Le Paléolithique du Nahr el Kébir

1. Outline Chapter I – This deals with the history of previous prehistoric research in the Kebir area, and the methods of processing and analysis of the finds from our survey. Chapter II – Description of the assemblages, our database. Each site is listed in order of geochronological age of the formation on which it was found, oldest first (Table 1). The artifacts of each site are described and classified, whether contemporary with the formation or not, in series. Chapter III – The most important Acheulean tool-types, the bifaces, are subjected to a further and detailed attribute analysis, starting with their context, the degree of rolling, weathering or alteration of the surface of the artifact, and including morphological and functional statistics. Chapter IV – Division of the Nahr el Kebir Acheulean into phases. The assemblages are amalgamated into contemporary blocks according to the known geological age of the formation. A sequence of four phases, two with sub-phases, is presented based on the typology and the chronological changes through time. Chapter V – The Acheulean sequence of the Kebir is compared to those in other parts of the Levant, phase by phase. 2. Summary of chapters I – IV In condensing Chapters I – IV, the same order of subjects as outlined above will not necessarily be followed, and references to the bibliography and to be illustrations are omitted. The Nahr el Kebir valley has been known for the past 50 years to contain Lower Paleolithic material. Enough artifacts and geomorphological data had been collected during brief visits or partial surveys to indicate that the prehistoric potential had not been fully explored. In 1976 a team under P. Sanlaville, consisting of himself and J. Besançon (geomorphologists) and L. Copeland and F. Hours (prehistorians) spent the month of October, 1976 in the area, identifying the Quaternary formations and collecting and studying assemblages of artifacts found in and on them. a. Terminology : The sequence of Quaternary formations distinguished in the Kebir Valley form a local chronology ; however, since the names will be unfamiliar to the reader, the regressive (fluviatile) periods are referred to in these chapters using the Alpine sequence (Günz, Mindel, Riss and Würm) ; we speak of 'glacials' and 'intergla- cials' rather than of 'pluvials' and 'interpluvials', and the interglacial periods can be correlated, as shown on Table 1 , with a not-too-distant shoreline sequence, that of Sanlaville (1977) for the Lebanon : Chaabian, Zaqrounian, Jbailian and Enfean. The last three mentioned correspond to the Günz-Mindel, Mindel-Riss and Riss-Würm transgressive phases. b. Method : The factor of crucial importance to the study of the prehistoric artifacts was their association with a succession of formations, the relative age and chronology of which had been established on geological criteria. Therefore, when collecting from the surface, a drastic first sorting was always made to eliminate fresh, unpatinated pieces. Only pieces with a physical condition (patina, degree of rolling, chemical alteration, etc.) matching that of the pebbles of each marine beach or fluviatile terrace, were retained and assumed (on this and other bases) to be contemporary with the formation. The raised beaches and river terraces were seen either in sections cut by river, sea or road, in the sides of wells or cisterns dug into the conglomerate (into which one could descend by ladder), or, as was most often the case, in the form of cemented or partially dismantled gravel conglomerates covering agricultural surfaces alongside both banks of the Nahr el Kebir. These conglomeretes survive on the summits of buttes, separated from eachother by the valleys eut by tributaries of the Kebir. The artifacts were dispersed in the surface soil of fields of pebbles. An exhaustive study of the condition of the artifacts, individually and in assemblages, was undertaken (to see if the most rolled (because re-worked) of one terrace related typologically to the least rolled of an earlier terrace for example). All the tests gave negative results, no typological differences being found between pieces of the same formation grouped as 'heavilly rolled', 'rolled', 'moderately rolled' or (relatively) 'fresh'. We finally came to assume that in each site the rolled artifacts in it could be considered as one contemporary assemblage. Classification of the artifacts was done by the Bordean method, and in addition the bifaces were studied by the method of D. Roe, which is designed to distin- guish differences in the morphology of assemblages of bifaces. The results were confirmatory. Other biface attributes examined included : dimensions, morphology of tip, base, cutting-edges, the kind of face-retouch and edge-retouch, blank. The results showed that we were dealing with a very consistent tradition of biface production which changed very gradually through time ; thick ovates were morphologically dominent throughout, but the proportion of amygdaloid forms increased ; overall size decreased ; sinuous cutting-edges gave way to straight edges. Sof-hammer techniques had apparently been used from the outset, but increased in competence through time. These changes were proportional, and only visible when the sample was large. The biface attribute study was considered together with the analyses of the other artifacts (cores, flakes, chopping-tools, picks, cleavers, retouched flake-tools). Here too, proportional changes through time could be detected but there was clearly no break in the tradition. At this stage the 17 most numerically large sam- ples were chosen for further analysis, and the sample for each phase was improved by amalgamating the artifacts of sites of the same age. The results are therefore based on the 2,796 artifacts, 692 of which are tools, from these 17 assemblages. By combining typological criteria and the geochronological framework, the Acheulean of the Kebir Valley may be divided as follows : Early Acheulean Phase I (type-site Sitt Markho) ; Early Acheulean Phase II (Jabal Idriss and Cheikh Mohammed) ; Middle Acheulean Phase I (Berzine) ; Middle Acheulean Phase II (Jibtaa and Khellale) : Late Acheulean (Roudo) ; Final Acheulean (Mchairfet es Samouk). c. Early Acheulean, Phases I and II Sitt Markho, perched at 130 m above present sealevel, is the only pre-Mindel terrace site found on the survey. It produced a sparse assemblage of large flakes many of which were transverse, two cleavers reminiscent of African types, and two heavy bifaces. The Sitt Markho fluviatile terrace precedes some high (120 m) raised beaches left by a marine transgression corresponding to the Zaqrounian (Günz-Mindel), and they are seen in exposures on both banks of the river. At two sites, Jabal Idriss and Cheikh Mohammed we were able to collect artifacts contemporary with the formation, and these represent our Early Acheulean Phase II. As at Sitt Markho, relatively few artifacts were recovered in spite of intensive search ; although of limited statistical reliability, we retained these samples on account of their age. To take ail three Early Acheulean sites together, they are characterised by a good number of chopping-tools, by the relative scarcity of bifaces, and by the low number of light-duty tools. As to debitage, the large size of the flakes differentiates them from those of later assemblages, as does the total absence of Levallois technique from their manufacture. As seen in Chapter V, the Kebir Early Acheulean does not correspond to contemporary chopping-tool facies in other regions. Nevertheless, since the earliest African Acheulean (dating at Bed II, Olduvai, to 1.3 - 1.4 million years) is older than Sitt Markho, we need not be surprised to find a biface industry in Syria at a slightly later date. d. The Middle and Upper Acheulean, considered together The three phases involded here are represented by 10 assemblages : Middle Acheulean Phase I by Berzine, from a Mindelian formation ; Middle Acheulean Phase II by Jabal Jibtaa and Khellale 4 and 5, from Mindel-Riss formations ; Late Acheulean by Roudo, from a Rissian formation ; the sample consiste of 1,707 artifacts. This sample allowed us to analyse in detail the developments summarised in Tables 23 and 24 ; the most useful elements were the bifaces and the debitage (cores and flakes). The biface studies have already been described ; for the cores and flakes the System used by A. Leroi-Gourhan was employed, as it distinguished types of core according to the desired flake-form and flakes according to their technical origin. One resuit was that a gradual shift towards prepared-flake cores could be seen. All the individual attribute scores are tabulated, and the trends are most clearly seen when contemporary assemblages are grouped. — The Middle Acheulean The Phase I site of Berzine was found in a terrace remnant of Mindelian date, and the assemblage was characterised by the predominence of thick and ovate bifaces, low number of laminar flake forms, and the absence of such types as trihedrals or polyhedrons (it is just these two types which characterise the contemporary sites in the Syrian interior, e. g. at Latamne and Joubb Jannine : see Chapter V). The Phase II sites of Jabal Jibtaa and Khellale 4 and 5 produced assemblages in the same tradition but slightly more evolved. This surprised us, as, before we went to the Nahr el Kebir, reports in the literature had led us to believe that a Middle Acheulean trihedral facies similar to that of Latamne, in the central graben or African Rift Valley extension, would be found there. This is certainly not the case, and it is to be recalled that at Berzine we could not be dealing with material later than a river flowing at 100 m altitude, in which the artifacts were (sometimes heavily) rolled ; the Khellale formation is also high 85 m. We have to assume that two different Acheulean facies were in existence at the same time in the northern Levant in Middle Acheulean times i. e. in the Mindel to Mindel-Riss phases. — The Late Acheulean Lower than the Khellale formation, deposits attributed to a cold phase, corresponding to the Riss, are found along the north bank of the Kebir, often seen in section, and sometimes attaining a thickness of 30 m. We retained four assemblages for study, Roudo being by far the richest. The traits of this phase are the same as Khellale and Jibtaa and also : an increase in Levallois debitage, which becomes quite typical both in cores and products ; an important increase in the number of light-duty tools ; decrease in large, broad-tipped ovate bifaces and increase of large or small, pointed and amygdaloid bifaces. In overview, in the Kebir Middle and Late Acheulean sequence, we see the dimensions of flakes and bifaces decrease through time (as is the case further South), and at the same time the proportions of biface types alter, ovoids giving way to amygdaloid and lanceolate types. Levallois technique appears during the Jbailian (Mindel-Riss). e . The Final Acheulean Corning to the Eemian - Enfean transgression, no raised beach of this age has been distinguished in the Kebir valley. The area was, we believe, occupied during this interglacial by the makers of what we are calling the Acheulean of Samoukian facies, a kind of Final Acheulean. The seven sites with similar material which we located are not in situ, but the Samoukians appear to have used older raised beach or terrace deposits as factories to produce assemblages with the following traits : there is a marked reduction in size of ail types of artifact, particularly bifaces, the lengths of the latter falling between 45 and 90 mm. Cores increase to form a quarter of the assemblage. Minuscule chopping-tools appear, made on pebbles, and form about a third of the tools, and another third is formed by the more numerous light-duty tools. Levallois debitage shows a substantial increase to 40 % of cores, 20 % of products. The patina is homogenous in each assemblage, and is not deep ; the pieces are not rolled. At the type-site, Mchaïrfet es Samouk, a raised beach of pre-Günz age, the Samoukian occured alo- ne. At the other sites, material contemporary with the formation (deeply patinated and rolled) was also present. The simultaneous presence of evolved Levallois techniques, miniature chopping-tools and small bifaces, suggests that the Samoukian could be the Kebir Valley equivalent of the industries of the same age in central Syria (Yabroud I or Tulul Defai, for example), Lebanon and Palestine. f . The Middle Paleolithic The fluviatile deposits of Würmian date produced artifacts, in the main of Mousterian typology, often with Levallois debitage, but we found no assemblage consistent enough to use as a basis of comparison with the well-known Levalloiso-Mousterian of the Levant. 3. Summary of chapter V: regional comparisons It is difficult to synthesise the Levant Paleolithic in the absence of an universal terminology, either for Quaternary phases, for cultural or industrial divisions, or even for flint tool-types. However, we will try to place the Kebir assemblages in the context of the Levant, phase by phase. Pre-Günz : We know of no artifacts of this age in the known early Pleistocene beaches up to 320 m altitude reported by workers such as Sanlaville, Ruske and Guerre. None were present at Mchaïrfet es Samouk, where ail the artifacts found were of a later date. The Günz : The only sites contemporary with Sitt Markho are found in the central graben. In a recent survey by ourselves in the middle Orontes area, a Günz terrace, the el-Khattab formation, was found. Sparse artifacts were obtained from several exposures, but they were not the same as those of Sitt Markho ; instead of flakes and bifaces, there were polyhedrons and chopping-tools, and these would seem to refer to the assemblages known from Ubeidiya in the Jordan Valley. In the same part of the Orontes valley, the sites of Sharia and Rastan, once thought to be of Villafranchian date, are now re-dated to a later phase, the Mindel. As regards the date of Ubeidiya, if, as the polarity changes suggest, it is earlier than the Mindel date now proposed, part of it could be contemporary with Sitt Markho. However, the industry is quite different, so that Sitt Markho remains so far without contemporary Levant analogues. The Zaqrounian (Günz-Mindel) : A good number of raised beaches of this age are known the length of the Levant coast, but only two of the six Kebir exposures and only two other sites (in South Lebanon) have produced artifacts. However, the assemblages at the latter sites of Borj Qinnarit and Dahr el Aassiye have no bifaces, and are quite unlike the assemblages of Jabal Idriss and Cheikh Mohammed ; they, too, seem to refer to the Ubeidiya tradition, but the sparseness of artifacts in all four cases is a point in common. The Mindel : At the Nahr el Kebir, we place here the change from Early to Middle Acheulean. None of the many exposures of contemporary fluviatile deposits along the Levant littoral are artifact-bearing, except the Nahr Beirut valley, and here the material is mixed. Inland, at this time appear the very distinctive 'descendants' of the Ubeidiya facies, at Latamne and Joubb Jannine, mentioned earlier. The elongated bifaces, tri- hedrals, chopping-tools, limestone spheroids and flint polyhedrons are entirely different from the thick ovate biface and flake tradition of Berzine, so that, for the Mindel phase of the Kebir, Berzine remains unaligned. The Jbailian (Mindel-Riss) : The main sites of this period in the Kebir Valley (e. g. Khellale) have produced comparable material for Passemard, Wendt, Van Liere and others ; the Phase II Middle Acheulean, described above as being slightly more evolved than Berzine, has a somewhat higher proportion of elongated and amygdaloid forms and this includes rare trihedric picks. (One such was found by Passemard and attributed by him to the 'Chalossian' or 'Abbevillian' ; the former is now dated geologically in France to the Riss according to Thibault. It is curious that at this time the trihedric tradition disappeared from the central graben – or at any rate, it was not present in the fluviatile terrace succeeding the Latamne formation.) Clear resemblances to the Kebir Middle Acheulean are to be found on the littoral, especially at Ras Beirut, on a fossil raised beach at 52 m, Ras Beirut IB, The same thick ovate biface tradition is present, and the same techniques of flake production are know to both, even if used in different proportions. The Ras Beirut IB material is sparse, but seems more like that of Berzine than Khellale or Jibtaa, and this may mean that it slightly pre-dates the beach containing it. Other Jbailian sites on the littoral include Middle Acheulean at Wadi Aabet and the Tayacian of the Cordon Littoral (Ras Beirut IB). The artifacts in Jbailian sites are generally at the same level technically, even if other differences can be seen. [Ν. Β. An exception must be made in the case of artifacts found on the north Syrian littoral in the region of the Jebleh Plain, which is a Jbailian abrasion platform covered with colluvium. Artifacts found here could date anywhere from the Mindel-Riss to the present (e. g. Hraïsoun).] The Riss : We attribute Roudo and its contemporary sites (e. g. Souayate) to the Late Acheulean of the Nahr el Kebir, an exception being Jinndiriye, which produced only flakes, sparse and dispersed. Although these flakes were not sufficient to form an assemblage, it is worth noting that flake facies proliferate in Riss times in the Levant (Umm Qatafa E, Bahsass, Tabun G, perhaps Yabrud IV). As to littoral biface industries, the sample from the 'Brèche de Pente' at Ras Beirut is very small but seems to have some traits similar to those of Roudo. Even the rough facies of Tranchée Bergy (Ras Beirut IV) has some rather crude picks which could be matched by the poorest specimens at Roudo. However, some of the now much more numerous inland industries such as Ma'ayan Barukh seem very close to Roudo, a situation in contrast to that obtaining in Mindelian times, when two different traditions occupied the central graben and the northern littoral ; now, there seems to be a similar technical and stylistic tradition all over the Levant from Dülük in the Hatay to the north, as far as Kissufim to the South. Since it was this tradition which we have seen developing through time in the Kebir area, we may ask what mechanism could be responsible for the relatively uniform biface morphology during Rissian times in the Levant. The Enfean or Eemian : Although no stratified sites of this phase were found in the Kebir area, it seems to have been occupied by the Samoukians, who we believe form the local equivalent to a variety of Final Acheulean variants seen all over the Levant, immediately preceeding (when in strata) the well-known Levalloiso-Mousterian. Examples are the Yabrudian and Acheulean of Levallois debitage at Yabrud I, the Amudian and Pre-Aurignacian, the erstwhile 'Micoquian' of Tabun E, the Acheuleo-Yabrudian of Bezez C, as well as others (unstratified) on the Golan and in southern Israel, and in the Orontes valley (Tulul Defai). Variants using Levallois methods seem more widely distributed than the Yabrudian, and the Samoukian would belong with the former. A site with close analogies to the Samoukian may be Evron Zinat, on the northern Israel littoral. In all these sites, the diminished size of the majority of bifaces seems to presage the disappearance of this tool-type altogether at the end of the Enfean. No good Levalloiso-Mousterian site being located in the Kebir area surveyed, no comparisons with other assemblages can yet be made. Conclusions : the Nahr el Kebir has produced evidence for: 1) the existence of an Early Acheulean, rolled in pre-Cromerian and Cromerian times, which may be contemporary with some levels at Ubeidiya, but of a different facies. 2) the existence of a Middle and Late Acheulean flint-working tradition of remarkable stability, which lasted from a glacial corresponding to the Mindel, until a glacial contemporary with the Riss. The Levallois method came into use as a definable element during the Mindel-Riss. 3) the proof that in the Near East there were at least two different, contemporary, Middle Acheuleans which appear to have occurred in different geographical and probably ecological zones. 4) the existence of what seems to be a local Final Acheulean facies, the Acheulean of Samoukian facies, corresponding to chronologically for example, the Acheulean of Yabrudian facies in the central Levant.Copeland Lorraine, Hours Francis. Le Paléolithique du Nahr el Kébir. In: Quaternaire et préhistoire du Nahr el Kébir septentrional. Les débuts de l'occupation humaine dans la Syrie du Nord et au Levant. Travaux de la RCP 438, sous la direction de Paul Sanlaville. Lyon : Maison de l'Orient et de la Méditerranée Jean Pouilloux, 1979. pp. 29-119. (Collection de la Maison de l'Orient méditerranéen. Série géographique et préhistorique, 9