So, You had Two Sisters, Right? Functions for Discourse Markers in Alzheimer’s Talk

Abstract

Naturally occurring conversation with Alzheimer’s speakers, including embedded or co-constructed narrative, can be used to support or augment clinical findings on features of Alzheimer’s discourse, which are as highly variable as is the disease. For example, not only can speakers with moderate to moderately severe Alzheimer’s disease maintain some level of politeness (Sabat & Collins 1999; Temple et al. 1999; Rhys et al. 2000); and interaction (Hamilton 1994a, 1994b; Ramanathan 1997), their pragmatic skills can sustain more fine-tuned analysis, particularly as those skills sustain or simulate fluency. Ellis (1996) comments that problems Alzheimer’s speakers have in organizing and concentrating information are both communicative and cognitive. In the early stages, as grammatical modes of processing deteriorate for whatever reason, those “cohesion ties that structure topicality will begin to fade”; in later stages, more severe problems with maintaining topic will surface (see for example, Kempler 1995). Alzheimer’s speakers typically depend on lexical cohesion, or word-based means of holding the elements of a sentence together, as opposed to grammatical because, adds Ellis, lexical cohesion relies on “meaning,” not on grammatical structures. They lose the ability to “ground” the discourse for the hearer, or to organize meaning with thematic information via pronouns (see Almor et al. 1999)