This project begins with the central hypothesis that some residual capacities for communication are preserved in severely non-fluent aphasic patients, i.e., patients who have severely impaired language secondary to neurologic injury such as stroke. C-VIC, a computerized iconic communication system, allows for investigation and treatment of language processing in those patients who are inappropriate for more traditional experimental investigations and treatment. Results from assessments and therapy will serve to test several contemporary theories of language comprehension and production in normal humans and provide additional insight into the pathophysiology of disruptions of language. A major focus of the proposed work will be to determine which aphasic deficits limit generalization of patients' therapy to spontaneous, everyday, language use. The work for this project will target therapy at specific, discrete aspects of language production. After each phase in therapy, evaluation of the targeted process as well as untreated processes will be carried out. Training patients to produce C-VIC will allow for the evaluation of the integrity of pre-linguistic cognitive representations; this level of representation is otherwise inaccessible in severely aphasic patients. C- VIC training has already been shown to improve patients' production of simple English sentences. Systematic investigation of the extent to which patients can generalize production of vocabulary learned in C-VIC training will be undertaken to determine how limitation of access to specific vocabulary items affects generalization to untrained vocabulary, and, ultimately, to spontaneous speech. In addition, the success and limitations of patients' abilities to learn abstract syntactic rules in C- VIC will be established, as will the extent to which these rules generalize to English production. Finally, the generalization of treatment aimed at language production to language comprehension (both in C-VIC and English) will be assessed. The issues involved are at the heart of current controversies in our understanding of normal language processing and are fundamental to the development of more effective therapies for aphasia.
