Autistic adults display different verbal behavior only in mixed-neurotype interactions: Evidence from a referential communication task.
Autistic adults get wordy only with non-autistic partners, so watch pair chemistry, not just individual goals.
01Research in Context
What this study did
Geelhand et al. (2025) paired adults in a picture-description game over Zoom. Some pairs were both autistic, some both non-autistic, and some mixed.
No one knew the partner’s diagnosis. The researchers counted words and time to see how talk changed with partner type.
What they found
Mixed pairs talked the longest. Autistic speakers used the most words when paired with a non-autistic partner.
Same-neurotype pairs finished faster. Partner type mattered more than autistic traits alone.
How this fits with other research
Boo et al. (2022) saw a similar pattern in kids. During VR chat, autistic and ADHD children used simpler language as the game got harder. Both studies show task plus partner shapes talk.
Marsack-Topolewski et al. (2025) looked at the listening side. Autistic adults learned new word sounds just like non-autistic peers. Philippine’s extra talking is not about poor hearing; it is about social fit.
Northup et al. (1991) and Cohen et al. (1990) first noted odd word use and partner effects in children. The new study shows these early quirks continue into adulthood and appear only in mixed pairs.
Why it matters
If you run social-skills groups, balance pairings. An autistic client may flood a non-autistic peer with words, not lack of skill. Teach both partners to share airtime and use clear stop signals. Track talk time as a data point, not just correct answers.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Time a mixed pair for one minute each way; teach a ‘pause’ card if talk ratio tops 70-30.
02At a glance
03Original abstract
Recent accounts of social difficulties in autism suggest that autistic and non-autistic individuals mutually misunderstand each other. This assumption aligns with findings that mixed-neurotype interactions are less efficient than same-neurotype interactions. However, it remains unclear whether different outcomes between mixed- and same-neurotype interactions are due to contact with a different neurotype or to inherently different communication styles, specific to each neurotype. A total of 134 adult participants were divided into three same-sex dyad types: 23 autistic dyads, 23 non-autistic dyads, and 21 mixed-neurotype dyads. Participants were unaware of their partner's neurotype. Dyads completed an online referential communication task where a "Director" guides a "Matcher" to rearrange abstract images, using both written (chat) and oral (microphone, no video) communication modes. Interaction outcome measures were task duration and verbosity of the Director. Across both communication modes, non-autistic dyads completed the task faster than autistic and mixed dyads, indicating that dyads with at least one autistic partner were generally slower. Notably, in mixed dyads, autistic Directors were more verbose than non-autistic Directors across both communication modes. These results, in conjunction with partners' unawareness of each other's neurotype, suggest that even in the absence of non-verbal cues neurotype mismatch triggers autistic adults to display different verbal behavior.Lay abstractRecent research shows that in conversations, both participants influence the outcome. More specifically, conversations do not go as smoothly when autistic and non-autistic people talk together compared to when people of the same neurotype (either all autistic or all non-autistic) talk to each other. In studies finding a "same-neurotype communicative advantage", interaction partners knew about each other's neurotype. Because of this methodological choice, it is unclear whether mixed-neurotype interactions go less smoothly because participants knew they were interacting with a different neurotype or because each neurotype really has a distinct communication style. In our study, 134 adults were grouped into same-sex pairs: 23 autistic, 23 non-autistic, and 21 mixed-neurotype pairs. The pairs did not know if the other person was autistic or not. They completed an online task where the "Director" instructs the "Matcher" to reorder abstract pictures. Pairs did this task in two ways: by typing in a live chat and by speaking into a microphone without video. The study looked at how long the task took and how much the Director talked/wrote. Results showed that non-autistic pairs were faster to complete the task than autistic pairs and mixed pairs, meaning pairs with at least one autistic person were slower in general to complete the task. Interestingly, in mixed pairs, only autistic Directors produced more words than non-autistic Directors, in both typing and speaking. These findings suggest that even without knowing about their partner's neurotype and seeing/hearing their partner, autistic adults communicate differently when they interact with a non-autistic person.
Autism : the international journal of research and practice, 2025 · doi:10.1177/13623613241298376