A Large-Scale Study on the Relation Between Multilingualism and Social Traits in Adults With and Without Autism Diagnoses.
Margreet Vogelzang, Sean Dong Wei Gan, Jamie Setch
Greater diversity of language use was associated with lower social interaction difficulty in autistic adults, but no effect on communication skills.
The link was stronger in the diagnosed group. This is one questionnaire study, not proof that multilingualism causes the difference.
Key findings
1In the full sample of 459, childhood language entropy was associated with lower Social Interactions scores (ηp² = 0.04, p < 0.001) and current language entropy with lower overall CATI scores (ηp² = 0.02, p = 0.001). No entropy measure predicted Communication scores.
2In the 132 participants with an autism diagnosis, current language entropy was associated with lower overall CATI scores (ηp² = 0.14) and lower Social Interactions scores (ηp² = 0.18). The best model for Communication was empty.
3Age of second-language acquisition and L2 proficiency did not improve any model beyond language entropy; the diversity of language use, not timing or fluency, was the relevant factor.
Still to come
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Cross-sectional design: the study shows an association, not a cause. People who grew up multilingual may differ from monolinguals in many ways beyond language exposure. Autism diagnosis was self-reported, not confirmed by clinical assessment. The 139 'suspected' participants were not included in the diagnosed-subset analysis. All language-use data were self-reported; childhood entropy is subject to recall bias. The questionnaires were verbal and not suitable for non-verbal individuals, so the sample excludes a portion of the autistic population. UK-based sample; the authors note the CATI has been validated in other languages but this is the first study with a substantial multilingual sample. Effects in the full sample are small (ηp² = 0.02–0.04, explaining 2–3% of variance). The larger effects (ηp² = 0.14–0.18) are in a subset of 132. No a priori power calculation was performed. The authors state that replication in different samples is needed before any clinical recommendations can be made.
Declared interests
Funded by an ESRC IAA Ideas to Impact Award (Economic and Social Research Council). The authors declare no conflicts of interest.
The easy way to misread this
Do not read the ηp² = 0.14–0.18 in the diagnosed subset as evidence that multilingualism reduces social difficulties in autism. The full-sample effects are small (ηp² = 0.02–0.04), the design is cross-sectional so direction of causation is unknown, the diagnosis is self-reported, and the authors themselves state that replication is needed before clinical recommendations can be made. The association may reflect unmeasured differences between multilingual and monolingual families rather than a causal effect of language diversity.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →