Le développement d'aspects phonético-phonologiques du français chez des enfants bilingues simultanés et successifs : Le VOT et la liaison dans une étude de cas multiples

This dissertation describes the development of Voice Onset Time (VOT) and liaison in child second language (cL2) French (L1: Swedish) as compared to simultaneous Swedish-French bilinguals (2L1) and monolingual (1L1) French children. VOT is studied in both languages but liaison only exists in French. Differences between L1 and L2 phonology have previously been attributed to a sensitive period for language learning or to L1 influence. According to recent extensive studies on morpho-syntax L2 acquisition with an age of onset (AoA) between 4 and 8 years constitutes a separate mode of acquisition (cL2), where some aspects pattern with L1 and others with adult L2 (aL2) development. Such results have been attributed to a combination of a sensitive period and the level of L1 development at onset. In comparison cL2 phonological development is under-researched and evidence for a cL2 mode of acquisition in phonology is still scarce. A longitudinal corpus was collected over 2.5 years. Eight children participated from age 3 to 6: three cL2 learners (AoA: 3;0-3;5), three 2L1s and two 1L1s. VOT and liaison were targeted using specific tests. The L1 phonological development was assessed through tests used in speech-language pathology. Results on the development of VOT in French 1L1 showed influence form consonants and vowels, as in adult speakers, and were found consistent with suggestions of an upside-down U-curve peaking at age 10 or later. The 2L1s did not differ significantly from 1L1 development in French, but in Swedish their development indicated a delay compared to 1L1s reported in the literature. This was also the case for the cL2 learners’ Swedish VOT. For French the cL2 learners initially displayed influence from Swedish (with 66% of French voiceless stops being produced with a long lag VOT) but seemed to converge with the (2)L1 children at later stages. Both 2L1 and cL2 separate the two languages. For liaison, 1L1 and 2L1 development did not differ in terms of rate of acquisition. The initial cL2 stage resembled that of aL2 learners, i.e. no liaison consonant was produced (e.g