Recent theoretical work on word-level (i.e., lexical) and sentence-level processes in the normal population suggests that these two processes are not independent of each other, as was previously thought, but rather that syntactic processes are strongly guided by knowledge about individual words and combinations of words, particularly verbs. The proposed research investigates the extent to which this hypothesis can account for language disorders in dementia of the Alzheimer Type (DAT). Specifically, we hypothesize that damage to lexical representation in DAT results in impaired processing of both nouns and verbs, and furthermore that this lexical damage has a direct effect on DAT patients' abilities to compute syntactic structure, thus affecting their sentence-level language comprehension and production abilities. This research assesses the effects of lexical organization on semantic, morphological, and sentence processing in three populations: young healthy adults, older healthy adults, and patients with DAT. A first set of studies involves a detailed investigation of semantic and morphological knowledge concerning nouns and verbs in these populations. A second set of studies addresses sentence-level processes that are hypothesized to be driven by lexical knowledge. Both sets of studies are guided the common theme, that the frequency and idiosyncrasy of the underlying lexical features predicts ease of processing in normal populations and patterns of impairment in DAT. Processing lexical items and morphological and syntactic structures is harder and more prone to damage if the underlying features are idiosyncratic and infrequent. Finally, computational """"""""connectionist"""""""" models are developed in order to test the hypotheses concerning the common basis of semantic, morphological, and syntactic processes. These models allow us to test the claim that the widespread brain damage associated with DAT has non- random effects, with lower frequency words and structures that are characterized by idiosyncratic features suffering greater impairments that other words and structures. Together, the empirical and modeling results will provide a detailed account of lexical organization and the role of lexical organization in the impairments associated with aging and Alzheimer's disease.
