Unbiased Cultural Transmission in Time-Averaged Archaeological Assemblages
Abstract
Unbiased models are foundational in the archaeological study of cultural transmission. Applications have as- sumed that archaeological data represent synchronic samples, despite the accretional nature of the archaeological record. I document the circumstances under which time-averaging alters the distribution of model predictions. Richness is inflated in long-duration assemblages, and evenness is "flattened" compared to unaveraged samples. Tests of neutrality, employed to differentiate biased and unbiased models, suffer serious problems with Type I error under time-averaging. Finally, the time-scale over which time-averaging alters predictions is determined by the mean trait lifetime, providing a way to evaluate the impact of these effects upon archaeological samples.
Keywords
Cite
@article{arxiv.1204.2043,
title = {Unbiased Cultural Transmission in Time-Averaged Archaeological Assemblages},
author = {Mark E. Madsen},
journal= {arXiv preprint arXiv:1204.2043},
year = {2012}
}
Comments
Submitted to the Journal of Anthropological Archaeology. 4 figures, 3 tables. This paper was originally presented in a symposium titled "Recent Developments in Cultural Transmission Theory and its Applications" at the 2012 Annual Meeting of the Society for American Archaeology, Memphis, TN