Two fundamental timescales organize human vocal behavior

Humans produce a remarkable diversity of vocal behaviors throughout life, from innate sounds such as cry, laughter, and babble to learned, sophisticated sequences such as speech and song. Whether these diverse vocalizations share basic principles in their temporal organization remains unknown. Here, we show that, despite their distinct functions and extraordinary cultural variation, these vocalizations consistently occupy two characteristic rhythmic regimes supporting distinct functionalities. Corpus analyses revealed that, by approximately three years of age, the acoustic rhythms of speech and song settle into two distinct frequency ranges: one between 4 and 8Hz, the other below 2Hz. Critically, these two temporal regimes are already evident in infant cry, laughter, and babble, prior to the emergence of language and song. Sensorimotor synchronization experiments, where participants vocalized or tapped to sound sequences presented at different rates, revealed that individual vocal production exhibited maximal resonance between 4 and 8Hz, whereas synchronization across individuals was strongest below 2Hz. Together, these findings identify two functionally distinct sensorimotor timescales that organize diverse human vocal behavior: a faster timescale supporting stable individual vocal production and a slower timescale facilitating interpersonal synchronization.