Recall as a self-limiting process

categorized list enhanced recall of the cued categories relative to free recall, but reduced recall of the noncued categories (Experiment 1). This recall interference from recall of cued categories increased with the number of category names presented (Experiment 2), and was not due merely to delaying the subject with an interfering interpolated task (Experiment 3). The results argue against theories assuming that categories are recalled via associations from previously recalled categories. Rather, the results suggest that recall of categories acts in a manner to limit recall of additional categories, in agreement with the characterization of recall as a sampling-with-replacement process as described by Rundus. The central importance of retrieval processes in human memory has been widely recognized since publication of Tulving and Pearlstone's (1966) landmark experiment. Briefly, they showed that the amount of information recalled depends critically on the nature of retrieval cues embedded in the test situation. When conditions affecting storage of material are hel