Semantic representation and naming in children with specific language impairment.

Abstract

When 16 children with SLI (mean age = 6;2) and 16 normally developing age-mates named age-appropriate objects, the SLI cohort made more naming errors. For both cohorts, semantic misnaming and indeterminate responses were the predominant error types. The contribution of limited semantic representation to these naming errors was explored. Each participant drew and defined each item from his or her semantic and indeterminate error pools and each item from his or her correctly named pool. When compared, the drawings and definitions of items from the error pools were poorer, suggesting limited semantic knowledge. The profiles of information included in definitions of items from the correct pool and the error pools were highly similar, suggesting that representations associated with misnaming differed quanlitatively, but not qualitatively, from those associated with correct naming. Eleven members of the SLI cohort also participated in a forced-choice recognition task. Performance was significantly lower on erroneous targets than on correctly named targets. When performance was compared across all three post-naming tasks (drawing, defining, recognition), the participants evinced sparse semantic knowledge for roughly half of all semantic misnaming and roughly one third of all indeterminate responses. In additional cases, representational gaps were evident. This study demonstrates that the degree of knowledge represented in the child's semantic lexicon makes words more or less vulnerable to retrieval failure and that limited semantic knowledge contributes to the frequent naming errors of children with SLI.