Teaching grammar to children with Specific Language Impairment

Teaching children with Specific Language Impairment can be challenging. Professor Mabel Rice of the University of Kansas discusses some details

Specific Language Impairment (SLI) is a language impairment that delays the mastery of language in children who have no hearing loss or other developmental delays1,2.  A hallmark characteristic of children with SLI is a protracted delay in acquiring requirements for tense marking in English and related languages.  Tense marking has several components:  One component is the cognitive recognition of tense to mark time, shown in examples such as “Patsy likes cookies” versus “Patsy liked cookies.”  A second component is the requirement to provide tense marking in sentences where required, such as “Mary did not eat cake” instead of “Mary ___ not eat cake” where “did” or “does” is required to mark past tense (either present or past).  This requirement is referred to as “finiteness” of clauses, i.e., a finite clause has a place in the sentence that must have a tense marker, even if it requires inserting “do” which does not contribute content meaning.  A third component is to know how tense marking is expressed phonologically, whether in regular patterns such as “walk/walked” or irregular patterns such as “run/ran” or “fall/fell.”  Children with SLI show they understand the concept of past tense by using words such as “yesterday,” yet they sometimes omit the past tense markings required by the grammar, in sentences such as “Patsy __ not like cake” and “yesterday Patsy like_ cake.”  Young typically developing children sometimes over-generalize regular past tense morphology, as in “walked,” to irregular forms to generate “runned” instead of “ran” or “falled” instead of “fell.”  Children with SLI persist in omitting tense markers for years beyond the time when typically developing children use these forms consistently, as required in the adult grammar.  Parallel to these grammatical errors are the phonological errors of over-generalizations.  As with finiteness omissions, children with SLI persist beyond the expected ages in over-generalizing regular patterns of phonology to irregular verb forms.

Figure 1
Figure 1

These patterns are depicted in Figure 13, which shows English-speaking children’s longitudinal performance on a task that asked them to label past actions with verbs that follow irregular phonological forms for past tense, such as “fell.”  In this figure the children are given credit for over-generalized forms, such as “falled” so we can see they know to mark the past concept even when they do not always avoid over-generalizations.   The mean percentage correct on the Y axis indicates progress toward the adult grammar, which is about 95-97% on this task.

The blue line in Figure 1 shows the longitudinal performance of typically developing children, relatively high when they were in kindergarten (~5-6 years) and at adult levels by second grade (~7-8 years).  The red and green lines show two groups of children with language impairment, who are unlikely to have been enrolled in speech/language pathology services and even if receiving clinical intervention it is very unlikely these forms were goals for intervention.    The red line shows an SLI group, documenting late acquisition of the past tense marking as we track them over time. They eventually catch up to the typically developing group at 3rd grade (~8-9 years), about 3 years behind the typically developing group.  The green line shows a group labelled as “Non-specific language impaired” (NLI), defined by a nonverbal IQ of less than 85, whereas the SLI group was required to have nonverbal IQ scores of 85 or higher. The green line differs from the red line, although the two groups started at the same accuracy level in kindergarten.  The improvement in the NLI group is more linear than for the SLI group who rapidly accelerate from grade to grade to catch up with the typical age peers at 3rd grade.  The NLI group, in contrast, makes steady gains on average from year to year but have not yet caught up at 4th grade (9-10 years).  This means that during this age period the SLI group has moved from errors such as “Yesterday Patsy fall down” to possible errors such as “Yesterday Patsy falled down,” showing that they were mastering the finiteness requirement even though they might not yet have the phonological requirements fully mastered.  In contrast the NLI group was slowly and steadily mastering the finiteness requirement, showing steady change per year which was not enough to change the gap even at 4th grade.

Highlighted complexities

It is reasonable to assume the data in Figure 1 capture children’s untutored acquisition of the details of past tense marking on verbs (reports of speech/language pathologist services suggest that, at most, 29% of children received such services, with no evidence that these details of grammar were targeted for intervention).  The results highlight the complexities of implementing a teaching program4First, in order to catch up to their age peers, children with SLI and NLI must learn faster, with a steeper slope of acquisition in order to catch up, although their history indicates a flatter slope for learning prior to kindergarten.  The SLI group met this requirement better than the NLI group.  How they managed to do this is unknown, as are the reasons for the delayed onset of the acceleration to catch up to their age peers.  Further, there is no explanation available for why the NLI group does not accelerate the rate of acquisition.  Second, it is important to differentiate between the 3 components of meaning (past time), grammar requirements for well-formed sentences (a finite form in each clause), and phonological rules for morphemes (regular vs irregular phonological patterns for marking past tense).  Recall that data in Figure 1 tracks the second component because the first component is known early on when children express past time by other means such as words like “yesterday,” and because the phonological requirements are not yet fully worked out for the NLI group.   Third, if we were to study in necessary detail how to teach finiteness marking on past tense verbs, we would want to know when to introduce a focus on finiteness in training.  It looks like the SLI children would be ready at, or before, kindergarten but the NLI group may not be at the same level of readiness.  The study design should also include assessments of nonverbal IQ to help plan for individualized treatment.  It could be argued that because both the SLI and the NLI groups make progress without training, perhaps the conclusion is that no training be provided.  The counter observation is that this period is when children are taught to read and it is an advantage for a child to be able to predict that “did” must proceed “not” in a sentence like “Patsy did not like cookies” when learning to decode written texts, or that “he fall down” is not the same as “he fell down” or that “he falled down” is highly unlikely in printed text.

The differences between “fall” and “fell” or “fall” and “falled” may seem to be a trivial detail when considering how to teach a child with SLI or NLI. Upon closer inspection, however, we see that it makes a difference in how children learn the grammatical requirements of their language and that such differences must be considered essential in order to improve methods for teaching these children.


Mabel L. Rice

University of Kansas



  1. Rice, M.L. Specific Language Impairment in Children. https://www.openaccessgovernment.org/specific-language-impairment-in-children/40152/
  2. Rice, M.L Specific Language Impairment in Children.  Part II. https://www.openaccessgovernment.org/specific-language-impairment-in-children/2018.
  3. Rice ML, Tomblin JB, Hoffman L, Richman WA, Marquis J. Grammatical tense deficits in children with SLI and nonspecific language impairment: Relationships with nonverbal IQ over time. J Speech Lang Hear Res. 2004;47:816-834.
  4. Rice, M.L.  Teaching children with Specific Language Impairment:  Who, where and how?


Please enter your comment!
Please enter your name here