| Résumé |
No infants in any primate species cry as much, and with as much diversity, as humans do. Two hypotheses compete to explain the development of such acoustical complexity in early life (Soltis, 2004). For one, the cry is a graded signal that develops to reflect the amount of e.g. neediness or pain felt by the infant. For the other, the cry is a categorical signal used to convey discrete types of needs. The present research proposes that it is, in fact, a mixture of both. We collected acoustic recordings of the cries of 35 infants, as well as verbal reports by their caregivers, at frequent and regular intervals during the infants' first year of age. Regression analyses of verbal categories identified in the reports indicate that, with development, caregivers increasingly attribute the cause of the cry to the infant's social (loneliness, shyness, opposition, etc.), rather than physiological (pain, sleepiness, hunger, etc.) needs; and that the cues utilized to infer one or the other type of need also change with time: the vocal content of the cry is used less and less to identify the infant's physiological needs, and more and more for its social needs. In parallel, acoustical analyses of the cries reveal a progressive differentiation of the patterns of expiration rate and pitch, in which each infant develops to adopt distinct signatures for different social needs. Together, these findings suggest that the infant's physiological states influence a graded-signal portion of the cry which is relatively stable through development, while social interactions influence a categorical-signal portion of the cry, a developping code which is probably learned jointly by the caregiver and the infant(Okanoya, 2007). |
