Discrimination of Chinese pronunciations of aspirated dental and retroflex syllables according to breathing power and its frequency dependency during VOT

Akemi Hoshino , Akio Yasuda

Retroflexing of aspirates in Chinese is generally considered to be difficult for Japanese students learning pronunciation, because the Japanese language has no such sounds. In particular, discriminating utterances with aspirated dental and retroflex syllables is the most difficult. As we observed in our study, a classroom of Japanese students of Chinese uttered aspirated retroflex sounds modeled after examples uttered by a native Chinese instructor, however, the utterances sounded like dental syllables to the instructor, and many students could not produce the correct sounds. Warping the tongue was not enough to produce the correct articulation, because there is no retroflex sounds amongst the Japanese syllables. In the present paper, the authors extracted the features of the correct pronunciation of the aspirated dental syllables ca[ʦ‘a]，ci[ʦ‘i] and ce[ʦ‘ɤ], and the aspirated retroflex ones cha[#D‘a]，chi[#D‘i]，and che[#D‘ɤ], by analyzing breathing power and its frequency dependency during VOT (Voice Onset Time) of sounds uttered by nine Chinese native speakers and ten Japanese students. We will develop an automatic discrimination system by using the deduced evaluation standards.

Mots clés: retroflex pronunciation; dental pronunciation; Chinese

Domaines: Acoustique [physics.class-ph]

Akemi Hoshino, Akio Yasuda. Discrimination of Chinese pronunciations of aspirated dental and retroflex syllables according to breathing power and its frequency dependency during VOT. Acoustics 2012, Apr 2012, Nantes, France. ⟨hal-00810643⟩