Examining the coherence of young children's understanding of causality: Evidence from inference, explanation, and counterfactual reasoning

As adults, we recognize that the world has causal structure. The goal of this dissertation is to investigate the coherence of children's understanding of this causal structure. Chapter 1 explored the coherence of children's inferential abilities. Experiment 1 demonstrated that children could make inferences about new causal relations based on indirect evidence. Further, children produced novel interventions that reflected their understanding of that inference. This suggests that their understanding of a new causal relation is coherently structured and not simply based on recognizing patterns of variance and covariance between events. Experiments 2 and 3 demonstrated that children are not engaging in causal inference by appealing to mechanisms that calculate the associative strength between event types. Instead, children behaved as though they incorporated information about the overall pattern of events to make causal inferences.; Chapter 2 investigated the coherence of children's causal knowledge by considering at what point children are able to make predictions about future hypothetical and counterfactual events. Previous research on children's counterfactual reasoning has treated it as a domain-general ability. The hypothesis that children engage in theory-formation—that children are born with a set of domain-specific theories that they revise throughout development—predicts that understanding the causal structure of a domain (i.e., having a developed theory) allows a child to engage in prediction, explanation, and counterfactual reasoning (Gopnik & Meltzoff, 1997). Experiments 4 and 5 examined children's ability to engage in prediction and counterfactual reasoning over a variety of domains. Children showed little difference between their ability to engage in prediction and counterfactual reasoning within each domain, but dramatically differed in those abilities across the domains. This suggests that once children understand the causal structure of a domain, they can engage in both prediction and counterfactual reasoning.; Following the predictions of a theory-formation hypothesis, Chapter 3 examined whether the same children could provide explanations and reason about counterfactual alternatives to sets of events. In Experiment 6, children were asked to provide explanations for either possible or impossible events that involved physical, biological, or psychological principles. The same children were also asked to generate counterfactual alternatives for similar events. Although children were more capable of explaining an impossible event than correctly recognizing that no counterfactual could be generated for an impossible event, these abilities correlated with each other. Only children who could explain why impossible physical and biological events were impossible could correctly recognize that no alternative could be generated when the character wanted to accomplish an impossible event. Like the results of Chapter 2, these results suggest that once children understand the causal structure of a domain, they can provide explanations and reason about counterfactual events. Overall, the results of each chapter suggest that children's understanding of causal structure is coherent and supports the hypothesis that children engage in theory-formation.