Grammar Reaches Into Reading

What the research found

For decades, the dominant account of reading difficulty has centred on phonology: the ability to hear, hold, and manipulate the sounds inside words. This study asks a quieter question. Once you set phonology aside, does a child’s grasp of grammar still shape how accurately they read? In Italian, the answer appears to be yes.

The researchers studied 42 Italian children in grades two to five. Fourteen had a diagnosis of developmental dyslexia; twenty-eight were typically developing peers of similar age. Every child completed three kinds of task. One measured how accurately and quickly they read text aloud. One measured whether they understood sentences of varying grammatical complexity, by matching a spoken sentence to the correct picture. One measured whether they could produce a specific and demanding grammatical form: the third-person direct-object clitic pronoun.

That last item deserves a plain-language gloss. A clitic pronoun is a small, unstressed word that stands in for something already mentioned. In English, think of the “it” in “I saw the dog and I fed it.” In Italian, these pronouns are even smaller and carry more grammatical weight at once, marking person, gender, number, and grammatical role. They are quiet in the mouth but heavy in structure. That combination is exactly why the authors found them useful.

The reading gap between groups was clear. Children with dyslexia read less accurately than their peers, and the difference was statistically reliable [F(1,40) = 8.58, p = 0.006]. They also read more slowly, and that gap was larger still [F(1,40) = 16.73, p < 0.001]. None of this is surprising; it is what a dyslexia diagnosis describes.

The novel result sits in what predicted reading accuracy. When the researchers modelled accuracy against grammatical ability, both syntactic comprehension [F(1,38) = 14.14, p = 0.001] and the production of clitic pronouns [F(1,38) = 6.72, p = 0.013] emerged as significant predictors. In other words, children who understood complex sentences better, and who produced these demanding little pronouns more successfully, read more accurately, regardless of which group they were in. Reading speed told a different story: there, only group membership predicted the outcome [F(1,38) = 12.25, p = 0.001], with grammar dropping out.

This matters because Italian is what linguists call a shallow orthography. Its letters map onto sounds with unusual consistency, so a child who has cracked the sound-to-letter code has, in principle, most of what phonics can offer. That grammatical ability still predicted accuracy in such a language is the study’s central point. It suggests that morphosyntactic knowledge, a child’s working sense of how grammar is built, contributes to accurate decoding through a route that runs alongside phonology rather than through it.

What it means for you and your child

Read carefully, this is a modest and honest study, and it is worth being clear about its limits before drawing lessons from it. Forty-two children, spread thinly across four school grades, is a small sample. The authors say so plainly and call for larger, longitudinal work. They also tested a narrow set of grammatical structures under favourable conditions. Most importantly, the design is correlational. It shows that grammar and reading accuracy travel together; it cannot prove that stronger grammar causes better decoding. The relationship could run in either direction, or a third factor could drive both. The authors are careful here, and so should we be.

There is also the question of language. Italian’s clean sound-to-letter mapping is precisely what made the finding striking, but it also means the result does not transfer automatically to English, whose spelling is famously irregular. What holds in a shallow orthography may look different in a deep one. Treat this as a signal from one language, not a settled fact about all reading.

With those cautions in place, the practical takeaway is a shift in attention rather than a new programme. If your child reads inaccurately, the instinct is to reach for more phonics. This study is a gentle reminder that decoding accuracy may also lean on oral language, on the grammatical scaffolding a child brings to the page. A child who stumbles over the small connective machinery of sentences in speech may carry that difficulty into reading.

The authors themselves draw a measured clinical implication. They suggest that when a child is assessed for dyslexia, it is worth looking at the linguistic profile as well as the phonological one, to see whether some reading errors trace back to grammatical rather than sound-based difficulty. Where they do, targeted work on recognising the small function words and grammatical markers in text might help accuracy. This is a hypothesis the study points toward, not a proven treatment.

For a parent, the sensible response is neither alarm nor a new regimen. It is a wider lens. Rich conversation, being read to, and unhurried talk about how sentences fit together are not separate from reading; they may feed it. If a professional is assessing your child, it is reasonable to ask whether oral language and grammar have been looked at alongside phonics. One small study cannot tell you what to do. It can, usefully, tell you where else to look.