Proceedings of the Annual Meeting of the Cognitive Science Society

Current causal theories argue that the statistical normality or abnormality of an action makes a difference to people’s causal judgements. In this paper, we present two experiments that explore the role of statistical norms in causal cognition. In our first experiment, we provide a preliminary test of two competing theories that aim to explain the effects of normality in causal cognition – the actual causal strength measure (Icard Kominsky & Knobe, 2017) and the correspondence hypothesis about causal judgements (Harinen, 2017). In addition, we control for an often neglected factor, the epistemic states of agents. Our second experiment investigates the effect of statistical normality in the same context, but with a probabilistic rather than deterministic causal structure. Our results favour Icard et al.’s (2017) model of causal strength, but show that the statistical normality of an action loses its influence when the occurrence of the outcome is probabilistic. We discuss the implications of our findings for current causal theories