Multiple Memory Systems in the Processing of Speech: Evidence from Aging

Abstract

Young and old adults gave verbatim recall of recorded prose passages that varied in average word predictability and rate of presentation. Subjects were allowed to interrupt the speech passages at points of their choosing for recall of what they had heard on a segment-by-segment basis. For both age groups, the sizes of the segments selected were affected by level of predictability but not by the speech rate of the spoken passages. Subjects tended to interrupt the passages for recall at linguistic constituent boundaries. Recall of the segments was poorer for the elderly adults than for the young adults, with larger age differences for faster speech rates and for passages that were lower in average word predictability. Results are discussed in terms of the recent suggestion that multiple memory representations of a speech message may co-occur briefly in time.