It is commonly believed that expressions visually signal rich diagnostic information to human observers. We studied how observers interpret the dynamic expressions that people spontaneously produced during a real-life high-stakes televised game. We find that human observers are remarkably poor at recovering what events elicited others' facial and bodily expressions. Beyond simple inaccuracy, people's causal reasoning exhibits systematic model-based patterns of errors. We show that latent emotion representations can explain people's reasoning about the unseen causes of expressions. A hierarchical Bayesian model simulates which events people infer to be the cause of others' expressions by comparing the emotions inferred from the expressions against the emotions people were predicted to experience in various situations. This causal model provides a close, parameter-free fit to human causal judgments, suggesting that humans interpret expressions in the context of emotion predictions generated by a causally-structured mental model of other minds.

Much of social cognition involves reasoning about others' minds: predicting their reactions, inferring their feelings, and explaining their behavior. By representing mental contents like beliefs, desires, and emotions, Bayesian theory of mind mod- els have made progress in capturing how humans manage these cognitive feats. But social life is not merely observation: hu- mans must also plan to intervene on these same mental con- tents. The present work models how people choose interven- tions to influence others' emotions. Building on a prior model of people's intuitive theory of emotions, we model how people use their intuitive theory to evaluate and simulate the effects of different interventions. We apply our model to data from behavioral experiments requiring counterfactual and joint in- terventions, and show a close alignment with human choices. Our results provide a step towards a potentially unifying expla- nation for emotion prediction and intervention, suggesting that they could arise from the same underlying generative model.