A new approach to identify heat treated silcrete near Pinnacle Point, South Africa using 3D microscopy and Bayesian modeling

•Novel method to identify heat treated silcrete using 3D microscopy.•Surface roughness measurements are lower in heated silcrete than unheated silcrete.•Silicon peels precisely replicate the surface roughness of silcrete flakes.•Our Bayesian model predicts heat treatment with an average accuracy of...

Full description

Saved in:
Bibliographic Details
Published in:Journal of archaeological science, reports reports, 2020-12, Vol.34, p.102622, Article 102622
Main Authors: Murray, John K., Harris, Jacob A., Oestmo, Simen, Martin, Miles, Marean, Curtis W.
Format: Article
Language:English
Subjects:
3D microscopy
Bayesian model
Heat treatment
Silcrete
Silicon peels
Surface roughness
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Novel method to identify heat treated silcrete using 3D microscopy.•Surface roughness measurements are lower in heated silcrete than unheated silcrete.•Silicon peels precisely replicate the surface roughness of silcrete flakes.•Our Bayesian model predicts heat treatment with an average accuracy of 84%.•The model performed better in a blind test with averaged roughness data (90%). The heat treatment of stone to enhance flaking attributes was an important advance in the adaptive toolkit of humans and a major step in pyrotechnology. The earliest evidence for this is the heat treatment of silcrete ~164 ka at the Middle Stone Age site, Pinnacle Point 13B in South Africa. Heating stone prior to knapping alters the physical and chemical composition of the stone. This study investigates whether surface roughness, as measured by a 3D microscope, can be used as a proxy to identify the presence of heat treatment in the archaeological record. We record values for multiple surface texture parameters on a sample of experimentally created stone tools from paired heat-treated and unheated silcrete nodules. A Bayesian probability model, trained on the experimental sample, was then used to evaluate the probability individual samples have undergone heat treatment. Furthermore, we tested whether an industrial silicon compound can be used to record and preserve surface roughness for analysis. This research provides a novel, probabilistic, and non-invasive technique for identifying heat treatment from three sources near Pinnacle Point.
ISSN:2352-409X
DOI:10.1016/j.jasrep.2020.102622