-M. and TASKIRAN, H. 1993 The excavations at Karain Cave, South-Western Turkey: an interim report. In: D.I. Olszewski et H.L. Dibble dir., The Paleolithic Prehistory of the Zagros-Taurus, The University Museum, University of Pennsylvania, University Monograph 83, University Museum Symposium Series volume V, pp. 101-117.

DIBBLE, H.L. 1984. The Mousterian Industry from Bisitun Cave (Iran). Paléorient 10 (2): 23-34. DIBBLE, H.L. 1993. Le Paléolithique moyen récent du Zagros, Bulletin de la Société Préhistorique Française, t. 90, n° 4, pp. 307-312. DIBBLE, H.L. and HOLDAWAY, S.J. 1993 The Middle Paleolithic industries of Warwasi. In: D.I. Olszewski et H.L. Dibble (Dir.), 1993. The Paleolithic Prehistory of the Zagros-Taurus, The University Museum, University of Pennsylvania, University Monograph 83, University Museum Symposium Series volume V.

YOUNG, T.C. Jr. and SMITH P.E.L., 1966. Research in the Prehistory of Central Western Iran. Science, 153: 386-91. 27