Narrative role-taking in autism.

Geurts et al. (2008) asked autistic and neurotypical kids to tell stories. They wanted to see who could hop inside different characters’ heads while talking.

Each child heard short tales. Then they retold them using the right point of view. The team scored how well kids switched among ‘I’, ‘he’, or ‘she’ voices.

Autistic children knew the rules but still scored lower. They stayed stuck in one voice and rarely shifted to another character’s thoughts.

Neurotypical kids moved smoothly between voices. The gap stayed even when both groups understood the story equally well.

Baixauli et al. (2016) looked at 24 similar studies. Their meta-analysis backs M et al.: autistic children almost always tell weaker stories.

Llanes et al. (2020) moved from oral to written work. They found the same pattern in personal essays. Theory-of-mind scores predicted the writing gaps.

Callenmark et al. (2014) seems to clash at first. They saw no group difference on explicit social tests. But their hidden, spontaneous tasks still caught autistic teens slipping—just like M et al. caught slips in natural storytelling.

If you run social-skills or language groups, don’t stop at teaching story parts. Build fast voice-switch drills: ‘Now tell it like the dog, now like the owner.’ Quick perspective hops build flexible social thinking that shows up in real conversation.