Based on his finding that subjects can show an affective preference to previously seen stimuli that they fail to recognize, Zajonc (1980) claimed that affective processing operates separately from cognitive processing. Over four experiments, we replicated and extended the finding that mere exposure to a briefly presented stimulus can increase positive affect through familiarity without enhancing the recognition of that stimulus. Among our findings, lateralized presentation of the irregular polygon stimuli showed that affect judgments were best for stimuli presented in the right visual field (left hemisphere), whereas recognition judgments were best for stimuli presented in the left visual field (right hemisphere). These effects were found only when the study stimuli were shown for 2 msec and were unmasked or for 5 msec and were pattern masked; when the stimuli were shown for 5 msec and were energy masked, target selection by affect or recognition was not greater than chance. These data, along with results from contingency probability analyses, indicate that affect and recognition judgments are different. Rather than viewing the difference between affect and recognition in terms of different features that might reside in the stimulus, the difference in judgments may reflect the manner in which a stimulus representation has been accessed. When viewed in terms of different retrieval processes that access different information, target selection by affect in the absence of recognition can be interpreted in terms of existing models of recognition memory.

This study found that repeated exposure to briefly presented stimuli increased positive affect through familiarity without enhancing recognition of the stimuli. Following exposure, subjects selected previously shown target stimuli on the basis of affect in the absence of stimulus recognition. Interpreted in terms of the manner in which information can be accessed in long-term storage, this study extends earlier research by showing that the ability to select target stimuli by affect can occur undiminished over a delay of 1 week between study and test. Repeated processing during study can produce a form of perceptual learning, called perceptual fluency, that can serve as the basis for stimulus discrimination in the absence of recognition at the time of test. The present results of familiar, but unrecognized, stimuli are analogous to the memory phenomenon of deja vu. Kunst-Wilson and Zajonc (1980) presented subjects with a series of 10 irregular polygons that were shown five times each at a brief exposure duration. In subsequent forced-choice judgments of affect (Which stimulus do you like?) or recognition (Which stimulus did you see before?), subjects selected previously seen stimuli at a better-than-chance level only for affect judgments. This increase in positive affect is called the exposure effect, and it results from repeated presentations of unfamiliar stimuli. To Zajonc (1980), affective preference for stimuli that cannot be remembered constituted evidence for the separation of "affective and cognitive" processes. Seamon, Brody, and Kauff (1983) replicated and extended the Kunst-Wilson and Zajonc (1980) finding that repeated exposure to briefly presented stimuli increased positive affect without enhancing stimulus recognition. But, rather than viewing the difference between affect and recognition judgments as evidence for separate channels of processing, Seamon et al attributed the difference in judgments to different ways in which stimulus representations might be accessed in memory. Target selection by affect in the absence of recognition is an unusual phenomenon, but it can be explained by existing models of recognition memory (e.g.,

No abstract