Entropology: an information-theoretic approach to understanding archaeological data

Abstract

Abstract This paper presents new tactics for characterizing the relationship between archaeological assemblages, based on entropy and its related attributes, primarily diversity, borrowing heavily from ecology. Our starting premise is that diachronic change in our data is a likely, albeit distorted, reflection of social processes and that spatial difference in data reflects cultural separation. To explore this, we have adopted a null model for comparing assemblage profiles. The modelling is tested on i) a Late Bronze Age Cretan data set compiled by one of us (PG) and ii) a 4th millennium Western Tripolye data set that was analysed earlier. The contrast between the strongly geographically and culturally heterogeneous Bronze Age Crete and the strongly homogeneous Western Tripolye culture in the Southern Bug and Dnieper interfluve show the successes and limitations of our approach. As such, this paper is not primarily about Late Bronze Age (LBA) Crete or Western Tripolye culture per se, although the modelling contributes to our understanding of Cretan archaeology of this period. A fuller discussion of Cretan archaeology and LBA datasets will be given elsewhere. Rather, we use the paper to exemplify problems with archaeological data. Even though we have ‘lots of Cretan data’ (originally 13,000 + artefacts) we cannot consider this as ‘big data’. Due to poor statistics, they only permit non-semantic analysis, particularly important when our aggregation protocols depend on how representative our data is, and whether our assemblages are treated as censuses or samples.