A prevailing question in the study of emotion has involved the number and identity of human emotions. Theorists have sliced the emotional pie in a variety of ways, but most fall into one of two camps. Advocates of categorical approaches have identified a relatively small number of “basic” emotions – such as anger, fear, joy, sadness, disgust, and surprise – that cannot be reduced to other, more fundamental states (e.g., Ekman, 1992; Izard, 1991; Plutchik, 1980; Tomkins, 1962). These theorists suggest that all emotional experiences can be defined as mixes, blends, or hybrids of these basic emotions.

In contrast, proponents of dimensional models have argued that emotions are not divisible into discrete units. Rather, they suggest that much of the variance in emotional experience can be captured by a small number of primary dimensions. Some theorists endorse two-dimensional models characterized by the valence of the emotion (pleasant vs. unpleasant) and the degree of arousal or activation involved (aroused vs. tranquil; see Larsen & Diener, 1992; Russell, 1980; Watson & Tellegen, 1985). Others suggest that the data are better explained by a three-dimensional model defined by dimensions of valence, potency, and activity (Shaver, Schwartz, Kirson, & O'Conner, 1987).