Operational Closure, Predictive Coding, and Psychosis: An Evolutionary Framework for Delusion, Hallucination, and Social Cognition

Abstract

Introduction: Predictive coding is an influential framework for psychosis, but most accounts begin at cortical inference and leave implicit the problem that makes prediction necessary: organisms must act in real time while operating only on internally generated states.Methods: This theoretical review integrates theoretical biology, predictive processing, computational psychiatry, the cognitive neuropsychiatry of delusions and social cognition.Results: Nervous systems evolved to coordinate action, not to reconstruct the world. Because perturbations become organism-specific neural states, the nervous system is operationally closed and causally open, and requires an anticipatory model constraining perception, action and belief. Self-monitoring follows as the generic solution for any closed system that acts, grounding comparator accounts. The model is updated in content and calibrated in confidence through two routes of correction: direct sensorimotor coupling, and interpersonal exchange, which operates only on what can be shared. As an appraisal accumulates structural load it becomes less shareable, and so beyond interpersonal correction.Conclusions: Hallucinations, delusions and social-cognitive disturbances appear as disruptions at different levels of a single system that updates and calibrates the model of the world. The account accommodates domain-general updating abnormalities, predicts that structural centrality explains additional variance in resistance to revision, and grounds psychosocial intervention.