The proposed research seeks to advance the understanding of sentence production deficits in aphasia. We will focus on an early stage of sentence construction, where semantically specified lexical items are assigned grammatical functions (cf. """"""""functional level representation"""""""" of Garrett, 1980, 1982). Breakdown at this level of sentence processing has received little attention, due in large part to the underdeveloped nature of psycholinguistic models. To advance this area of research, a model is presented that extends earlier work by Garrett (1980, 1982) and Bock (1987), by integrating formal work in lexical semantics (Levelt, 1989; Pustejovsky, in press-a; Grimshaw, 1990). This model, which will be referred to as the Lexical-Semantic Functional Processing (LSFP) model, is used to construct the experimental variables that form the foundation of this proposal, and will be employed to guide data analysis and interpretation. In accordance with the central role that the LSFP model gives to the verb for establishing basic sentence structure, this project will focus on the contribution of verb-based deficits to sentence construction difficulties in aphasia. A heterogeneous group of aphasics will be examined with a battery of tests that focuses on three aspects of verbs that are critical for planning basic sentence structure: (1) argument structure (thematic roles associated with a verb), (2) selectional restrictions (semantic constraints on nouns that can appear with a verb), and (3) cospecification (a frequency metric of word co-occurrences in sentences; Pustejovsky, in press-a). Two basic sentence production paradigms will be used: (1) verb-based Sentence Generation, and (2) Sentence Completion following the main verb. Additional tests will be administered to facilitate the interpretation of the sentence production performance: (1) Sentence Judgment for lexical and syntactic errors, (2) Production of Verb Agreement, and (3) Cognitive-Semantic Judgment. Aphasics will be tested in an acute and chronic stage to examine recovery and the development of adaptive strategies to deficits. Results will be examined at the case and/or group level to: (1) identify specific verb-based deficits during sentence construction, (2) determine their possible relationship to agrammatism/paragrammatism, (3) determine possible lesion correlations, and (4) apply findings to rehabilitation by combining results on a subset of aphasic subjects with their performance on an ongoing treatment program in sentence production at the primary performance site of the proposed project. It is anticipated that verb-based deficits in speech planning are widespread in aphasia, and that their understanding will not only help evaluate current models of normal sentence processing, but improve the precision of aphasia diagnosis and help tailor speech-language rehabilitation to individual patients.
