Outstanding Puzzles for Predictive Coding Explanations of Delusions
One of the most exciting breakthroughs in recent cognitive science has been its newfound recognition of a fundamental organising principle of cognition: namely, that the brain is fundamentally in the business of inferring what’s going on around it, by predicting its own sensory inputs. At all scales of neural processing, a wealth of top-down prior expectations are brought to bear on inferentially ‘explaining away’ the pattern of bottom-up activity. Bayes’ theorem gives us a mathematical rule for optimising inference, by combining prior knowledge with new information according to the relative reliability of each – and dopamine gives us a neurochemical mechanism by which this relative reliability is coded.
Disturbances to dopamine-mediated predictive coding may cause psychosis – a condition in which the individual becomes trapped in a prison of their brain’s own inferences, which have somehow become unyoked from external reality. But since predictive coding theories of psychosis appeal to a pervasive dysfunction in a general, overarching principle of cognition, they struggle to account for the surprising regularity with which certain highly-specific themes and features crop up again and again in the phenomenological content of psychotic experience. I will describe the many explanatory successes of predictive coding theories of psychosis, before discussing how my own research attempts to tackle the pressing challenges they still face in accounting for the facts that delusions overwhelmingly tend to be unpleasant, social, and bizarre.
Bio: I studied Natural Sciences at Cambridge, specialising in Experimental Psychology for my BA in 2013 and completing the MSci in History and Philosophy of Science the following year. In 2014 I applied to do a PhD investigating the development of psychotic symptoms (particularly delusions) in the context of schizophrenic illness. I received funding from the Wellcome Trust NeuroScience in Psychiatry Network, and started my PhD later that year under the supervision of Prof Paul Fletcher and Dr Graham Murray.