This study examined whether the atypical speech style that is frequently reported in autistic adults is underpinned by an inflexible production of phonetic targets. In a first task, 20 male autistic adults and 20 neuro‐typicals had to read and produce native vowels. To assess the extent to which phonetic inflexibility is due to an overall fine‐grained control of phonetic behavior or to a lack of flexibility in the realization of one's phonological repertoire, the second task asked participants to reproduce artificial vowel‐like sounds. Results confirmed the presence of a greater articulatory stability in the production of native vowels in autistic adults. When instructed to imitate artificial vowel‐like sounds, the autistic group did not better approximate the targets' acoustic properties relative to neuro‐typicals but their performance at reproducing artificial vowels was less variable and influenced to a greater extent by the articulatory properties of their own vocalic space. These findings suggest that the greater articulatory stability observed in autistic adults arises from a lack of flexibility in the production of their own native vowels. The two phonetic tasks are devoid of any pragmatic constraint, which indicates that phonetic inflexibility in autism is partly independent of register selection. Lay Summary Autistic and neuro‐typical adults took part in two tasks: one in which they produced vowels from French, their native tongue, and the other where they imitated unfamiliar vowels. Autistic adults displayed significantly less variation in their production of different French vowels. In imitating unfamiliar vowels, they were more influenced by the way they pronounce French vowels. These results suggest that the atypical speech style, frequently attested in autistic individuals, could stem from an unusually stable pronunciation of speech sounds.

Cette étude vise à examiner le développement phonético-phonologique du français et plus précisément la progressive organisation du système vocalique chez des enfants bilingues préscolaires exposés à l’une des trois combinaisons linguistiques suivantes : français-italien, français-arabe et français-mandarin. Les productions orales des enfants ont été recueillies longitudinalement via une tâche de dénomination originale en français. La dominance linguistique des enfants ainsi que leur développement lexical ont été évalués via des questionnaires parentaux. L’évolution de l’organisation du système vocalique a été évaluée via l’indice PHI (Huet & Harmegnies, 2000) permettant de générer une mesure de dispersion des valeurs formantiques dans l’espace vocalique. L’impact de la combinaison et de la dominance linguistiques ainsi que du développement lexical sur l’organisation du système vocalique a été investigué. Les résultats indiquent : (1) un système vocalique initialement mieux organisé chez les bilingues français-italien, (2) une augmentation parallèle de l’organisation du système vocalique et de la compétence lexicale chez les bilingues français-italien et français-arabe et (3), aucun effet de la dominance linguistique. Ces résultats enrichissent les connaissances sur le développement de la parole en français dans des contextes de bilinguisme simultané contrastés.

Purpose: This study conducted a transcription-based and spectral moments' analysis of alveolar and alveopalatal fricatives in monolingual and bilingual French-speaking children, aged 2;6–6;10 (years;months). We measured the percent accuracy of fricatives and investigated whether young children could distinguish alveolar and alveopalatal fricatives on the basis of spectral moments. In addition, we examined which child- (i.e., age, gender, bilingualism, and alveopalatal fricative inventory size) and word/sound-related (i.e., place-of-articulation [PoA], voicing, vowel quality, and word position) factors influenced spectral moments and fricative duration. Method: Children ( N = 89) participated in a picture-naming task in which they produced words containing alveolar /s, z/ and alveopalatal /ʃ, ʒ/ fricatives in word-initial, -medial, and -final positions. The words were transcribed and analyzed acoustically, and the first and third spectral moments (i.e., centroid and skewness) and the duration of fricatives were calculated. The data were subject to mixed-effects linear regression. Results: Percent accuracy results indicated effects of age on alveopalatal fricatives and effects of word position on voiced fricatives. Statistical models indicated that age, gender, and alveopalatal fricative inventory size influenced spectral moments. Age and inventory size interacted significantly with PoA. Children as young as age 2;6 distinguished alveopalatal and alveolar fricatives on the basis of centroid but not skewness values. The distinction between the two sets of fricatives increased with age. Bilingual children who spoke languages with greater numbers of alveopalatal fricatives distinguished alveopalatal and alveolar fricatives less well than monolinguals and bilinguals who spoke languages with fewer numbers of alveopalatal fricatives. Girls had higher centroid and lower skewness values than boys. Models also revealed a significant influence of word/sound-related factors (voicing, vowel quality, and word position) on spectral moments and fricative duration. Conclusions: Findings indicated that multiple factors influence the spectral moments and duration measures of children's alveolar and alveopalatal fricatives. In particular, we found that spectral moments were sensitive to gender and bilingualism effects.

Notre recherche vise à étudier le développement phonologique et phonétique, en production de parole, d'enfants bilingues préscolaires ayant différentes combinaisons linguistiques. Notre objectif est d'évaluer l'évolution des habiletés de production et des réalisations phonétiques des enfants dans une approche translinguistique et écologique. Dans cette optique, nous avons élaboré un protocole expérimental original et adapté. Les productions orales d'enfants bilingues sont recueillies longitudinalement via une tâche de dénomination en français. Le corpus cible des structures phonologiques spécifiques et implique un ordre de présentation particulier des items, par âge d'acquisition croissant et complexité phonologique progressive. Soumises à des procédures de codage spécifiques dans Praat, les productions récoltées font l'objet d'une transcription phonétique fine permettant d'isoler des phénomènes particuliers. Nous présentons les résultats préliminaires des premières analyses qualitatives pour trois enfants bilingues français-italien sur base de deux recueils de production de parole pour chaque participant.

No abstract