Abstract

Research on the neurocognitive underpinnings and disruptions of the sense of agency has been greatly inspired by a simple computational account of agency: the comparator model. However, this account faces several shortcomings in explaining agency in healthy and neuropsychiatric subjects and in explaining the distinction between feelings of agency and judgments of agency. This chapter suggests that integration of agency cues might be governed by a specific weighting mechanism, namely a weighting mechanism that follows the principles of optimal cue integration. According to this notion, internal action-related cues, external action-related cues, and action-independent cues are each constantly weighted according to their reliability in a given situation. Also prior information on one’s action is incorporated, probably along Bayesian principles. Such a process would provide the basis for a much more flexible and at the same time more robust mechanism than proposed by previous neurocognitive accounts of agency.