Social Cognition, metacognition, and psychosis
Munich Psychiatry Lecture Series | May 23rd 2017
Metacognition can be defined as the monitoring and control of lower-level cognitive processes. There are at least two levels of metacognition. Implicit metacognition, an automatic, sub-personal process, can often display optimal levels of control. This is not the case for conscious, explicit metacognition. Many features of schizophrenia, such as thought insertion and lack of insight, reflect disorders of metacognition. However, experimental studies suggest that these disorders only emerge at the explicit level of metacognition.
In comparison with implicit metacognition, even in the normal case, explicit metacognition is fragile and error prone. But it has great importance since it provides a channel through which people can tell each other about their cognitive processes. Such two-way discussions can change the way that metacognitive signals are interpreted and used to modify behaviour. These effects require, first, that people can describe their cognitive processes to others and, second, that the reports of others can change these processes. Strikingly, both these mechanisms are impaired in psychosis.
Recent computational models of metacognition involve an architecture in which a higher-level system ‘observes’ the behaviour of a lower-level system and estimates its hidden states. Although discussed in terms of a system that is observing itself, the same mechanisms would be equally applicable when attempting to uncover the hidden states of other people. This close relationship between metacognition and mentalising can also be observed in psychosis where many symptoms reflect false inferences about the mental state of others as well as the self. Studies of explicit metacognition will help us to better understand the symptoms of psychosis.