Listening as active process – probing a rhythmic mode of listening

• Applicant: Professor Dr. Christoph Kayser
• Subject Area: Biological Psychology and Cognitive Neuroscience
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 502207424

Human listening behaviour supposedly relies on a so-called rhythmic mode of perception. This notion has been introduced based on two lines of evidence: first, based on neurophysiological recordings in animals demonstrating rhythmic changes in the excitability of auditory cortices; and second, based on psychoacoustic studies showing that our abilities to detect or discriminate sounds can vary with an apparent rhythmicity relative to a reference time point. The notion of a rhythmic listening mode rests on the general idea that perception is an active process that can switch between distinct modes of operation and which relies on the interaction between sensory and motor systems. Yet, and despite its prominence in the literature, previous work leaves central questions around the emergence and robustness of a rhythmic listening mode and its relation to active behaviour unresolved. We here address the following timely questions using psychoacoustic studies in humans, eye tracking and computational modelling. First, and using previously established tasks, we will provide robust and reliable data on the rhythmicity of perceptual reports in challenging listening tasks in the absence of entraining sounds, in order to understand the conditions under which a putative rhythmic mode emerges. Then, we test the prediction that a rhythmic mode can be actively engaged or disengaged using experimental manipulations, and use pupillometry to link listening modes and listening performance to participant’s vigilance. We then probe how spontaneous and instructed oculomotor behaviour relates to listening modes and performance, in order to understand whether listening modes are indeed directly linked to active behaviour as postulated by active sensing accounts. And finally, we employ computational modelling of the behavioural data to dissociate sensory- and decision-level origins of presumed rhythmic patterns in behavioural data. The proposed work will consolidate our understanding of listening as active perception. This knowledge is paramount for understanding the basic mechanisms underlying perception in general, but is also critical for our understanding of speech comprehension and related deficits, a highly relevant topic for an aging society.

• DFG Programme: Research Grants