Acoustical studies in perception and production of long vowel by L2 learners and some considerations from the point of view of interlanguage theory

通过对日语作为第二语言的超级学习者的录音进行研究，得出了以下结论。(1)即使对于超级学习者来说，精确地实现相对较少使用的单词的音高模式似乎也很困难。(2)在超级学习者的日语语音生成中，他们似乎更倾向于缩短词尾位置的长元音，而不是非词尾位置的长元音。这与以日语为母语的人所表现出的趋势相同，这一事实表明，即使在所谓的关键时期之后，隐性语言规则也可能是可以学习的。通过研究发现，即使使用Multi Speech或Sugi Speech等声学分析工具，也很难确定短元音和长元音的分类。建立一种公认的判断语音长短的方法是学习者语音分析研究者的重要任务。第二个问题是中介语理论在各种语音习得现象中的应用是否充分。根据本研究揭示的内容，有必要开始纵向收集更可靠的数据，以科学地检验数据与理论之间的关系。

By examining the recordings of super learners of Japanese as L2, the following implications were obtained.(1)Even for the super learners the precise realization of pitch patterns of words which are relatively less frequently used seem to be difficult.(2)In the production of Japanese speech sounds by super learners, it seems that they tend to shorten the long vowels in word final positions more than non-word-final position ones. This is the same tendency that the native speakers of Japanese show and this fact implies that implicit linguistic rule might be learnable even after so called the critical period.Through the research was found quite hard to definitely categorize short vowels and long vowels even using acoustical analysis instruments such as Multi Speech or Sugi Speech. It is an important task for researchers of learner speech analysis to establish a certain commonly accepted way of judging whether the vowel is short or long phonologically.Second question that remained was the adequacy of the application of interlanguage theory to the various phenomena of speech sound acquisition. Following what was brought to light through this research, it is necessary to start collecting more reliable data longitudinally to scientifically examine the relation between the data and the theory.