Accurately reconstructing past human population dynamics is critical for explaining major patterns in the human past. Demand for demographic proxies has driven hopeful interest in the “dates-as-data” approach, which models past demography by assuming a relationship between population size, the production of dateable material, and the corpus of radiocarbon dates produced by archaeological research. However, several biases can affect assemblages of dates, complicating inferences about population size. One serious but potentially addressable issue centers on landscape taphonomy — the ways in which geologic processes structure the preservation and recovery of archaeological sites and/or materials at landscape scales. Here, we explore the influence of landscape taphonomy on demographic proxies. More specifically, we evaluate how well demographic proxies may be corrected for taphonomic effects with either a common generalized approach or an empirically based tailored approach. We demonstrate that frequency distributions of landforms of varying ages can be used to develop local corrections that are more accurate than either global corrections or uncorrected estimates. Using generalized scenarios and a simulated case study based on empirical data on landform ages from the Coso Basin in the western Great Basin region, we illustrate the way in which landscape taphonomy predictably complicates “dates-as-data” approaches, propose and demonstrate a new method of empirically based correction, and explore the interpretive ramifications of ignoring or correcting for taphonomic bias.

准确重建过去的人口动态对于解释人类过去的主要模式至关重要。对人口代理的需求激发了人们对“日期即数据”方法的兴趣，该方法通过假设人口规模、可测定日期的材料的生产以及考古研究产生的放射性碳日期语料库之间的关系来模拟过去的人口统计学。然而，一些偏差可能会影响日期的组合，使关于人口规模的推断复杂化。一个严重但可能解决的问题集中在景观埋藏学上，即地质过程在景观尺度上构建考古遗址和/或材料的保存和恢复的方式。在这里，我们探讨了景观埋藏学对人口代理的影响。更具体地说，我们评估使用通用的通用方法或基于经验的定制方法可以在多大程度上纠正人口代理的埋藏学效应。我们证明，不同年龄地貌的频率分布可用于开发比全局校正或未校正估计更准确的局部校正。我们利用广义情景和基于大盆地西部地区科索盆地地貌年龄经验数据的模拟案例研究，说明了景观埋藏学如何使“日期即数据”方法变得复杂化，提出并论证了一种新的方法。基于经验的校正方法，并探讨忽略或校正埋藏学偏差的解释后果。