In his target article, Vigil suggests that sex differences in emotion are an adaptation to a presumed social structure exhibited by early hominoids, in which females migrated from their natal group and males tended to stay in their natal groups. As evidence for extant sex differences, Vigil reviews literature showing that women are superior to men at perceiving emotion in others, and then uses such evidence as the basis for inferring that women also produce more emotion than do men. In this commentary, we first discuss evidence that production-based (perceiver-independent) measures of emotion reveal few consistent sex differences. Next, we review evidence that perceiver-based measures, which do suggest women are more emotional, evoke retrospective biases that highlight gender stereotypes, or dispositional biases that attribute emotional responses to a women's nature. We end by discussing an alternative view that women might appear to be more emotional because they are more facile with emotion language.

Perceiver-independent measurements provide very little consistent evidence that women are “more emotional” than men (with the exception that women do cry more often). For example, some facial electromyographic (EMG) studies show sex differences in facial muscle activity in response to emotional stimuli (reviewed in the target article), but many do not (Lundqvist Reference Lundqvist1995, experiment 2; Lundqvist & Dimberg Reference Lundqvist and Dimberg1995; Kelley et al. Reference Kelley, Forsyth and Karekla2005; Sloan et al. Reference Sloan, Bradley, Dimoulas and Lang2002). Even among papers that report sex differences, explanations based on orienting responses (Lang et al. Reference Lang, Greenwald, Bradley and Hamm1993) or facial imitation (Dimberg & Lundquist Reference Dimberg and Lundquist1990) cannot be ruled out. There is also inconsistent evidence for sex differences in smiling: some studies show that women smile more than men (LaFrance et al. Reference LaFrance, Hecht and Paluck2003), but in others women smile less (Ansfield Reference Ansfield2007). Similarly, there is no evidence for sex differences in psychophysiological responding to emotional stimuli (Kelley et al. Reference Kelley, Forsyth and Karekla2005), nor in the acoustics of emotional vocalizations (Viscovich et al. Reference Viscovich, Borod, Pihan, Peery, Brickman, Tabert, Schmidt and Spielman2003). Likewise, a recent meta-analysis of imagining studies found no major sex differences in how the brain responds to emotional stimuli (although males tend to exhibit more lateralized activation compared to females) (Wager et al. Reference Wager, Phan, Liberzon and Taylor2003).

Results from perceiver-based measurements of emotion generally tell a different story. Women, compared to men, report that they are more emotionally expressive (Barrett et al. Reference Barrett, Robin, Pietromonaco and Eyssell1998; Kring & Gordon Reference Kring and Gordon1998). Women also report experiencing more intense emotions than do men (Allen & Haccoun Reference Allen and Haccoun1976; Allen & Hamsher Reference Allen and Hamsher1974; Balswick & Avertt Reference Balswick and Avertt1977; Larsen & Diener Reference Larsen and Diener1987), and perceivers typically agree (Kring & Gordon Reference Kring and Gordon1998). However, these sex differences are observed primarily when self-report measures draw upon memory for prior emotional experiences (Barrett et al. Reference Barrett, Robin, Pietromonaco and Eyssell1998; Robinson & Clore Reference Robinson and Clore2002b; Robinson et al. Reference Robinson, Johnson and Shields1998). When men and women report their momentary emotional experiences in everyday life using experience-sampling procedures, they appear equally emotional (Barrett et al. Reference Barrett, Robin, Pietromonaco and Eyssell1998). Thus, one reason for perceiver-based sex differences is that women are simply better at recalling the information needed on global or dispositional self-report measures of emotional experience. Consistent with this suggestion, women also have more sophisticated emotion concepts that can serve as retrieval cues (Barrett et al. Reference Barrett, Lane, Sechrest and Schwartz2000; Seidlitz & Diener Reference Seidlitz and Diener1998). In addition, women recall emotional memories more quickly and frequently (Davis Reference Davis1999), intensely (Seidlitz & Diener Reference Seidlitz and Diener1998), and ruminate more on negative events compared to men (Nolen-Hoeksema et al. Reference Nolen-Hoeksema, Parker and Larson1994; Wood et al. Reference Wood, Saltzberg, Neale, Stone and Rachmiel1990).

A second reason that women appear more emotional in perceiver-based measurements is that memory for emotional events is infused with gender stereotypes (Grossman & Wood Reference Grossman and Wood1993; Robinson & Clore Reference Robinson and Clore2002b; Robinson et al. Reference Robinson, Johnson and Shields1998). Specifically, Robinson et al. (Reference Robinson, Johnson and Shields1998) showed that gender stereotypes are employed as a heuristic when people lack easy access to the target and situation-specific information (such as perceiving emotion in another person). Robinson and Clore (Reference Robinson and Clore2002a, Study 3) showed that women reported more intense emotion, and men less intense emotion, when participants were not concurrently experiencing emotion but gender-based beliefs about emotion were primed. Thus, people might believe women are the more emotional sex because they are engaging in retrospective biases that highlight gender stereotypes.

Recent work within our laboratory highlights a third reason that perceivers experience women as more emotional: sex differences in emotionality might stem not from what men and women actually do, but from the explanations that perceivers give for those behaviors. Specially, Barrett and Bliss-Moreau (under review) found evidence that people are more likely to assign a dispositional cause to female displays of emotion, whereas a situational cause is more frequently assigned to male displays of emotion. Thus, people might believe women are the more emotional sex because they treat women's emotional behavior as evidence that women have an emotional nature, whereas men's emotional behavior is interpreted as evidence that the situation warrants such behavior.

Lastly, women, compared to men, might also report more emotion because they are more likely to conceptualize basic affective changes as emotional. This might be the result of women having a broader and more facile emotion vocabulary than do men. Consistent with this suggestion, parents tend to discuss emotions differently with their daughters and sons. Mothers elaborate about emotion more with their daughters than with their sons, and place emotions in a more interpersonal context with their daughters (Fivush et al. Reference Fivush, Berlin, McDermott Sales, Mennuti-Washburn and Cassidy2003). Mothers also use more emotion labels during conversations with their preschooler-aged daughters than with sons. Women consistently use more emotion words when describing their own and others' reactions to interpersonal conflicts (Barrett et al. Reference Barrett, Lane, Sechrest and Schwartz2000). Simply stated, then, more facile emotion language to which females are exposed might provide an internal context that shapes emotion perception.

In fact, a recent review (Barrett et al. Reference Barrett, Lindquist and Gendron2007) summarized a number of different lines of evidence that support the idea that language is a key component in the conceptualization of emotion. Language might not only help determine the emotion categories people acquire but also how variable instances of core affect become conceptualized as a discrete emotion. More precisely, conceptual knowledge that is supported by language might explain why emotions are perceived as discrete entities even when the majority of production-based measures (including peripheral nervous systems responses, facial EMG, and neuroimaging) do not robustly and unambiguously differentiate among emotions (for a review, see Barrett Reference Barrett2006b; Barrett et al. Reference Barrett, Lindquist and Gendron2007; Wager et al. Reference Wager, Barrett, Bliss-Moreau, Lindquist, Duncan, Kober, Joseph, Davidson, Mize, Lewis, Haviland-Jones and Barrett2008). As a result, we suggest that conceptual knowledge is a powerful tool that not only might explain sex differences in emotion, but also might shed light on the very nature of emotion (for further discussion, see Barrett Reference Barrett2006a; Reference Barrett2006b).