Abstract
Hasher, Rose, Zacks, Sanft, and Doren (1985) report an experiment in which subjects had to learn emotionally positive, negative and neutral items. Subjects’ mood (either depressed or non-depressed) was not found to facilitate the learning of mood-congruent information. In this comment, we offer an alternative explanation for the negative results obtained, hy suggesting that the manipulation of the affective nature of the learning material was overridden by a context factor, i.e. subjects’ mood at encoding and retrieval, which was essentially the same on both occasions. When one considers this aspect of the procedure used, the findings are perfectly compatible with the larger body of evidence demonstrating mood-dependent memory. Our discussion rests heavily upon the notions of “episodic” and “semantic” memory (Tulving, 1983).