Auditory-Motor Computations for Timing, Sequencing, and Learning

Auditory perception is tightly coupled to action: we speak and sing with precise timing and continuously adapt our motor output using auditory feedback. This talk will discuss recent work on the auditory–motor loop as a computational system for prediction, sequencing, and control, emphasising how motor planning and movement-related signals shape auditory representations even when the primary task is “listening.” I will outline a framework in which the auditory cortex not only encodes acoustics, but also participates in sensorimotor inference—generating predictions about upcoming sensory input, comparing these predictions to feedback, and updating internal models that support learning and robust behaviour in noise.

We will present evidence and modelling ideas linking (i) fast auditory feature tracking, (ii) temporal expectation and rhythm, and (iii) feedback-based error signals to coordinated interactions among auditory cortical fields and motor-related regions. Finally, I will discuss implications for speech and music, and for NeuroAI/BCI applications where decoding and closed-loop stimulation can benefit from explicitly modelling perception–action coupling.