Can autistic children read the mind of an animated triangle?

Salter et al. (2008) showed autistic children short cartoons of two triangles moving on a screen.

The kids had normal IQ and could speak.

Researchers asked the children to describe what the triangles were doing and thinking.

The autistic children used mental-state words like "want" or "think" just as often as typical peers.

Yet their stories about the triangles were far less accurate and social.

They missed the subtle social plot the shapes were acting out.

Kerr et al. (2004) saw a similar pattern with thought-bubble pictures: autistic preschoolers could talk about thoughts but still failed classic false-belief tasks.

Ohan et al. (2015) and Sharp et al. (2010) later showed that Mind Reading software plus practice can boost emotion-recognition scores, suggesting the gap can be narrowed.

Shearn et al. (1997) tried direct theory-of-mind lessons and found no real-world conversation gains, reminding us that lab skills do not automatically transfer to the playground.

A child who can say "he thinks" may still miss the real social story.

Probe social understanding with dynamic scenes, not just vocabulary lists.

Pair mental-state language goals with live peer practice so the skill shows up where it counts.