We Can’t Just Lie About Sex Differences
Steve Stewart-Williams on Evolutionary Psychology, Pseudoscience, and the Politics of Gender
In this conversation with Timur Sevincer and Nikil Mukerji, Steve Stewart-Williams, author of the new book A Billion Years of Sex Differences, discusses what evolutionary psychology actually claims, how the field can deepen our understanding of human nature, where the field has gone wrong, why the charge of pseudoscience misses the target, and how skeptics should respond when real science becomes politically unwelcome.
Common Misunderstandings About Evolutionary Psychology
Timur Sevincer: As skeptics, we care about pseudoscience, but also about exaggerated, distorted, or plainly false scientific claims. Which claims about sex differences or evolutionary psychology are most often misrepresented?
Steve Stewart-Williams: The list is long!1 A common view is that evolutionary psychology is a way to justify existing social roles, especially traditional sex roles. Another is that it’s a way to excuse misbehavior, especially male misbehavior. Both claims are false. Explaining a behavior is not the same as morally justifying it. That men prefer more sexual variety and partners than do women, on average, says nothing about what men or women should do. We have to decide for ourselves what’s right and wrong. We shouldn’t let natural selection dictate our ethics. Evolution can explain behavior without excusing it.
Another common claim is that evolutionary hypotheses are unfalsifiable.2 That may have been true in the 1980s when such accusations were hurled at the nascent field (then called sociobiology), but it no longer applies. Evolutionary hypotheses are falsifiable, and plenty have been falsified. More of my own hypotheses have been falsified than I’d care to admit. I’ve come up with hypotheses and conducted research on them, but sometimes my scientific colleagues respond “No, you’re wrong about that.”
Most human sex differences are statistical differences with large overlap.
A further misunderstanding is that evolutionary psychologists think men and women are virtually different species: women are monogamous, men are promiscuous; women are choosy, men will sleep with anything that moves; men care only about a woman’s looks, women only about a man’s resources. None of this is what evolutionary psychologists claim. Most human sex differences are statistical differences with large overlap. Both sexes can form committed relationships and both can be interested in casual sex. Both sexes are choosy about long-term partners and both care about looks. Sex differences are documented statistically by comparing averages and extremes, while acknowledging overlaps and similarities.
Finally, many people assume that if a trait has an evolutionary component, it must therefore be fixed or biologically predetermined. That’s not the claim. Genes may push development in certain directions, but they’re not the only influences. Learning, culture, incentives, and institutions also matter.
Timur Sevincer: I teach open-minded, curious, and intelligent students. However, some of these misunderstandings seem to be very common before they engage with evolutionary psychology in more depth. In rare cases, some course evaluation comments continue to portray evolutionary psychology as pseudoscience even after extensive discussion of overwhelming empirical evidence. Those ideas must come from somewhere.
Explaining a behavior is not the same as morally justifying it … Evolution can explain behavior without excusing it.
Steve Stewart-Williams: I’ve encountered that too, though less in Malaysia, where I now teach, than in the UK. Some people appear to associate evolutionary psychology with outdated or harmful ideas before they’ve really looked at the evidence. They’ve picked up widespread but mistaken views about the field. I think there’s been a successful misinformation campaign against it.
Is evolutionary psychology pseudoscience?
Nikil Mukerji: Pseudoscience is not simply unfalsifiability. Some pseudosciences, such as homeopathy, are eminently falsifiable and have been falsified. The core issue is that they fake doing science. The philosopher Harry Frankfurt, for example, has famously defined bullshit as indifference to truth.3 Science, by contrast, is governed by truth-directed norms: heeding evidence, testing claims, trying to find out what is the case. Pseudoscience fakes participation in that truth-seeking practice. That is why the pseudoscience charge against evolutionary psychology is more than insulting. One has to say what exactly is wrong with the claim or field. Does it avoid falsification? Does it ignore counterevidence? Does it only imitate the surface of science?
Steve Stewart-Williams: Exactly. Anyone can say, “You’re a pseudoscientist!” It’s like that Spider-Man meme where the two Spider-Men point accusingly at each other. The label alone doesn’t settle anything. You have to say what, specifically, is wrong with each claim.
Claims in evolutionary psychology aren’t all equally strong. Some are weak, some have been falsified, but some are excellent. The right approach is to avoid thinking “evolutionary psychology is all good” or “evolutionary psychology is all bad,” and instead to evaluate each hypothesis or theory on its merits.
A Field That Can Be Wrong
Timur Sevincer: Could you give examples of how evolutionary hypotheses can be falsified?
Steve Stewart-Williams: A famous example is the kin-selection explanation of same-sex sexual orientation, associated with E.O. Wilson. Same-sex sexual orientation is puzzling in evolutionary terms. Wilson suggested that it might be a form of kin altruism: instead of having children of one’s own, someone might help siblings or other relatives to have children, thereby passing on their genes indirectly through them.4 But people did the research, and the verdict is that this explanation is false.5
Timur Sevincer: Another example might be the dual-mating or ovulatory-shift idea: the idea that women might seek investment from one kind of man and “good genes” from another, with preferences shifting around ovulation. There were T-shirt smell studies and other work that initially attracted considerable attention,6, 7 although today some of the literature contests this hypothesis.
Steve Stewart-Williams: I’m very skeptical of the ovulatory-shift hypothesis. To be fair, the debate isn’t over—there are still people on both sides of the issue. But I’m much more persuaded by the people who say that the evidence hasn’t borne it out.
Timur Sevincer: How does evolutionary psychology fare after the replication crisis?
Steve Stewart-Williams: It hasn’t emerged unscathed, but I think it fares fairly well. It does better than social psychology, though not as well as behavior genetics. The most robust claims in the field are those grounded in well-established theories from evolutionary biology: sexual selection theory, parental investment theory, parent-offspring conflict theory, kin selection theory. These aren’t just free-floating speculations. They’re grounded in principles that have been demonstrated across many species. This is especially true for many sex differences.
Another reason the field has done well in certain areas is that it was seen as controversial from the outset. To persuade people, evolutionary psychologists have often had to clear a higher evidential bar. David Buss’s work on mate preferences, which found that men and women differ on average in how strongly they prioritize certain traits in long-term partners, used extensive cross-cultural evidence.8, 9 Martin Daly and Margo Wilson’s work on homicide and step-parenthood, the so-called “Cinderella effect,” that children are at greater risk of abuse and violence from stepparents than from biological parents, drew on large archival datasets.10 The sex difference in interest in casual sex has multiple converging lines of evidence: cross-cultural evidence, real-world dating data, cross-species patterns, and links to prenatal hormonal exposure.11 This convergence increases our confidence in the hypothesis.
Nikil Mukerji: There is an interesting complication here. A low replication rate can be a warning sign. But in some pseudoscientific or ideological literatures, one finds suspiciously high positive-result rates: everything seems to confirm the preferred story. So a perfect replication rate can be suspicious as well.
Steve Stewart-Williams: Yes. To do replication properly, you need open science practices, preregistration, stated methods that could falsify the hypothesis, and ideally multiple research groups, including people with different prior commitments. Even if a claim is falsifiable in principle, one isn’t really doing science unless one uses methods that could in principle falsify it.
The sexes do differ on average, and I think there’s an evolutionary contribution to some of those differences. We can’t just lie about that. The right response is to tell the truth carefully and responsibly.
Social psychology deserves credit here. The fact that we know about its replication problems shows that many in the field are testing their claims and trying to improve. A low replication rate isn’t fatal if researchers are actively checking their findings and weeding out false positives. The problem is that, until recently, they often weren’t.
The Ethical Trap
Nikil Mukerji: Some findings in evolutionary psychology are bound to annoy people, especially in certain political milieus. Are you concerned that this kind of science might play into the wrong hands? How do you deal with the ethics of communicating sex differences?
Steve Stewart-Williams: Claims about sex differences can certainly be misused if people exaggerate the size of the differences, or if people infer that, because a difference has an evolutionary basis, it’s therefore good or should be preserved. That can do harm. If we exaggerate differences, we may lower our expectations for one sex in certain domains.
To this issue I have two responses. First, the sexes do differ on average, and I think there’s an evolutionary contribution to some of those differences. We can’t just lie about that. The right response is to tell the truth carefully and responsibly.
That means emphasizing that most human sex differences aren’t huge.12, 13 There’s a lot of overlap between the sexes. Everyone should get a fair chance. And the fact that something is natural, genetic, or partly shaped by evolution has no direct moral implications.
Second, denial can also do harm. It’s not the case that exaggerating sex differences is dangerous while minimizing them is harmless. We can get this wrong in both directions.
Nikil Mukerji: So denial is wrong in principle, and one can also make a consequentialist case against it.
Steve Stewart-Williams: Exactly.
Timur Sevincer: This is also a communication problem. Many people hear “sex differences” and immediately think one is defending misogyny or fixed social roles. In my own teaching, I often begin by saying: there is usually substantial overlap between the sexes, most differences are small, evolutionary influences don’t make differences fixed or unmalleable, and a trait’s evolutionary origins don’t make it morally good or bad.
Nikil Mukerji: And acknowledging average differences doesn’t necessarily make people conform to them. In some cases, it could even have the opposite effect: if people are aware of an average tendency, they may consciously decide not to follow it. Recognizing statistical differences and encouraging individuality are not incompatible.
Steve Stewart-Williams: I think that’s right. It’s also important to take people’s concerns seriously. Some people worry that evolved sex differences will be used to revive old claims about male superiority or fixed gender roles. Given our history of anti-female sexism, that worry is understandable. But the right response isn’t to shut down the discussion. The right response is to discuss the evidence carefully.
Socialization, Stereotypes, Patriarchy
Timur Sevincer: When people think about sex differences, they often default to three explanations: socialization, stereotypes, and patriarchy. Boys and girls are raised differently; stereotypes become self-fulfilling; women are forced into roles by structures created by men. You deal with these explanations in A Billion Years of Sex Differences. Where do they fall short?
Steve Stewart-Williams: Well, let me start by saying that I think “fall short” is the right phrase. It’s not that they have no merit. It’s impossible to believe that socialization has no impact on behavior. Stereotypes have an impact. Cultural factors matter. But they’re not enough on their own.
Take parental treatment. It’s often invoked as an explanation for sex differences. But there’s research suggesting that, at least in the West, parents don’t treat sons and daughters particularly differently in many important ways.14 Behavior genetics also suggests that, within the normal range, the shared family home doesn’t have a massive impact on how people turn out.15 This doesn’t mean genes explain everything. The environment matters too. But the simple parental-socialization story is weaker than many people assume.16
On patriarchy, the gender-equality paradox is a serious problem for simple versions of the story. For many traits, sex differences are larger, not smaller, in wealthier, more individualistic, more gender-equal nations.17, 18 That’s the opposite of what one would expect if the differences were due to strict sex roles or patriarchy. I’m not saying that every difference follows this pattern, or that culture is irrelevant—in fact, the gender-equality paradox is itself a cultural effect. What I am saying, though, is that it challenges certain traditional sociocultural explanations for differences between the sexes.
On stereotypes, the work of Lee Jussim has been a major influence on me. Stereotypes can be self-fulfilling to some extent, but the main causal direction is usually from social reality to stereotype, not from stereotype to social reality.19 In the case of sex differences, stereotypes are partly shaped by actual average differences that people perceive in the world, and they correspond reasonably well to the actual sex differences.20 In some cases, those stereotypes might then amplify or galvanize the differences. But they don’t create them out of nothing.
Nikil Mukerji: There is a habit of reasoning that often gets in the way here. Instead of asking whether a claim is true, one asks: “What makes you say that?” and “Who does this help?” Truth recedes into the background.
Steve Stewart-Williams: That can become a form of conspiracy theorizing. If someone says, “You only believe that because you’re a man,” or “You only believe that because you’re defending your group,” that’s a claim about motive. But even if the motive attribution were true, it wouldn’t show that the claim is false. It’s essentially just an ad hominem argument.
What Denial Costs
Timur Sevincer: Many people understand that exaggerating sex differences can do harm. What is less obvious to many is that downplaying differences can also do harm. Can you give examples?
Steve Stewart-Williams: One is that, if we minimize sex differences in psychological disorders, we may fail to spot them in the less affected sex. ADHD often presents differently in girls than in boys, and the standard description fits boys better. That means girls with ADHD may be overlooked. Autism may also be overlooked in girls because it presents differently than in boys—for example, it tends to involve fewer repetitive behaviors. Conversely, depression in men more often involves substance use and aggression than in women, so male depression may be missed.
It’s not the case that exaggerating sex differences is dangerous while minimizing them is harmless. We can get this wrong in both directions.
There are social and occupational examples as well. If we ignore average differences in occupational interests, we may assume that every gender gap in a field is due entirely to bias. Bias exists, but sex differences in interests also matter. Men are, on average, more interested in things and things-related jobs; women are, on average, more interested in people and helping professions.21 If we ignore that, we may misdiagnose the cause of gaps and pour resources into interventions that don’t address the real causes.
It can also lead to coercive policies. If one reserves positions for one sex, or chooses a less qualified candidate to achieve numerical parity, one may reverse bias rather than remove it. On top of that, it can cast doubt on the competence of the supposedly benefited group. Even beneficiaries themselves may wonder whether they really deserved the position. And portraying STEM as saturated with sexism may deter some girls and women who would otherwise enjoy those fields.
Discrimination exists (although far less today than decades ago). But exaggerating it, and ignoring the progress that has been made, can backfire.22 And we shouldn’t overlook the fact that there’s also evidence of discrimination against men in some hiring contexts.23, 24 Everyone focuses on the lower percentage of women in STEM fields as a sign of discrimination, but no one makes the charge that men are being discriminated against in fields dominated by women.
The Slippery Word “Gender”
Nikil Mukerji: In A Billion Years of Sex Differences you are quite critical of the concept of gender. Why?
Steve Stewart-Williams: I don’t think it’s a very useful concept, partly because it’s not just one concept. Sometimes “gender” is just a synonym for sex. Sometimes it refers to masculinity and femininity. Sometimes it refers to sex differences supposedly due to nurture rather than nature. Sometimes it refers to social roles, stereotypes, gender identity, or clusters of personality traits.
The nurture-based definition is especially problematic. It often assumes what should be demonstrated: namely, that nurture is the sole cause of the difference. Many recurring male-female differences involve both nature and nurture. If one defines gender as the nurture part and sex as the nature part, most female-male differences become partly sex differences and partly gender differences. That’s an awkward way to frame things.
Other definitions have other problems. If “gender” means personality profiles, for instance, then the claim that there are many different genders collapses into the trivial observation that every person has a unique personality profile. At that point, we’re talking about personality, not something clearly linked to sex.
We should be able to say: Yes, some past claims were wrong and sexist; no, that doesn’t mean every claim about sex differences is wrong or sexist.
For these and other reasons, I prefer to talk about sex rather than gender. In biology, sex is defined by gametes: females are organized around the production of eggs; males around the production of sperm. Sex-linked traits such as chromosomes, genitals, hormones, masculinity, femininity, sex-typical behavior, and personality traits don’t define sex. Instead, they’re among the many traits that evolved to facilitate reproduction via either eggs or sperm.
Why This Topic Can Set Rooms on Fire
Timur Sevincer: Why do sex differences, sexual behavior, and evolution sometimes provoke such strong emotions? In principle, one could discuss them scientifically.
Steve Stewart-Williams: The topic is inherently combustible. A big part of the reason is our history of sexism against women. Early scientists made false and sexist claims about women’s intellectual abilities, so it’s understandable why people worry. If we open the door to evolved sex differences, are we going to slide back into Victorian ideas about male superiority and fixed roles?
But the science of sex differences actually refutes those early sexist claims. Shutting down the discussion is the wrong reaction. We should be able to say: Yes, some past claims were wrong and sexist; no, that doesn’t mean every claim about sex differences is wrong or sexist.
Unfortunately, a modern trend on campus and elsewhere is to fight unwelcome ideas not with arguments and evidence, but with accusations, complaints to authorities, protests, occupations of classrooms, and so on. These are intimidation tactics as much as anything else. It’s anti-intellectual. In principle, anyone with any view can use intimidation tactics. Thus, intimidation can’t sort out who’s right and who’s wrong.
I had my own taste of this. A paper I wrote on sex differences in STEM25 caused an online controversy and led to calls for retraction. The issue reached the diversity, equality, and inclusion (DEI) office at the University of Nottingham’s UK campus. They did not like the paper! To their credit, though, they defended my academic freedom, and I was promoted to full professor despite listing the paper as one of my main pieces of work. But the experience was still worrying. I didn’t seriously think I’d lose my job, but I did wonder whether expressing perfectly reasonable views about sex differences might make it difficult for me to get another academic job.
Some people worry that evolved sex differences will be used to revive old claims about male superiority or fixed gender roles … That worry is understandable. But the right response isn’t to shut down the discussion. The right response is to discuss the evidence carefully.
That was part of why I started my Substack, The Nature-Nurture-Nietzsche Newsletter. I wanted a fallback. If the worst came to the worst, I wanted another way to make a living. And it’s worked out well: I’m no longer financially dependent on my academic job. It’s a great position to be in. I’ve always tried to write honestly and openly about the topics I cover, even when knowing it might not be popular. But now I do so with a lot less trepidation.
Can persuasion still work?
Nikil Mukerji: Some of these controversies are not only about evolutionary psychology itself, but also about deeper disagreements concerning evidence, reasoning, and the interpretation of scientific claims. How does one talk to people who have been taught that logic is a tool of oppression? They often fail to separate the truth of a claim from alleged motives for making it.
Steve Stewart-Williams: Even people who officially reject reason and evidence may still be moved by logical arguments. They might not change their minds immediately, but they may eventually. Social psychologists call that a sleeper effect, and it’s a well-documented phenomenon.
Nikil Mukerji: My suspicion is that another causal factor is needed. Some people change their minds after reality punches them in the face. Then they remember that someone had explained the matter differently.
Steve Stewart-Williams: That could well be right. But persuasion is still worth attempting. It won’t work on everyone. Nothing will work on everyone. But across history, humanity has moved away from superstition toward more scientific ways of thinking. That suggests that slow persuasion is possible.
Can evolutionary science inform moral reasoning?
Nikil Mukerji: In A Billion Years of Sex Differences, you reject the inference from “natural” to “good,” but David Sloan Wilson and colleagues26 argue that evolutionary psychologists often overcorrect here: Hume’s point about not inferring an ought from an is, is only that ought cannot be derived exclusively from is, not that evolved facts have no ethical relevance. So when you say that the science of sex differences cannot answer policy questions because those are ultimately about values, how do you avoid turning a valid warning against “natural = good” into a stronger and false separation between facts and ethics? More concretely: should evolved sex differences alter our moral and political reasoning about fairness, coercion, individual freedom, and the costs of eliminating gender gaps—and if so, exactly how?
Steve Stewart-Williams: I don’t think anyone argues that evolved facts have no ethical relevance. They just argue that the fact that something has an evolutionary origin doesn’t necessarily mean it’s good, bad, or anywhere in between.
Example: It might be the case that, if a given sex difference has an evolutionary origin, it would be hard to eliminate, and that doing so would do more harm than good. As a result, we might decide that we shouldn’t try to eliminate it. But the reason for that decision wouldn’t be that the sex difference is natural; it would be that trying to eliminate it would do more harm than good. In other words, we’d be dealing with a consequentialist argument, not an argument from nature.
Importantly, this cuts both ways. If eliminating a socially constructed sex difference would do more harm than good, then we shouldn’t do that either, even though the difference isn’t “natural.” And if eliminating a natural difference would boost general well-being, we should do it despite its natural origins. In itself, the nature-nurture question is irrelevant to the ethical question.
We can turn the tables here as well: Why should we exclude certain facts from our ethical reasoning just because they happen to have an evolutionary origin? To do so is to risk falling prey to an anti-naturalistic fallacy!