Predictive Processing and its Role in Explaining Cognition

Predictive processing has become one of the most influential explanatory frameworks in philosophy of mind, neuroscience, and the cognitive sciences. Following the tracks laid by the revisionist ideas of von Helmholtz (1821–1894), the brain is taken as being a prediction machine first and foremost. The brain generates memory-based predictions that are adjusted on the fly. If the need arises, according to concurrent sensory sampling, the brain minimises a specific type of error, resulting from a mismatch between the expected sensory input and actual information impinging on the senses. Under this perspective, top-down predictions in fact establish perception, action and learning, whereas bottom-up sensory input assumes the secondary role of providing feedback to correct and fine-tune predictions as needed. Theories of predictive processing offer a unifying framework for understanding the full range of behavioural and cognitive phenomena, which are seen as falling under the umbrella of minimising prediction error. As such, the predictive processing framework has become central to psychological and neuroscientific investigations, and has generated a vast amount of empirical research. The philosophical implications of this framework are currently being hotly debated, including the idea of bottom-up formation of representation, the free will problem and the enactive/extended/embedded/embodied approach to studying mind and cognition.