SENTENCE PROCESSING BY ADOLESCENTS WITH AND WITHOUT MENTAL-RETARDATION

A more thorough test of the hypothesis that persons with mental retardation are less likely to construct semantically integrated representations of sentences that they bear than are subjects without mental retardation (Merrill & Bilsky, 1989; Merrill & Mar, 1987) was provided. A series of sentences were presented to adolescents with and without mental retardation. Their memory for the object nouns of the sentences was then tested when they were provided with either the subject noun, the verb, or the subject plus verb of the sentence as a retrieval cue. The two-word cue was relatively better if an integrated semantic representation was constructed. Manipulations included decreasing the processing time given to subjects (expected to inhibit the construction of integrated representations) and presenting a picture with the sentence (expected to facilitate the construction of integrated representations). The reduction in time decreased performance for the subjects without mental retardation to the level normally observed for those with mental retardation; presenting a picture increased performance of subjects with mental retardation to the level of comparison subjects. Results support the suggestion that previously observed differences in sentence processing between individuals with and without mental retardation may be due to differences in generating integrated representations of the sentences during processing.