Levallois lessons: The challenge of integrating mathematical models, quantitative experiments and the archaeological record

Contemporary scientific archaeology has an array of methodological approaches at its disposal. In addition to a multitude of specialist techniques, this endeavour may, more generally, draw upon data from field survey and excavation, experimental approaches and mathematical modelling. However, the extent to which these different strands of enquiry are adopted may vary widely from researcher to researcher. Lip service is often paid to the notion of 'integrating' different approaches but whether this is genuinely achieved is debatable, while some may ignore one or more particular approach entirely. The study of Levallois artefacts (e.g. flakes and cores) has been an important topic within Palaeolithic archaeology for more than a century. Studies of these artefacts have been implicated in major debates concerning cognitive and behavioural aspects of evolution in hominins. Here, we discuss something of the history of investigation into Levallois, and consider whether insights that have been gained by applying data from artefactual studies, experiment and mathematical modelling might point toward ways in which such alternative approaches might be integrated more closely. Key to this, we argue, is exploitation of the concepts of 'internal validity' versus 'external validity', which are possessed by these contrasting data sources to varying degrees. By emphasizing both the strengths, but also the weaknesses of these different avenues of enquiry, these validity concepts may enable a better sense of how the links between them can be strengthened in archaeological enquiry.