Arithmetic methods for exploring archaeological source provenance using geochemistry – a short reply

Abstract

In this reply, we respond to the comment by Pearce and colleagues regarding our recently proposed arithmetic method for assessing geochemical similarity between lithic artefacts and potential source outcrops. We thank the authors for their critique, which highlights several important limitations and opportunities for methodological refinement. We agree that issues relating to analytical precision, numerical rounding, certified reference material selection, and the weighting of geochemical variables warrant further consideration. We note, however, that several aspects of our work have been misrepresented. In particular, Pearce and colleagues imply that our study advocates the use of geochemical similarity calculations in isolation from petrographic and other geological evidence. As we demonstrate, we explicitly rejected such an approach and instead presented the method as an exploratory tool for identifying geochemical similarity. We also show how some of the criticisms made by the authors are inconsistent with practices adopted in their own studies, including source provenance interpretations based primarily on geochemical datasets. We defend the use of trace element/Zr ratios for silcrete provenancing, discuss the rationale behind element selection in our worked examples, and clarify our treatment of compatible and incompatible elements in igneous rocks. Using case studies, including investigations into the provenance of Stone 62 and the Altar Stone at Stonehenge, and Mediterranean volcanic artefacts, we illustrate the problems arising from the use of selective bivariate plots for source determination – as recommended by Pearce and colleagues – and highlight the value of multivariate approaches. We conclude that our arithmetic method remains a useful and adaptable framework for provenance assessment when applied critically alongside complementary geological and archaeological evidence.