Paleolithic archaeologists study regional variation among assemblages of stone tools in order to delineate cultural boundaries and reconstruct mechanisms of cultural transmission in the deep past. Structured population models are especially suited to aid in this endeavor, for they teach us how cultural evolutionary forces—copying error, intergroup transmission, drift, and selection imposed by functional constraints or biased cultural transmission—affect regional cultural variation. We use an agent-based model to address how copying error, intergroup transmission, and time-averaging affect the degree to which regional archaeological assemblages differ at a selectively neutral discrete trait passed from “experienced” to “naïve” individuals via one of four mechanisms of cultural transmission in a structured population of toolmakers. The results of our simulation experiment illustrate why researchers who use time-averaged archaeological data to identify past cultural boundaries or infer mechanisms of cultural transmission should be more mindful of the nature of the cultural trait(s) available for study. In light of our results, we discuss seven questions archaeologists ought to address before attempting to infer cultural boundaries or cultural transmission mechanisms from between-assemblage variation.

旧石器时代考古学家研究石器组合之间的区域差异，以划定文化边界并重建深远过去的文化传播机制。结构化的人口模型特别适合帮助这项工作，因为它们告诉我们文化进化力量——复制错误、群体间传播、漂移以及功能约束或有偏见的文化传播所施加的选择——如何影响区域文化差异。我们使用基于代理的模型来解决复制错误、群体间传递和时间平均如何影响区域考古组合在选择性中性离散特征的差异程度，该特征通过结构化工具制造者群体中的四种文化传播机制之一从“经验”个体传递到“天真”个体。我们的模拟实验结果表明，为什么使用时间平均考古数据来识别过去的文化边界或推断文化传播机制的研究人员应该更加注意可供研究的文化特征的性质。根据我们的结果，我们讨论了考古学家在试图从组合之间的差异中推断出文化边界或文化传播机制之前应该解决的七个问题。