Abstract
The vocalizations of eight infants were recorded longitudinally in relation to different social and nonsocial contexts. The infants were observed bi-weekly from 3 to 25 weeks. At each visit they were presented with their mother, a female stranger and a doll, who were alternately active and passive. The results showed that by 7 weeks the infant vocalizations could be categorized from the social perception of adults into relatively long sounds containing variable pitch contours (melodic); short, nasal-like sounds containing uniform pitch (vocalic); and sounds such as crying, laughing and fussing (emotional). The infants modulated these sounds depending on the context. They produced significantly more melodic sounds when the women conversed with them than in any other context, and significantly more vocalic sounds when the adults were unresponsive. Overall the emotional sounds were produced significantly more to people than to objects. The results indicated that the various sounds served different purposes for the infants since they were used differentially in the social and nonsocial contexts.