The general aim of the current proposal is to test and extend recent findings supported by an NIH FIRST Award to the PI on phonological output deficits in fluent aphasia. This work has identified distinct error patterns that correspond to breakdown at two major points in the normal phonological output system: (1) retrieval of entries from the phonological lexicon; (2) application of the operations used to spell out the full phonemic inventory of the stored forms, in preparation for articulatory-phonetic encoding. The present grant proposes to further explore such deficits, by adopting a more detailed model of phonological production that is motivated independently by recent phonological theory (e.g., Underspecification Theory--UST), yet is able to capture the levels of breakdown revealed in the PI's earlier work. This model, referred to as the UST Processing Model: (1) is used to motivate patterns of expected behavior with major deficits to the system, (2) serves as the basis for designing our experimental tests, and (3) will guide data analysis and interpretation. The project is divided into 7 studies. In Studies 1-4, the single-word repetition and picture naming of fluent aphasics will be examined at the group level (using subgroups based on a subset of the variables from PI's previous work) and single case level (including longitudinal analysis), to determine how the segmental and syllabic structure of words affect patterns of correct and incorrect performance. Study 5 will examine the extent to which findings on the proposed studies map onto the deficits identified in the PI's earlier work. Study 6 will present Alzheimer's patients with a subset of the planned tests as a means for broadly validating the aphasia findings and learning more about patterns of lexical-phonological dissolution. Study 7 will study the nature of neighborhoods in the phonological lexicon, a key variable affecting patterns of phonological breakdown, by examining formal paraphasias (i.e., sound-related word substitutions) in picture naming and repetition. Our study will have both theoretical and clinical implications. By adopting a production model which incorporates current phonological theory, we can achieve a clearer and more detailed understanding of the nature of phonological output disorders in aphasia. In turn, this understanding will provide the basis for more precise diagnosis and effective treatment of these disorders. Moreover, it provides a basis for testing important theoretical issues in cognitive psychology and theoretical linguistics.
| Kohn, S E; Melvold, J (2000) Effects of morphological complexity on phonological output deficits in fluent and nonfluent aphasia. Brain Lang 73:323-46 |
| Glosser, G; Kohn, S E; Sands, L et al. (1999) Impaired spelling in Alzheimer's disease: a linguistic deficit? Neuropsychologia 37:807-15 |
| Glosser, G; Friedman, R B; Kohn, S E et al. (1998) Cognitive mechanisms for processing nonwords: evidence from Alzheimer's disease. Brain Lang 63:32-49 |
| Glosser, G; Kohn, S E; Friedman, R B et al. (1997) Repetition of single words and nonwords in Alzheimer's disease. Cortex 33:653-66 |
