Imagine a hunter-gatherer society in which men hunt, facing dramatic surprises and life-threatening situations regularly. Men with facility in emotion regulation would be better hunters, promoting higher overall fitness; in contrast, women would face weaker selection pressure for emotion-regulation abilities. Such an evolutionary construction could predict why a functional magnetic resonance imaging (fMRI) study of cognitive reappraisal found neurophysiological evidence that men were more effective than women at down-regulating emotional responses to negative stimuli (McRae et al. Reference McRae, Ochsner, Mauss, Gabrieli and Gross2008).
This evolutionary explanation is plausible. But, so is the following cultural explanation: Western societal norms and gender stereotypes differentially encourage men to down-regulate emotional responses to negative events (Brody Reference Brody1997). Thus, adherence to societal norms of expressivity, which pervade everyday life (e.g., Simpson & Stroh Reference Simpson and Stroh2004), rather than prolonged evolution favoring sensitivities, could also easily account for the fMRI findings.
Evolutionary accounts do provide a provocative lens through which to view modern human behavior. However, as the foregoing vignette illustrates, a concern with most evolutionary psychological theories, such as Vigil's socio-relational framework for expressive behaviors (SRFB), is that their hypotheses rely on a number of assumptions that are difficult, if not impossible, to examine empirically in human populations.
We question the validity of basic assumptions of the SRFB, specifically (1) evidence of patrilocality in the ancestral populations that gave rise to contemporary humans, (2) the extent to which patrilocality led to purported differences in emotional expressivity in ancestral populations, and (3) the likelihood that the selection pressures mediating these hypothesized sex differences have persisted across large human evolutionary time scales to result in modern sex differences. In light of these concerns, we question the SRFB's utility as an integrative framework for understanding emotion and sex differences. We propose that current social-cognitive models of human behaviors provide a more parsimonious explanation of emotional expressivity and any purported sex differences.
1. How prevalent is patrilocality across cultures?
The SRFB's explanation of sex differences in emotional expressivity rests on the assumption that women and men faced different social ecologies, which imposed different evolutionary constraints. However, in nearly one-fourth of human societies included in Murdock's (Reference Murdock1967) ethnographic database, which includes data from a myriad of societies, including preindustrial ones, the residence pattern in which men stay with kin and women move with non-kin (patrilocal residence) is not observed. Thus, these findings cast doubt upon the SRFB's assumed universality of patrilocality and patrilocality-induced sex differences in emotional expression.
2. Did patrilocality lead to adaptive sex differences in emotional expressivity in ancestral populations?
Even assuming that the majority of ancestral human populations exhibited patrilocal residence patterns, the adaptive value of Vigil's purported sex differences in emotional expressiveness is unsubstantiated. According to the SRFB, women had to advertise trustworthiness to non-kin through submissive emotions. However, other evolutionary arguments (Cosmides & Tooby, Reference Cosmides, Tooby, Lewis and Haviland-Jones2000) suggest that such displays might have also been associated with costs. As a result, the cost of expressing emotions in distant (non-kin) relationships might have been relatively more costly than expressing emotions in close (kin) relationships; in the latter, costs arising from emotional expressions might have been offset because of incurred inclusive (shared) fitness benefits. Thus, based on this account, it is unclear why women, who were moving into distant relationships, did not limit their emotional expressiveness, and why men, who remained near kin, did not exhibit greater emotional expression with kin and limit expression of vulnerabilities to competitors.
3. Is there evidence that directional selection favoring sex differences in ancestral populations persisted throughout modern human evolution?
The SRFB rests on the assumption that men experienced prolonged selection pressures that favored less expressiveness, and that women experienced prolonged (and opposing) directional selection favoring more expressiveness. Prolonged directional selection is unlikely, because the environment for which this trait has evolved has changed over the long course of human evolution. However, neither hypothesis can be directly tested in extinct populations.
Moreover, prolonged directional selection would have resulted in relatively large sex differences in emotional expressivity (Grant & Grant Reference Grant and Grant1992; Kocher Reference Kocher2004). This is clearly not the case. The empirical reality is that substantial sex differences in emotional expressivity are not observed; One comprehensive review of research on emotion as expressed through behavior, self-report, or physiology, unequivocally concluded that “sex differences in emotionality are small, inconsistent, or limited to the influence of specific situational demands. … Reviews do not support belief in sex-based affective differences” (Wester et al. Reference Wester, Vogel, Pressly and Heesacker2002, p. 639, emphasis in original).
Furthermore, because sex differences in emotion facilities, when they appear, tend to be small (e.g., Montagne et al. Reference Montagne, Kessels, Frigerio, de Haan and Perrett2005; see also Brody Reference Brody1997; Wester et al. Reference Wester, Vogel, Pressly and Heesacker2002), between-sex variability in emotion expressivity is actually smaller than within-sex variability. Indeed, men's and women's distributions of scores on a measure of emotional expressivity, assuming a small effect size of r=.1, overlap by 84.3%. Applied to the SRFB, this suggests that a substantial proportion of women display “masculine” patterns of capacity and trustworthiness cues, and a substantial portion of men display cues in “feminine” patterns. The high variance of this behavioral trait does not fit with expectations of prolonged, directional selection favoring sex-specific patterns of expressivity, as proposed by the SRFB.
4. Social-cognitive models of human behavior: A parsimonious account of emotional expressivity and sex differences in emotional expressivity
Key assumptions of the SRFB remain speculative. Specifically, the adaptive significance of sex differences in expressivity in ancestral human populations and the conservation of such purported differences both across cultures and throughout modern human evolution cannot be validated. Moreover, extant research suggests women and men are much more alike than different in their emotional expression. The large within-sex individual differences, relative to small between-sex differences, suggest that emotional displays are strongly influenced by contemporary context (e.g., Ambady & Hall Reference Ambady and Hall2002; Callahan et al. Reference Callahan, Hasler and Tolson2005) rather than ancestral sex differences in sensitivities (see Brody Reference Brody1997).
A more parsimonious account of emotional expressivity, as well as any possible sex differences in emotional expressivity, is offered by current social-cognitive models of individual differences and human behavior (e.g., Zayas et al. Reference Zayas, Shoda and Ayduk2002; see Mischel & Shoda Reference Mischel and Shoda1995; Shoda & Mischel Reference Shoda, Mischel, Read and Miller1998). Such models highlight the adaptive value of flexible emotional expressivity for both women and men, and the importance of culture and contemporaneous situational influences in guiding appropriate emotional displays and behaviors.
By accounting for evolutionary constraints and empirical and theoretical contributions from broad areas of psychology and neuroscience, such social-cognitive models construe a person's behavior as a function of his or her processing system (e.g., sensitivity to displays) and the particular contingencies present in the situation. This position is in stark contrast to Vigil's current assumptions that sex differences in emotional expressivity reflect differences in ancestral selection pressures for men and women. Additionally, because social-cognitive models allow the generation of falsifiable hypotheses, they have broader potential for empirical scrutiny.
ACKNOWLEDGMENTS
We gratefully acknowledge Emre Selcuk for his assistance and Cindy Hazan for her helpful comments and suggestions on earlier drafts of this commentary.
