Speech perception in early infancy: Perceptual constancy for spectrally dissimilar vowel categories

Abstract

While numerous studies on infant perception demonstrate the infant's ability to discriminate individual speech-sound pairs, very few demonstrate the infant's ability to recognize the similarity among phonetic units when they occur in different phonetic contexts, in different positions in a syllable, or when they are spoken by different talkers. In two studies, six-month-old infants demonstrated the ability to distinguish two spectrally dissimilar vowel categories (/a/ and /i/) in which the vowel tokens were generated to simulate tokens produced by a male, a female, and a child talker. In experiment I, the infants were initially trained to discriminate the /a/ and /i/ tokens produced by the computer-simulated male voice. They were then gradually exposed to a number of novel tokens in a progressive transfer-of-learning task. In experiment II, the infants were initially trained to discriminate the same vowell contrast, but were then immediately tested with all of the tokens in both vowel categories. In both experiments the infants demonstrated rapid transfer of learning from the training tokens produced by the male talker to the tokens produced by female and child talkers. Both experiments provide strong evidence that the six-month-old infant recognizes acoustic categories that conform to the vowel categories perceived by adult speakers of English.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 316 citations exist upstream.