Carcase utilization patterns

To demonstrate clear patterns in bone element frequencies of the kind documented by Chase at Combe Grenal and at some of the other sites discussed above, is one thing. To formulate clear-cut explanations for such patterns in human behavioural terms is a far more difficult challenge, especially with the very limited data at present available on such questions as the seasonal use of different species, the age and sex distribution of animals, the overall spatial patterning of the faunal remains in the sites and so on. These issues have generated much discussion in the literature leading to a range of divergent interpretations (Binford 1978, 1981, 1984, 1988, 1991; Speth 1983; Chase 1986a, 1988, 1989; Stiner 1991a, b; Klein 1986, 1989b, c; Turner 1989; O'Connell et al 1988a, b). All that will be attempted here is to focus on some of the most striking and well documented patterns which have emerged from recent studies at Combe Grenal and elsewhere, and to evaluate some of the alternative behavioural models which could possibly account for them. More in-depth analysis of these questions will be possible when some of the current studies of the faunal assemblages from Combe Grenal and elsewhere have been completed and fully published (see Binford 1992).

Red-deer processing patterns

One of the most consistent patterns documented by Chase (1986a) in the Combe Grenal data relates to the assemblages of red deer remains recorded in both the Würm I and the earlier part of the Würm II levels on the site. These assemblages have three characteristics (Fig. 7.5): (1) a strong representation of heads and teeth; (2) an equally strong representation of lower limb remains; and (3) a general under-representation of remains from the major meat-bearing parts of the main axial skeleton and upper limbs. Interestingly, this pattern has been documented not only in the remains of red deer from both the Würm I and Würm II levels at Combe Grenal but also in the remains of reindeer for the earliest levels of the Würm II sequence (levels 35-26). Similar patterns have also been recorded in red deer remains from several other sites in the same region (Pech de l'Aze II and IV, Le Moustier, Grotte Vaufrey).

In assessing the behavioural significance of these patterns, there are at least three quite separate models which could possibly be advanced to account for the bone element data:

Scavenging

Chase accepts that in many ways the most obvious interpretation for these patterns would be in terms of systematic scavenging of abandoned carcases from either natural animal deaths or carnivore kills. All the recent research on carnivore predation has emphasized that carnivores tend to consume initially the meatiest parts of the carcase (i.e. mainly the axial region and upper parts of main limbs) and to leave behind the least meaty portions represented by the head, lower limbs and feet (Binford 1978, 1981; Blumenschine 1986, 1987; Stiner 1991a, b). It is those parts of the skeleton left over at carnivore kills which are therefore most frequently available to human scavengers (Fig. 7.14) and it is precisely these elements which are best represented in the various assemblages of red deer bones from Combe Grenal.

Whilst acknowledging the general strengths of this argument, Chase has rejected virtually the whole of the scavenging model for the faunal material from both these and other occupation levels at Combe Grenal (Chase 1986a, 1988, 1989). The detailed arguments will be discussed more fully later but in essence Chase feels that the combination of other arguments in favour of systematic hunting throughout the Combe Grenal sequence can be used to dismiss any simple scavenging theory to explain these bone frequencies. He argues for example that patterns of cut marks and breakage patterns on many bones are not consistent with large-scale utilization of remains from partially desiccated or frozen carcases (Fig. 7.11). He also argues that several other lines of evidence appear to document competent and systematic hunting of relatively large animals well before the time range of the Combe Grenal sequence. Although not discussed specifically by Chase in this context, he could also have drawn attention to other aspects of the faunal assemblages to argue against a simple scavenging hypothesis, including for example the relatively strong representation of pelvic bones of red deer (a relatively meaty bone which one might expect to have been removed by carnivores well before any scavenging of carcases by human groups) and the weak representation of phalanges and other foot bones (which one would normally expect to be brought into the site attached to lower limbs) (Fig. 7.5). Significantly, similar patterns of bone-element frequency can be documented for red deer remains in several of the later Würm II levels at Combe Grenal (levels 20-25), where there is almost conclusive evidence for systematic hunting in the exploitation of reindeer (see below). To account for these particular patterns it would be necessary to suggest quite different strategies (i.e. scavenging for red deer and hunting for reindeer) were used for these two species of deer during precisely the same periods of occupation.

On-site butchery and processing of animal carcases

The interpretation advanced by Chase himself (1986a, 1988) to account for the patterns discussed above is related directly to the processing and utilization of red deer carcases on site. Chase argues in essence that the Combe Grenal site served principally as a 'primary' hunting and butchering location in which largely complete carcases of hunted animals were taken for systematic processing, fairly close to the location of the kills. His hypothesis is that the majority of the red deer carcases were dismembered and butchered on site and most of the prime, meaty parts removed from the site - perhaps after drying to aid storage - for use and consumption in another complementary type of activity or settlement location. According to this hypothesis, the animal parts that were abandoned at Combe Grenal consisted of the less economically useful parts of the carcases (i.e. mainly heads and lower limbs) which were processed on site to extract certain delicacies such as the brains and tongues, together with marrow from the main marrow-bearing bones (metapodials, ulnae, tibiae and phalanges). To use his own words:

In either case it would appear that the meatiest portions of the carcass were being transported elsewhere for consumption, while the less meaty portions were being consumed on the spot. This would imply that the site served as a hunting camp for red deer, and that there existed another site or sites whose occupation depended to some extent upon the hunting carried out in the vicinity of Combe Grenal

On the face of it, this model is no less coherent or internally consistent than that of

% Soft tissue Marrow volume

Figure 7.8 Frequencies of different skeletal elements of red deer recorded in levels 50A-54 at Combe Grenal, compared with the relative amounts of'soft tissue' (i.e. meat) (left) and volumes of marrow (right) associated with the different bones. After Chase 1986a. The symbols for the different bone elements are: CR cranium; F femur; H humerus; M mandible; MC metacarpal; MT metatarsal; P pelvis; PI first phalange; P2 second phalange; R radiocubitus; SC scapula; T tibia; V vertebrae.

% Soft tissue Marrow volume

Figure 7.8 Frequencies of different skeletal elements of red deer recorded in levels 50A-54 at Combe Grenal, compared with the relative amounts of'soft tissue' (i.e. meat) (left) and volumes of marrow (right) associated with the different bones. After Chase 1986a. The symbols for the different bone elements are: CR cranium; F femur; H humerus; M mandible; MC metacarpal; MT metatarsal; P pelvis; PI first phalange; P2 second phalange; R radiocubitus; SC scapula; T tibia; V vertebrae.

the scavenging hypothesis discussed above. When examined in detail, however, it reveals a number of obvious problems. First, it leaves open the question of the precise location and character of the supposedly complementary parts of the settlement and exploitation system of which Combe Grenal, by implication, formed only one specialized part. Where exactly are these other sites which hypothet-ically should contain all the more meaty, economically useful parts of red deer carcases lacking at Combe Grenal? Chase of course could well dismiss this objection on the grounds that other economically specialized sites could well exist, for example, at open-air locations on the plateaux of the region or at other as yet undiscovered or unexcavated sites. While one can hardly dismiss these arguments, the model does rely heavily on missing data and on the existence of specialized consumption or home-base sites for which direct evidence is, as yet, lacking.

There are other problems with Chase's hypothesis. As he points out (1986a: 47) the theory cannot easily account for the relative abundance of pelvic bones of red deer at Combe Grenal, nor for the apparent under-representation of phalanges and other foot bones - unless this can be attributed entirely to heavy fragmentation and accordingly poor recovery of phalanges. More significantly it fails to explain the striking contrasts between the patterns of red deer bones compared with those of reindeer in the later Würm II levels at Combe Grenal (layers 20-25) (Fig. 7.5). Again, one would need to explain why very different patterns of carcase utilization were applied to two similar species in the same occupation levels. In other words, why should the same groups practise a systematic policy of transporting most prime parts of red deer carcases away from the site, whilst apparently processing and consuming most of the same components of reindeer carcases on the site itself?

Reindeer Layers 23-25

Reindeer Layers 23-25

% Soft tissue

Figure 7.9 Frequencies of different skeletal elements of reindeer in layers 23-25 (upper) and layers 20-22 (lower) at Combe Grenal, compared with the amounts of soft tissue and marrow volumes associated with the different bones. Abbreviations for the different bone elements follow those in Fig. 7.8. After Chase 1986a.

% Soft tissue

Figure 7.9 Frequencies of different skeletal elements of reindeer in layers 23-25 (upper) and layers 20-22 (lower) at Combe Grenal, compared with the amounts of soft tissue and marrow volumes associated with the different bones. Abbreviations for the different bone elements follow those in Fig. 7.8. After Chase 1986a.

Transport-distance effects The third model could be described by what Perkins and Daly (1968) referred to as the 'schlepp effect'. This asserts that the extent to which particular parts of animal carcases are brought back to occupation or home-base sites depends partly on the distances over which the remains have to be carried, partly on the overall costs (in terms of work and effort) involved in transporting them, and partly on the perceived economic value of different parts of carcases. Where remains of relatively large animals are involved, such as those of large ungulates, one would normally

Figure 7.10 Frequencies of different skeletal elements of reindeer in layers 23-25 and 20-22 at Combe Grenal, compared with the amounts of soft tissue and percentages of oleic acid (as a reflection of fat content) associated with the different bones. Abbreviations for the bone elements follow those in Fig. 7.8. After Chase 1986a.

Figure 7.10 Frequencies of different skeletal elements of reindeer in layers 23-25 and 20-22 at Combe Grenal, compared with the amounts of soft tissue and percentages of oleic acid (as a reflection of fat content) associated with the different bones. Abbreviations for the bone elements follow those in Fig. 7.8. After Chase 1986a.

expect them to be dismembered and partially butchered at or very close to the point of killing, and only selected, especially valued parts of carcases transported over substantial distances to the main base camps or occupation sites.

This hypothesis could be invoked to account for several aspects of the Combe

Grenal assemblages. It assumes that most red deer were killed at some distance from Combe Grenal - perhaps on the various plateaux areas adjacent to the Dordogne valley -and that the particular parts of red deer carcases most frequently transported back to the site were perceived to be of some special value. As Chase (1986a: 46-50) has sugges ted, this might well be applicable to both head remains and many lower limb bones. The value of heads presumably depended on the use of the brains and tongue - both relatively tasty and fat-rich foods which frequently have been reported in the ethnographic literature as special delicacies among hunting and gathering groups (Speth & Spielmann 1983; Speth 1987, 1990; O'Connell et al. 1988a, b; Stiner 1991a). The same observation could apply to the major marrow-yielding bones, which provide an equally rich and nutritious source of food. In addition, Chase points out that the particular distal limb segments brought back to Combe Grenal seem to be those with the highest concentration of fatty acids and which also retain substantial amounts of marrow throughout the annual cycle of deer growth and development (Figs 7.9, 7.10) (Chase 1986a: 51; Speth and Spielmann 1983; Speth 1987, 1990). Viewed in these terms Combe Grenal would have served primarily as a special food-processing location where these selected parts of the red deer carcases would be introduced for intensive processing and extraction of the food delicacies - procedures which apparently involved time-consuming breakage of the main marrow bones, similar processing of the crania and mandibles to extract brains and tongue and probably (in the case of heads) cooking or roasting over a substantial hearth (Binford 1984: 160-1; 1992). All these activities might well have involved the use of special processing locations, possibly located some distance away from the actual procurement sites of the game.

This hypothesis could of course be accommodated equally well within either the deliberate hunting or opportunistic scavenging models. In the case of hunting, the implication would be that most of the immediately accessible and edible parts of the carcases were consumed either on the site of the kills or perhaps at other closely associated butchery or carcase processing locations. If so, this would present an obvious challenge to future archaeological researchers to identify examples of these primary kill and butchery locations. For scavenging, the pattern could be much simpler. In this case, the kind of intensive head and marrow-processing sites documented in these early Würm I levels at Combe Grenal might well represent the only substantial occupation or special processing sites potentially visible within the archaeological records of the Middle Palaeolithic.

Reindeer processing patterns

The second distinctive pattern documented by Chase at Combe Grenal relates to the assemblages of reindeer remains in some of the later Würm II levels - notably those from the main sequence of Quina-Mousterian levels in layers 25-21 (Fig. 7.5). Bone-element frequencies of reindeer recorded in these levels contrast in several major respects with those of red deer remains from both the Würm I and Würm II levels of the site. In the reindeer remains from levels 25-20, heads (including both upper and lower jaws) are poorly represented, whilst the main upper limbs (humerus, femur and scapula) are more strongly represented than those in any of the earlier levels in the site. In most respects this pattern is almost the reverse of the red deer bones in showing an increase of principal meat-bearing bones and a reduction of more marginal parts, especially those of the head and lower jaws. The only parts of the main meat-bearing skeleton which are conspicuously under-represented are the ribs and vertebrae. As discussed earlier, however, this may have more to do with poor survival of these particular bones than their original absence from the site.

In this case there seems to be a general concensus shared by both Chase (1986a, 1988) and Binford (1984, 1985, 1991) that the exploitation of reindeer in these later Würm II levels must have involved some component of systematic hunting. Binford has suggested, from a preliminary analysis of the age distribution of the remains, that most of the bones were from young individuals, probably killed shortly after birth or as yearling calves, and probably within their late spring and early summer calving grounds, which could have been located directly in front of the Combe Grenal site. According to this hypothesis the heads would probably have been removed at the site of the kills (to reduce the weight of the carcases) while the rest of the carcase was brought into the site, apparently for consumption on the spot. The overall distribution of the bone elements is regarded by both Chase and Binford as providing effectively conclusive evidence for the deliberate hunting of reindeer in this particular segment of the Combe Grenal sequence.

Was this article helpful?

0 0

Post a comment