Partner-dependent communication without dynamic adaptation in autism

Autism is a neurodevelopmental condition characterized by persistent communication challenges, often compounded by social anxiety. Although differences in perspective-taking, cognitive flexibility, and social motivation have been implicated in these challenges, their influence on live interpersonal interactions remains unclear. In this study, we quantitatively examined how autistic and non-autistic individuals with varying levels of social anxiety adapted their communication during experimentally controlled interactions with two ostensibly distinct partners-a child and an adult-both portrayed by the same role-blind confederate. Autistic participants were equally motivated and capable as non-autistic participants in adjusting their communication to stereotypical assumptions about a partner's abilities, spontaneously using greater emphasis when addressing the presumed less capable child. However, they were less likely to modify these stereotype-driven behaviors in response to interaction-based evidence of partners' equal competence. While non-autistic participants dynamically adapted their communication to treat both partners equivalently, autistic participants maintained their stereotype-driven adjustments throughout the interaction. Preregistered analyses further linked non-autistic individuals' adaptive responses to early social exposure, a developmental factor not observed in autistic participants. Together, these findings highlight a core interactional capacity, shaped by early social experiences and operating on interaction-based evidence, as central to understanding communication challenges in autism.Lay abstractEveryday communication can be challenging for autistic individuals, particularly when social anxiety is involved. Research suggests that differences in understanding and adapting to others may contribute to these challenges, but it remains unclear whether and how these differences affect real-time interactions. This study invited autistic and non-autistic participants with varying levels of social anxiety to interact online with two "partners." One partner was introduced as a child and the other as an adult, although in reality the same actor played both roles without knowing which role he was assigned. All participants initially emphasized their communication more with the presumed child, whom they assumed was less capable. Over time, however, non-autistic participants adapted their approach, treating both partners equally as they gathered evidence that the child was just as competent as the adult. In contrast, autistic participants continued to treat the child as less capable throughout the interaction. Moreover, non-autistic participants who adapted more quickly tended to have experienced greater early social exposure in daycare, a relationship that was not observed in autistic participants. These findings suggest that while autistic individuals are willing and able to adjust their communication based on initial assumptions about others, they are less likely to revise these adjustments in response to evidence of a partner's actual understanding during interaction, a skill that appears to develop differently for them from an early age.