Realization of the Laryngeal Contrast by Native and Korean L1 Speakers of Quebec French

This study investigates the acoustic realization of the laryngeal contrast by 10 late Korean/French bilinguals in Quebec, compared to 19 native speakers of Laurentian French. French distinguishes voiced /b d ɡ/ from voiceless /p t k/ primarily through vocal fold vibration during closure, whereas Korean maintains a three-way contrast (lenis, fortis, aspirated) that does not rely on distinctive voicing. We examined five acoustic dimensions, namely Voice Onset Time (VOT), proportion of voicelessness, closure duration, fundamental frequency (F0), and voice quality (H1*−H2*), in reading and repetition tasks, using generalized linear mixed-effects models. Native speakers predominantly prevoiced voiced stops (95.4%), with marginal but community-wide devoicing and no evidence of aspiration. Korean speakers frequently produced unvoiced realizations of voiced stops, with substantial inter-individual variation. While VOT values were largely comparable across groups, Korean speakers showed a reduced VOT distinction between unvoiced and voiceless stops in intervocalic position, produced voiceless stops with significantly longer closures, displayed a larger F0 difference between series in word-initial position, and used breathier voice quality for voiceless stops, particularly word-initially and for lower-frequency words. No facilitative effect of Korean intersonorant voicing was found for voiced stops in intervocalic position, nor did the repetition task yield more target-like productions than reading. Taken together, these findings highlight the importance of examining multiple acoustic dimensions and positional variation in L2 phonetic acquisition.