EUROSPEECH 2001 Scandinavia
7th European Conference on Speech Communication and Technology

Aalborg, Denmark
September 3-7, 2001


Native vs Non-Native Production of English Vowels in Spontaneous Speech: An Acoustic Phonetic Study

Kimiko Tsukada

Macquarie University, Australia

This study aims to examine acoustic characteristics of English vowels produced by 1 Australian English talker and 3 Japanese learners of English in spontaneous speech. Primary stressed vowels in multi-syllabic words were extracted from five 15-minute interview sessions. While there was a considerable overlap between different vowel categories both in native and non-native vowel spaces, centroids were more clearly separated in the former than in the latter. All three Japanese learners' vowel spaces were widely spread in the F2 direction. The Australian talker showed a moderate spectral distinction in two pairs /i - sci/ and /a invv/. Although this appears contrary to the spectral overlap commonly reported for these pairs in Australian English, it is consistent with the notion that short vowels are more susceptible to reduction than their long counterparts which are less likely to be undershot in various consonantal contexts.

Full Paper

Bibliographic reference.  Tsukada, Kimiko (2001): "Native vs non-native production of English vowels in spontaneous speech: an acoustic phonetic study", In EUROSPEECH-2001, 305-308.