Behavior-relevant periodized neural representation of acoustic but not tactile rhythm in humans

Abstract

Music makes people move. This human propensity to coordinate movement with the rhythm of music requires multiscale temporal integration, allowing fast sensory events composing rhythmic input to be mapped onto slower internal templates such as periodic beats. Relatedly, beat perception has been shown to involve a selective enhancement of the beat periodicities in the neural response to rhythm. However, the extent to which this ability to move to the beat, and the related periodized neural representation, are shared across the senses beyond audition remains unknown. Here we addressed this question using acoustic and tactile rhythms, while recording separately the electroencephalography (EEG) responses and finger tapping to these rhythms in healthy volunteers. Consistent with previous studies, EEG responses to the acoustic rhythm featured significant enhancement of the beat, and this periodized neural representation was specifically concentrated within a low-frequency range below 15 Hz. In contrast, the same rhythm conveyed with tactile inputs elicited responses over a broader frequency range, up to 25Hz, with no significant periodization, and resulted in less stable tapping. Together, these findings indicate that low-frequency neural activity preferentially supports behavior-relevant internal representation of rhythmic input. However, this neural representation is not necessarily shared across the senses, as well as the ability to move to the beat, corroborating multimodal differences in beat perception. This low-frequency neural representation may thus reflect a process of multiscale temporal integration allowing the auditory system to go beyond mere tracking of onset timing and support higher-level internal representation and motor entrainment to rhythm. Significance statementIntegrating the fast sensory events composing music into slower temporal units is a cornerstone of beat perception and social interaction through music. The current study shows that this ability critically relies on brain activity concentrated in a lower frequency range - below the recurrence of sensory events - in response to acoustic rhythm. In contrast, when the rhythm is conveyed through touch, brain responses comparatively exhibit higher frequency activity corresponding to the faithful tracking of each individual events of the tactile rhythm. Most importantly, these auditory-specific slow fluctuations feature a periodization of rhythmic inputs, compatible with behavior. This higher-level neural processing of rhythmic input could thus reflect internal representations of the beat that are not necessarily shared across sensory modalities, in line with the concept of auditory dominance in temporal event perception and motor entrainment to rhythm. The current study thus opens a promising avenue to gain fundamental knowledge on high-level multimodal perception and motor entrainment processes specific to humans.