Acoustic regularities in infant-directed speech and song across cultures

Abstract

AbstractAcross taxa, the forms of vocal signals are shaped by their functions1–15. In humans, a salient context of vocal signaling is infant care, as human infants are altricial16,17. Humans often produce “parent-ese”, speech and song for infants that differ acoustically from ordinary speech and song18–35, in fashions that are thought to support parent-infant communication and infant language learning36–39; modulate infant affect33,40–45; or credibly signal information to infants46. These theories predict a universal form-function link in infant-directed vocalizations, with consistent differentiation between infant-directed and adult-directed vocalizations across cultures. Some evidence supports this prediction23,27,28,32,47–50, but the limited generalizability of individual ethnographic reports and laboratory experiments51 and small stimulus sets52, along with intriguing reports of counterexamples53–60, leave the question open. Here, we show that infant-directed speech and song are robustly differentiable from their adult-directed counterparts, within voices and across cultures. We built a corpus of 1615 recordings of infant- and adult-directed singing and speech produced by 410 people living in 21 urban, rural, and small-scale societies and played the recordings to 45,745 people recruited online from many countries. We asked them to guess whether or not each vocalization was, in fact, infant-directed. The patterns of inferences of these naïve listeners, supported by acoustic analyses and predictive modelling, demonstrate acoustic cues to infant-directedness that are cross-culturally robust. The cues to infant-directedness differ across language and music, however, informing hypotheses of the psychological functions and evolution of both.