Too Rational: How Predictive Coding’s Success Risks Harming the Mentally Disordered and Ill

Abstract

The so-called predictive coding or predictive processing theory of mind has attracted significant attention in the brain and behavioral sciences over the past couple of decades. We aim to discuss an important ethical implication of the theory’s success. As predictive coding has become influential in the study of mental disorder and illness, particularly on autism spectrum disorder (ASD) and schizophrenia, we point out a significant risk of further harming an already stigmatized population. Specifically, because predictive coding is undergirded by Bayesian inference, and Bayesian inference is often thought to imply ‘rationality’, the cognitive framework engenders a risk of strengthening existing negative attitudes towards individuals having mental disorders and illnesses by associating such individuals with also having ‘irrational brains.’