| The ability to predict future events in the environment is an essential part of human cognition. We are continuously anticipating what we will hear, see, and feel next. For example, when listening to someone speak, we often infer what the person is going to say before he or she is finished speaking. Predictions occur at all levels, ranging from perception of simple visual shapes or sounds to complex processes like language and social cognition. It has been argued that prediction is not just one of the things the brain does, but rather its primary function. How the brain implements these predictive mechanisms is however still largely unknown. A recent account, termed predictive coding, suggests that the brain sets up prior expectations at each level within the cortical hierarchy. According to this model, incoming information traversing the brain is compared at each stage with the prior expectation. Predictions travel backwards from higher-order association areas to lower-level sensory areas, while prediction errors lead to enhanced forward flow from lower-level to higher-level brain regions. The aim of this project is to empirically test this hypothesis within the domain of visual perception. Using a multi-pronged approach, I propose to study how prior information influences perception. Five projects are centered on a single experimental paradigm that establishes a set of expectations using implicit learning of the regularities in a sequence of visual stimuli. Using this paradigm and a set of complementary brain imaging methods, I will investigate the localization and timing of brain regions that are involved in generating predictions, assess the direction of prediction-related activity and establish causal relationships between prediction and perception. The aim is to discover what may be the ultimate and most common of all brain functions: the capacity to predict the future. |
