ABSTRACT This article examines whether conflation, a term introduced by Nihalani, occurs among English vowels produced by four Japanese native speakers reading a short English text. On the basis of perceptual, acoustic, and statistical methods, it is argued that conflation is not key to understanding the structure of vowels in Japanese English (JE). Having inspected vowel variants, we conclude the following. First, since conflation is not guaranteed, singular emphasis on suprasegmental features is untenable. Second, vowel space in JE tends to be centralized. As a result, speakers have more difficulty producing open back vowels. Third, the centralized vowel [a], substituted for five vowels (/æ/, /ə/, /ʌ/, /ɜ/, /ɑ/), is not statistically speaking a single sound. Fourth, JE speakers produce near-native vowels to different degrees. Statistical test results showed no significant difference between the pronunciation of the open-mid vowel [ʌ] produced by JE speakers and an American English speaker.

中文翻译：

---

有混杂吗？日语英语元音的声学分析

摘要 本文研究了 Nihalani 引入的术语混淆是否出现在四位以日语为母语的人阅读一篇简短的英语文本时产生的英语元音中。基于感知、声学和统计方法，有人认为合并不是理解日语英语 (JE) 元音结构的关键。在检查了元音变体之后，我们得出以下结论。首先，由于不能保证合并，因此单一强调超段特征是站不住脚的。其次，JE 中的元音空间趋于集中。结果，说话者更难发出开后元音。第三，中心元音[a]代替五个元音（/æ/、/ə/、/ʌ/、/ɜ/、/ɑ/），在统计上不是一个单一的声音。第四，JE 使用者产生不同程度的接近母语的元音。