This study investigated the perception of Canadian French word-final vowels by English-dominant and French-dominant bilinguals living in Montreal. In a modified identification task, listeners selected the response that rhymed with the target word, embedded in a carrier sentence. Minimal sets of real and nonsense target words were used, contrasting in word-final vowel /i, e,ε, a, o, u, y, ø/ preceded by labial, coronal and back stop consonants and /R/, as well as morphosyntactic verb minimal pairs contrasting in word-final /e-ε/ following /R/ (e.g., parlerai “I will talk” v. parlerais “I would talk”). Differences in accuracy and speed for identifying experimental vowels /e, ε, o, u, y, ø/ were investigated, relative to control vowels /i, a/, expected to be easiest and fastest to identify by both groups. Of interest was performance on front-rounded vowels /y, ø/, which are non-phonemic in English, and /e-ε/, which are phonologically contrastive in both languages, but /ε/ is disallowed word-finally in English.

Both groups were highly accurate in identifying experimental vowels, although the French-dominant group was comparatively more accurate and faster. For the English-dominant group, overall accuracy correlated with participants’ perceived accentedness and fluency in French. The French-dominant group was more accurate than the English-dominant group in identifying /e--ε/ in both lexical and morphological tokens. The English-dominant group was slower than the French-dominant group in identifying /y/ and /e/. Mouse cursor movements were captured trial-by-trial to analyze online perceptual processing. For instance, mouse track patterns of both groups was less-direct when identifying /y/, and participants of both groups often moved the cursor towards the response button for /u/ before correctly identifying /y/. The findings in this study contribute to the understanding of phonological processing by bilinguals, in exploring automaticity in vowel perception relative to language dominance. The fast, accurate perception of /ø/ by English-dominant participants is evidence of the development and stability in perception of this phoneme. The English-dominant group demonstrated less-automatic perception of most experimental vowels. However, even performance speed and mouse patterns of the French-dominant group varied among native vowel categories, implying possible interactions between automaticity and auditory salience of particular vowel contrasts.