An investigation of the effect of source memory on the use of natural fluency cues in recognition judgments

Dual process theories of recognition associate familiarity with implicit memory; the more fluently an item is perceived or mentally processed, the more likely that it will be experienced as familiar (Jacoby, 1991). A recent puzzle for dual process theories has been that amnesics sometimes demonstrate chance recognition performance, even when perceptual fluency is spared. One explanation for amensic's chance recognition performance is that amnesics do not rely upon item familiarity (i.e., they utilize inappropriate recognition strategies) for recognition tests, and the reason they fail to do so is that they cannot remember having studied a list (Cermak et al., 1993). A temporary amnesia paradigm was used to examine whether memory for the study list affects recognition performance (i.e., use of perceptual fluency cues). In experiment one, half of the participants were administered midazolam (i.v., 2 mg, with pethidine, 50 mg) followed immediately by a study list (forming the "list source naive group"). A second experimental group was presented with two study lists, one study list 30 minutes prior to administration of midazolam, and another immediately afterwards. For these subjects ("study list aware group"), a single recognition test was administered that contained words from both lists. There was no significant difference in recognition performance between the study list aware group and the study list naive group, even though all participants in the study list aware group remembered having seen a study list. Finally, it was found that midazolam may have also affected perceptual fluency to some degree. Priming on an implicit stem completion task for a list studied under the influence of midazolam was evident but did not quite reach significance (p < .07). Results are discussed in relationship to lorazepam, which appears to spare familiarity, even though it interferes with both recollection and perceptual priming. The results support the view that memory for the study list (i.e., list source) does not necessarily influence subjects' use of familiarity in explicit tasks. Subjects do not appear to rely on processes that subserve implicit memory for list items studied under the influence of midazolam, whether or not they remember having studied a list.