Modelling cultural evolution processes in Dressel 20 amphora production during the Roman Empire

Abstract

How people interact socially and transmit their knowledge from generation to generation leads to different habits and traditions. When these traditions are about pottery-making techniques, these differences may have left a quantifiable trace in the remaining cultural material. However, linking the differences observed in the archaeological record with social learning strategies and how production techniques were transmitted is extremely difficult. In archaeology, this has been studied in handmade production but barely analysed in large-scale production implying long-distance social interaction during an extended period. The goal of this study is to explore the transmission of technical skills among potters within the Roman Empire. Specifically, our case study has focused on the Dressel 20 amphora production processes based in Baetica province (currently Andalusia). To achieve this, we (1) compute the covariance matrix between morphometric measurements and use it to define the core rules of pottery-making techniques, (2) define scenarios in an Agent-based model that represent hypotheses about how techniques were shared between potters, and (3) quantify how likely each scenario is to reproduce the correlation between morphometric similarity of the pottery and the distance between the workshops observed empirically. Our analysis highlights that, while the spatial distribution of workshops influences social interaction among potters, it does not act as a strict barrier. Even workshops that are relatively distant can exhibit some morphometric similarities, suggesting that knowledge and techniques can be transmitted over larger distances. However, to obtain the level of correlation between distance in space and closeness in shape observed in the real world, geographical constraints – represented in our case by riverine connections – need to play a role in limiting interactions. Finally, we believe this method provides a versatile framework to explore morphometric differences related to the transmission processes in a large-scale production.