The development of prosodic contrastivity during the first year of life

The relationship between the infant's early vocal development and subsequent speech and language development has been generally a matter of speculation, largely based upon anecdotal evidence, despite the voluminous literature on the subject. Included in the existing data on infant prosodic development are descriptions of differentiated crying behaviour ostensibly expressive of different internal states such as pain, hunger and pleasure. Instances of differential vocalization to objects and people have likewise been alluded to in the literature, but no systematic studies of the phenomenon have been made. The present study examines certain aspects of vocal differentiation and differential vocalization, primarily with respect to familiar environmental objects. Nineteen normal, healthy, 'first-born' infants served as subjects. All were being raised in an exclusively monolingual English home environment, where all data were collected at biweekly intervals from five weeks to approximately one year of age. The vocalizations of ten of the infants were studied longitudinally over an entire year, the remaining subjects from five weeks to approximately six months. Intonation patterns of non-cry vocalizations occurring in two basic situations (i.e., infant alone and infant in the context of various objects) were analyzed spectrographically for fundamental frequency, within-utterance range, duration and contour. Each variable was examined longitudinally and in different contexts. A number of age trends are evident: Duration and within-utterance range increase with age, whereas fundamental frequency remains relatively stable over the first year. Females exhibit a higher F₀ than males at all age levels examined. Peak values of the variables are commonly observed at 4-6, 9 and (to a lesser degree) 11 months. The RF intonation contour increases in frequency of occurrence during the first year, while the other contours demonstrate little change. Contrasts of utterances occurring in different categories reveal essentially no difference between contexts for the variables studied. Examination of the distributions of the different intonation contours for each context indicates that the infants could be manipulating contour differentially in a given context. In sum, it is felt that the infants exhibit a very real, albeit circumscribed, capacity for vocally differentiating environmental events. It is felt that there exists sufficient evidence to refute the parochial view that linguistic acquisition can only be relevantly discussed when the child's segmental phonetic output begins to resemble that of the adult standard. The evidence presented corroborates the hypothesis of continuity from babbling to speech.Medicine, Faculty ofAudiology and Speech Sciences, School ofGraduat