Analysis Of Speech Input In Early Language Acquisition

Speech input plays a key role in language acquisition. If there is no speech input, language cannot be acquired. However, most researches deal with the description and analysis of child language development, namely, language production. There is a lack of research on speech input in early language acquisition, especially the study on the quantity of speech input. Therefore, the focus of language acquisition research now should be put on the study of speech input.The theoretical framework of this paper includes Chomsky's Universal Grammar, Lenneberg and Werker's research on Critical Period,CDS (Child-Directed Speech) and Shan Xingyuan's Non-macro Language Acquisition Model. Then the author reviews the first language acquisition, second language acquisition, bilingualism and speech input.Three hypotheses are formulated in this paper. They are as follows. (1) The MLU (Mean Length of Utterances) of the CDS is smaller than that of the adult-to-adult language. The MLU of the CDS slowly increases with the growth of children during the early period of language development. The MLU of caregivers seems to be larger than the children's MLU. (2) Adult's quantity of speech input per minute is always larger than the child's quantity of speech output per minute. And the children tend to use less word types and they have a higher repetition rate of words. (3) With Shan Xingyuan's Non-macro Language Acquisition Model, the children can become bilingual as long as there is enough quantity of speech input. But the precondition is that it must happen within the Critical Period.Quantitative method is used to analyze the corpuses from the CHILDES (Child Language Data Exchange System) database. The data used to prove the first two hypotheses are from the CHILDES database. To analyze the data, the CLAN (Computerized Language Analysis) is used to measure MLU, word types and amount. Besides, case study is used to prove the third hypothesis. Mandarin Chinese-Korean bilinguals in China are examples.Through analyzing the data, we find out that mother's MLU is always larger than child's MLU. And both mother's MLU and child's MLU in CDS increase during the early period of language development. There is a sharper increase of MLU during the first 3 years of language development than the later two years (from age 3 to 5).We also find out that the adult's quantity of speech input is about twice as much as the child's quantity of speech output both in Chinese and English during the first three months of the third year. There is more English input than Chinese input per minute. And the child's English output is also more than the Chinese output. It has also been found out that the child's type/token ratio tends to be lower than the adult's, which means that the children use less word types and they have a higher repetition rate of the words.At the same time, through case studies of Korean children (Chaoxian nationality), the author proves that Shan Xingyuan's non-macro language acquisition model is critical for language acquisition.