Séminaire de Mme Tran Thi Thuy Hien, doctorante du laboratoire GIPSA-Lab, actuellement invitée au Centre MICA - Date : jeudi 4 juin 2009 - 14h00 - Lieu : salle polyvalente Centre MICA
Intervenant :
Mme Tran Thi Thuy Hien
doctorante du laboratoire Gipsa-Lab, actuellement chercheur invité du Centre MICA
Date : jeudi 4 juin 2009 - 14h00
Lieu : Salle polyvalente, bâtiment C10, Centre MICA, IP Hanoi
Interprète traducteur : le séminaire sera présenté en français
Résumé/Abstract:
A great difficulty encountered by Vietnamese subjects who learn French is that consonant clusters, which do not exist in Vietnamese, are mispronounced and this problem persists even after several years of practicing (Nguyễn, 2000), even if specific consonant combinations are found in both languages. Whereas Vietnamese has specific phonetic realization of unreleased final stops (Trương, 1970; Cao, 1985; Đoàn, 1999, Michaud, 2004), it differs from French which in general allows released final stops.
Several strategies of cluster replacement are observed at learning. They are often replaced by one of the units which exist in the native phonological system and generally by neglecting the first consonant of the cluster. Also, consonant cluster can be pronounced via the insertion of a vowel which facilitates their articulation. They are moreover pronounced with the Vietnamese syllabic structure (e.g. ‘cyclo’ adapted ‘xích lô’) (with /k/ unreleased).
What are the main reasons for this barrier to French cluster acquisition by Vietnamese learners? To answer this question, the framework posed by Schwartz & al. (2002), namely the Perception for Action Control Theory (PACT), seems to provide an adequat way of understanding the mispronounciation of consonant clusters. According to this theory, the speaker’s ability to control vocal gestures is connected to the speaker’s perception. The principle at the core of this theory is obvious: Without auditive perception, deaf people do not correctly acquire motor control for speech production.
The general aim of our project is to identify the factors which are the main cause of this problem. In this paper, we present a study which examines the perception of syllable-final stops in Vietnamese (/p/ - /t/ - /k/ and /m/ - /n/ - /N/) by 20 Northern-Vietnamese native listeners (10 males, 10 females). The 55 natural auditive stimuli were either monosyllabic words (CVC) (e.g. ‘các’, ‘mát’, ‘táp’) or the first syllable of compound words (CVC.CVC) (e.g. ‘pháp’ in ‘pháp luật’, ‘mát’ in ‘mát dạ’, ‘ác’ in ‘ác liệt’).
Our results show a clear difference in the perception of final stops in CVC words compared to CVC.CVC word structures. Our findings suggest that specific acoustic characteristics of final consonants (plosives and nasals) indicate to the listener if this consonant is located at word boundary or at syllable boundary in a dissyllabic compound word and are consequently relevant to the discussion on the role of the syllable in phonetic theory.
