Experimental analysis of syntax training in Broca's aphasia: a generalization and social validation study.

A multiple baseline design across responses was used to examine the effect of syntax training on the sentence production of 4 individuals with chronic Broca's aphasia. Subjects were trained to produce five exemplars of five sentence types from Helm's Elicited Language Program for Syntax Stimulation (Helm-Estabrooks, 1981). Generalization and maintenance of trained sentence types to novel exemplars and novel stimulus conditions served as dependent measures. In addition, five naïve judges rated subjects' responses before and following the treatment in terms of their "adequacy." Generalization to novel exemplars was demonstrated sequentially by 3 subjects (i.e., following the teaching of all five forms), and the remaining subject demonstrated generalization for three of five sentence types trained. Maintenance was variable across subjects and sentence types. Generalization across stimulus conditions was limited for all subjects. Adequacy judgments revealed improved communication skills for wh-questions but limited changes in the perceived adequacy of subjects' declarative responses. These findings indicate that the effects of syntax training procedures are limited to those grammatical constructions taught, that generalization of learned forms to novel stimulus conditions is not an automatic consequence of acquisition, and that the effect of such training on the adequacy of subjects' responses may be limited.