Arthur Wingfield

The memory problem in normal aging has its roots in reduced efficiency in acquiring new information, and it is largely this limitation that later translates into memory failures. Our approach to this question is focused on rapid speech comprehension, and memory for what has been heard. A major factor we examine is the effect of reduced hearing acuity, as hearing loss, whether mild, moderate, or more severe, often accompanies normal aging. In addition to these sensory changes are age-sensitive reductions in the capacity of working memory and speed of perceptual processing that would paradoxically seem to predict far more serious decrements in spoken language comprehension than one actually sees in healthy aging. At the same time, the perceptual effort due to even a mild hearing loss may bring a cost to successful speech perception in the form of a draw on attentional resources that would otherwise be available for understanding speech with complex syntax, or encoding the speech in memory. We then use computer editing of speech to add structural coherence, prosodic contour and linguistic constraints to the speech to explore how these features are used by older adults to bring their performance to a level more closely approaching that of the young. In this way we are able to examine the delicate interplay between "top-down" contextual support (at both the acoustic and linguistic levels) as it may be used to supplement the declining sensory, or "bottom-up" analysis of the acoustic signal itself. Dr. Wingfield and his collaborators also use functional brain imaging as an added tool in this exploration.