Do autistic adults spontaneously reason about belief? A detailed exploration of alternative explanations.
Autistic adults notice social cues but do not spontaneously predict others’ false beliefs, pointing to a core mentalizing deficit.
01Research in Context
What this study did
The team asked autistic and neurotypical adults to watch short videos. A character put an object in one box, then left. Another person moved the object. Viewers’ eyes were tracked to see where they expected the first person to look.
No one told the adults to predict anything. The test measured if they automatically reasoned about the character’s false belief.
What they found
Neurotypical adults quickly glanced at the original box, showing they expected the character to look there. Autistic adults kept their eyes still. They noticed the social scene but did not use it to guess the character’s next move.
The result points to a specific mind-reading gap, not a general attention problem.
How this fits with other research
Bradford et al. (2018) saw the same pattern with a computer task: autistic adults were slower to switch between real and believed locations. Together the studies form a steady line of evidence.
Boucher (2012) warned that classic false-belief tasks may miss real-life skills. The new eye-tracking method answers that worry by testing belief reasoning without questions or prompts.
Fernandes et al. (2022) found dampened brain responses to social cues. van der Miesen et al. (2024) now show the behavioral outcome: the cue is seen but not used for prediction.
Why it matters
When an autistic client seems uninterested in social hints, the gap may lie in automatic prediction, not in attention or will. Teach belief reasoning explicitly: state what the other person knows, show the hidden move, ask where they will look. Model the update out loud. Over time pair the lesson with natural cues so the skill transfers beyond the clinic room.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Before a hidden-object game, narrate the belief out loud: “John thinks the keys are in the drawer.” Point, pause, then let the client predict where John will search.
02At a glance
03Original abstract
Southgate <i>et al.</i>'s (Southgate 2007 <i>Psychol. Sci.</i> 18, 587-92 (doi:10.1111/j.1467-9280.2007.01944.x)) anticipatory-looking paradigm has presented exciting yet inconclusive evidence surrounding spontaneous mentalizing in autism. The present study aimed to develop this paradigm to address alternative explanations for the lack of predictive eye movements on false-belief tasks by autistic adults. This was achieved through implementing a multi-trial design with matched true-belief conditions, and both high and low inhibitory demand false-belief conditions. We also sought to inspect if any group differences were related to group-specific patterns of attention on key events. Autistic adults were compared with non-autistic adults on this adapted implicit mentalizing task and an established explicit task. The two groups performed equally well in the explicit task; however, autistic adults did not show anticipatory-looking behaviour in the false-belief trials of the implicit task. Critically, both groups showed the same attentional distribution in the implicit task prior to action prediction, indicating that autistic adults process information from social cues in the same way as non-autistic adults, but this information is not then used to update mental representations. Our findings further document that many autistic people struggle to spontaneously mentalize others' beliefs, and this non-verbal paradigm holds promise for use with a wide range of ages and abilities.
, 2024 · doi:10.1098/rsos.231889