| Literature DB >> 11708528 |
L J Altmann1, D Kempler, E S Andersen.
Abstract
Researchers studying the speech of individuals with probable Alzheimer's disease (PAD) report that morphosyntax is preserved relative to lexical aspects of speech. The current study questions whether dividing all errors into only two categories, morphosyntactic and lexical, is warranted, given the theoretical controversies concerning the production and representation of pronouns and closed-class words in particular. Two experiments compare the speech output of 10 individuals with Alzheimer's disease to that of 15 healthy age- and education-matched speakers. Results of the first experiment indicate that the pattern of errors in the speech of participants with mild PAD reflects an across-the-board increase in the same types of errors made by healthy older speakers, including closed-class and morphosyntactic errors. In the second task, participants produced a grammatical sentence from written stimuli consisting of a transitive verb and two nouns. Only adults with Alzheimer's disease had difficulties with this task, producing many more closed-class word errors than did healthy older adults. Three of the participants with PAD produced nearly agrammatic speech in this task. These 3 people did not differ from the rest of the PAD group in age, education, working memory, or degree of semantic impairment. Further, error rates on the two tasks were highly correlated. We conclude that morphosyntax is not preserved in the speech output of individuals with PAD, but is vulnerable to errors along with all aspects of language that must be generated by the speaker. We suggest that these results best support a model of speech production in which all words are represented by semantic and grammatical features, both of which are vulnerable to failures of activation when there is damage or noise in the system as a result of pathology, trauma, or even divided attention.Entities:
Mesh:
Year: 2001 PMID: 11708528 DOI: 10.1044/1092-4388(2001/085)
Source DB: PubMed Journal: J Speech Lang Hear Res ISSN: 1092-4388 Impact factor: 2.297