Sum things are not what they seem: Problems with point-wise interpretations and quantitative analyses of proxies based on aggregated radiocarbon dates

Radiocarbon-date assemblages are commonly used as proxies for past human and environmental phenomena. Prominent examples of target phenomena include past population levels and sea level fluctuations. These processes are thought to have affected the amount of organic carbon deposited into the archaeological and/or palaeoenvironmental record. Time-series representing through-time fluctuations in the frequency of radiocarbon samples are, therefore, often used as proxies for such processes. However, there are critical problems with using radiocarbon ?dates-as-data? in point-wise comparisons and these problems have gone largely underappreciated. The key problem is that the established proxies are easily misinterpreted. They conflate process variation and chronological uncertainty, which makes them unsuitable for point-wise comparisons aimed at identifying rates of change, comparing variables directly, or estimating parameters in regression models. Here we explore the interpretive and analytical problems in detail in an effort to raise awareness and promote skepticism about the use of the established proxies in point-wise comparisons. We also provide suggestions for future research and point to potential methodological alternatives that may improve the viability of dates-as-data approaches.Introduction Interpreting proxies based on aggregated radiocarbon dates - SPDF - Average KDE models Discussion - Analytical problems for point-wise approaches - Implications - Directions for future work Conclusio