Phonetic-Phonem Comparison in Teaching Vietnamese Pronunciation: A Study of Final Sounds for Chinese Speakers

Abstract

This study investigates strategies for improving the pronunciation of Vietnamese codas among Chinese-speaking learners through a contrastive phonetic–phonological analysis. The pedagogical framework integrates: (1) contrastive analysis to enhance learners’ phonological awareness, (2) minimal pair exercises, and (3) experimental application of automatic speech recognition (ASR) systems, such as Google Voice Typing, for error detection. Conducted with 30 Chinese-speaking learners over 12 weeks, the study identifies key challenges, including confusion among nasal codas (/m, n, ŋ/), omission of the bilabial stop /p/, and difficulties in pronouncing low-falling glottalized tones (thanh nặng). The results indicate that ASR systems struggle to detect isolated phonemes, underscoring the need for visual-acoustic feedback tools (e.g., Praat) for effective pronunciation training. The study proposes pedagogical strategies for integrating AI-powered technologies and advocates for developing Vietnamese-specific ASR systems to support second language pronunciation instruction.