Longitudinal changes in consonant production in infant-directed speech and infants' early speech production from 6 to 12 months.

Abstract

Previous research suggests that acoustic features of infant-directed speech (IDS) might be beneficial for infants' language development. However, consonants have gained less attention than prosodic and vowel-based features. In the current study, we examined voice onset time (VOT) - a distinguishing cue for stop consonant contrasts - in IDS and adult-directed speech (ADS), and its relation to infants' speech production. We used a longitudinal sample of 48 Norwegian parent-infant dyads. Parents' IDS and ADS were recorded in-lab at three timepoints (infants' age: 6, 9, 12 months), and the VOTs of the stop consonants /b-p/, /d-t/, and /g-k/ were measured. In addition, at each timepoint, parents reported their infants' production of the same consonants, as well as their babbling. Hypotheses were preregistered, and we used full-null model comparisons to minimise type I-errors in the analyses. Our results demonstrate that, while controlling for speaking rate, in IDS, parents' VOTs were longer in voiceless stops, but shorter in voiced stops, resulting in overall less distinct consonant contrasts compared to ADS. Further, VOTs in IDS approached ADS values with infants' age. However, we found no relationship between parents' VOTs and infants' consonant production or babbling. Consonants, like vowels, appear to be less distinct in IDS than ADS, thus reinforcing the interpretation that IDS may serve an attentional and/or affective aim, rather than a didactic purpose.