The Acquisition Mechanism of Perception and Pronunciation of Plosives by East-Asian Learners of Japanese : A Study of Interlanguage

基于以下三项调查，准备了四篇论文和五次学术会议上的口头报告，以展示下面(1)到(5)的研究结果。韩国日语学习者爆破词感知与发音的纵向研究（2003.07—2004.02）2004年2月至3月，在中国天津外国语大学、上海外国语大学和韩国庆熙大学进行了基于语音开始时间（VOT）合成变化的炸药分类感知的语际调查；庆熙大学（2004年9月）和上海同济大学（2005年1月）爆破音和重音的研究。(1)东亚日语学习者普遍难以掌握爆破音的感知和发音。对于韩国的学习者（说首尔话的人）来说，从感知和发音的角度进行的研究较少，他们在发音方面遇到了严重的问题。这说明，他们可能认为日语单词的浊音爆破音发音很松散，就像韩语单词的首字母一样。在感知上，他们听到了日语单词首字母的无声爆破音，这似乎与韩国语的强音相似。在此基础上，成功地阐明了日语学习中包含主要混淆因素的习得机制。(2)对韩国人的北京话、上海话和首尔话的声学-语音分析表明，感知习得和发音习得存在差异。说上海话的人首先获得了感知和发音；说北京话的人遇到的困难最大。在感知上，北京话和上海话的人将单词中间的不发音爆破音误认为是浊音爆破音，而韩语的人则将单词首字母中的不发音爆破音误认为是浊音爆破音。在发音方面，所有的说话者都很难发出单词首字母中的元音。说上海话的人和说韩语的人，在他们的母语中有发声的爆破音，他们分别把它们读成母语中不发音的不送气音和松散的单词首字母。(3)在韩国学习者对爆破音的感知和发音习得过程中，也存在与口音相关的母语干扰。使用基本频率进行检查的结果显示，一个单词的开头包含一个不发音的爆破音，其发音方式是高低音调。然而，一个包含浊音爆破音的单词，则以中高音调模式发音，第一个莫拉音较低，随后的莫拉音升高。(4)通过不同起音时间（VOT）的综合变化考察了北京话、上海话和韩语日语学习者对日语爆破音的范畴感知，发现北京话和上海话的范畴感知与日语母语者在单词首字母中包含爆破音的范畴感知相似。但是，当爆破音出现在单词中间时，韩国语使用者与母语为日语的人对应。这些结果揭示了不同语言之间不同的范畴感知。(5)通过包括干扰和近似在内的发展过程考察了爆破语习得过程中中介语的形成过程。结果发现，Mayor（1997）提出的第二语言习得个体发生模型（Ontogeny Model）并不能适用于所有情境，并被认为对中介语理论的构建做出了贡献。少

Based on the following three investigations, four papers and five oral presentations at academic conferences were prepared to show research results (1) through (5) below.Longitudinal study on perception and pronunciation of plosives by Korean learners of Japanese (July 2003--February 2004) ; Interlingual investigation on the categorical perception of plosives using the synthesized variations of voice onset time (VOT) at Tianjin Foreign Studies University and Shanghai International Studies University, of China, and Kyung Hee University, of Korea (February--March 2004) ; Research on plosives and accents at Kyung Hee University (September 2004) and Tongji University, in Shanghai (January 2005).(1) It is commonly difficult for East-Asian learners of Japanese to acquire perception and pronunciation of plosives. As for learners from Korea (Seoul-dialect speakers), where less research has been performed from the standpoints of perception and pronunciation, they experience a severe problem in … More pronunciation. This reveals the possibility that they understand the voiced plosives of Japanese words to be lax in pronunciation, much like the word initials of Korean. In perception, they hear voiceless plosives in the word initials of Japanese that appear to be similar to the forced sounds of Korean. Based on these data, the mechanism of acquisition containing the major factors of confusion in the learning of Japanese was successfully clarified.(2) Acoustic-phonetic analysis of the Beijing-dialect, the Shanghai-dialect, and the Seoul-dialect of Korean speakers revealed that the acquisition of perception and pronunciation differ. Shanghai-dialect speakers acquired both perception and pronunciation first ; Beijing-dialect speakers had the most difficulty. In perception, Beijing- and Shanghai- dialect speakers misheard voiceless plosives in the middle of words as voiced plosives, but Korean speakers misheard voiceless plosives in word initials as voiced plosives. In regards to pronunciation, it was difficult for all speakers to pronounce voiced plosives in word initials. Shanghai-dialect and Korean speakers, who had voiced plosives in their native languages, pronounced them as the voiceless unaspirated sounds and as the lax word initials of their native languages, respectively.(3) In the acquisition by Korean learners of perception and pronunciation of plosives, there was also first language (L1) interference related to accents. As a result of an examination using fundamental frequency, it was revealed that a word including a voiceless plosive at its initial was pronounced in the high-low pitch pattern. A word including a voiced plosive at its initial, however, was pronounced in the middle-high pitch pattern, with the first mora low in pitch and the subsequent mora rising.(4) When the categorical perception of Japanese plosives by Beijing-dialect, Shanghai-dialect, and Korean speakers studying Japanese was examined by means of synthesized variations with different voice onset times (VOT), it was found that the categorical perception of Beijing- and Shanghai-dialect speakers was similar to that of native speakers of Japanese when a plosive was included in a word initial. However, Korean speakers corresponded to native speakers of Japanese when a plosive was in the middle of a word. These results revealed different categorical perceptions among languages.(5) The process of shaping an interlanguage in the acquisition of plosives was examined through the developmental processes, such as including interference and approximation. As a result, it was found that the Ontogeny Model for second language (L2) acquisition developed by Mayor (1997) could not be applied to all situations, and it was believed to have contributed to the construction of the interlanguage theory. Less