Abstract
<p>Pain is a private, subjective experience, yet observers routinely make consequential judgments about another person’s pain (e.g., whether it warrants treatment, accommodation, or whether it’s real at all). How do observers make these judgments? We propose that people attribute pain through an intuitive theory—the same kind of theory they use to infer other hidden mental and physical states. We explore the structure of people’s intuitive theory of pain by formalizing competing conceptions of pain: from pain resulting from bodily damage observable through physical evidence to pain as a subjective experience known only via self-report. Each theory is formalized as a causal model that predicts how people interpret naturalistic vignettes drawn from Reddit posts and medical training materials. Three experiments probe classical signatures of Bayesian causal reasoning: predicting pain from physical evidence (Experiment 1), inferring and treating pain (Experiment 2), and inferring a hidden cause (ulterior motives) that explains away inflated pain reports (Experiment 3). Across all three, a causal model that integrates over physical and mental states best captures human behavior and generalizes across eight categories of bodily pain, from headaches to finger burns. These results are consistent with our proposal that people have an intuitive theory of pain arising from problems in the body, and this theory supports quantitative causal reasoning. As a direct consequence of this intuitive theory, lay people view pain reports that lack corroborating physical evidence with suspicion. This work is a first step toward a quantitative, scientific model of folk intuitive theories of pain.</p>