Autistic children's use of semantic common sense and theory of mind: a comparison with typical and mentally retarded children.

The team tested the preschoolers: 20 with autism, 20 with Down syndrome, and 20 typical kids. All could speak in short sentences. Each child heard short stories and answered two kinds of questions. One set checked if the child knew real-world facts (semantic common sense). The other set tested false-belief understanding (theory of mind).

Autistic kids scored just as high as controls on the real-world fact questions. They knew, for example, that 'we wear shoes on our feet.' On false-belief tasks, however, only 30 % of the autistic group passed, compared with 80 % of the other two groups. There was no link between how well a child did on the two tasks.

Carr (1994) saw the same split in older, higher-functioning children. Even those who passed classic false-belief tests still failed naturalistic story questions about feelings and thoughts. The pattern holds across ages.

Ploog et al. (2007) later showed that autistic children remember semantic details just fine. Together the papers build a clean picture: semantic encoding and comprehension stay intact, while theory-of-mind reasoning lags.

Vierck et al. (2015) seems to disagree. Their high-functioning Mandarin-speaking sample passed scalar-implicature tasks that also tap pragmatics. The difference is task demand: Esther used simple picture-choice; Mika used open false-belief questions. Kids can succeed when the response is easier.

You can quit drilling basic vocabulary once a child shows they understand sentences. Instead, target the missing piece: perspective-taking. Use visual supports, role-play, and explicit false-belief stories. The 2023 follow-up by Hsiu-Man et al. shows early ToM gains forecast later social success, so start now.