Recall Using Well Known And New Relational Information

Abstract

The thesis examined the use of two kinds of relational information in the process of recall. Previously, there has been separate study of the recall of items that share well known relations in permanent knowledge, and of items for which relations have been newly learned during study. Two approaches to explaining recall have evolved, each emphasizing the use of a different type of relational information. The thesis attempted to synthesize the two approaches into a single view of the process of recall. Experiments 1 and 2 established that the two kinds of relations are independent sources of information in memory, and that both are used in recall. Experiment 3 demonstrated that well known and new relations are used differently in recall even when they are processed in the same study episode. Experiment 4 examined how the two sources of information can be used together to maximize recall. Experiment 5 extended the findings to a different way of establishing new relations between previously unrelated items. The results suggested that the recall process makes use of whatever information is currently available in memory, including what was known previously and what was added in the course of study. It depends primarily upon any relevant permanent knowledge. If there are no permanently known relations to mediate recall, the process can occur by accessing the record of the new relation as it was formed at study. In addition, the different sorts of relations can combine forces to allow the recall of items that would not otherwise be accessible.