Paul Fletcher: Perception and Deception in the Psychotic Brain, Part 1/6

Summary

Paul Fletcher argues that understanding delusions requires connecting measurable brain activity with the subjective experience of holding an unusual belief. He highlights two related challenges:

• The explanatory gap: Finding an unusual pattern of brain activity does not, by itself, explain how that activity gives rise to a particular delusion—for example, believing that a television is talking about oneself.
• Interpreting brain changes: An observed brain difference might be a cause of a delusion, a consequence of it, or a compensation for the person’s experience. Brain activity and belief therefore cannot be interpreted in isolation.

Fletcher describes competing explanations of delusions. One view holds that unusual sensory experiences lead to unusual—but otherwise intact—inferences. Another proposes that perception is normal but inference or belief formation is impaired. A two-factor account combines both: abnormal sensations help explain a delusion’s content, while abnormal inference helps explain why the belief is maintained. Fletcher argues that perception and inference should instead be understood as interacting processes.

Perception, inference, and prediction error

• What a person already knows and expects influences what they perceive; perception is not simply a passive recording of the world.
• Prediction error—a mismatch between expectation and incoming information—helps the brain revise its conclusions.
• Fletcher proposes that:
  • When prediction error is low, prior knowledge and expectations have more influence on perception.
  • When prediction error is high, perception has more influence on what the person comes to believe.

Problems in processing prediction errors may provide a useful way to investigate delusions.

Fletcher presents Bayesian inference as a useful model for how the brain works out the causes of sensations. Sensory signals are noisy and do not provide direct access to their causes. The brain must therefore combine the likelihood that a given cause would produce a sensation with the prior probability that the cause is present. For example, paw prints in central Stockholm are more likely to come from a dog than from a cheetah.

Main conceptual points

• Sensations are imperfect signs of events, not direct or complete representations of the world.
• Because the brain needs to predict what will happen—including what may bring reward or punishment—it must infer causes from uncertain sensory information.
• Bayesian theories, Fletcher suggests, help explain how prior beliefs and new evidence jointly shape perception and belief.
• Studying the interaction of perception, inference, and prediction error may be more informative than treating perception and belief formation as separate processes.

Fletcher says he will illustrate the framework with published findings and newer ketamine studies in healthy participants, though those results have not yet appeared in this excerpt.

Speakers and sources mentioned

• Paul Fletcher — the speaker.
• Fredrik — thanked by Fletcher; apparently involved in the event, but does not speak in the excerpt.
• Predrag — referenced as having just discussed related material; does not speak in the excerpt.
• David Lewis — cited in connection with the warning to “beware of the Three Cs.”
• Maher — cited for the view that unusual sensory experience can lead to delusional beliefs.
• Richard Bentall — cited in relation to abnormal inference accounts.
• Filippo Gareti — named in the subtitles as an expert associated with the Institute of Psychiatry, in connection with abnormal inference accounts.
• Max Coltert — named in the subtitles for the two-factor explanation of delusions.
• Hermann von Helmholtz — cited for the idea that sensations are signs of events rather than direct representations of the world.
• Stephen Lorre — mentioned as the lecturer for another event attendees might wish to attend; no contribution to this excerpt is shown.