Encoding specificity: Relation between recall superiority and recognition failure.

The results of four experiments show that (a) recall superiority over recognition is reversed by the use of unrelated word pairs in the study list, and (b) the reversal of recall superiority leaves intact the phenomenon of recognition failure of recallable words. These results extend the generality of encoding specificity and suggest that although recall superiority is a sufficient condition for recognition failure of recallable words it is not a necessary condition. In memory experiments in which people are required to remember the items of a list, the most effective retrieval cues are almost always nominal copies of to-be-remembe red items. If such cues are ineffective, it is unlikely that any other cues will succeed. Under certain experimental conditions, however, copies of list items are not the most effective cues for retrieval—in the presence of other cues, people do recall learned list words that they fail to recognize. This phenomenon of recognition failure of recallable words has been manifested in a number of experiments in which the same procedure, with only minor variations, has been used (e.g., Tulving, 1974; Tulving & Thomson, 1973; Watkins & Tulving, 1975; Wiseman & Tulving, 1975). In these experiments, learners studied a list of pairs of ,associatively related words. One word of each pair was designated as the to-be-remembered (TBR) or target word; the other word was