Fixed and flexible perceptual rhythms.

Abstract

Our sensory inputs are never identical across time and contain temporal structure. Cognitive scientists have recently been fascinated by how these sensory rhythms interact with neural oscillatory rhythms to dictate perception. However, there are parallel lines of enquiry that propose clear, but apparently incompatible, answers to this question. One largely suggests that neural oscillations are flexible and adjust to the temporal structure of the external world. Another suggests that they are fixed by intrinsically determined processes, like our own motor routines or local neural architectural constraints. Here, we highlight how the literature must now crucially ask how rhythmic sources are combined to determine sensory and, in turn, motor processing. To this end, we offer a model based around statistical (Bayesian) learning principles.