Statistical modelling of artefact compositional data

Model-based methods for clustering artefacts, given their chemical composit ion, often assume sampling from a mixture of multivariate normal distributi ons and/or make explicit assumptions about the way in which a composition i s formed. It is argued that, analysed within a modelling framework, several important and apparently competing methodologies are more similar than wou ld initially appear. The opportunity is taken to note that models for popul ations are often not compatible with models for compositions, and that dilu tion correction--which can be accomplished in a variety of ways--can be int erpreted as an attempt to resolve this problem.