From Affective Valence to Motivational Direction

The affective-valence hypothesis is among the most successful theories in the science of emotion. It proposes a frontal asymmetry, with the left prefrontal cortex (PFC) subserving approachrelated positive affect, and the right PFC subserving withdrawalrelated negative affect (Davidson, 1984). However, even in the early years of the theory, occasional anomalies in experimental findings suggested that the left PFC processed the negative emotion anger. Although such findings fit the approach-withdrawal dimension of the affective-valence model, because anger prepares the organism for the prototypical approach response of aggression (Fox, 1991), they are irreconcilable with the positivenegative dimension. In recent years, evidence that anger is processed by the left PFC has been accumulating (HarmonJones, 2004). Consequently, Harmon-Jones (2004) parsimoniously removed the positive-negative distinction from the affective-valence model and built his motivational-direction model on the remaining dimensions of approach and withdrawal. Thus, in this model, the dimensions of approach and withdrawal exclusively define the frontal asymmetry of emotion. Recently, Harmon-Jones and his coworkers (Harmon-Jones, Lueck, Fearn, & Harmon-Jones, 2006) reported further evidence supporting his model. Left PFC activation in response to personally relevant anger-provoking stimuli was demonstrated during approach-related action expectation. In this case, as in most of Harmon-Jones's research, evidence was gathered by experimentally manipulating mood or emotion. In contrast, the evidence that favors the affective-valence model is weaker correlational evidence relating positive emotion to the left PFC. To reconcile the affective-valence model with the findings regarding anger, Davidson (2004) recently suggested that some kinds of anger might be processed alongside positive emotions, such as happiness, within the left PFC. However, this notion is contrary to the idea that the frontal asymmetry of emotion evolved to preclude conflicts among action tendencies (Davidson, 2004), because happiness works antagonistically to both anger and anxiety (Wegner & Pennebaker, 1993). Given their inherent approachand withdrawal-related properties, anger and anxiety seem the best candidates for asymmetric frontal processing, but there seems to be no direct reason for happiness to be processed asymmetrically in the PFC (Harmon-Jones, 2004). Happiness more likely is a state of psychobiological wellbeing that requires a balance between affective approach and withdrawal (Schulkin, 2003; Van Honk & Schutter, 2006). A technique that can provide definitive answers in controversies regarding the frontal asymmetry of emotion is repetitive transcranial magnetic stimulation (rTMS). Indeed, a recent rTMS study supported the motivational-direction model in demonstrating greater processing of anger when left-sided dominance was induced in the PFC than when right-sided dominance was induced in the PFC (d'Alfonso, van Honk, Hermans, Postma, & De Haan, 2000). This study, however, lacked the sham rTMS control necessary to directly pinpoint anger to the left PFC. In the present study, brain activity in the left and the right PFC of healthy subjects was locally decreased by applying low-intensity/low-frequency rTMS; in the control conditions, sham rTMS over the left and right PFC was used instead. Next, subjects performed an emotional memory task that taps into the attention and working memory abilities of the PFC and indexes the processing of anger and happiness (Van Honk et al., 2003). According to the motivational-direction model, deactivation of the left PFC by rTMS should decrease the processing of anger but not necessarily happiness, relative to sham rTMS or deactivation of the right PFC by rTMS. The affective-valence model predicts the opposite pattern, with deactivation of the left PFC causing decreased processing of happiness but, in general, not anger.