Studies on the Malleability of Emotion Perception

Current theories of emotion perception posit that basic facial expressions signal categorically discrete emotions or affective dimensions of valence and arousal. In both cases, the information is thought to be directly ‘‘read out’’ from the face in away that is largely immune to context. In contrast, the three studies reported here demonstrated that identical facial configurations convey strikingly different emotions and dimensional values depending on the affective context in which they are embedded. This effect is modulated by the similarity between the target facial expression and the facial expression typically associatedwith the context.Moreover, bymonitoring eye movements, we demonstrated that characteristic fixation patterns previously thought to be determined solely by the facial expression are systematically modulated by emotional context already at very early stages of visual processing, even by the first time the face is fixated. Our results indicate that the perception of basic facial expressions is not context invariant and can be categorically altered by context at early perceptual levels. Whether on a first date, in a fight at the local pub, or in a poker game, humans continuously attempt to decipher social and emotional cues from each other. A particularly important source of information for decoding such cues is the face and its expressions (Adolphs, 2002, 2003; Darwin, 1872/1965; Ekman, 1992, 1993; Russell, 1997). Consequently, extensive research has investigated how facial expressions are processed and perceived. Twomajor views have evolved. According to the discretecategory view (Ekman, 1992), basic facial expressions convey discrete and specific emotions. For example, Buck (1994) stated that ‘‘the receiver has, literally, direct access to the motivational-emotional state of the sender’’ (p. 104). In the extreme formulation of this view, the readout of specific emotions from facial expressions is largely unaffected by their context (e.g., Ekman & O’Sullivan, 1988; Nakamura, Buck, & Kenny, 1990). According to the second, dimensional view (Russell, 1980, 1997), facial expressions are not categorized directly into specific emotion categories, but rather convey values on the dimensions of valence and arousal. These values are read out from the facial expression and are subsequently used to attribute a specific emotion to the face. Although the final attribution of a specific emotion to the face entails the integration of the aforementioned dimensions and situational information, the initial reading out of affective dimensions from the facial expression is assumed to be unaffected by context (Carroll & Russell, 1996; Russell, 1997; Russell & Bullock, 1986). Hence, although the discrete-category and dimensional frameworks hold different views about the information conveyed by facial expressions, they share the notion that affective information (specific emotions or affective dimensions, respectively) is read out from the face by a process that is relatively immune to context. Yet, in real life, faces are rarely encountered in isolation, and the context in which they appear is often very informative. Therefore, regardless of which of these two frameworks one accepts, there is reason to believe that the interplay between facial expressions and their context may prove to be an important determinant of emotion perception (e.g., Trope, 1986). Early studies examined this topic by pairing facial expressions with verbal vignettes that conveyed emotional information. Participants then judged what emotion was felt by the target person. Unfortunately, results proved inconsistent: Some studies demonstrated negligible contextual effects (e.g., Nakamura et al., 1990), others demonstrated strong contextual effects (e.g., Carroll & Russell, 1996), and still others failed to show dominance for either contextual or facial information (e.g., Fernandez-Dols, Sierra, & Ruiz-Belda, 1993; Goodenough & Tinker, 1931). Furthermore, the relevance of those studies to the Address correspondence to Shlomo Bentin, Department of Psychology, Hebrew University of Jerusalem, Jerusalem, 91905, Israel, e-mail: [email protected]; to Hillel Aviezer, e-mail: hillel.aviezer@ mail.huji.ac.il; or to Ran Hassin, e-mail: [email protected]. PSYCHOLOGICAL SCIENCE 724 Volume 19—Number 7 Copyright r 2008 Association for Psychological Science at The Hebrew University Library Authority on January 25, 2011 pss.sagepub.com Downloaded from perception of facial expressions is indirect at best, because participants were asked not to describe the emotion expressed in the face, but rather to attribute emotion to the target person. More recently Meeren, van Heijnsbergen, and de Gelder (2005) have shown that incongruence between body language and facial expressions (e.g., a picture incorporating an angry face on a fearful body) may cause Stroop-like interference effects when participants engage in speeded categorization of briefly presented faces. Interestingly, although participants in that study were slower and less accurate in the incongruent than in the congruent condition, they were still much more likely to categorize the faces correctly and ignore the context than to incorporate it into their face categorizations. Hence, in line with previous studies (e.g., Ekman, Friesen, & Ellsworth, 1982; Ekman & O’Sullivan, 1988), the study by Meeren et al. suggests that although the perception of basic facial expressions might be modulated by context, the emotional categorization of facial expressions is largely unaffected by it. As we noted, however, there are reasons to suspect that facial expressions might be more sensitive to context than previous studies suggest. Specifically, prior studies might have resulted in equivocal results because they did not take into account the perceptual similarity among facial expressions. One goal of the present study was to address this factor and unveil rules that govern contextual effects on the perceptual processing of facial expressions and on the mapping of facial expressions into emotion categories. In the first two experiments, we investigated the malleability of emotion perception by systematically manipulating the similarity between the target face (i.e., the face to be judged) and the face that would typically be associated with the emotional context. In Experiment 1, we took a discrete-category view, asking participants to categorize the emotion expressed by faces. In Experiment 2, we took a dimensional stand, asking participants to rate the valence and the arousal expressed by faces. Finally, in Experiment 3, we monitored eye movements to determine if context can modulate early stages of face processing, or only later processes that influence categorization.