Hartman and Hasher (1991) used a garden-path task in which younger and
older adults generated the final word for each of a series of high-cl
oze sentences. Under instructions to remember the final word, the expe
riment included critical sentences for which the generated word was re
placed by a new, to-be-remembered target. Using an implicit priming ta
sk, the first experiment replicated a basic finding: Younger adults sh
owed priming only for the target words, whereas older adults showed pr
iming for both the generated and target words. Two experiments explore
d boundary conditions. One showed that an additional sentence that int
erpreted the new target word enabled older adults to narrow access to
only the target word. The provision of additional time following the i
ntroduction of the new target word did not. Specific information, not
more time, is required for inefficient inhibitory mechanisms to clear
the recent past from memory.