3. It now seems to be generally accepted that it is extremely difficult to find convincing origins for the distinctive patterns of Aurignacian technology within most regions of Europe. This has been emphasized by Kozlowski (1982, 1992) for the Balkans and southeast Europe; by Allsworth-Jones (1986,
1990), Otte (1990) and others for central Europe; by Mussi (1990), Goia (1990) and Broglio (1993) for Italy; by Bordes (1968), de Sonneville-Bordes (1960), Demars (1990), Rigaud (1993) and others for western France; and by Bischoff et al. (1989) for northern Spain. In all these areas the earliest Aur-
ignacian industries appear as an abrupt break in the local patterns of technological development, with no apparent links with the immediately preceding Mousterian industries in the same regions. Only rarely has the possibility of purely local origins been suggested for Aurignacian technology in Europe, for example by Cabrera Valdes and Bernaldo de Quiros (1990) for the succession at El Castillo in northern Spain and by Valoch (1983) for some of the Czecho-slovakian industries. Both suggestions, however, have been contested by other workers (e.g. Kozlowski 1993; Djindjian 1993) and have since been withdrawn by Valoch himself (1990) for the Czechoslovakian industries.
At present the most plausible origins for Aurignacian technology seem to lie in some of the Middle Eastern industries, most notably in the long sequence of Aurignacian, proto-Aurignacian and so-called transitional industries recorded at Ksar Akil in the Lebanon (Copeland 1976; Marks & Ferring 1988; Ohnuma & Bergman 1990). Significantly, it is there, in contrast to the various regions of Europe, that the earliest Aurignacian industries were preceded by a long succession of demonstrably earlier Upper Palaeolithic technologies, apparently extending back to 45-50,000 BP (Marks & Ferring 1988; Mellars & Tixier 1989; Bar-Yosef 1994).
4. The relative and absolute chronology of the earliest stages of the Aurignacian within the different regions of Europe still remains to be documented in secure terms - largely owing to the inherent limitations of radiocarbon dating within this age range. The overall pattern of available dates nevertheless appears to suggest a pattern of successively younger dates extending from east to west across the continent, from around 43-45,000 BP in eastern and southeastern Europe, through to ca 40,0000 BP in central Europe, northern Spain and the Mediterranean coast, to around 35,000 BP in the classic region of southwestern France (Fig. 13.13: Kozlowski 1992, 1993; Mellars 1992b; Hahn 1993b). There is an urgent need for more dates to confirm this pattern, preferably with methods other than radiocarbon. As the evidence stands at present, however, it could be seen as at least consistent with the hypothesis of a progressive spread of Aurignacian technology from east to west across the continent.
5. Finally, the character and scale of the various technological and other behavioural innovations associated specifically with the earlier stages of the Aurignacian in different regions of Europe should be re-emphasized -ranging from innovations in the technology and typology of stone tool production, through extensively shaped bone, antler and ivory artefacts, to the effective explosion of symbolic artefacts in the form of notched and incised bonework, various forms of personal ornaments and remarkably varied and sophisticated representational art. Even if we set aside evidence for apparent shifts in the densities of human population, patterns of animal exploitation and the sizes of local social and residential groups, this is an impressive range of behavioural innovations which, as argued earlier in Chapters 11 and 12, almost certainly reflects equally radical changes in the social, cognitive and most probably linguistic patterns of the associated populations. Of course, radical and wide-ranging innovations of this kind cannot be taken as an automatic indication of episodes of population dispersal or replacement in the archaeological record, since it is clear that under certain conditions episodes of rapid behavioural change can occur through processes of either cultural diffusion or purely internal cultural change. Nevertheless the close association of all these behavioural innovations with the first appearance of Aurignacian technology - and apparently fully modern skeletal anatomy - in the different regions of Europe, is at least consistent with the hypothesis of a population dispersal at this point in the archaeological sequence, even if archaeological evidence alone cannot be held up as conclusive proof.
Viewed as a whole, therefore, the archaeological evidence for the Aurignacian in the different regions of Europe coincides closely with the patterns that one would predict from the implications of current population-dispersal scenarios of modern human origins. Whether the same data could be held to be equally consistent with the population-continuity or multiregional evolution hypothesis is much more doubtful. How in this case would one account for the striking uniformity of Aurignacian technology over such a vast area of Europe and the Middle East, superimposed on so much diversity in the technology of the immediately preceding Middle Palaeolithic populations in the same regions? How would one explain the sudden and abrupt appearance of this technology in so many different regions, without any convincing origins or antecedents in preceding technologies in the same areas - or the sheer range, diversity and magnitude of the various cultural and behavioural innovations involved? In the classic region of western France there can be no serious doubt that the appearance of the Aurignacian reflects the intrusion of an essentially new human population, not only in the sudden and abrupt appearance of this technology (clearly later than in the neighbouring areas of northern Spain and the Mediterranean coast: Fig. 13.13), but also in the clear evidence that the earliest Aurignacian communities in this area persisted and apparently coexisted for some time alongside the latest Neanderthal populations in the same region (see below: cf Mellars 1989a; Demars & Hublin 1989; Rigaud 1993). If we accept such population intrusion in southwestern France, we should be prepared to give the same hypothesis equal consideration in the other regions of Europe where the overall spectrum and character of the archaeological evidence appears to show a similar pattern.
One of the most intriguing questions posed by current studies of the origins and dispersal of biologically modern human populations is how far we can identify evidence for any contact or interaction between the final Neanderthal populations and the earliest, hypothetically intrusive populations of anatomically modern humans in the different regions of Europe? If the population-dispersal scenario is valid, then this particular issue cannot be avoided. The inescapable implication of this model is that some kind of contact and interaction between the intrusive, expanding populations of anatomically modern hominids and the local, indigenous populations of archaic Neanderthals must have occurred repeatedly, and over the whole of the geographical range occupied by the expanding modern populations. This scenario has provided the inspiration for several popular novels, such as William Golding's The Inheritors and Jean AueFs The Clan of the Cave Bear, but remains surprisingly poorly studied from the perspective of the archaeological evidence.
Over the last decade, evidence for such chronological overlap, contact and apparent interaction between the final archaic and earliest anatomically modern populations has been claimed from several different regions of Europe (e.g. Allsworth-Jones 1986, 1990; Kozlowski 1988, 1990, 1993; Harrold 1989; Otte 1990; Mussi 1990; Goia 1990; Valoch 1990; Demars 1990; Demars & Hublin 1989; Hublin 1990; Mellars 1989a, 1991; Rigaud 1993; Djindjian 1993). The clearest evidence once again comes from the extreme western fringes of Europe, centred on the Perigord and adjacent provinces of southwestern France and resides in the demonstrable contemporaneity of two quite distinct and sharply contrasting archaeological assemblages, represented by the classic Aurignacian industries on the one hand and those of the Chatelperronian or Lower Perigordian group on the other. The juxtaposition of these two industries raises a number of intriguing issues which are worth examining closely.
1. On the basis of simple technological and geographical criteria, there can be no doubt that the Aurignacian and Chatelperronian industries were products of separate human populations in the southwestern French sites. The distinctive type-fossils which define the two industries (Chatelperron points for the Chatelperronian and various forms of nosed and carinate scrapers, Aurignacian blades, Dufour and Font Yves bladelets, split-base bone points etc. for the Aurignacian: cf. Figs 13.1, 13.14) show mutually exclusive distributions in material from the most recently excavated sites, and there is evidence that both the basic techniques of flake and blade production and the specific sources exploited for lithic raw materials in the two variants were different (de Sonneville-Bordes 1960; Harrold 1989; Demars 1990; Demars & Hublin 1989; Pelegrin 1990; Leveque et al 1993; Rigaud 1993). Most significantly, the geographical distributions of the two industries are radically different: whereas the Aurignacian has a distribution extending over virtually the whole of western, central and eastern Europe, the Chatelperronian is restricted to a small zone confined entirely to the western and central parts of France (to the west of the Rhone valley) and penetrating for a short distance into the adjacent parts of the Pyrenees and northern Spain (Fig. 13.10).
2. The existence of a substantial period of overlap between the Aurignacian and Chatelperronian populations canbe demonstrated from several different aspects of the chronological data. In addition to correlations based on detailed climatic and vegetational sequences recorded at different sites (Leroyer & Leroi-Gourhan 1983; Leroyer 1988) we now have evidence from three sites in southern France and northern Spain where discrete levels of Châtelperronian and Aurignacian industries occur directly interstratified within the same stratigraphie sequences -notably at the Roc de Combe and Le Piage in southwestern France and at El Pendo in Can-tabria (Bordes & Labrot 1967; Champagne & Espitalié 1981; Harrold 1989; Demars 1990). The available radiocarbon evidence admittedly remains rather sparse and potentially ambiguous for the southwestern French sites. From the immediately adjacent areas of the Mediterranean coast and Cantabria, however, there is now clear radiocarbon evidence that typically Aurignacian industries were being manufactured by at least 38-40,000 BP, preceding by at least 4000-5000 years dates for typical Châtelperronian industries at sites such as Les Cottés and Arcy-sur-Cure in western and central France (Fig. 13.13: Bis-choff et al. 1989; Cabrera Valdes & Bischoff 1989; Harrold 1989; Farizy 1990). From the combined palaeoclimatic, stratigraphie and radiocarbon evidence, there can be no doubt that the time ranges of the Aurignacian and Châtelperronian industries must have overlapped within these extreme western parts of Europe for several thousand years.
3. The critical importance of this chronological overlap lies in the fact that there is now almost conclusive evidence that these two technologies were the product of contrasting biological populations. All the available skeletal evidence from France and other regions of Europe suggests that the Aurignacian industries were the product of fully anatomically modern populations (Howell 1984, 1994; Stringer et al 1984; Smith 1984; Gambier 1989, 1993; Demars & Hublin 1989; Hublin 1990). By contrast, there is explicit evidence from the hominid remains recovered from Saint-Césaire (Fig. 13.8), as well as from the series of human teeth recovered from earlier excavations at Arcy-sur-Cure, that the populations responsible for the Châtelperronian industries were of distinctively archaic, essentially classic Neanderthal type (Lève-
que & Vandermeersch 1980; Vandermeersch 1993b; Stringer et al. 1984; Leroi-Gourhan 1958). If this evidence is accepted, then it confirms the coexistence of these two biologically contrasting populations wTithin the western fringes of Europe over a very substantial span of time.
What has not always been so clearly recognized in the earlier literature is that these archaic associations of the Châtelperronian industries had already been predicted, several decades before the discovery of the Saint-Césaire skeleton, purely on the basis of the technology of the industries. As long ago as 1954 François Bordes argued that many of the distinctive technological features of the Châtelperronian industries, such as the character of the steeply backed Châtelperron points, as well as the occurrence of typical side scrapers, denticulates, and even small, bifacial hand-axe forms (Fig. 13.14), showed obvious links with the preceding Mousterian industries of the same region, especially with those of the MTA group (Bordes 1954-55, 1958, 1968, 1972). Later, I added further strands to these arguments, by pointing to the closely similar geographical distributions of the Châtelperronian and MTA industries (both confined to areas to the west of the Rhône valley in France and both extending into the adjacent areas of northern Spain) and arguing that the MTA industries appeared from several lines of evidence to represent the final stages of the local Mousterian sequence in southwestern France, immediately preceding the emergence of the Châtelperronian industries (Mellars 1969, 1973). In short, the arguments for believing that the Châtelperronian industries are the product of entirely indigenous, Neanderthal, populations in western Europe can be supported strongly on the basis of both the direct skeletal associations of the industries (at Saint-Césaire and Arcy-sur-Cure) and the basic technology, chronology and spatial distribution of the industries themselves.
4. The final and in many ways most interesting point is that this period of overlap between the Aurignacian and Chatelperron-ian populations seems to be reflected in the various forms of interaction or acculturation between the two populations. As discussed in detail elsewhere (e.g. Harrold 1989; Mellars 1989a, 1991; Farizy 1990, 1994; Leveque 1993) it is now clear that while the basic technological roots of the Chatelperronian industries lie in the immediately preceding Mousterian industries, many of their specific features are of distinctively Upper Palaeolithic type. This applies not only to the strong component of typical blade technology apparent in most Chatelperronian assemblages but also to the presence of highly typical and abundant forms of end scrapers and burins and - in at least some sites - simple but extensively shaped bone and antler tools and even personal ornaments, in the form of carefully grooved and perforated animal teeth (Fig. 13.15: Harrold 1989; Farizy 1990; Leveque 1993; Leroi-Gourhan & Leroi-Gourhan 1964). The crucial point to recognize is that all these specifically Upper Palaeolithic elements in the Chatelperronian appear to have developed at a relatively late stage, certainly long after the initial appearance of fully Aurignacian industries in northern Spain (Fig. 13.13) and probably while Aurignacian populations were already present in the southeastern parts of France (Leroyer & Leroi-Gourhan 1983; Leroyer 1988; Cabrera Valdes & Bischoff 1989; Mellars 1992b). Exactly how such interaction and apparent acculturation between the final Neanderthal and earliest anatomically modern populations should be visualized remains more controversial (see Graves 1991 and associated comments for further discussion). There seems little doubt, however, that the emergence of typically Upper Palaeolithic technological features amongst the final Neanderthal populations of western Europe can be explained more economically by vari-
ous contact and acculturation processes than by a purely spontaneous invention of Upper Palaeolithic technology by the final Neanderthal communities themselves.
Exactly how this kind of coexistence between the two populations could be maintained is more difficult to answer from the available archaeological evidence. We are still extremely ignorant about many of the most basic adaptive and organizational features of the Châtelperronian populations, mainly due to the poverty of faunal material recovered from most of the sites and the lack of detailed studies of the available economic data. One possibility is that the Châtelperronian and early Aurignacian groups were adapted to significantly different foraging and subsistence strategies - with the Aurignacian perhaps focusing mainly on specialized hunting of reindeer herds along major migration trails (such as the valleys of the Dordogne and Vézère) while Châtelperronian groups were adapted to more generalized animal exploitation, perhaps still dependent partially on scavenging rather than on deliberate and strategic hunting of game. The very generalized faunal assemblages recovered from Châtelperronian levels at Saint-Césaire, Roc de Combe, Châtelper-ron, Trou de La Chèvre etc. (in each case showing fairly balanced frequencies of horse, red deer, bovids, reindeer etc.) could be taken to support this suggestion (Delpech 1983, 1993; Patou-Mathis 1993). Another possibility is that the overall levels of population density and highly mobile patterns of seasonal and annual foraging strategies practised by the two groups were such that there was rarely any direct competition between them for particular economic resources or for the simultaneous occupation of the same territories. Evidence for close interstratification of Aurignacian and Châtelperronian levels documented at the Roc de Combe and Le Piage (Bordes & Labrot 1967; Champagne & Espitalié 1981; Demars 1990) might perhaps be seen as a direct reflection of these highly mobile foraging patterns. Most probably it was only when the population density of the Aurignacian groups built up to relatively high levels during the middle and later stages of the Aurignacian that any strong economic and social competition for the use of particular resources or social territories would have emerged in some of the more ecologically favoured areas such as the Dordogne and Vezere valleys (de Sonneville-Bordes 1960; Mellars 1989a; Demars 1990, Rigaud 1993). It is at this time, significantly, that evidence for Chatelperronian occupation seems to be restricted mainly to the peripheral zones of western and central France, such as the Arcy-sur-Cure caves and some of the areas to the north and south of the Perigord region, in the sites of Les Cottes, Fontenioux and Quingay in the Vienne, or at Roc de Combe and Le Piage in the Lot (Ler-oyer & Leroi-Gourhan 1983; Leroyer 1988; Leveque 1993). Seen in these terms it is reasonable to suggest that the process of eventual population replacement of the Chatelperronian by the Aurignacian groups was a relatively gradual and progressive phenomenon, probably reflecting more of a gradual shift in population numbers and the occupation of specific territories rather than any outright confrontation between the two groups (Zubrow 1989).
How far similar interaction and acculturation patterns between final Neanderthal and earliest anatomically modern populations can be recognized in other regions of Europe remains a topic of continuing debate. Alls-worth-Jones (1986, 1990), Kozlowski (1988, 1990, 1993), Valoch (1990) and several others have put forward this argument for the emergence of the Szeletian and related leaf-point industries of central and eastern Europe (Figs 13.10,13.16), pointing out that the time-range occupied by these industries almost certainly overlaps with that of the, apparently intrusive, Aurignacian industries in the same regions, and that strictly local roots for these distinctive industries can be seen in both the
technology and spatial distribution (Fig. 13.10) of the archaeological assemblages. Mussi (1990), Goia (1990) and others have presented similar arguments for the emergence of the Uluzzian industries of the Italian peninsula - again almost certainly contemporaneous with the presence of typical Aur-ignacian industries in the adjacent areas of the Mediterranean coast, and again showing a restricted geographical distribution in the Italian sites (Fig. 13.10). Further to the east, similar patterns may be reflected in the dichotomy between the Streletskaya and Spitsinskaya industries of the south Russian Plain (Soffer 1985; Hoffecker 1988).
To summarize, recent research into the earliest stages of the Upper Palaeolithic now seems to reveal a broadly similar pattern in different regions of Europe. In each area there is evidence for the presence of typical and seemingly intrusive Aurignacian industries, apparently associated with fully anatomically modern hominids, and appearing in most regions between ca 40,000 and 35,000 BP. Alongside these industries, and at a broadly similar date, there is evidence for the emergence of a range of sharply contrasting forms of early Upper Palaeolithic technology, each restricted to a limited geographical area and each showing strong and obvious links with the latest Middle Palaeolithic technologies in the same regions. It is only in western Europe that these local technologies have been found in association with well documented human skeletal remains but in this particular case (i.e. the Chatelperronian) the skeletal remains are of distinctively archaic, Neanderthal form (Fig. 13.8). Proponents of the population dispersal hypothesis would argue that this pattern coincides closely, if not exactly, with the pattern one would predict from the scenario of a rapid dispersal of new human populations over different regions of Europe, combined with varying degrees of chronological overlap, contact and eventually acculturation with the local, indigenous Neanderthal populations.
Was this article helpful?