Neurolinguistic Investigations of Aphasia and Recovery

The present project is proposed to continue work toward establishing the effects of treatment for sentencedeficits in agrammatic (Broca's) aphasia. In this research we examine the relevance of linguistic theory tounderstanding sentence breakdown and recovery patterns in aphasia and gather data to further developingaccounts of normal sentence production. We plan to build on previous work highlighting the important rolethat linguistic structure can play in guiding treatment for aphasia. In particular, we examine and manipulatelexical and syntactic properties of sentences that are the most difficult for agrammatic aphasic subjects. Wefocus on (a) sentence structures (e.g., filler gap structures) in which noun phrases (NPs), and other elements(e.g., verbs and auxiliaries) have been moved out of their original d-structure positions, (b) functionalcategories (e.g., verb tense morphology and complementizers), and (c) verb and verb argument structure.We also continue to test the hypothesis that training more complex rather than simple structures will result ingreater improvement in aspects of sentence production and comprehension. We define sentence complexityin terms of the (a) type of movement involved in sentences, i.e., wh- and NP-movement, (b) phrasal treestructure and the relation between nodes in the syntactic tree, and (c) the number and type of embeddings.In addition, we consider the lexical properties of verbs and how they impact syntactic operations indetermining complexity.In this period we also extend our work to begin to understand mechanisms that generate observed recoverypatterns. Using head-mounted eyetracking, we examine filler gap processing in normal participants and inaphasic patients prior to and following treatment. FMRI studies also are included to examine the neuralmechanisms of recovery. While neuroimaging studies have begun to elucidate the neural correlates oflanguage, few have examined brain sites recruited by aphasic patients, and even fewer have studied theneural mechanisms supporting recovery. FMRI studies undertaken in the previous cycle showed importantchanges in activation patterns from pre- to post-treatment. We, therefore, continue this effort in ourcontinuing studies.