Social and sensory influences on linguistic alignment

Abstract

Previous research has demonstrated that speakers adapt individual characteristics of speech production to the social context, for example via phonetic convergence. Studies have measured the impact of social dynamics on convergence in typical speakers, but the impact of individual differences is less well-explored. The present study measures phonetic convergence before and after a cooperative interaction with an undergraduate student by comparing teens with a history of autism spectrum disorder (ASD) and with typical development. Results revealed a small temporal convergence effect for typically developing dyads; in contrast, dyads including autistic teens showed no convergence. Furthermore, the degree of convergence was negatively related to severity of ASD symptomatology and atypical sensory profiles. While social and sensory processing abilities may play a role in phonetic convergence, high covariance between these measures makes it difficult to differentiate their contributions in isolation.

Phonetic entrainment in L2 human-robot interaction: an investigation of children with and without autism spectrum disorder

“…In addition to the findings of vowel formants, autistic children entrained their mean f0 comparably to their TD peers. This result was inconsistent with some previous studies, where neither autistic nor TD children showed prosody entrainment (Hogstrom et al, 2018;Wynn et al, 2018). As most studies computed the differences between conversation partners to indicate phonetic entrainment, it is very likely that their findings about entrainment or lack of entrainment was actually driven by the adjustment of their interlocutors.…”

“…In addition, children as young as 9 years old were found to converge in mean f0 in "spot-the-difference" games in Lehnert-LeHouillier et al (2020). Hogstrom et al (2018) also reported convergence of phoneme duration from children aged from 12 to 18 in a cooperative map searching task.…”

“…Studies on phonetic entrainment of autistic children showed mixed results. Hogstrom et al (2018) found that TD children converged in their phoneme duration in the post-interaction period while autistic children showed a trend of divergence. However, some studies reported a tendency of similar unchanged phonetic adjustment between autistic and TD children, for example, similar unchanged adjustment in speech rate (Wynn et al, 2018) and f0 range (Lehnert-LeHouillier et al, 2020).…”

“…In the study of Wynn et al (2018), although they found significant differences of speech rate convergence between autistic and TD adults, autistic children and TD children's speech rate did not show significant differences. Hogstrom et al (2018) compared the phoneme duration of keywords before and after a conversation task with a TD interlocutor. They reported that autistic children tended to diverge in phoneme duration from the interlocutor after the task, compared with the pre-task production, while TD children showed convergence.…”

Phonetic convergence of children with and without autism in robot interaction

“…Particularly, we carried out a study on phonetic convergence between children with Autism Spectrum Disorder (ASD) and typical developing (TD) children. Atypical social behaviour is one core feature of ASD population (American Psychiatric Association, 2013), and it has been reported that they lack phonetic convergence in conversations with human beings (Hogstrom et al, 2018;Wynn et al, 2020). Therefore, we raised the following research questions: 1) do ASD and TD children show phonetic convergence in a conversation with a social robot?…”