Audiovisual speech perception in 4-month-old infants

Previous research indicates that for adults and children the perception of speech can be significantly influenced by watching a speaker's mouth movements. For example, hearing the syllable /bi/ while watching a speaker mouth the syllable /vi/ results in reports of a 'heard' /vi/. Some evidence suggests young infants also may be able to integrate heard and seen speech. One theory suggests that an innate link between perception and production (Liberman & Mattingly, 1985) accounts for this phenomenon while another theory suggests that experience (e.g., producing speech sounds) may be necessary into order to develop fully the underlying representation of visible speech (Desjardins, Rogers & Werker, in press; Meltzoff & Kuhl, 1994). My dissertation addresses the above controversy by examining whether the integration of heard and seen speech is obligatory for young infants as it is for adults. In Experiment 1, 4-month-old female infants habituated to audiovisual /bi/ showed renewed visual interest to an auditory /bi/-visual NM suggesting that they may have perceived the auditory /bi/-visual /vi/ as /vi/, as do adults. In Experiment 2, neither male nor female infants showed renewed visual interest to a dishabituation stimulus which represents only a change in mouth movements. In Experiment 3, male infants looked longer to an audiovisual /bi/ than to an audiovisual /vi/ following habituation to an audio /bi/-visual /vi/, while female infants tended to look only slightly longer to an audiovisual /vi/ than to an audiovisual /bi/. Taken together these experiments suggest that at least some infants are able to integrate heard and seen speech, but that they do not do so consistently. Although an innate mechanism may be responsible for integration, a role for experience is suggested as integration does not appear to be obligatory for young infants as it is for adults.Arts, Faculty ofPsychology, Department ofGraduat