Wh interrogative production in agrammatic aphasia: an experimental analysis of auditory-visual stimulation and direct-production treatment.

The effects of auditory-visual stimulation treatment derived from principles associated with a stimulation approach for aphasia treatment and direct-production treatment derived from a behavioral or learning approach were examined in 4 neurologically stable agrammatic aphasic subjects. Subjects were trained to produce selected exemplars of wh interrogative morphemes in complete sentence contexts, while the acquisition, response generalization (both within and across interrogative forms), stimulus generalization (to language samples), and maintenance effects of the two treatments were assessed. An alternating treatments design (ATD) in combination with a multiple-baseline design across behaviors and a multiple-baseline design across subjects was employed. Interrogative constructions were counterbalanced across subjects and treatments, and probes were administered daily to assess treatment effects. Results indicated that direct-production treatment was consistently more effective than auditory-visual stimulation treatment in facilitating acquisition of target responses for all subjects. Response generalization within interrogative forms paralleled acquisition regardless of treatment approach. Stimulus generalization to the elicited language-sample condition was not evident, however, trained responses were maintained subsequent to treatment. These data provided support for using direct-production treatment for interrogative intervention with agrammatic aphasic patients and indicated that training a selected number of exemplars of target interrogatives results in generalization of that question form to novel language responses. However, the lack of generalization across interrogatives indicated that wh interrogatives do not constitute a response class and, thus, pointed out a need for programming generalization to untrained members of that linguistic class and to spontaneous language.