Bayesianism and the perception-cognition divide

Perceptual experience and belief are frequently treated as distinct kinds of mental states. A belief might prompt a new perceptual experience, and new experience can confirm or trigger a belief. Despite causal influences of this sort, it was commonly held that perceptual experience is insulated from the information contained in beliefs. However, recent scientific evidence shows that this picture is mistaken: perception is routinely influenced by beliefs and expectations. This evidence of cognitive penetration thus erodes a strict perception-cognition divide. Two recent approaches to the mind, Bayesianism and Predictive Coding, do further damage to the divide. According to these approaches, influences from cognition on perception are not just pervasive, but integral to its functioning. In this talk I’ll argue that if these two approaches are correct, there is no use in saving divide. Perception and cognition do not exist. Understood as paradigm changes, Bayesianism and Predictive Coding imply eliminativism with respect to belief and experience. They constitute a real revolution in the philosophy of mind, and it is time for philosophers to embrace the change