Blogs About Reading

Shanahan on Literacy

Timothy Shanahan

Literacy expert Timothy Shanahan shares best practices for teaching reading and writing. Dr. Shanahan is an internationally recognized professor of urban education and reading researcher who has extensive experience with children in inner-city schools and children with special needs. All posts are reprinted with permission from Shanahan on Literacy.

Why We Need to Teach Sentence Comprehension

October 5, 2020

For the want of a nail the shoe was lost;

For the want of a shoe the horse was lost;

For the want of a horse the battle was lost;

For the failure of battle the kingdom was lost —

And all for the want of a horseshoe nail.

— Proverb

This oft-used litany reminds me of reading:

For the want of phonemic awareness the decoding was lost; for the want of phonics the fluency was lost … you get the idea. The abilities that comprise reading are hierarchical, each nested in the other (though it is not as linear as the horseshoe nail formula — we don’t completely accomplish a reading step before the onset of the later ones, and those later steps can enhance the earlier ones; phonemic awareness, for example, is easier to accomplish when the phonics that it enables, is itself being taught).

Over the years, I’ve written a lot about letters and phonemes, decoding, fluency, vocabulary, and the like. Recent research (Sorenson, et al., 2020) reminds me of an important step in the learning sequence that we tend to skip. Reading researchers have assiduously explored the importance of vocabulary and text structure in reading comprehension, as well they should; these are important aspects of language that have been found to facilitate the ability to understand text. But between these two linguistic extremes (the smallest chunks and the largest), there is the seemingly unloved sentence.

Correlational studies have long demonstrated that one’s ability to negotiate the meaning of sentences is connected to reading comprehension. This connection has been shown by comparing performances with texts that vary in their sentence complexity (think of all the studies of readability), by correlating the results of grammar tests and reading comprehension tests, and by evaluating good and poor reading comprehenders’ ability to understand particular oral sentence structures (as in the recent study, that explored passive and active sentences). Sorenson and colleagues reported that passive sentences were markedly harder for fifth graders to understand.

Despite the long history of such research, that has not translated into substantial efforts to improve students’ comprehension through sentence instruction.

Part of the reason for that may be due to the long-noted failure of explicit grammar instruction to improve writing quality or reading comprehension (e.g., Braddock, Lloyd-Jones, & Schoer, 1963). If grammar instruction doesn’t help, then why pursue the issue?

At one time, I would have agreed with that. I can’t say I took to formal grammar instruction much as a boy, and in fact, I considered it to be quite a pain in my nether reaches. As children we were tortured with sentence diagramming exercises that I still don’t really understand when one gets much beyond the declarative sentences of the Hemingway variety.

But I’ve come to believe that the issue is more subtle and that the expectation that general grammar instruction should enhance reading or writing for native speakers is somewha simplistic. Readers must be able to understand sentences, but they must do so like proficient language users, not linguists. If a student can construct sentences that make sense and tease out the meanings of those sentences they confront in texts, then I don’t care much whether they can explain the difference between an infinitive and a participle or know what a gerund is.

I’m not rejecting the value of formal grammar instruction altogether either. It clearly helps when one is studying a second language, at least with regard to sentence constructions that differ across languages. For instance, in English a simple sentence may follow the sequence: Subject – Verb – Direct Object… while in French, it would be Subject – Direct Object – Verb. It can help to have somebody point that out. (If you’re French and trying to learn English you don’t want to say, “John him called”).

Steve Graham helpfully pointed out in his meta-analyses on writing instruction that while formal grammar had a negative effect size (meaning the comparison groups outperformed the grammar groups), unlike the other approaches to instruction, grammar was always in the role of control group. What this means is that grammar was never tested in a circumstance in which the researchers were striving to make it work. All the new materials, professional development, classroom visits, and the like were showered on the alternative approach being touted by the researchers. Perhaps if someone set out to make formal grammar teaching work, it might fare better in such studies.

But even if not, it strikes me that instruction in how to make sense of sentences could play an important role in reading comprehension.

We don’t monitor students’ comprehension of text especially closely. Oh, we evaluate comprehension both formally (e.g., standardized tests) and informally (e.g., classroom discussions, teacher questions). But we aren’t especially attentive to the potential sources of the misunderstandings. Where did the students go wrong?

If we recognize that students may struggle with sentences written in the passive voice, then it would behoove us to teach reading with some texts that use this difficult construction. Our monitoring of student success in this case would not simply pursue general questions about the ideas in the text. They would zero in on the ideas expressed in those passive voice sentences to see if that was part of the problem. Obviously the same could be done with all kinds of grammatical constructions (several problematic ones have been identified in the research literature).

When students fail to understand such sentences, it would make sense not just to tell them they got it wrong. We’d want to show them how to make sense of those kinds of sentences. A student who easily understands, “The cat chased the dog” may be confused by, “The dog was chased by the cat.” Teaching students to keep their eyes open for that kind of sentence and how to either translate it to its active form or to question who was doing the chasing seem to be in order.

Of course, that kind of teaching cannot be beneficial in an instructional environment in which students are protected from language complexity (e.g., the instructional level). If students are to spend their instructional time reading texts they can already understand easily, then teaching them to make sense of complicated sentences won’t improve their performance and kids will soon learn to disregard what for them would be unproductive teaching.

We do something like this with vocabulary; intentionally introducing words we think students may not know and supporting them with vocabulary instruction. (As with grammar, the value of such instruction varies to the extent that comprehension turns on the meaning of those words. Vocabulary instruction has greater effects when comprehension is evaluated with texts containing the taught words than with texts that don’t).

Reading instruction should intentionally place students in situations in which their understanding of a text will depend upon their ability to surmount some particular conceptual or linguistic barriers. As noted, vocabulary instruction often does that. We should also be doing it with morphology, sentence grammar, cohesive links, text structure, and the like.

For the want of a word a sentence was lost;

For the want of a sentence the text was lost;

For the want of a text the learning was lost;

For the failure of learning the kingdom will truly be lost.


Sorenson Duncan, T., Mimeau, C., Crowell, N., & Deacon, S. H. (2020). Not all sentences are created equal: Evaluating the relation between children’s understanding of basic and difficult sentences and their reading comprehension. Journal of Educational Psychology. 

See comments here >

Add comment

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
"Reading should not be presented to children as a chore or duty. It should be offered to them as a precious gift." — Kate DiCamillo