Similarity of movement in recognition of self‐performed tasks and of verbal tasks

Abstract

Recognition performance is strongly enhanced when subjects not only verbally learned actions, but also performed the actions during learning. In the multimodal memory model, this effect is explained by assuming that learning by enactment provides subjects with motor information which helps recognition of old items. However, the same type of information should impair recognition and increase the number of false alarms when the distractors are motorically similar and subjects reprocess motor information during testing. We tested this prediction. Subjects learned actions either by verbal means or by performing the denoted actions. Recognition performance was tested in a single‐item old/new decision. The distractors were constructed by changing the objects or the verbs of the old items so that the new items were either conceptually and motorically similar or dissimilar. Half of the items were tested verbally, whereas the other half were also performed during testing. The results supported the predictions. Hit rates were higher in the enactment group than in the verbal group. Similar distractors were more often erroneously accepted as old than dissimilar items. This effect was strongest for those subjects who enacted the actions during learning and testing. In addition, it was observed that object changes were more easily detected then verb changes, which effect did not interact with any other factor.