The linguistic register known as """"""""motherese,"""""""" or child-directed talk (CDT), it notable for its prosodic differences from adult-directed talk (ADT), in particular its high mean F0 and an expanded F0 range. The proposed research has two goals: (1) to investigate the effect of these prosodic characteristics on the phonetic perception and production of nonnative suprasegmental and segmental speech contrasts by adult listeners, and (2) to examine the speech of children between the ages of 7 and 10 years for evidence of """"""""motherese"""""""" like prosody, in particular, its high mean F0 and an expanded F0 range. The first three of the four experiments in this proposal are designed to achieve the first goal, i.e., to test the hypothesis that the prosodic characteristics of CDT, as opposed to those of ADT, improve adult listeners' phonetic perception and production. In each of the three experiments, adult subjects will identify or make category judgments on a phonetic contrast produced as either CDT or ADT, and another group of subjects will attempt to produce the phonetic contrast using either CDT or ADT tokens as models. These productions will be subjected to acoustic analysis and given to native listeners who will judge the accuracy of the imitations. In the first experiment, adult listeners' phonetic perception and production of Mandarin tones, primarily cued by F0 information, will be assessed. In the second experiment, adult listeners will be tested on the phonetic perception and production of the duration and amplitude patterns of French bisyllabic words. In the third study, adult listeners' phonetic perception and production will again be tested, this time on a segmental contrast from French. The results of this investigation may lead to the development of techniques for enhancing phonetic perception and production in several populations, including second-language learners, certain groups of aphasics, and the hearing impaired. The fourth experiment is designed to achieve the second goal, i.e., to test the hypothesis that the speech of children between the ages of 7 and 10 years shows evidence of """"""""motherese""""""""-like prosody. This fourth study focuses on the speech of two groups of children, aged 7 and 10 years, in three different contexts --talking to an adult, to another child, and to a child who does not yet know English. The middle two minutes of each child's productions in each context will be acoustically analyzed for mean F0 and F0 range. The results of this experiment should provide information on the suprasegmental characteristics of children's speech that will help identify normal patterns of prosodic development.
