The acoustic structure of vowels in mothers' speech to infants and adults

Research has shown that exposure to a specific language alters infants' perception of vowel sounds by the time they reach 6 months of age. This raises the important question of how infants develop language-specific patterns of vowel perception from the language-general pattern they appear to be born with. Language spoken to infants may exert an important influence in this regard. The present study compares the acoustic structure of vowels in the words "sheep" and "shoes" produced by 10 mothers in conversation with their infants, with their acoustic structure when produced by the same 10 women in conversation with an adult. Mothers were instructed to play with their infants using a toy sheep and pair of shoes. They were asked to use the same words in conversation with an adult. The infant-directed tokens exhibited higher F0, greater pitch excursions and longer duration than the adult-directed tokens, as is typical of motherese. Although F0 was significantly higher in infant-directed vowels, F1 remained at essentially the same frequency as in adult-directed /i/ and /u/. In contrast, F2 was significantly higher in infant-directed /i/ and significantly lower in infant-directed /u/ than in the adultdirected tokens. Thus, the infant-directed tokens reached more extreme acoustic targets. The formant structure also indicated an equivalent or slightly greater degree of coarticulation of infant-directed /u/ with the preceding fricative in "shoes." Overall, mothers in this study consistently hyperarticulated vowels in speech to infants. Hyperarticulation may contribute to the acquisition of native language vowel categories by increasing the degree of acoustic separation between vowel categories.