Clues to stone tool function re-examined: Comparing starch grain frequencies on used and unused obsidian artefacts

The identification of plant residues observed on prehistoric stone artefacts has often been used to inform on tool function without adequate consideration of potential contamination due to post-depositional processes. The paper redresses this balance by proposing a methodology which combines use-wear and residue analyses with systematic testing of depositional context to distinguish residues formed during tool use from accidental contamination. A case study involving obsidian artefacts from an open site in Papua New Guinea is used to illustrate the new approach. Starch grains, a class of residue which has had little attention outside Australia, were chosen as the focus of study because of their abundance at this site. The frequency of starch grains extracted from residues on stone artefacts was compared with those in sediments adhering to and adjacent to the artefact. A use-wear study conducted as a blind test provided an independent measure of whether the artefacts had been used. The frequency of starch grains was shown to be significantly correlated with used artefacts and not correlated with unused artefacts or the sediments. The case study demonstrates the importance of testing for contamination and also illustrates the benefit of studying starch grains as a class of plant residue.