Neural Nets for Style: A Method for the Examination of Material Culture Variation

Summary

This is an abstract from the "SAA 2019: General Sessions" session, at the 84th annual meeting of the Society for American Archaeology.

The cause of morphological variation in material culture has long been debated. This investigation into archaic projectile point variation from the Gault site in central Texas looks through the lens of social learning to suggest that different teaching and learning strategies represent the root cause of variation. These strategies may in turn reflect part of the social function of the items being made. Specifically, some of these objects may serve the purpose of signaling information to other people, while other forms may simply be utilitarian with no symbolic value. It is proposed that projectile point forms that show relatively less variation may represent objects with a symbolic purpose. The chronologically controlled forms are quantified into ideational units by training an artificial neural network to recognize patterns in the morphological data. These patterns can be compared to the data of each individual specimen in the groups to determine a level of intra-group variability for each group relative to each other group. It is hypothesized that the variation among these groups will display a bimodal distribution with one mode representing a group with a symbolic aspect, while the others is simply utilitarian.

This Resource is Part of the Following Collections

• Society for American Archaeology 84th Annual Meeting, Albuquerque, NM (2019)

Keywords

Spatial Coverage

Record Identifiers

Abstract Id(s): 24991