VERBALIZING NONVERBAL MEMORIES: SOME THINGS ARE BETTER LEFT UNSAID

It is widely believed that additional processing, i.e. rehearsal, elaboration, or delayed recall, of a previously presented stimulus will improve subsequent memory performance. However, in a series of five experiments verbal processing of non-verbal stimuli impaired subsequent recognition performance. In Experiment 1 subjects viewed a videotape including a salient individual. Later some subjects described the individual's face. Subjects who verbalized the face performed less well on a subsequent recognition test than control subjects who did not verbalize the face. Experiment 2 examined the role of visual recall in mediating this effect. As in Experiment 1, subjects who verbalized the face exhibited poorer recognition performance than control subjects. However, subjects who simply visualized the face performed the same as controls, indicating that recognition impairment is not simply due to recalling the face but rather to some process associated with explicit verbalization. Experiment 3 and 4 explored the hypothesis that verbalization impairs memory for stimuli that are difficult to put into words. Experiment 3 used a different difficult-to-verbalize stimulus: color. Subjects viewed a color chip and then either described the color, visualized the color, or engaged in an unrelated activity. As with faces, subjects who described the color performed less accurately on a subsequent recognition test than visualize or control subjects. Experiment 4 included a stimulus that could be precisely verbalized: the statement spoken by the individual in the video tape. While verbalization once again impaired face recognition, statement recognition was marginally improved, thus suggesting that verbal processing differentially affects stimuli that can be described precisely. Experiment 5 demonstrated that the effects of verbalization persist for two days and probably much longer. This collection of results is consistent with the notion that subjects access their verbalizations at the expense of their original memories. When the verbalizations are reasonably precise, as in the case of statements, memory performance is unimpaired, however, when the verbalizations are impoverished, as in the case of faces or color, memory performance suffers.