Explicit contamination in 'implicit' memory for new associations

In one view of implicit memory, priming arises from modification of preexisting representations; however, the role of such representations is currently in doubt following findings of implicit memory for newly formed associations. Closer consideration of studies reporting this effect, and of others that have failed to obtain it, suggests that such priming might result from the employment of explicit memory strategies. With measures designed to permit exclusion of such strategies, three experiments using lexical decision and stem-completion tasks found no evidence of truly implicit memory for unrelated pairs. Instead, priming was found only in those subjects (50% of the total in one experiment) who reported using explicit memory in stem completion. Contrary to previous conclusions, the results indicate a role for established representations in explaining implicit memory.