Identity-Error: How Social Self-Evidencing Exposes Vulnerability to Falsehoods and Conspiracy Beliefs

ongoing

Understanding how people discern the “truthfulness” of information is of fundamental importance to science. However, the complexity of dynamic information processing in social environments has made it convenient to “silo” research at different levels of analysis (neurological, psychological, social). To overcome this, we need end-to-end modeling that explains when and why people are vulnerable to falsehoods and mistruth. IDENTITY-ERROR pursues a new methodological framework that integrates a social-identity approach with recent advances in computational modeling to socially embed the emotional and cognitive elements of epistemic processing. I propose that the experienced “truthfulness” of information is determined by a recursive process from individual (affect, cognition) to social (identity-structured information). A process of social-self-evidencing fits evidence to models in an identity-congruent manner by error-tuning, either rejecting evidence (common) or models (less common). Identity-imbued models of reality manage error, and in some conditions, objectively incorrect models are very effective error sinks (i.e., they allow good evidence to be easily rejected). This breakthrough will be advanced by harnessing the identity structure of social information to make predictive processing models scalable at the social level. This allows the parallel computation of (i) information reliability alongside (ii) the social costs of accepting it as true (i.e., social evidencing). Importantly, error dynamics in these models enable the integration of emotional influence. IDENTITY-ERROR will establish a new field of psycho-social identity modeling, optimized for understanding the social functions of maintaining falsehoods and conspiracies, and providing a firm basis for socially grounded health and science communication that is resilient to conspiracy theory and misinformation.