Online Event
Predictive Processing Models of Rhythm Perception and Production
Thomas Mark Kaplan
Presenter: Thomas Mark Kaplan / Queen Mary University of London
Abstract: Rhythm perception and production appear to be reciprocally related, but the underlying neurophysiology and role of the motor system are still poorly understood. Cognitive models offer a predictive processing perspective of rhythm perception, but often do not consider the complications of action (rhythm production). In my work I am developing predictive processing models to analyse the relationship between rhythm perception and production, focusing on (1) the reference structure and (2) expressive timing in musical rhythms. Firstly, I will describe an action-oriented predictive processing model of how overt rhythmic actions (as in motor entrainment) influence meter perception, and present some preliminary analysis against empirical observations. Secondly, I will briefly describe a dynamic predictive processing model inspired by sensorimotor synchronisation studies, which will be used to test theories of nuanced beat-based expectations---where the distribution of beat onsets at the micro level of a rhythm influences metric expectations. (edited)