Do recall and recognition rely on qualitatively different processes?

A fundamental question in memory is whether recall and recognition reflect qualitatively different processes or whether they both involve essentially the same processes. To explore this question, two variables that have typically dissociated free recall and recognition were tested at different levels of a cued-recall task. Zero (free recall), two, three, four, five, six, and eight (recognition) letters of eight-letter words were presented as cues in the experiments. The two variables that were manipulated were study instructions (elaborative rehearsal and directed-forgetting instructions in Experiment 1) and word frequency (high- and low-frequency words in Experiments 2 and 3). In Experiment 1, to-be-rehearsed words were more likely to be freely recalled than to-be-forgotten words, but the two types of words were about equally likely to be recognized. In the two-letter cued-recall condition, fragments of to-be-rehearsed words were more likely to be completed than those of to-be-forgotten words (resembling the pattern in free recall), whereas in the six-letter cued-recall condition, fragments of to-be-rehearsed and to-be-forgotten words were equally likely to be completed (resembling the pattern in recognition). The patterns of performance in the other cued-recall conditions fell in between those in the two- and six-letter cued-recall conditions. In Experiment 2, high-frequency words were more likely to be freely recalled than low-frequency words, whereas low-frequency words were more likely to be recognized than high-frequency words. With two-, four-, and six-letter cues, fragments of low-frequency words were more likely to be completed than those of high-frequency words (resembling the pattern in recognition). In Experiment 3, the results of Experiment 2 were reversed in the cued-recall conditions even though the same variable (word frequency) was used. At all levels of cued-recall, fragments of high-frequency words were more likely to be completed than those of low-frequency words (resembling the pattern in free recall). The main difference between Experiments 2 and 3 was that in the former subjects tried to complete all fragments presented, whereas in the latter they tried to complete only fragments of previously studied words. When evaluated with respect to the transfer-appropriate processing framework, the present experiments suggest that recall and recognition involve similar processes.