Overview
Circular inference, not to be confused with circular reasoning, is a mathematical model of information-processing defects in the brain where internal predictions are processed as external sensory evidence instead of balancing each other. The concept is used often in interpreting the neurological basis of disordered thinking patterns found in conditions like schizophrenia.
In a typical brain, bottom-up sensory information is gathered and passed on as raw data to be interpreted by regions responsible for holding a model of the world based on memory and context, like the prefrontal cortext. If we step outside after a long day and our senses tell us that the sky is dark and the air is chilly, our higher-level models of the world inform us that this is because it’s nighttime and the sun has set.
Conditions like schizophrenia lead to a breakdown in this process as the neurons responsible for inhibiting a message after its processing fail to function correctly. Excitatory pathways that amplify messages and signals continue to pass input around. If someone steps outside and expects the sun to be up because they believe it should still be daytime, the brain interprets this as a fact rather than an expectation from an internal model, so that person might be more likely to think that something has gone wrong with the sun or passage of time itself. As a less extreme example, someone who feels anxious in a room might generate expectations to locate the source of the stimulus creating anxiety, like the suspicion that someone across the room taking notes is actually logging their movements. Rather than being evaluated against context in the way cognitive therapies encourage challenging irrational distortions, this message loops repeatedly through the brain and registers as empirical evidence each time. This can produce a very strong and sincerely held belief as the brain struggles to distinguish thoughts generated through its internal models from sensory perceptions.
History and Etymology
The term was formally defined by Renaud Jardri and Sophie Denève in a 2013 paper published in the journal Brain called Circular Inferences in Schizophrenia.