ABSTRACT
In 1981, at the virtual peak of the information processing revolution, Gordon Bower published his famous paper on mood and memory (Bower, 1981). Ironically, nearly 15 years later, it now seems that this article, so confident in its extension of semantic network theory to affective memory, actually defined the limitations of purely cognitive theories as applied to emotion and personality. Although many studies have supported an associational congruency between mood state and remembered material (e.g., Madigan & Bollenbach, 1982; Natale & Hantas, 1982; Snyder & White, 1982; Teasdale, Taylor, & Fogarty, 1980; Wright & Mischel, 1982), other researchers have found congruency between positive mood and positively toned material but not between negative affect and negatively toned memories or cognitions (e.g., Clark & Waddell, 1983; Macht, Spear, & Levis, 1977; Mischel, Ebbesen, & Zeiss, 1976; Teasdale & Fogarty, 1979). Bower himself (Bower & Mayer, 1989) was able to demonstrate reliable effects for the learning of mood congruent material but not for mood dependent retrieval.
