Looking back to plan ahead: Causal judgments as a sampling approximation for action effectiveness

Throughout human thought and discourse, we make judgments of how much certain particular events caused others: For instance, we judge that a product sold because of its viral ad campaign more than because of its celebrity endorsement, or vice versa. Yet, the precise functional role of such judgments remains elusive. Why do people make and care about these judgments so much, and why do those judgments follow the patterns they do? We show that core properties of actual causal judgment—its focus on rare antecedents with common counterparts, its centralization of necessity, and its dependence on actual as opposed to hypothetical features of the target situation—make it an ideal candidate for a particular approximate planning algorithm. Specifically, we propose that judgments of whether something (e.g. a viral ad campaign) was the cause of an outcome (e.g. robust sales) in past contexts help cumulatively identify effective ways of bringing about an outcome in general future contexts. We offer a formal account of this process and show how it unites seemingly disparate patterns of actual causal judgment in a common functional role.