Philosophy, Psychology and Neuroscience

Explaining Hallucinations Computationally

In this talk, computational explanations of episodes of hallucination are analyzed from the perspective of the mechanistic account of explanation. To make the discussion more specific, I focus on visual hallucinations occurring in people with Charles Bonnet Syndrome. Even if computational explanations, as I argue, need not be representational, and representations are not reducible merely to computational phenomena, there are numerous features of representations that can be explained computationally. To substantiate this claim, I briefly introduce a recent computational model of this hallucination, which relies on generative models in the brain, and argue that the model is a prime example of a representational and computational explanation. I conclude by arguing that computationalism is a natural ally of explanatory pluralism.