Abstract
Frizelle, P., & Fletcher, P. (2014). Relative clause constructions in children with specific language impairment. International Journal of Language and Communication Disorders, 49, 255-264.
Common Core Standards require that students read more complex text. Such texts employ less familiar, more abstract vocabulary and more complex syntax. A primary way that syntactic complexity during the schools increases is through the addition and embedding of adverbial, adjectival/relative, and nominal dependent clauses. Students with specific language impairment (SLI) characteristically have morphological and syntactic deficits. The authors of this article ask the following questions:
Do children with SLI differ from their age-matched counterparts and from younger typically developing children in their performance on relative clause constructions overall?
Is the performance of the children with SLI better on single- than dual-proposition relative clause constructions?
Does the difficulty level of relative clauses revealed by the performance of the group of children with SLI reflect the profile of difficulty seen in the relative clause performance of English-speaking typically developing children?
Answers to these questions can provide valuable information to speech-language pathologists (SLPs) to develop appropriate goals for students with SLIs that will promote their academic achievement.
The Study
Participants
Thirty-two children between 6;0 and 7;11 years with SLI participated in the study. As controls, there were 32 typically developing age-matched children (AM-TD) and 20 typically developing children who were 2 to 3 years younger (YTD). Children were in the SLI group if they scored below −1.25 SD on the receptive language composite of the Clinical Evaluation of Language Fundamentals (CELF-4; Semel, Wiig, & Secord, 2006). Standard scores ranged between 46–81. The Raven’s Test of Progressive Matrices (Raven, 2008) was administered to assess cognitive ability. Children had to achieve a standard score of 85 or greater; scores ranged between 85–115.
Task
The authors devised a sentence-recall task, including 52 relative clause (RC) constructions.
Children were asked to repeat RC constructions containing one of six types of relative clause, attached to either a presentational matrix clause (PN, single proposition) or to the direct object of a transitive matrix clause (DP, dual proposition). See examples in Table 1 that are similar to those used by the researchers. All relative clauses were introduced by one of the relative markers, who, that or whose. In each condition, two of the four sentences included a copular matrix clause (PN) and the other two sentences included a transitive matrix clause (DP).
Some Types of Relative Clauses
Analysis
A scoring system was devised to allow for a detailed description of both correct and incorrect responses. Unfortunately, the authors do not give a complete explanation of the scoring system. Children’s responses were assigned a score ranging from 10 to 0, with a higher score representing a more accurate performance. The sum of these scores resulted in a total sentence-recall (SR) score on which the groups could be compared. An additional score was computed by recording the total number of sentences that accurately reproduced the target relative clause construction—that is, both maintained the structure of matrix sentence and relative clause, even though the wording of the sentence might not be accurate in all respects. Sentences that were repeated with complete accuracy received a score of 10. If the repeated sentence maintained the syntactic structure of the target but had a lexical substitution, it was given a score of 9 (e.g., repeating the sentence “Joe rubbed the cat that the goat stood on last week,” repeated as “Joe rubbed the cat that the goat
Results
All three groups achieved higher scores on the PN than DP relatives, suggesting that PN construction is easier. The SLI group exhibited the greatest differences in scores between the PN and DP relatives.
Children with SLI showed the same pattern of performance difficulty as AM-TD and YTD groups, but their performance was significantly lower than both the AM-TD and YTP groups on both PN and DP relatives.
Children with SLI performed best on subject relatives (with intransitive subject relatives better than transitive subject relatives), followed by object relatives, considered overall. Their performance on oblique and indirect object relatives is at too low a level to draw firm conclusions aside from the obvious ones that these forms are highly problematic for them.
Discussion
This study demonstrates that there is not one type of relative clauses; and some types of relative clauses are acquired earlier than others, even in typically developing children. The earliest relative clauses are embedded in copular constructions (e.g., there is…, this is…). Although these sentences consist of two clauses they reflect only a single proposition and can be paraphrased by a simple sentence (“There is the boy who the teacher helped in the gym” can be paraphrased as “The teacher helped that boy in the gym”). The copular clause has a formulaic construction that functions as an attention-getter which is combined with a second component such as a relative clause. This structure is pragmatically useful in parent–child speech. Typically the referent is visible in the child’s environment and the demonstrative (that’s…) or adverbial (here’s…) that introduces it is linked to the context.
In facilitating development of relative clauses in children with SLI, SLPs should consider the types of relatives, if any, students do use. Interventions could begin with teaching the presentation relations first, then subject relatives, followed by object relatives. Oblique and indirect object relatives would be the last to be taught.