Bovids

of red deer discussed above, through with an even stronger emphasis on heads and jaws and a weaker representation (at least in the case of bovids) of extreme lower limbs. The possible explanations for these patterns would also seem to be similar. Binford sees this pattern as strongly indicative of scavenging activities, with particular emphasis on the parts of carcases (especially heads) most frequently available in the leftovers of carnivore kills. Against this, Chase (1986a, 1988, 1989) has argued that the main emphasis was on upper rather than lower parts of limbs of bovids and horse (although this does not emerge clearly from his own data - with the possible exception of levels 23-25 at Combe

Grenal) and that there is no evidence for heavy chopping or hacking through the bones that might be expected to result from the dismemberment of partially desiccated or frozen carcases (Fig. 7.11). Chase also draws attention to Levine's (1983) studies of the age distribution of horse teeth for three different levels at Combe Grenal (Fig. 7.12), which appear to indicate an essentially 'catastrophic' profile in the age distribution of the animals killed (see below) and which is arguably much more consistent with deliberate hunting than with any kind of scavenging activities. These objections have been countered by Binford (1991) who argues that even in the case of scavenged animals, some

Age in years

Figure 7.12 Estimated age distribution of remains of horses in three different stratigraphic levels at Combe Grenal, based on crown-height measurements of the molar teeth. Allowing for the underrepresentation of very fragile teeth from the youngest age classes, the distributions seem to reflect a 'catastrophic' age profile, comparable with that to be expected in a living herd (see Fig. 7.13). After Levine 1983.

Age in years

Figure 7.12 Estimated age distribution of remains of horses in three different stratigraphic levels at Combe Grenal, based on crown-height measurements of the molar teeth. Allowing for the underrepresentation of very fragile teeth from the youngest age classes, the distributions seem to reflect a 'catastrophic' age profile, comparable with that to be expected in a living herd (see Fig. 7.13). After Levine 1983.

degree of selection of the more useful parts of the carcase would probably be involved. Similar objections could be raised against the interpretation of the horse and bovid remains in terms of primary butchery locations from which selected parts of carcases were systematically removed for further processing or consumption elsewhere. Why in this case should the pattern of horse remains (notably in levels 25-20 at Combe Grenal) be significantly different from that of reindeer in the same levels? And why is there such a conspicuous scarcity of the feet and lower limbs of large bovids which one would normally expect to be abandoned at the primary kill or butchery location as the least useful parts of the carcase? As Chase points out, perhaps the most economical interpretation for the remains of large bovids, at least, would be in terms of the 'schlepp effect' model discussed earlier. In view of the massive differences in the body weights of large bovids and reindeer this may be the most plausible way of accounting for the striking differences in the patterns of bone-representation of these two species - as well as those of horse - in the southwestern French sites. But any such interpretation would leave entirely open the question of whether the basic exploitation of these large species involved systematic hunting or simply opportunistic scavenging of remains from either natural animal deaths or carnivore kill sites.

Head processing patterns

Finally, some interesting details which have emerged from the studies of both Chase and Binford relate to the variable representation of upper and lower jaw remains of different species. From the data assembled by Chase (1986a) it can be seen that whether calculated in terms of numbers of teeth from the upper and lower jaws or from the surviving fragments of maxillary and mandibular bones, there is much variation in these ratios both between different species and in different levels of the Combe Grenal sequence (Figs 7.5, 7.6). In most levels (both at Combe Grenal and elsewhere) there seems to be a strong tendency for lower jaw remains to be represented more frequently than upper jaw remains - sometimes by a ratio of three or four to one (for example in the reindeer remains from levels 11-21 at Combe Grenal and red deer remains from levels 31-35). But in other contexts these ratios may be reversed; thus a predominance of upper over lower jaws and teeth is apparent in the red deer remains from levels 54-43 at Combe Grenal (Fig. 7.5) and from the contemporaneous levels at Pech de VAze II (layer 4B) and at Pech de l'Aze IV (layer J3a). The same can be seen in large bovid remains from layer C at La Ferrassie and levels G and H at Le Moust-ier. Since this cannot be attributed simply to poor survival or incomplete recovery of the remains during excavation, the pattern must have some significance in economic or behavioural terms.

Recently, Binford (1984, 1991) has argued that these patterns probably reflect deliberate economic strategies by Palaeolithic groups, related to different approaches to processing animal brains and tongues as especially nutritious and favoured food resources. From a study of modern ethnographic sources, he points out that while some groups deliberately separate the mandible from the remainder of the skull at the site of the kills, other groups tend to transport complete heads back to occupation sites for systematic processing there (Binford 1984: 222-3). In some cases complete skulls may be cooked over substantial hearths, leaving distinctive patterns of burning on the lower parts of the mandibles. In other cases the crania may be cooked separately from lower jaws (i.e. with mandibles detached) producing evidence of burning selectively on the maxillary teeth (Binford 1984: 160-1).

Most of these data remain to be systematically analysed for the Mousterian faunas under discussion here. It may be significant, however, that all the samples from Combe Grenal which show a clear pattern of over-representation of upper teeth relate to the assemblages of red deer bones in the Würm I levels of the site. Remains of reindeer, red deer, horses and bovids from all of the documented Würm II levels, by contrast, show an emphasis on teeth from the lower jaws, in some cases with three or four times as many lower as upper teeth. As noted above, the only other samples which show a predom inance of teeth from the upper jaws are those of large bovid remains from the principal Ferrassie-Mousterian level at La Ferrassie and from the two MTA levels (layers G and H) at Le Moustier. Thus there could well be a significant element of behavioural patterning in this data, relating either to different industrial variants of the Mousterian or to different chronological stages in the last glacial sequence.

Binford has provided a few interesting additional details. From his own analyses of the Combe Grenal data he has suggested that skulls and mandibles may have been treated differently according to the age of the animals. Thus in the Denticulate Mousterian levels of layers 16-14 the majority of lower jaws of horses seem to come from relatively old (or at least adult) animals, while the upper jaws are mainly from young individuals (Binford 1984: 222-3). He suggests by analogy with the behaviour of recent Nuna-muit eskimo that this may indicate a deliberate preference for processing the brains of young animals and the tongues of older and larger animals. More detailed information on these patterns will no doubt be presented in the forthcoming monograph on the Combe Grenal material (see Binford 1992). From the data provided by both Chase and Binford it is clear that there is some fascinating behavioural information still to be derived from some of the more detailed features of the faunal assemblages from Combe Grenal and other sites.

Was this article helpful?

0 0

Post a comment