Bone grease processing is frequently used in archaeology to investigate human diet breadth because it constitutes a costly mode of lipid procurement. However, problems of equifinality often complicate the identification of this activity. This paper develops new criteria focused on morphology, the presence of micro-inclusions, and forms of damage that were derived from a bone grease rendering experiment that involved red deer(Cervus elaphus)long bones. Because they are poorly represented in a distinct experiment focused exclusively on marrow extraction, the criteria presented here appear to provide robust signatures of bone grease processing. A survey of the actualistic literature shows that certain bone processing activities, such as stewing and soup making, mostly involve coarse spongy fragments. Because these fragments are too large to be ingested, grease can be extracted from them only through heating or boiling. In contrast, bones ingested in bone meal or as flour necessitate pulverization. A high percentage of coarse fragments may therefore provide a proxy for cooking technology or, minimally, the use of fire, if other patterns are consistent with grease extraction. Given the evolutionary significance of these innovations, the criteria presented here may help to strengthen arguments about dietary shifts during the Paleolithic and later periods.

Recent excavations at the cave site of Roc de Marsal (in the Dordogne region of SW France) have yielded several Mousterian assemblages rich in well-preserved faunal remains. The Layer 4 faunal assemblage, associated with a rich Quina Mousterian occupation, provides an opportunity to investigate Neandertal prey selection, transport decisions and reindeer carcass processing strategies. One of the most striking characteristics of the Roc de Marsal Layer 4 faunal assemblage is an apparent lack of spongy parts that cannot be explained by taphonomic processes or carnivore activities. Limb elements were selectively transported to the site, where they underwent considerable further processing, resulting in a notably high degree of fragmentation of almost all elements. Comparison of reindeer carcass processing strategies employed by Neandertals during the Quina Mousterian with those identified for the later periods of the Upper Palaeolithic gives rise to some unexpected interpretations of Neandertal consumption behaviours. We show that this kind of intensive fragmentation is associated with bone grease rendering in ethnographic and Upper Palaeolithic contexts, in most such cases with the aid of fire. However, evidence of fire is remarkably scarce in Roc de Marsal Layer 4. Layer 4 could reflect a systematic Neandertal practice of sucking and chewing on the spongy portions of bones in order to extract as much of the highly nutritious bone grease as possible. Additionally, Roc de Marsal Layer 4 shows examples of use of animal resources for nonalimentary purposes by Neandertals. The abundance of cutmarks on the anterior surfaces of reindeer metapodials seems to indicate the acquisition of tendons, and bones were frequently used as retouchers.

No abstract