Gaslighting—emotional manipulation that makes individuals doubt their reality—may stem from normal learning processes gone awry, according to a theoretical framework published in Personality and Social Psychology Review. The study draws on prediction error minimization, a theory describing how the brain updates beliefs when expectations conflict with experience, and was developed by lead author Willis Klein, PhD, and colleagues of McGill University and the University of Toronto.
In close relationships, individuals develop epistemic trust—confidence that the other person accurately interprets reality. Klien and colleagues suggest that gaslighters exploit this trust by repeatedly creating “prediction errors,” then offering false explanations suggesting that the target is irrational or incompetent. Over time, the victim internalizes this feedback, loses confidence in their perceptions, and becomes increasingly dependent on the gaslighter.
The model integrates attachment theory, shared-reality theory, and social learning to explain how repeated manipulations erode agency. Researchers identified risk factors for victimization, including anxious attachment, low self-esteem, fragile self-concept, rejection sensitivity, and social isolation. Biologically, the authors highlight possible roles for dopamine, oxytocin, and cortisol—neurochemicals linked to learning, bonding, and stress—that may amplify dependence and blunt self-correction.
Recovery involves rebuilding epistemic self-trust through mindfulness, creative activities, and supportive relationships.
The authors conclude that gaslighting arises as a “function of typical social-cognitive mechanisms operating in atypical social situations, that is, the tendency to prioritize the testimony of close others combined with an epistemically manipulative close other.”
They call for empirical testing using computational modeling and neurophysiological methods to clarify how manipulation reshapes belief and agency.
The study was supported by the Fonds de Recherche du Québec – Société et Culture and McGill University grants. The authors reported no conflicts of interest.
