Bidirectional connectivity changes in the frontotemporal areas reflect figure–ground reversal in multivoiced music

Abstract

The neural processes underlying the perception of prominently heard voices (figure) and less prominently heard voices (ground), particularly when this prominence is naturally reversed while listening to music, remain inadequately investigated. We investigated changes in frontotemporal connectivity in response to figure-ground reversal within multivoiced music using Mozart's variation KV 265, including the "Twinkle, Twinkle, Little Star (TTLS)" melody based on our previous study of frontotemporal connectivity (TTLS connectivity), influenced by the emergence of the TTLS melody in the upper voice at the beginning of each variation, and independent of lower voices. We examined the consistency of the TTLS connectivity pattern across repetitions of the same phrases within each variation. Notably, TTLS connectivity changed only for the final repetition between variations sharing the same TTLS melody. This result indicates a perceptual reversal, processing the TTLS melody initially perceived as figure into the ground. Our data effectively illustrate how the brain dissects voices within the multidimensional structures of continuously changing music, reconstructing them through a momentary switch in figure-ground processes. These findings provide valuable insights into the neural processes associated with the dynamic experience of listening to music, a ubiquitous aspect of human life.