On the Use of Parsimony in Archaeological Reasoning

Abstract When evaluating competing hypotheses in archaeology, researchers frequently invoke the principle of parsimony, which states that simpler hypotheses should be preferred. However, the criteria for measuring simplicity and the rationale for labeling a hypothesis as most parsimonious often remain unclear. More broadly, the epistemic merit of parsimony and its relevance to archaeological reasoning are generally assumed but rarely clarified. This article explores how archaeologists use parsimony in ethnographic analogy and formal model selection. In the first context, it is usually uncertain how simplicity should be measured or why increased simplicity raises the plausibility of an ethnographic analogy. We contend that discussions of ethnographic analogy are better characterized under inference to the best explanation, where parsimony is only one heuristic among others. In the second context, simplicity is assessed by the relative complexity, rather than the quantity, of postulates in a model. This approach to parsimony, which does track plausibility under certain conditions, helps prevent false positives in archaeological interpretation. However, it also heightens the risk of rejecting alternative, complex causes. We argue that parsimony can aid in evaluating the relative likelihood of competing models and, more importantly, serve as a guide to clarify the complex histories of archaeological phenomena.