Profiling fluent aphasic spontaneous speech: a comparison of two methodologies.

Grammatical features of fluent aphasic speakers have not received as much attention as those exhibited by non-fluent aphasic speakers. In part, this neglect reflects the difficulty of applying consistent analytical procedures to fluent aphasic speech. For the analytical process to be meaningful, the methodology must be robust, replicable and capable of dealing with the voluble and non-specific output which is often typical of these speakers, and yet be sensitive enough to reveal subtle grammatical differences between aphasic and normal speech. In this paper the methodological problems encountered in applying analytical procedures to spontaneous speech data are discussed. Two different procedures for analysing speech data are compared and applied to samples of fluent aphasic and normal speech. Both procedures reveal that (compared with normal speakers), although these aphasic speakers are capable of producing some well-constructed grammatical structures, the distribution of these structures may differ compared with the distribution exhibited by normal speakers. These results suggest that complex aspects of grammatical organisation may be compromised in some fluent aphasic speakers.