Bayesian inference of material culture phylogenies using continuous traits: A birth–death model for Late Neolithic/Early Bronze Age arrowheads from Northwestern Europe

Abstract

The twentieth century saw two modes of thinking spread through biological systematics: population thinking and tree thinking. Many archaeologists investigating ancient technologies have adopted the population thinking approach for archaeological systematics, as evident from the rise of work on quantifiable, attribute-based patterns of diversity in studies of past technological systems, much of it driven by application of geometric morphometrics. Yet, tree thinking, as one of the central concepts of phylogenetic biology, remains rare in the archaeological sciences. We review some of the obstacles that have impeded the uptake of this concept by archaeologists. Some of these are conceptual, but we also identify one specific methodological obstacle: the prevailing use of discrete character traits in phylogenetic analysis. This is a major challenge for archaeological applications where standardised trait-analytical protocols are scarce. To address this challenge, we present a case study that demonstrates a Bayesian framework for inferring phylogenies using continuous traits derived from artefact shape coefficients obtained via outline based geometric morphometrics. We use a previously published sample of Late Neolithic/Early Bronze Age arrowheads from Northwestern Europe to demonstrate the efficacy and accessibility of our approach. We also sketch out the potential for phylogenetic comparative methods to address archaeological questions. From https://github.com/benmarwick/bayesinfmatcultphylo

A macroevolutionary analysis of European Late Upper Palaeolithic stone tool shape using a Bayesian phylodynamic framework

“…By the same token, and as already noted by [29], Late Upper Palaeolithic armature outlines may in general be relatively poorly suited for these kinds of analyses due to the overall simplicity of the shapes and the seeming lack of normativity in form imposition at this time. In contrast, the results of a similar study using the outline shapes of Late Neolithic/Early Bronze Age arrowheads deliberately shaped in the context of prestige display, which lead to indicate overall higher posterior clade probabilities [58].…”

“…Building on previous work [58], we here take initial exploratory steps towards the Bayesian phylodynamic inference of stone tool phylogenies. A vital contribution here lies firstly in the benchmarking of these computationally demanding methods, the different models, the data, and their interaction.…”