Self-Other Prediction Error: How Individuals Process Others’ Behavior Based on Perceived Alignment

The present report argues that the pursuit of social alignment drives all aspects of how individuals evaluate and respond to another person’s behavior. Social/moral judgment – whether you see another person as good and their behavior as appropriate – can be framed as an evaluation of whether the person’s values align with yours and whether they acted the same way you would. Following negative judgments, your responses can be understood as attempts to reestablish alignment, such as by conforming to this person’s behavior or punishing it. From this perspective, the present research proposes the Self-Other Prediction Error (SOPE) model, which conceptualizes these mechanisms within frameworks of predictive coding, Theory of Mind, and dual-process decision-making. Recent studies experimentally testing this model validate its core arguments and show that SOPE predicts participants’ reactions to others’ behavior more accurately than earlier theories focusing on norms or utility. Furthermore, reinterpreting older studies suggests that the proposed model successfully reconciles a wide array of phenomena concerning moral psychology, behavioral economics, and other social cognition topics. Hence, the proposed SOPE account is posed as a foundation for understanding social information processing generally and its computational underpinnings.