Why Evolutionary Psychology Is Not Feminist: Assessing the Core Values and Commitments of the Evolutionary Study of Gender Differences

4.1 Empirical Adequacy

Scientific output is considered virtuous if it is empirically adequate, i.e., if it is in accordance with the greater body of observational and experimental data. This virtue reflects work in the feminist philosophy of science that has previously shown that research based on linking differences between men and women to a set of biological causes does not fulfill minimal standards of empirical adequacy, either due to faulty research design or improper statistical methodology (Longino 1996, 45). EP’s scientific output about gender is not in accordance with other findings relevant to the study of gender, and the methodology employed to test hypotheses about the different psychologies of men and women is considered flawed. First, the conception of sex dimorphism that underlies EP research has been deemed empirically inaccurate by many. At the core, EP is committed to a particular interpretation of sex dimorphism: Humans are said to have evolved a universal mind, with the only source of variation being biological sex (see Section 2). EP, therefore, seems to assume two dimorphic sets of modules in the absolute sense. However, according to Blackless et al. (2000, 162), sex dimorphism can come in degrees: Absolute sex dimorphism means that the two forms of one species do not overlap, while moderate sex dimorphism is characteristic of species in which differences between the two forms only occur on average. The authors demonstrate that humans are moderately dimorphic, with an overlap in the size or composition of reproductive glands, genital morphology, chromosomes, and hormonal physiology close to two percent on the bimodal continuum of biological sex (i.e., intersex people). EP disregards this empirical finding altogether.^[2] Some EPs might argue that moderate sex dimorphism would also be sufficient to retain their theoretical core, i.e., that average differences in physiology can be causally linked to universal gender differences in behavior. However, as will become clearer in my next point, the claim that absolute sex dimorphism is a causal factor in universal gender differences is empirically inadequate, making the argument that a weaker form of dimorphism can have the same causal effect on psychological differences even less plausible.

From an empirical adequacy standpoint, it is not only the basis for gender differences in EP that has been criticized, but also a number of the assumptions underlying theories and hypotheses about adaptive problems and reproductive strategies. For instance, Parameswaran (2014) explores evidence from historical records on several communities pre and post-European colonization. She demonstrates that EP falsely assumes fairly uniform practices of human mate selection and child-rearing across the world to be evidence of adaptive behavior. The author argues that this uniformity, which EPs take to bolster their claims about a universal human nature, is the result of European, Judeo-Christian values imposed on or adopted by colonized populations. Colonized areas of the world initially displayed great diversity in mating practices, family arrangements, gender roles, and child-rearing practices. Parameswaran (2014) criticizes that this counterevidence is either downplayed by EP or not considered altogether. How EP downplays counterevidence is highlighted in more detail by Nash (2014), who discusses the example of self-reporting data to study EP hypotheses about jealousy and promiscuity in men and choosiness in women. There are many possible factors influencing this self-reported data, such as societal gender roles and gendered expectations. However, data in accordance with EP hypotheses is taken at face value and believed to indicate ultimate causes when it is used to bolster EP hypotheses but only taken to reveal proximate causes when the results do not align with EP theories (321f).

Additionally, Fausto-Sterling (1997) criticizes that EP theories and hypotheses tend to be too vague. In particular, she is referring to the theory that ancestral women faced the adaptive problem of acquiring resources to survive pregnancy and lactation in Buss (1995) and the author’s claim that “all people are descendants of a long and unbroken line of women who successively solved this adaptive challenge” (Buss 1995, 164). It lacks specifics, Fausto-Sterling contends, such as when exactly the described reproductive behaviors are imagined to have started to appear, at what point in time the relevant traits are postulated to have become a fixed part of human genealogy,^[3] in which types of environments this development might have taken place, and if there were any genetic bottlenecks^[4] during evolution (1997, 241f). Moreover, she suggests that EPs should consider relevant information during theory building, such as data on the culture of prehistoric humans and protohominids, the expansion of human populations to different continents, and the literature that uses molecular evidence to track human evolution.

Not only are EP theories and hypotheses considered to be built too broadly, but they are also considered poorly tested. A critical example of how EP’s scientific output is typically tested is a large-scale international survey conducted by Buss (1994) intended to analyze EP hypotheses about women’s mate preferences. The idea is that successfully solving adaptive problems gave rise to women desiring specific traits in potential male mates. Buss (1994) hypothesizes that this is the cause of women’s behavior in mating today and accepts his hypothesis because, in the conducted surveys, women tended to favor characteristics such as good financial prospects, high social status, older age, ambition, emotional stability, intelligence, similarity in character, tallness, good health, and expressing love in a partner. Fausto-Sterling (1997) argues that this is methodologically unsound because the study uses data about contemporary humans to explain past events, but current behavior does not necessarily illuminate anything about past behavior (Fausto-Sterling 1997, 244). This is corroborated by the idea of exaptation,^[5] which refers to the possibility that a trait evolves for one function but might be adapted for another, or the idea that there exist traits without any function (Meynell 2020, 307). This makes it even harder to argue that current traits can be causally traced to the behavior of our ancestors. Other methods for hypothesis testing include experimental methods as found in psychology, such as Singh’s (1993) study on male mate preferences, which was conducted by asking participants to rank pictures of women with different waist-hip ratios. Downes (2021) points out that results from such experiments cannot be taken as causal evidence that mate preferences are the result of an evolutionary mechanism. I additionally argue that testing evolutionary hypotheses using recent survey data or psychological experiments is methodologically flawed because even if the postulated psychological mechanisms exist in humans, it would be very difficult to isolate their impact on human behavior from being driven by unconscious heteropatriarchal gender roles. Thus, as it stands, EP does not fulfill the virtue of empirical adequacy. One should, therefore, be wary of any scientific output that aims to explain gender-based differences with (purported) biological differences.

4.2 Novelty

Scientific pursuit or output is considered novel if it departs in significant ways from currently accepted theories, models, and hypotheses, with the goal of either supplementing or completely replacing them. The novelty criterion is motivated by skepticism about mainstream, androcentric frameworks that potentially fail to account for current problems (Longino 1996: 46). Recently, Buss (2020) claimed that EP “is a scientific revolution, providing a fundamental paradigm shift” (316), indicating that EP is not, prima facie, motivated by conservatism. In light of this, assessing if EP research fulfills the novelty criterion is particularly interesting. In fact, EP has been criticized by feminists precisely because it does not offer novel explanations and does not concern itself with novel phenomena.

In EP, any postulated psychological differences between men and women are followed by claims about equality between the genders, such as equality of intelligence, indicating more progressive views at the surface. This stands in contrast to traditional theories of male superiority and female inferiority. However, Meynell (2012) criticizes such ‘different but equal’ explanations of men and women. She argues that “we are encouraged to think that this “new” account of dimorphic human nature is not the pernicious sex essentialism of the past” (2). ‘Essentialist’ is a term used to characterize research that relies on the idea that “certain phenomena are natural, inevitable, universal, and biologically determined” (DeLamater and Hyde 1998: 10). Meynell (2012) seems to be relating to this type of research: After closer inspection of EP theories, one finds explanations of gender psychology as biologically determined, a strategy considered one of the most conservative attempts at explaining gender. So, while claims about gender equality seem compatible with feminist thought at first glance, the method underlying the study of psychological sex differences is not novel.

Moreover, EP aims to explain phenomena that have been studied previously. Lloyd and Archer (1981) explain that psychological sex differences became an exciting field of study in the 1960s, way before EP emerged. Social psychologists started to pay attention to the subject by the 1970s and explained differences between men and women not based on biology but on social processes (287). Moreover, traditionally feminist or social scientific issues, such as violence against women, have also been tackled before EP emerged (Kelly 2014). In conclusion, EP does not fulfill the novelty criterion.

4.3 Ontological Heterogeneity

An ontologically heterogeneous theory grants equality to different kinds of entities, as opposed to ontologically homogeneous theories, which tend to erase individual differences by treating them as (inferior) variations of a particular standard entity. A well-known example of an ontologically heterogeneous scientific method is the explicit preservation of differences in the study of primates as conducted by Altmann (Longino 1996, 46f). Does EP’s scientific output grant parity to different sorts of entities without positing a standard? The sex and gender binaries present in all EP theories of gender come directly to mind. EP theories of mating and sexuality that start with anisogamy as the root of psychological differences between men and women tend to posit a dimorphic distribution of traits in absolute terms, both biologically as well as socially. Any other type of human that does not fit neatly into these binaries is erased from their ontology. Indeed, Meynell (2012, 2) criticizes that EP ignores the differences within the broad groups of men and women, respectively, and instead posits standard types, which contrasts with focusing on particularity and individuality. In Meynell’s (2012) words, EP relies on “the idea that women simpliciter are not merely of a particular type but are innately, naturally, or biologically so” (2f). A similar claim might be made about men.

So far, the argument has been about differences between men and women, respectively, but one might take it further. Relying on the idea that females and males are the standard types means being negligent towards variations of those types as previously described. Even more, it means completely erasing humans who are biologically neither strictly female nor male. EP is silent about the psychological mechanisms of intersex people because, in their theories, intersex people practically do not exist. EPs might argue that intersex people are ‘negligible’ because they are an inferior variation of a typical male or female. However, it is exactly this line of thinking that goes against ontological heterogeneity as a virtue: Difference should not be seen as something negligible or inferior, but should be treated equally to what are often considered ‘standard’ types or entities.

Another standard in EP is evidently heterosexuality. Theories about mating preferences rely on the idea that human sexuality is, for the vast majority and at least primarily, comprised of attraction to the opposite sex. The ontological commitment to heterosexuality as a standard seeps into theories about the psychology of homosexual individuals, too. For instance, EP hypothesizes that butch lesbians have similar mate preferences to heterosexual men and femme lesbians are similar to heterosexual women (Smith, Konik, and Tuve 2011: 658). This highlights the idea that homosexuality is an inferior variation of heterosexuality in EP since at the core of any relationship, EP posits typically male and female roles. We see this even more explicitly in how EP theorizes about homosexuality as “an extreme of an adaptive trait of bisexuality” (Bártová and Valentová 2012: 61) potentially adaptive for ensuring the survival of offspring. This line of reasoning shows how non-heterosexuality is conceptualized as something that bolsters scripts of heterosexual pairing and is thus treated as an inferior variation. All in all, EP posits a homogeneous ontology, and individual variation and particularity are challenging to find in EP.

4.4 Mutuality of Interaction

Mutuality of interaction can be understood as a call for a science of complex interaction: It requires that scientific output treats relationships between entities and processes as mutual rather than unidirectional and as positing multiple rather than single factor models. Because not one factor controls an outcome, mutuality of interaction reflects the rejection of dominant-subordinate relationships in scientific explanations (Longino 1996, 47). One example of how complex interaction manifests in science is Fausto-Sterling’s (2021) framework for conceptualizing gender/sex development in infants as a dynamic process consisting of phases with more or less stability in identity. Rather than positing one defining factor, the theory highlights the interaction of different factors such as lived body experience, mental life, and the environment.

Does EP posit models and theories of complex interaction? In relation to this, it is worth discussing an article by Kenrick, Trost, and Sheets (1996) on the phenomenon of so-called ‘trophy wives.’ When older, wealthy, and powerful men marry young and conventionally attractive women after divorcing their same-aged spouses, the new wife is typically called a ‘trophy wife’ (31). The authors criticize that “sociocultural views [of trophy wives] are inadvertently sexist (…) [and] imply that females are helpless pawns and males are omnipotent despots in relationships” (29f). Instead, an evolutionary perspective is deemed more suitable. More specifically, the authors claim that it is helpful to relate the phenomenon of trophy wives to sexual selection since it “assumes female choice as its most powerful driving force” (30). Regardless if their accusations against sociocultural analyses of the discussed issues are warranted, it is worth evaluating if the evolutionary perspective can offer a theory of mate choice based on complex interaction. The authors explain that female choice is the more important driver of sexual selection than male choice. The idea that female choice primarily drives sexual selection goes all the way back to Darwin’s theory of mate choice selection (Buss 1994). In the 1970s, it was argued following up on that theory that females, because of their higher minimum parental investment, are “selected to be more discriminating in choosing a mate” (240), leading to competition among males for access to females. Thus, based on this theory, multiple males might compete for sexual access to one female, but the female makes the ultimate choice (41f). As an outcome of this process, women tend to desire older, more powerful men as mates, and men prefer younger, conventionally attractive women (44f). The idea is that from an evolutionary perspective, it is advantageous for young women to pursue older, wealthy men and that they are by no means ‘helpless pawns’.

Indeed, it seems that in the theorized ancestral environment posited in EP theories, women and men are both given equal amounts of agency: Males can choose and compete for a female, and a female can accept or reject a willing male so that both females and males are assigned goals of their own. In this context, it is worth discussing Longino’s (1996) explanation of single factor models in more detail:

“(…) simple models of single factor control make one party (the dependent as distinct from the independent variable) in interaction a passive object rather than an agent. This has been the fate of female gametes in accounts of fertilization and of female organisms in accounts of social structure. Asymmetry of agency in the physiological context is used to naturalize asymmetry in the social. These naturalizing arguments are explicit in sociobiological stories attributing the presumed docility of females and activity of males to anisogamy (…)” (47, own emphasis).

Notice that even though females and males are assigned their own goals, and even though female choice is technically thought to be the primary driver in mate choice, the theory that even a seemingly feminist contribution such as Kenrick, Trost, and Sheets (1996) relies on is one that presupposes a biological asymmetry between males and females, and this asymmetry is, as Longino explains, used to argue that social behaviors are in fact natural. The agency of the respective gender is limited by its respective (imagined) reproductive role. Further, it seems that Kenrick, Trost, and Sheets (1996) attempted what Meynell (2012) refers to as a ‘different but equal’ explanation of men and women, as discussed in the subsection on novelty. At the end of the day, docile females and active males are absolutely a product of EP theories, fitting relatively neatly into the described criticism of single factor control models. Thus, EP’s scientific output does not display mutuality of interaction all the way down.

4.5 Applicability to Current Human Needs

Applicability to current human needs means that scientific output ought to be assessed based on the effects of its adoption, and research programs that can generate applicable knowledge to improve economic and social outcomes are preferred (48f). EPs might argue that this virtue is inadequate for evaluating EP’s scientific output since EPs are motivated by value-free scientific exploration (Fehr 2012, 58). In contrast to the virtues discussed above, which concern empirics, theory, and methodology, applicability to current human needs might be seen as more directly political, which might be considered problematic. However, it is widely accepted that science is not and cannot be value-free (Douglas 2009; Dupré 2007; Kitcher 2011; Putnam 2002; Rudner 1953). As outlined in the last several sections, EP can by no means be understood as a value-free, ‘purely scientific’ endeavor. EP might not explicitly claim to follow any political goals but displays a commitment to a particular worldview as a research program. This is in line with Nelson’s (2017, 259) analysis of EP as containing both empirical and normative claims. Thus, it is vital to evaluate its political impact.

So, can insights from EP be applied to improve the material conditions of humans in a way that benefits as many as possible? Criticisms entail that EP research based on absolute sexual dimorphism and biological essentialism can be politically dangerous by undoing feminist work to improve the material conditions of humans. Any effort toward the emancipation of women and queer people from gender roles would be thwarted by theories that purport to naturalize such roles. This, in essence, is Meynell’s (2020, 324) standpoint on EP’s gender essentialism and its assumptions about the nature of men and women. Moreover, the public adoption of EP findings has been argued to be a problem. Jackson and Rees (2007) argue that the mainstream dissemination of evolutionary psychology accounts poses a problem to the adoption of well-founded sociological accounts of societal issues, especially because EP can masquerade as more ‘objective.’ Moreover, according to Ging (2017), in the manosphere, a conglomeration of anti-feminist online interest groups claiming that men need to be liberated from ‘feminist delusion,’ EP findings have been used to give rise to a “uniquely misogynist, heterosexist, and racist lexicon” (649).

So far, the feminist literature on EP speaks to the issue of practical application only indirectly: Instead of simply not being able to generate applicable knowledge, EP tends to produce knowledge that can be harmful when adopted by science and popular culture. In addition, it is worthwhile to evaluate what EPs themselves suggest about applying their scientific output for practical purposes. As a last attempt to investigate EP theories’ potential for practical application, one might analyze theories about the origins of men’s coercive behavior, such as Thornhill and Palmer’s (2001) evolutionary theory of rape. The authors claim that rape is not morally justified and is currently ‘maladaptive.’ They argue that teaching young men about the causes of such predispositions might be helpful in rape prevention (Thornhill and Palmer 2001 in Nelson 2017, 262). This might indeed be a practical application of EP scholarship, but preventative measures against male violence based on EP theories might not have the desired effect. According to Nelson (2017), the way in which Thornhill and Palmer phrase that male coercive behavior used to be adaptive but is now considered ‘maladaptive’ does not negate that such behavior would reduce reproductive fitness. Instead, the way they employ the term ‘maladaptive’ seems to suggest that the behavior in question is ethically concerning because it is harmful to women in the present and thus rather signifies ‘irrational’ or ‘unwanted.’ However, the same ethical concern is not extended to ancestral women since the same kind of behavior is not viewed as ‘maladaptive’ in that case (263f). Another point of critique not discussed by Nelson might be the following: Claiming that the predisposition to rape is something males have inherited from their ancestors might be unfair towards the men who never experienced the ‘urge to rape’ and would normally go on to never display any type of violence towards women. This is because such education would paint every man as an inherently violent creature who has to fight his ‘natural’ urges.

In conclusion, EP promotes contentious gender stereotypes instead of fighting them, the latter being the goal of the feminist movement. Thus, it is difficult to see how EP would be beneficial and thus fulfill the virtue of applicability to current human needs. Therefore, it remains that EP theories have little to no potential for practical application in the ameliorative sense, and the adoption of EP theories can even be harmful.