| The purpose of this study was to investigate the functions, language choice, developmental changes, and cognitive and social contexts of private speech for an initially monolingual Chinese--subsequently bilingual Chinese-English--child in natural settings. Data were collected through naturalistic videotaping of the child in nursery school and kindergarten for two years, using a radio microphone. Private utterances were identified, and their functional category, language choice, cognitive context, and social context were coded for statistical analysis. Intensive observation of the subject in school and at home, and interviews with him and his parents provided additional qualitative information.;Results indicated that private speech in natural settings displayed a greater variety of functional types than has been reported in experimental studies. A curvilinear developmental trend was found for the total amount of private speech, supporting Vygotsky's hypothesis. Linear increase in the category of incomprehensible/inaudible supported Vygotsky's contention that private speech "goes underground" to form inner speech. Vygotsky's hypothesis that self-direction would occur more frequently in cognitively demanding than in cognitively undemanding contexts was not confirmed. The small proportion of private speech produced that was strictly self-regulatory in nature ran counter to Vygotsky's claim that self-regulation is the primary function of private speech. Both L1 (native language) and L2 (second language) were used for all the functions examined in this study, with a gradual shift to L2 in most of the functions. The language choice in private speech was not correlated with the language spoken by others around at the onset of second language learning, and indeed was somewhat negatively related, suggesting that the child did not wish others to overhear him. As bilingual competence developed, the correlation increased dramatically near L2 speakers, and finally, the language of the setting was adopted almost entirely. Intensive observation suggested that physical settings, nature of activities, social relationships, and feelings about the setting had influence on the production of private speech. Naturalistic methodology is recommended for additional future research on private speech. |
