2. Two positions

The primary question when comparing the sexual desires of males and females is whether there are differences between typical male and female sexual desires, and if there are, what is the nature of those differences (strength of desire, heterogeneity in objects of desire, etc). Moreover, the question is whether such differences reflect deep facts about human nature. The standard view holds that male sexual desires are stronger and less choosy than female sexual desires. We saw in section 1 that the standard view has been asserted by various scientific programs, and it is of course widely asserted in folk discourse and popular culture. Yet, an alternative position denies the standard view.

Today the standard view is perhaps most prominently articulated by evolutionary psychology. Indeed, the standard view is a core commitment of evolutionary psychology, and the putative success of evolutionary psychology in explaining the standard view is held to be an argument in favor of evolutionary psychology. “Parental investment theory” of evolutionary psychology holds that the differential resources demanded of mothers and fathers to produce and raise offspring entail different optimal mating strategies for females and males. By their nature, goes this thinking, males tend toward causal sexual encounters and polygyny while females tend toward careful mate choice and monogamy.^{Footnote 3} This view was asserted in strong terms by the evolutionary psychologist Donald Symons:

To maximize evolutionary fitness, an organism must maximize the number of surviving offspring. According to this line of thinking, females and males have different optimal reproductive strategies to maximize fitness. Males are better off reproducing indiscriminately while females are better off being careful in their mate choices, goes this theory, and these facts of our long evolutionary history influence the nature of male and female sexual desire today. Proponents of the standard view have appealed to a massive volume of empirical evidence; I discuss an illustrative sliver of that in the text that follows. Critics have noted numerous biases in this evidence, and the inference from this evidence to conclusions about sex differences in desire is severely underdetermined.

An alternative position denies the standard view. To do so requires an explanation for the great volume of empirical evidence that has been amassed in favor of the standard view. Broadly construed, this is accomplished using various forms of constructivism about sexuality. The standard view assumes that there is a nature to sexual desire, precisely because the view holds that sexual desires of males and females are different by their nature, and therefore, sexual desires must have some natural basis. If, instead, sexual desires are constructed by social and cultural forces rather than our biological constitution, then the observed differences between males and females would be contingent and the standard view would be false.

Constructivism about sexuality has been argued for from various quarters. Shere Hite, the lead on the massive surveys of female sexuality that led to the Hite Report, claimed “our model of sex and physical relations is culturally (not biologically) defined, and can be redefined—or undefined” (Reference Hite2004 [1976], 335; emphasis in original). The philosopher Roger Scruton similarly claimed that “[s]exual desire, like the human person, is a social artefact, and can be built in many ways” (Reference Scruton1986, 15). And the historian David Halperin made the pithy claim that “sexuality is not a somatic fact; it is a cultural effect” (Reference Halperin1989, 257). Indeed, constructivism about sexuality is a staple among some historians and anthropologists, who emphasize historical and cultural variation in sexuality: homosexuality emerged in the nineteenth century, pederasty was normal in Ancient Greece and in some contemporary indigenous groups, females have been the libidinous sex in some times and places. If our sexuality is indeed so diverse and so sensitive to cultural context, then, goes this thinking, little general can be said about sex differences in sexual desire.

A specific version of constructivism with respect to sex differences in sexual desire was offered by Freud. Freud seemed to think that males and females had similar sexual natures through childhood, yet “[a]mong girls, the tendency to sexual repression generally appears greater” (Reference Freud1905, essay 3). Freud proceeded to argue that apparent differences between the sexes do not appear until after puberty: “[T]he sexuality of little girls is entirely masculine in character,” in which, by masculine sexuality, Freud meant simply that the libido is strong. He continued, “the libido is regularly and invariably masculine in nature, whether it occurs in men or in women, and irrespective of whether its object is a man or a woman.” In short, Freud seemed to hold that the strength of sexual desires of males and females is similar by nature, but females tend to experience more repression, which becomes more salient as females mature and is pronounced by teenage years.

Another version of constructivism that could call into doubt any empirical support for the standard view was defended by the feminist scholar Catherine MacKinnon, who urged that the study of female sexuality must consider inequality and sexual violence. There is no way to understand female sexual desire, argued MacKinnon—including what females in fact do and what they say they desire—without considering the constant threat of injustice and violence that females face: “This is why interpreting female sexuality as an expression of women’s agency and autonomy, as if sexism did not exist, is always denigrating and bizarre and reductive” (Reference MacKinnon1989, 153). Thus, we ought not interpret first-person claims about female sexual desire, or third-person observations of female sexual behavior, as authentic manifestations of female sexual nature, even if the females under study explicitly claim that their desires are autonomous features of their agency. If so, then empirical evidence based on first-person reports or observations of sexual behavior cannot confirm or disconfirm theories about the nature of female sexuality or natural differences between male and female sexual desires.

Thus, there are two profoundly different positions regarding the standard view: One position, often asserted by evolutionary psychologists and others, defends the standard view, while the other position, often asserted by the human sciences, denies it, holding instead that observed sex differences in sexual desire are the result of social forces.

3. Behavioral science

An enormous range of evidence has been gathered that has been taken to both vindicate and challenge the standard view, including evidence from anthropology, psychology, primatology, and sociology. To illustrate the sorts of methodological problems such work can face, and how scientists address these problems, I focus on a few lines of evidence from the study of human behavior.

Two forms of evidence that are relevant to the standard view are what people say they do and what people in fact do. For example, the evolutionary biologist Buss (Reference Buss2016) claims that “Women’s current mate preferences provide a window for viewing our mating past,” and thereby provides insight into the nature of female sexual desire, while MacKinnon (Reference MacKinnon1989, 138) argues that pornography viewing habits “provide an answer” regarding the nature of sexual desire. The study of sexual behavior has generated much evidence that appears to support the standard view. However, critics respond by noting that what people say they desire, their self-reported claims about sexual behavior, and sexual behaviors themselves, are heavily modulated by social contexts. To address this criticism, scientists have developed their observational techniques in a variety of ways, such as using fake lie-detector machines, studying sexual behaviors in a diverse range of social contexts, and asking subjects how they would behave in controlled or idealised conditions. The results of these methods suggest the presence of sex differences that are much smaller than earlier survey methods.

The first large-scale surveys about sex led to the Kinsey Reports (the report on males was published in 1948 and the report on females in 1953). Kinsey’s group interviewed thousands of people and gathered data on frequency of various kinds of sexual behavior and demographic data such as sex, age, income, and religion.

One phenomenon that seemed to emerge from such surveys is that males are more interested in casual sex than females. In a widely cited study, males and females in their early twenties on a college campus made one of three kinds of proposals to people of the opposite sex: whether they would go on a date with the proposer, whether they would come over to their apartment that night, or whether they would go to bed that night with the proposer (Clark and Hatfield Reference Clark and Elaine1989). Both males and females said yes to the proposed date roughly 50 percent of the time; in contrast, about 70 percent of males said yes to the proposed sex, while no females did. This article is deemed by some to be a classic in social psychology, and evolutionary psychologists have taken this result as convincing evidence for the standard view (see, e.g., Buss Reference Buss2016). However, alternative explanations to this result have been proposed and tested. One alternative way to interpret these findings is that females are more concerned about their safety with a stranger rather than less interested in casual sex. A second alternative is that females are just as interested in casual sex but less confident that they would enjoy it with the young males who approached them. Another alternative explanation is that males are merely more willing to say that are interested in casual sex compared with females, not that they in fact are more interested, and this difference in willingness to say they are interested in casual sex could arise because norms of masculinity nudge males’ willingness to say that they are interested in casual sex upward, or because norms of femininity nudge females’ willingness to say that they are interested in casual sex downward, or both.

So, the standard view is underdetermined by the empirical evidence from such surveys. But scientists find creative ways to manage this underdetermination.

Developing the survey approach, Conley (Reference Conley2011) asked subjects to imagine this scenario: While on vacation in Los Angeles you are visiting a trendy restaurant, and sitting a few tables away is a famous person, X, who catches your eye and approaches you and asks if you would join X in bed that night. For the male subjects, X was Christie Brinkley, Angelina Jolie, and Rosanne Barr. For the female subjects, X was Johnny Depp, Brad Pitt, and Donald Trump (this study was performed prior to Trump’s foray into politics). Female subjects agreed to the propositions from Depp and Pitt about as often as male subjects agreed to the propositions from Brinkley and Jolie. Barr was rejected by the male subjects, and the female subjects emphatically rejected Trump.

In another survey Conley asked subjects “what did you say the last time someone asked to have sex with you?” Female and male subjects both reported having said yes about half the time. Conley surveyed males and females about their willingness to have causal sex with a good friend of the opposite sex, and when perceived sexual competence was controlled for, females reported willingness as often as males (also reported in Conley Reference Conley2011).

Heterosexual males and females report vastly different sexual behaviors, some of which appear paradoxical (such as the mean numbers of lifetime opposite-sex partners, as suggested by the Kinsey reports and later surveys). One possible explanation for these reported differences is that the sexual desires and behaviors of males and females are in fact very different, thereby substantiating the standard view (though this would not explain difference in mean numbers of lifetime heterosexual partners). However, this observed difference has an obvious alternative explanation: Given social expectations such as the sexual double standard, males are more likely to honestly report or exaggerate particular sexual behaviors and desires (such as their number of lifetime partners) while females are more likely to underreport. Alexander and Fisher (Reference Alexander and Terri2003) posed questions to males and females about their sexual desires and behavior under three scenarios: one in which subjects observed that the experimenter could view their responses, one in which subjects’ responses were kept anonymous, and one in which subjects believed that lying could be detected because they had been connected to a (fake) lie detection machine. The differences between reports of sexual behavior and desire between males and females were greatest in the first scenario, modest in the second scenario, and all but disappeared in the third scenario.

In another genre of experimental design, researchers have subjects engage in a “speed dating” exercise to test the standard view’s prediction of differential selectivity between male and female choices of potential sex partners (Finkel and Eastwick Reference Finkel and Paul2009). Such settings are usually structured as follow: Females are seated around a room in a circle, and males are seated across them in an outer circle; after a set time (usually a few minutes), the “date” between pairs is over, and males stand and move one place over; at the end of the night subjects are asked who they would like to have a second date with, and mutual matches are connected. Psychologists have gathered data from such venues. An advantage of such venues for testing mate choice, compared to studying mate choice “in the wild,” is that the partners who we in fact settle with are the result not only of our preferences but also of our limited range of opportunities—to mate, you must meet.

One finding that seemed to consistently emerge from this genre is that females are far less likely than males to want a second date—roughly half as likely, by some estimates. However, there appears to have been a confound in some of these studies. Speed dating venues have adopted a norm in which it is the males who get up and move tables to greet the next female. Finkel and Eastwick (Reference Finkel and Paul2009) tested to see if the role of “approacher” versus that of “approached” modulates willingness for a second date: In their experiment half the time males approached females and half the time females approached males. They found that when females approached males, they were more likely than males were to say yes to a second date.

Commenting on the reliance of behavioral evidence for the standard view by evolutionary psychologists, John Dupré noted the problem of underdetermination suggested by the previously mentioned examples: “[T]he evidence, for example the answers to the questionnaires designed by Buss to elicit the sexual preferences of large numbers of men and women, do nothing to discriminate between these different kinds of explanation” (Reference Dupré2001, 57). That is, there are various alternative hypotheses that can explain the evidence generated by methods used by studies such as that reported in Clark and Hatfield (Reference Clark and Elaine1989), and this mitigates the empirical support that such evidence provides to the standard view. Yet, such underdetermination is not always insurmountable, as suggested by Conley’s clever experiments and Alexander and Fisher’s use of the fake lie detector. A conclusion that can be drawn, albeit tentatively, is that when questions about sexual desire and behavior are posed in ways that attempt to control for effects of social context, such as the influence of social norms or concerns that casual encounters will be dangerous or unpleasant, male and female sexual desires look more similar than in behavioral studies that do not control for those confounding factors.

Nevertheless, no matter how clever the behavioral methods have become for studying sexual desire, the approaches canvassed in the preceding text are fundamentally based on what people in fact do, what people say they do, or what people say they want to do. What people in fact do is constrained by all sorts of factors, such as social norms or sheer availability of opportunity. What people say they do and what people say they want to do is similarly modulated by social forces. Indeed, what people in fact want to do is plausibly modulated by social forces. That, in turn, motivates an alternative form of evidence pertaining to our behavior, one that is perhaps less constrained by social norms and availability of opportunity: pornography viewing habits.

Studies of pornography viewing habits, afforded by large datasets available from internet pornography venues, have found some male-female differences: Males consume more pornography, begin viewing at a younger age, use it more frequently when in relationships, and view more hardcore content. As noted in the preceding text, the feminist legal scholar Catherine MacKinnon claimed that pornography provides the “truth about sex,” specifically, the truth about male sexual desire: “If pornography has not become sex to and from the male point of view, it is hard to explain why the pornography industry makes a known ten billion dollars a year selling it as sex mostly to men” (Reference MacKinnon1989, 139). However, MacKinnon also claimed that “pornography is a means through which sexuality is socially constructed” (ibid.). That is, pornography can both reflect the nature of sexual desires and influence what is taken to be that nature.

So, when considering what we can learn about the nature of sexual desire by studying pornography, there are two opposing models of the direction of causation. One we can call the “microscope model”: Studying pornography viewing habits informs us about the real nature of sexual desire, which otherwise remains shrouded by factors such as self-censorship in surveys or by social control of behavior. On this model, the content of pornography is evidence about the real nature of our desires; the content of pornography is ultimately caused by, or responsive to, the content of sexual desires. Pornography is like a microscope that can provide evidence about the nature of sexual desire. The other model we can call the “advertisement model”: Studying pornography allows us to observe a way in which our sexual desires are shaped by cultural forces, just as advertisements and other media modulate our desires and preferences more generally. On this model, the content of pornography shapes the content of our sexual desires. Pornography is like an advertisement that influences our sexual desires, just as advertisements influence our desire for many other kinds of objects and activities. MacKinnon is appealing to both models and, indeed, both models might be true: Between the content of sexual desires and the content of pornography, the causal influence is plausibly bidirectional.

Yet, MacKinnon and some other commentators seem to emphasize the microscope model. Here, for example, MacKinnon cites Dworkin, who claims that pornography “reveals that male pleasure is inextricably tied to victimizing, hurting, exploiting” (Reference MacKinnon1989, 140), suggesting that pornography can indeed reveal the true nature of male sexual desires. As another example, MacKinnon writes “From pornography one learns that forcible violation of women is the essence of sex” (ibid.), implying that there is indeed an essence to (male) sexuality, and that pornography is a microscope to learn about it.

A problem with interpreting evidence from pornography viewing habits was suggested by MacKinnon: The direction of causation between contents of desires and contents of pornography is very likely bidirectional. Moreover, the relevance of pornography viewing habits to the standard view is less direct that evidence from behavioral studies. The standard view involves a cluster of claims about strength of desire, desired number of partners, degree of promiscuity, and so forth, and it is not obvious how to make an inference about those aspects of sexual desire from pornography viewing practices.