So what? NOW WHAT?! Sentence comprehension in children with DLD

Some sentences are harder than others to understand, especially for students with developmental language disorders (DLD). Comprehension suffers when meaning doesn’t follow the order of the words and when sentences are dense with details.

We have good evidence that it takes extra attention, memory, and effort for students with DLD to understand some sentence patterns. These include: 

  • Passives:  The order was issued by the king.
  • Adjectives: The angry colonists attacked the threatening soldiers.
  • Adjective clauses: Animals that live in the wild find their own food.
  • Object relatives:  This is the house that Jack built.

Students with DLD don’t seem to pick up on word patterns (i.e., syntax) as well as their peers. Also, they struggle to remember, attend to, and manipulate complex sentences in working memory.   

Students encounter these sentences all the time in school. Montgomery, Gillam, and Evans (2021) wanted to shed light on what goes into sentence comprehension so effective treatment can be designed. 

The study

To test comprehension accuracy, two groups of children (DLD and typically developing) age 7 through 11 listened to different kinds of sentences. Pointing to pictures, they identified which noun performed the action of each sentence they heard.  

Importantly, the sentences presented were syntactically intact but not meaningful. (For example, The square had changed the bed under the very dry key.) The authors reasoned this would allow them to see how word order alone affected sentence comprehension.  

To find out what contributes to sentence comprehension and how, they also measured:

  • Fluid reasoning: the ability to recognize and interpret language patterns to determine who did what. (Because better pattern detectors are better sentence comprehenders.)
  • Attention: the ability to attend to all of the words in the sentence and chunk them into multi-word units. (Because it’s way easier to remember 3 or 4 chunks than 28 words.) 
  • Language knowledge: the capacity to access long-term memory, recognize the grammatical function of multi-word chunks (i.e., phrases vs. clauses), and assign meaning to them. (Because people get better at this with age.)
  • Working memory:  the system responsible for remembering information while performing some kind of mental activity with it. (Because comprehension requires drawing on what you know about words and their meanings.)

Not surprisingly, all students did better when the noun that performed the action was the first one mentioned in the sentence and when sentences weren’t jam-packed with details. 

Also, younger typically developing children did better with some of the more challenging sentences than older students with DLD.  

So what? 

The findings suggest that children with DLD are slower to gain syntax knowledge to guide their sentence comprehension than their peers. Also, they struggle with understanding more cognitively demanding sentences. 

Knowing why helps us know how to help them, and we all want to do that, so check it out…

The results support a super sensible notion — that working memory is a conduit for sentence comprehension. The authors didn’t draw it this way, but my brain sees their model like this:

Vivido Sentence Model


Like three individual roads merging together, fluid reasoning, attention, and language knowledge feed information into working memory, which (hopefully) fuses them all and leads to “Aha!” 

This makes so much sense! We just have to give students what they need to get through the dark working memory tunnel and into the light of comprehension. 

Then comes this sentence, though: “The field has no evidence-based treatments specific to sentence comprehension (p. 459).”


NOW What?!

Once again, this leaves SLPs with no choice but to do the best we can with what we know and design research-infused interventions for students with DLD. It seems to me that three steps get us started:

  • Stop teaching grammar one word at a time and asking students to memorize parts of speech.  They get lost in the dark tunnel of working memory. Students must hold all the words in the sentence in mind individually, go rooting around long-term memory for the right part of speech name, then try to hook them up and figure out what the sentence means. It’s too much! We know the capacity of working memory is 4-5 chunks of information. There’s no way anyone can parse sentences like this — one that has 30 words in it — if they can’t chunk them up into word groupings that make sense.  
  • Teach students to spot multi-word chunks in sentences — how words team up in meaningful ways. Good pattern detectors chunk language intuitively. Students with DLD need to be taught to see those chunks explicitly. Start with the sentence structures we know students have trouble with. (I listed four up above, but there are more.)
  • Make it easy to talk about what each chunk does. It’s not imperative that students call chunks by their formal names. For the sake of understanding, “whole thought” may be easier to grasp than “independent clause,” so just go with it. To get through the dark tunnel of working memory, make sure students can spot the chunks, understand what they do, talk about them, and know how to use them meaningfully.  

Sadly, if you want an evidence-based method for teaching sentence comprehension, there aren’t any. If you want an evidence-infused method, though, we’ve got one. It’s called Unlocking Sentences. Angela Tamborella (a brilliant SLP) and I forged it from the three principles above and everything we know about how brains and language work.

We have a long way to go with it, but pilot data tells us we’re onto something. With instruction grounded by research, students can learn to unlock the meaning of sentences they read. 


Montgomery, J., Gillam, R. & Evans, J. (2021). A new perspective on the sentence comprehension deficits of school-age children with developmental language disorder: Offering implications for theory, assessment, and intervention. Language, Speech, and Hearing Services in Schools, 52, 449-466.