Proceedings of the Annual Meeting of the Cognitive Science Society

All causal relations are subject to exceptions. Theories of how causal cognition handles exceptions focus primarily on their frequency or probability, but does the nature of exceptions also influence reasoning? In two experiments, we presented adults with information about two candidate causes. Both produced the same effect in four of six cases. In the remaining two cases, one candidate produced no discernible effect while the other produced a similar but distinct effect. Participants were sensitive to this difference: despite the two candidates' identical probabilities of producing the majority effect, participants had strong preferences about which candidate was acceptable to cite in causal claims (Experiment 1) and employ in causal interventions (Experiment 2). These preferences held even in the absence of generic language and appear to be sensitive to complex and specific intuitions. This study offers novel insights into how causal reasoning understands and navigates exceptions in a rich, conceptual way.