Retour vers liste

Détail de la contribution

Auteur: Younghyon HEO

Co-Auteur(s): Emiko KANEKO, University of Aizu, Japan Ahrong LEE, York University, Canada

Asymmetry in the Choice of an Epenthetic Vowel during the Perception of a French and Russian Voiceless Alveopalatal Fricative by Japanese Listeners

Abstract/Résumé: This paper explores the role of low-level allophonic variations in the choice of an epenthetic vowel during the cross-language perception of a non-native syllable structure. In particular, it examines Japanese speakers’ choice of different epenthetic vowels ([ɯ] vs. [i]) after a word-final voiceless alveopalatal fricative in French and Russian, the source sounds of which differ in terms of lip protrusion. As is widely known, the alveopalatal fricative [š] is accompanied by lip protrusion in some languages such as English and French while it is not so in languages such as Japanese and Russian (Ladefoged & Maddieson 1996:148). In Japanese, [š#] is adapted as [šɯ] with the default epenthetic vowel [ɯ]. In order to test whether Japanese listeners would differentiate the word-final voiceless alveopalatal fricative in French and Russian by perceiving them with different epenthetic vowels, a perception experiment was conducted. After hearing 30 French and Russian nonce words with word-final [š] randomly presented, 44 linguistically-naïve Japanese listeners were asked to choose either [ši] or [šɯ] as the correct rendition. It was hypothesized that Japanese listeners would be more likely to perceive Russian word-final [š] with epenthetic [i] while French word-final [š] would be heard with [ɯ]. Even though neither [ɯ] nor [i] contains lip rounding, [i] is more [+spread] than [ɯ]; therefore, it is likely that Russian [š] would be matched with epenthetic [i], and French [š] with the default epenthetic vowel [ɯ]. Contrary to our prediction, the results indicate that there was no statistically significant difference in the perception of French and Russian [š]’s; the only factor that seemed to have significant impact was the frontness (or non-backness) of the preceding vowel. Namely, when preceded by a non-back vowel ([i], [e] or [a]), a high front vowel [i], instead of the default epenthetic vowel [ɯ], was inserted significantly more frequently than in the other environments. Furthermore, French front vowels caused the epenthesis of [i] more frequently than Russian front vowels. All these results imply that the hearers were not sensitive to lip protrusion in the source language, which is non-contrastive in Japanese, but to the aftersound of the preceding vowel and/or its impact on the production of the alveopalatal fricative in the stimuli. Further implications for perception of forbidden syllable structures will be proposed in the presentation, providing speculation and interpretation of the results in relation to controversial phonemic status of Japanese [š].