Dyadic Interaction Between Bilingual Parents and Their Young Children with Autism Spectrum Disorders

Background: Given concerns that dual-language exposure might confuse children with Autism Spectrum Disorders (ASD), bilingual parents may choose to restrict a child’s exposure to a single language. The empirical evidence to date suggests no particular negative effects of dual-language exposure for children with ASD (e.g., Reetzke et al., 2015; Valicenti-McDermott et al., 2013). However, there is also evidence that some professionals advise against dual-language exposure, and suggest parents should engage the child using the community dominant language (e.g., Ijalba, 2016; Yu, 2015).

Objectives: We investigated a potential side effect of the decision to restrict parent-child interaction to community-dominant English – the possibility that parental non-native language use might alter interaction behaviours usually considered to facilitate child development.

Methods: We recruited 39 dyads – each including a child with ASD (67% boys; M_age = 4 years 4 months) and their parent (87% mothers) – into two groups. Around half of parents (n=20) were monolingual English speakers, and 19 were bilingual with English as the common non-native language. The ADOS was administered to confirm child diagnosis, and the Mullen Scales of Early Learning were used to ascertain developmental level, with children in each group matched on ADOS Calibrated Severity Score and Non-Verbal Developmental Quotient. Key measures of interest for this study concerned the parent. English language competence was assessed via a standardised measure of expressive vocabulary knowledge. Free-play interaction samples were filmed with each dyad for later coding. One sample was filmed for monolingual adults (i.e., who spoke in English) while two samples were filmed for bilingual adults (i.e., one each using their native language and non-native English). Immediately following each interaction sample, parents rated their comfort on 7-point Likert-type scales. Off-line, we then coded parental communicative synchrony (e.g., Hudry et al., 2013) and scaffolding of child language (e.g., Haebig et al., 2013).

Results: The parents in each group were matched on chronological age and most were tertiary-educated. Nevertheless, bilingual parents had significantly poorer non-native English-language vocabularies compared to monolingual parents. Further, during free-play interaction, bilingual parents were less communicatively synchronous with their children, used less verbal imitation of child speech, and modelled grammatically simpler language than did monolinguals. This was apparent irrespective of which language bilingual parents used, with little evidence – in this sample – of specifically altered interaction behaviours when using non-native English versus a native language. However, this unanticipated finding may be explained by the apparent presence of subgroups within our bilingual sample – one reporting using English during everyday interactions with their child with ASD (n=11), and the other reporting maintained native-language use (n=8).

Conclusions: These novel empirical data provide partial support for some accounts provided in previous qualitative studies of the perceived impact of bilingualism and non-native language use for parental interaction with young children with ASD (e.g., Wharton et al., 2000; Yu, 2013, 2015). With rates of both ASD diagnosis and cultural/linguistic diversity growing internationally, understanding the potential impact of language-use choices by parents and language exposure patterns for children clearly warrants further dedicated attention.