Let’s cut to the chase and rely on fracture mechanics to identify projectile armatures

Abstract

Stone armatures are privileged archaeological evidence to inform us about Palaeolithic hunting technology, and macro- and microscopic traces have been demonstrated to be reliable indicators of projectile use. Researchers have used a broad range of criteria for the identification of projectiles, including morphometrics, diagnostic impact fractures or microscopic wear and fracture evidence. An identification as a projectile should not rely on purely morphological grounds, but opinions diverge concerning the minimal criteria that are required for reliable identification, and a unified approach is still lacking. As a result, projectiles are identified on criteria that strongly diverge between researchers hindering comparisons and an understanding of broader-scale patterning. In this paper, we seek to address this difficulty, and we propose a revised approach that relies on the principles of fracture mechanics. We propose to characterise fractures with attributes instead of fracture types and operate with a scoring system that allows the interpretation of the observed patterns. Drawing upon an experimental dataset that also includes blind tests, we demonstrate that the proposed approach provides a reliable way to identify projectiles. In comparison to approaches focused on diagnostic impact fractures or morphological criteria, our proposition offers an alternative that searches to improve both precision and reliability. The approach is not aimed at identifying all projectiles within an assemblage, but to isolate a reliably identified subset that permits subsequent explorations of other parameters such as the propulsion mode used. We argue that the approach is robust and provides a fundamental step towards a better understanding of the evolution of Palaeolithic weapon technology.