Sentence comprehension strategies in children with autism and specific language disorders.

Researchers compared how autistic kids, kids with language disorders, and younger typical kids figure out sentences. They used picture-choice tasks that pitted word order against real-world sense.

For example, the sentence 'The pencil kicks the dog' breaks real-world rules but follows English word order. Kids had to pick which picture matched.

Autistic children picked the picture that matched word order, even when it made no real-world sense. They leaned on syntax, not on what usually happens in life.

The pattern looked like the language-impaired group and like younger typical kids. All three groups used the same backup plan: follow the grammar when meaning is tricky.

Naito et al. (2004) seems to disagree. Their verbal preschoolers with autism used real-world sense just fine. The gap is age and language level: the 2004 kids were younger and had to show verbal skills to join, while the 1988 study mixed ages and did not screen for language level.

Ploog et al. (2007) also pushes back. They found autistic kids store and recall semantic info like peers, showing no deep semantic deficit. The 1988 task, however, tested quick choices under conflict, not memory.

Kritsotakis et al. (2026) extends the idea. They show that morphosyntax still predicts figurative gaps in older autistic kids. The 1988 finding—that syntax is the go-to lever—holds up even for metaphors a generation later.

When you give instructions, keep the grammar clear even if the context feels obvious. Say 'First touch the red car, then the blue block' rather than 'Touch the one you play with.' Check whether the child can use real-world cues by testing both types of sentences. If they default to word order, teach semantic cues explicitly with visual supports and contrastive trials.