Speech accommodations to dementia.

This study investigated whether spouses would adopt a specialized speech register when communicating with adults with probable Alzheimer's disease. A picture description task was used so that the effectiveness of such speech accommodations could be assessed. The AD subjects did not vary the syntactic complexity, semantic complexity, or content of their descriptions when they were describing individual pictures versus directing their spouse to choose one of four pictures in a barrier task. The spouses' picture descriptions were more complex syntactically and semantically than the AD subjects' and included more highly salient elements. The spouses also varied the complexity and content of their descriptions, reducing syntactic and semantic complexity and increasing references to highly salient picture elements during the barrier task. These accommodations appeared to facilitate the AD subjects' performance on the picture description task.

Syntactic complexity and adults' running memory span.

This study investigated age group differences in adults' running memory span for prose. College students and adults 60 to 94 years of age listened to a prose passage that was interrupted occasionally by pauses. At each pause, the adults attempted to recall the immediately preceding text. The pauses followed either two single-clause sentences, a two-clause right-branching sentence, or a two-clause left-branching sentence. There was a significant Age Group x Syntactic Form x Clause Order interaction such that the age group differences in verbatim recall were exacerbated by the effects of syntactic complexity. The elderly recalled 25% fewer words from the first embedded clause of the left-branching sentences than the college students, whereas they recalled only 4% fewer words from the first of two successive single-clause sentences. Performance on the running memory span task was also correlated with two measures of the adults' working memory: forward digit span and backward digit span. The pattern of correlations indicated that working memory limitations determine adults' running memory span for prose and contribute to age-group deficits in comprehension.