The Uncomfortable Science Of Sex Differences - Steve Stewart-Williams

2026-07-06 - 153 min - source - Read full transcript

Chris Williamson (host)Steve Stewart-Williams

Key insights

Sex is defined biologically by gamete size (anisogamy), not by any other trait.

Stewart-Williams argues the scientific definition of sex is definitional, not measured: males are whichever type produces the smaller gametes (sperm), females the larger (eggs). This holds across nearly all sexually reproducing species; only a minority of simple organisms are isogamous (same-size gametes), and even those are evolutionarily unstable, tending to split into a small-gamete/large-gamete system over time.

sex-differences

Six independent lines of evidence converge to support an innate component to many sex differences.

Stewart-Williams lists: sex differences appearing very early in development (often before culture could plausibly cause them); differences that emerge despite cultural pressure to suppress them (e.g., societies actively try to curb male aggression, yet it still rises post-puberty); persistence of the same differences across an entire century of measurement; correlation with prenatal hormone exposure (e.g., women with congenital adrenal hyperplasia show more male-typical traits); cross-cultural universality (some differences appear in 53/53 or 80/80 studied nations); and parallels in other species under similar evolutionary selection pressures.

evolutionary-psychology

The gender equality paradox: sex differences often grow larger, not smaller, in more gender-equal societies.

Both social-role theory and patriarchy theory predict that stricter, more patriarchal societies should produce bigger sex differences. The data often show the opposite - more egalitarian societies show larger gaps in personality, career interests, and some physical/cognitive traits. Stewart-Williams found this genuinely surprising even from an evolutionary standpoint, since he expected patriarchy to amplify innate differences further.

gender-politics

Most sex differences trace back to a difference in maximum possible offspring number, driven mainly by parental investment.

Because females typically invest more per offspring (pregnancy, birth, nursing) while a male's minimum investment can be a single sex act, males have a much higher theoretical ceiling on offspring number and higher variance in reproductive success. This selects for traits like male risk-taking, status-seeking, physical competitiveness, and interest in multiple partners, since being one of the few high-reproducing males mattered evolutionarily far more than being an average-reproducing one.

parental-investment

Humans are a mammalian outlier: our reproductive system resembles birds more than most mammals.

Only 5-10% of mammal species show pair-bonding with biparental care, versus roughly 90% of bird species. Because humans invest heavily in biparental care and pair-bonding, the male reproductive-variance advantage (and therefore many downstream sex differences) is present but substantially compressed relative to most other mammals.

parental-investment

The single largest, most obvious psychological sex difference is rarely named: which sex each sex is attracted to.

Stewart-Williams notes that scientific papers summarizing evolved sex differences routinely omit the fact that the overwhelming majority of men are attracted to women and vice versa - an enormous, clearly evolutionarily-grounded difference that gets overlooked because it seems too obvious to state.

mate-preferences

Physical attractiveness matters more to men than women for long-term mates, but that gap vanishes for short-term relationships.

In long-term mating, men place more weight on a partner's looks (linked to youth and fertility, given female menopause), while women weight resources and status more heavily. But in short-term/casual encounters, where no further investment is expected, women weight looks about as much as - or even more than - men do, because in that context looks are the clearest available signal of 'good genes.'

mate-preferences

Explaining behavior evolutionarily does not excuse it - the same standard should apply to sociocultural explanations.

Stewart-Williams repeatedly stresses that an evolutionary account of cheating, aggression, or sexual coercion is not a moral justification for it, any more than a sociocultural explanation would be. Both types of explanation locate a cause outside the individual's choice; if identifying a cause excused behavior, we would have to excuse everything, since all behavior has causes. He credits a line - 'analysis is not justification' - for capturing this.

evolutionary-psychology

Gay men and lesbians act as a natural experiment isolating each sex's true preferences from cross-sex compromise.

Because gay men don't need to negotiate with female preferences, they have more casual sex than straight men; because lesbians don't need to negotiate with male preferences, they have less casual sex than straight women. This is used as evidence that heterosexual behavior reflects a compromise between men's higher average interest in casual sex and women's lower average interest, rather than either sex's preference being 'the truth.'

sex-differences

Male violence scales with severity, and men commit over 90% of homicides in every studied society - a gap that mirrors chimpanzee data almost exactly.

The sex difference in aggression is modest for verbal aggression (roughly Cohen's d = 0.5), larger for low-level physical aggression, and largest at the extreme: men perpetrate more than 90% of homicides across every nation and dataset examined, closely paralleling the roughly 92% of chimpanzee 'chimpicides' committed by males, despite chimps being overall more violent than humans.

sex-differences

Men's jealousy centers more on sexual infidelity; women's centers more on emotional infidelity.

Stewart-Williams links this to paternity uncertainty (no woman has ever doubted a birth was hers, but a man historically could never be fully certain) versus women's greater risk from partner desertion (an emotional affair is more likely to end in the partner actually leaving, cutting off resources and co-parenting support).

mate-preferences

Denying or minimizing sex differences carries real costs, not just exaggerating them - and equality is not inherently good in itself.

Because symptoms and disorder rates differ by sex (e.g., women's heart attacks presenting as shortness of breath rather than arm pain, girls' autism presenting with fewer repetitive behaviors), ignoring sex differences leads to under-diagnosis. Assuming occupational gender gaps are entirely due to discrimination leads to interventions targeting the wrong cause, which then fail and breed resentment. Stewart-Williams uses male violence to show the underlying principle: shrinking that sex difference is desirable only because violence is bad, not because sex differences are inherently bad - you could technically 'close the gap' by raising female violence to match male violence, which would obviously make things worse.

gender-politics

Books referenced

A Billion Years of Sex Differences - Steve Stewart-Williams - The book being promoted and discussed throughout the episode, covering the evolutionary origins of sex differences and their modern social implications.
The Moral Animal - Robert Wright - Both host and guest say this book is what got them into evolutionary psychology; also the source of the theory that parental grief tracks a child's reproductive value.
When Men Behave Badly - David Buss - Cited for research (with Martie Haselton) on the sexual overperception bias, and for findings that men underestimate how upsetting sexual harassment is to women.
The Ape That Understood the Universe - Steve Stewart-Williams - Stewart-Williams' earlier book, referenced as one of a curated reading list of five essential books.

Media referenced

Nature, Nurture, Nietzsche (Substack newsletter) - article - Stewart-Williams' Substack, plugged at the end of the episode; known for monthly 'Linkfest' roundups of studies and insights.

Companies

Ashley Madison - Cited stat that the cheating website had ~20 million active male users versus ~1,500 active female users despite men paying and women getting free membership, used as evidence for the sex difference in interest in casual sex/infidelity.
Goldman Sachs - Referenced in passing as the setting for a viral incident where a man accused a senior female executive, discussed as a rare role-reversal harassment case.
Tinder - Cited for its height-filter feature (originally an April Fool's joke that became real), illustrating stated female height preferences in mates.
23andMe / Ancestry.com - Anecdote about a DNA test revealing a decades-old half-sibling from a one-night-stand affair, used to discuss paternity uncertainty and mate guarding.

Techniques and frameworks

Bateman's principle - The idea that reproductive variance is greater in males than females in most species; Bateman attributed this to anisogamy (gamete size difference).
Trivers' parental investment theory - Robert Trivers' refinement of Bateman's principle: the sex that invests more in offspring (gestation, nursing, care), not just gametes, becomes the choosier, more reproductively limited sex.
Gender equality paradox - The finding that many sex differences in personality, interests, and some abilities are larger, not smaller, in more gender-equal societies - the opposite of what social-role theory predicts.
Sexual overperception bias - Men are more likely than women to misread friendliness as sexual interest; explained via a 'smoke detector principle' where missing a real opportunity is costlier (evolutionarily) than an awkward false positive.
Six lines of evidence for innate sex differences - Stewart-Williams' framework for arguing a sex difference has an evolved component: early developmental emergence, persistence despite cultural pressure to suppress it, stability over decades of measurement, correlation with prenatal hormone exposure, cross-cultural universality, and parallels in other species under similar selection pressures.
Soft bigotry of male expectations - Chris Williamson's term for treating male behavior as the neutral standard, so that women doing something less than men (e.g., entering STEM) is framed as a problem to fix, while men doing something less than women (e.g., parenting) rarely gets the same treatment.
Reproductive value / grief theory - Drawn from The Moral Animal: parental grief after a child's death tracks the child's reproductive value, peaking a couple of years past puberty, illustrated by Darwin's own reaction to losing his 13-year-old daughter.

Summary

Chris Williamson hosts evolutionary psychologist Steve Stewart-Williams to walk through the evidence for and against innate sex differences, built around his new book. They open by addressing why the topic is so charged: a long history of genuinely sexist "science" (illustrated by an 1800s quote comparing intelligent women to two-headed gorillas) has made people wary that any discussion of evolved differences is a smuggled-in justification for sexism. Stewart-Williams' answer is that modern evolutionary science actually disproved those old ideas, and that a trait having an evolutionary origin carries no moral verdict either way - his four-word philosophy is "let people be themselves."

A large portion of the conversation is methodological: how do you know a sex difference isn't just socialization? Stewart-Williams lays out six converging lines of evidence - early developmental emergence, persistence despite active cultural suppression, stability across a century of data, correlation with prenatal hormone exposure (including the CAH natural experiment), cross-cultural universality, and parallels in other species under similar selection pressures. He's careful to note that none of these alone is conclusive, but that their convergence across a given trait makes a strong case. This sets up one of the episode's most counterintuitive findings, the gender equality paradox: many sex differences get bigger, not smaller, in more gender-equal societies, directly contradicting the prediction of social-role and patriarchy theories.

From there the two dig into the evolutionary mechanics: anisogamy (gamete size defines biological sex), Bateman's principle and Trivers' parental investment theory (the sex that invests more per offspring becomes the choosier, more reproductively-limited sex), and how this produces higher reproductive variance - and therefore more risk-taking, competitiveness, and interest in casual sex - in males across most species. Humans are notable as an outlier among mammals, resembling birds far more than typical mammals in their reliance on pair-bonding and biparental care, which compresses many sex differences relative to, say, gorillas.

The middle stretch is a tour through specific domains: sociosexuality and interest in casual sex (roughly one standard deviation apart on average, verified by Ashley Madison usage data and gay/lesbian "natural experiment" comparisons), visual arousal and porn/romance-novel consumption asymmetries, mate preferences (looks matter more to men long-term but the gap disappears short-term), aggression (scaling from modest for verbal aggression to men committing over 90% of homicides globally, mirroring chimpanzee data), sexual violence, parenting, sexual jealousy (sexual vs. emotional infidelity), personality (neuroticism and agreeableness are the two largest, most consistent Big Five differences), the people-vs-things interest divide (one of the largest and most cross-culturally consistent differences on record, stable for over a century), cognitive ability (no average IQ gap, but higher male variance and some abilities favoring each sex), and physical/mental health (men die younger everywhere, "88% of Darwin Award winners are men but so are 90% of Carnegie Hero Award recipients").

Throughout, Stewart-Williams returns to a repeated caveat: analysis is not justification. Explaining a behavior evolutionarily does not excuse it, any more than a sociocultural explanation would - both types of causal account externalize responsibility equally, so society still has to hold people accountable for behavior regardless of its ultimate cause. The two also spend real time on Williamson's own framing of a "soft bigotry of male expectations," the idea that male behavior is treated as the neutral standard so that women falling short of it (e.g., in STEM) is treated as a problem to fix, while men falling short of female-typical behavior (e.g., in parenting or empathy) rarely gets equivalent attention.

They close by weighing the risks on both sides: exaggerating and moralizing sex differences risks pushing people into traditional roles they don't want, while minimizing or denying them risks real-world harm - under-diagnosing heart attacks in women or autism in girls because symptoms present differently, or assuming occupational gaps are entirely due to discrimination and pursuing interventions that target the wrong cause, generate resentment, and sometimes tip into reverse discrimination. Stewart-Williams' proposed solution throughout is simple: tell everyone the full facts, remove genuine bias and barriers, and then respect the choices people actually make.

Notable Quotes

"Analysis is not justification." - Steve Stewart-Williams

"Let people be themselves." - Steve Stewart-Williams

"It's not because the sex difference is bad. It's because violence is bad." - Steve Stewart-Williams

"Whatever men do, that's the standard of what is right. And if women do it less, that's a bad thing, and we should make them more like males, but not so much the reverse." - Chris Williamson

"If male mortality rates were lowered to female rates... we would save more years of life than if we cured cancer." - Chris Williamson