INFERENCE FROM LARGE SETS OF RADIOCARBON DATES: SOFTWARE AND METHODS

Enrico R Crema,Andrew Bevan

doi:10.1017/rdc.2020.95

Enrico R Crema, Andrew Bevan

Open Access

https://doi.org/10.1017/rdc.2020.95

Copy DOI

Abstract

ABSTRACTThe last decade has seen the development of a range of new statistical and computational techniques for analysing large collections of radiocarbon (14C) dates, often but not exclusively to make inferences about human population change in the past. Here we introduce rcarbon, an open-source software package for the R statistical computing language which implements many of these techniques and looks to foster transparent future study of their strengths and weaknesses. In this paper, we review the key assumptions, limitations and potentials behind statistical analyses of summed probability distribution of 14C dates, including Monte-Carlo simulation-based tests, permutation tests, and spatial analyses. Supplementary material provides a fully reproducible analysis with further details not covered in the main paper.

Highlights

The last few years have seen a dramatic increase in the number of research projects constructing proxy time series of demographic change out of large lists of archaeological radiocarbon (14C) dates
This approach assumes that, given a large enough set of 14C dates taken on anthropogenic samples, the changing frequency of dates through time will preserve a signal of highs and lows in past human activity and, by extension, in human population
Rick’s (1987) work was pioneering in this regard, being the first to propose the key assumption that more people in a given chronological period would typically lead to more anthropogenic products entering the archaeological record in that period, implying more potential samples to date and more published 14C dates

Summary

Introduction

The last few years have seen a dramatic increase in the number of research projects constructing proxy time series of demographic change out of large lists of archaeological radiocarbon (14C) dates. In the case of SPDs, the probability distributions associated with each 14C date have different shapes depending on measurement error and the particularities of the relevant portion of the calibration curve.

Results

Conclusion