A corpus study of native and non-native vowel quality

Chen Huei Wu, Chilin Shih

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper investigates foreign accent by comparing vowel production of native speakers, heritage and non-native learners with data from a large corpus of spontaneous Chinese learner speech. Snippets were evaluated by untrained Mandarin raters on accent ratings and followed up with acoustic analysis of vowel qualities. The rating result showed a high correlation between accent and pronunciation. It is found that it is easier to improve the goodness of pronunciation, but the impression of accent is hard to change. Duration and formant studies reveal that L1 transfer has long-term impact on accent. The vowel [u] produced by second language learners was more fronted than that by native speakers. The vowel [y] is difficult for learners who associate front vowel with unrounding or, alternatively, whose performance falls between [y] and [u]. The formant space of Mandarin low vowels [a] and [a] produced by learners were in the opposite direction from the way native speakers did or were not distinguished. The research findings have implication on language teaching and pronunciation training method.

Original languageEnglish (US)
Title of host publicationProceedings of the 6th International Conference on Speech Prosody
PublisherTongji University Press
Pages234-237
Number of pages4
ISBN (Print)9787560848693
StatePublished - 2012
Event6th International Conference on Speech Prosody 2012, SP 2012 - Shanghai, China
Duration: May 22 2012May 25 2012

Publication series

NameProceedings of the 6th International Conference on Speech Prosody, SP 2012
Volume1

Other

Other6th International Conference on Speech Prosody 2012, SP 2012
Country/TerritoryChina
CityShanghai
Period5/22/125/25/12

Keywords

  • Foreign accent
  • Mandarin vowels
  • Second language acquisition

ASJC Scopus subject areas

  • Language and Linguistics
  • Computer Vision and Pattern Recognition
  • Human-Computer Interaction
  • Software
  • Mechanical Engineering

Fingerprint

Dive into the research topics of 'A corpus study of native and non-native vowel quality'. Together they form a unique fingerprint.

Cite this