New data, published in the Journal of Child Psychology and Psychiatry, suggest that pronoun use during natural conversation might inform us about clinically meaningful social function. Amber Song and colleagues set out to understand whether they could identify natural language markers of social characteristics exhibited by girls and boys with autism. To do so, they recruited 50 school-aged autistic children (17 girls and 33 boys who were matched for level of social impairment) and 37 typically developing children who were matched for age (mean age = 11.35 years) and IQ (mean IQ = 107). They then asked the children to engage in a short, informal conversation. From the conversation sample, the researchers analysed the use of first person (e.g., “we” and “us”) and third person (e.g., “they” and “them”) plural pronouns to determine whether autistic girls and boys talk differently about social groups.
“This finding is especially important because girls with autism are not as well understood as boys”
Overall, autistic children used significantly fewer plural pronouns than their typically developing peers, indicating that they talked less about social groups. However, autistic girls talked more about social groups than autistic boys. Yet compared to typically developing girls, autistic girls talked more about groups they were not a part of; this was indicated by more frequent use of the pronouns “they” and “them” and less frequent use of “we” and “us”. The researchers speculate that this finding might demonstrate potential awareness of social exclusion in affected girls.
Finally, Song et al. assessed whether pronoun use could predict ADOS-2 scores in children with autism. After controlling for age and IQ, they found that “we” and “they” variants significantly predicted ADOS-2 social affect but not repetitive behaviour scores in girls. Thus, pronoun use might predict individual differences in social function in autistic girls.
“This finding is especially important because girls with autism are not as well understood as boys”, explains study author Julia Parish-Morris. “When we learn more about how girls with autism behave in the real world, it helps us develop personalized supports to address their unique needs most effectively”. Although this study was relatively small, the researchers believe the data serve as a proof-of-concept that natural conversation samples might inform us about clinically meaningful social similarities and differences between girls and boys with autism.
Song et al. now plan to expand their work to include more participants and to collect data using semi-structured question-and-answer formats. From here, they hope to determine whether the identified patterns hold true for other age groups and those with lower, or higher, IQ scores. Future research might then test whether subtle linguistic clues can help improve identification of autism in girls.
Song A et al (2020). Natural language markers of social phenotype in girls with autism. J Child Psychol Psychiatry, 62, 949–960.
- CAMHS around the Campfire: Trajectories of autistic social traits in CYP and disordered eating at age 14 years
- Webinar for parents: Tackling low diagnosis for autism in girls, A Ozsivadjian & M Murin,
- JCPP Video Abstract: Anorexia nervosa and autism: a prospective twin cohort study, L Dinkler
How you can help
Do please share with your networks and get involved by;
- completing this 2min survey
- following us on Twitter @TheBridgeACAMH
- to contribute, and to give feedback, please email TheBridge@acamh.org