The Role of Linguistic Input in Person-Reference of Preschoolers with ASD

Background: Numerous studies have examined the effects of linguistic input on children with autism’s expressive language (e.g.,Fusaroli et al.,2018). In the domain of person-reference, or how a speaker refers to themselves and their conversational partner: using pronouns (e.g., I, you) or nouns/names (e.g., kiddo, mom, John), one study showed that the parental input of toddlers at high risk for autism was characterized by a higher use of the child’s name, especially in the context of attracting the child’s attention, compared to the input of low-risk toddlers (He et al.,2018). Furthermore, studies have provided evidence for children with autism’s avoidance of pronouns and preference for using their own name (e.g.,Shield et al.,2015). Nevertheless, no study has directly examined the effect of parental input on person-reference in autism.

Objectives: We aimed to examine the relation between preschoolers with autism and their parents’ person-reference. We focused on amount of person-referential language and on preference for one form of reference, pronouns, over the other, nouns/names, both concurrently and across time.

Methods: We collected language samples from children with ASD (N=38; 7female) and their parents during free play at three time points: T1:M=27.13months, T2:M=39.63, and T3:M=51.68. The samples were transcribed following standard SALT transcription procedures, and measures of person-reference were extracted for both children and their parents: proportion of person-reference words out of total words to reflect amount of person-reference, and proportion of pronouns and of nouns/names out of total words, and proportion of pronouns used over total number of person-referential words (pronoun and nouns/names) to capture the trade-off between the two forms of referential language. At each time point, children’s ASD diagnosis was confirmed with the ADOS, and their cognitive and language ability was assessed with the Mullen(Table-1).

Results: There were no significant correlations between children’s person-referential language out of total number of words and that of their parents either concurrently or across time points. Children used more pronouns than nouns/names out of total words at Time 2 and 3, and the proportion of pronouns out of total person-referential words did not change over time (Table-2). Parents used significantly more pronouns at each time point as well, and their preference strengthened across time (χ²(2)=7.913,p=.019). There were no significant correlations between children’s proportion of pronouns and that of their parents either concurrently or across time after controlling for children’s language ability and correcting for multiple comparisons.

Conclusions: In contrast to past research showing general influence of linguistic input on expressive language in autism, we found no associations between parents’ and children’s person-reference either in amount or in preference for one form of reference over the other. Children used significantly more pronouns and this preference did not change across time. These results combined with the children’s relatively strong expressive language as evidenced by their Mullen scores suggest that the factors contributing to the use of person-reference in ASD may be primarily non-linguistic after a certain baseline language ability is achieved. These findings can inform clinical advice to parents about ways of addressing their children with ASD.