A Study Of Intervals Of The Reduplicated Words In Southern Mandarin Young Children’s Speech

This thesis is a phonetic study of the temporal aspects of the reduplicated words according to a longitudinal study of two Southern Mandarin children. They are a boy and a girl aged01;09-02;09. The purpose of this study is to provide a systematic phonetic approach to the early speech development of Southern Mandarin children and to get a deep understanding of the process of children’s phonetic acquisition, as well as to examine the Independence Hypothesis (Lenneberg,1967) and the Hypothesis of Early Interaction (Boysson-Bardies et al.,1984). This thesis would address the following research questions:1) Do Southern Mandarin children’s intervals of reduplicated words decrease with increasing age?2) Is the temporal variability of intervals greater when the children are younger?3) Does gender affect their interval change of the reduplicated words?The two subjects are from two families respectively whose parents are university teachers. Both of them are the only child in their families with no speaking, cognitive or hearing problems. Mandarin Chinese is their daily language. They got audio-(SONY ICD-SX35) and video-recorded (SONY HDR-HCI) once at approximately one-week intervals, and each recording lasted45-60minutes. All the data were recorded under natural language environment by postgraduate student at their homes. There are728tokens of the following reduplicated words:[pApA],[mAmA],[wAwA],[khεn khεn],[jεjε],[phi αu phi αu],[thA η thA η],[t(?)i t(?)i] and [t hiu t hiu]. Six age stages are studied respectively:01;09-01;10,01;11-01;12,02;01-02;02,02;03-02;04,02;05-02;06,02;07-02;09. I transferred the data into the PC, and did the segmentation by using Cool Edit Pro.(Version2.1). Then they were textgrided by Praat (Version5.1.37). By using a script, the absolute intervals were obtained. Finally, I normalized all the data. The transcription was completed by postgraduate students majoring in this field and checked by two professional researchers majoring in Mandarin Chinese phonetics. The intra-reliability was98.11%and the inter-reliability was91.35%I obtain the following major findings:1) Southern Mandarin young children’s intervals of the reduplicated words decrease with increasing age.Through a series of analysis of the two children, we have found that the young children’s intervals of the reduplicated words decrease with increasing age. A comparison between the children and their mothers are also done respectively. It can be concluded that the decrease of mean intervals is a process that children are getting closer to adult’s speech. This finding is consistent with the Hypothesis of Early Interaction (Boysson-Bardies et al,1984).2) The variability of intervals is greater when the children are younger, and the variability of intervals is smaller when the children are older.Via a systematic analysis, it is confirmed that the two children’s temporal variability of the reduplicated words is greater when the children are younger. This finding may be explained by Smith (1996) that decreases in interval would be greater when children are younger and interval decreases smaller when they are older. Via the comparison with their mothers, it is also found that their decrease of variability is approximating to their mother’s via more exposure to their ambient language.3) Gender affects the interval change of the reduplicated words.Via a comparative analysis, it is found that for most of the time, the girl’s mean intervals are relatively longer than the boy’s. Based on a comparison of their intervals across the six age stages, it is found that during the first four stages, more than a half of the girl’s intervals are longer; during02;05-02;06, the boy’s intervals begin to lengthen; and in the last stage, some of the boy’s intervals are even longer than the girl’s. The result indicates that the girl’s mean interval of reduplicated words reach the adult’s level earlier than the boy’s. However, both of them show the tendency to approximate the adults’speech concerning intervals.The findings are in support of the Hypothesis of Early Interaction (Boysson-Bardies et al,1984) and negate the Independence Hypothesis (Lenneberg,1967).