Explaining the 'Klasies Pattern': Kua ethnoarchaeology, the Die Kelders Middle Stone Age archaeofauna, long bone fragmentation and carnivore ravaging

The caves at Klasies River Mouth (KRM) in the southern Cape Province, South Africa, continue to play an important role in the study of human evolution and debates about the origins of 'modern' humans. As with most Palaeolithic sites, the behavioural interpretations offered for the Middle Stone Age (MSA) inhabitants of the KRM caves are based in large part on the interpretation of fragmentary faunal remains from the 'Main Site', or caves 1, 1A, 1B and 2 excavated in 1967-68 by Singer and Wymer (1982). Using zooarchaeological patterns detected in ethnoarchaeological data collected among Kua foragers of the eastern Kalahari, Central District, Botswana, as well as in data collected from Die Kelders Cave 1, Layer 10, South Africa, we offer an alternative explanation for a provocative pattern of skeletal part representation discovered in the KRM MSA archaeofauna over 20 years ago by Richard Klein (1976). The Klasies pattern is characterized by the underrepresentation of large bovid upper limb elements relative to more anatomically complete small bovids. To date, the pattern has been interpreted in three different ways, each with strikingly different implications for Middle Stone Age (MSA)/Middle Palaeolithic (MP) human behaviour. Understanding the nature of the pattern is not only important for reconstructing MSA human behaviour at KRM, but it is also important for understanding how the conduct of zooarchaeological analyses can profoundly influence our reconstructions of the course of human behavioural evolution. The Kua ethnoarchaeological data and experimental case studies unambiguously connect to site formation histories to zooarchaeological patterns. At Die Kelders Cave 1, Layer 10, the ethnoarchaeological expectations are expressed archaeologically in a context analytically relevant to KRM. Overall, by joining the Kua and the Die Kelders data, we show how uniformitarian/actualistic approaches can be productively employed in zooarchaeological interpretation. The 'Klasies Pattern' is likely to be an artefact of a widespread taphonomic sequence, followed by a particular style of analysis employed, for practical reasons, by many zooarchaeologists. We suggest that the situation is common and likely to be seriously distorting inferences about many archaeofaunas around the world. More generally we show how uniformitarian studies offer a coherent framework for developing methods and generating data useful in making behavioural inferences from faunal remains. Rather than a human behavioural signal in the KRM archaeofauna, the Klasies Pattern has more to do with the methods of zooarchaeology than with the behaviour of ancient people.