This op-ed piece in the San Francisco Chronicle is by two academics, Joan Roughgarden, whom people here might know (I reviewed her book Evolution’s Rainbow in the TLS), and Jaimie Veale, who are identified in the piece like this:

Joan Roughgarden is a professor emerita in the Department of Biology at Stanford University and author of “Evolution’s Rainbow: Diversity, Gender and Sexuality in Nature and People.” Jaimie Veale is a senior lecturer at the School of Psychological and Social Sciences at the University of Waikato in New Zealand and a past secretary of the World Professional Association for Transgender Health.

As Wikipedia notes, “In 1998, Roughgarden came out as transgender and changed her name to Joan, making a coming out post on her website on her 52nd birthday.”  Jamie Veale‘s gender is not public, but she (Wikipedia refers to Veale as “her”) is described as researching transgender health and other issues. But their own identification, while perhaps motivating the thesis of this short letter, is hardly relevant to discussing the issues below. Click the headline to get the archived version, as the Chronicle’s own version is paywalled. 

Roughgarden and Veale (henceforth “R & V”), make a number of statements, some of which I agree with and other which I don’t, but overall they made me think about the differences between (biological) sex and gender.

First, they agree that sex is defined by gamete size, something that Roughgarden, to her credit, has always admitted:

Zoologists, botanists, ecologists and evolutionary biologists generally define sex in this way: males make small gametes (sperm), females make large gametes (eggs) and hermaphrodites, such as most plants and many marine animals, make both.

Many animals change sex, such as coral reef fish that switch from making sperm to making eggs, or the reverse, during their lives. In turtles and other reptiles, sex is determined by the temperature at which eggs are incubated. Thus, sex may be determined well after conception according to social and environmental circumstances. And in humans, gamete production does not occur at conception. Various precursor stages appear in the fetus weeks after conception and gamete production awaits puberty.

I would note, though, that sex in humans, which is what everyone’s interested in, is determined at fertilization: whether or not the fetus has the Y-linked SRY gene that is the trigger male development.

But they also claim that every trait other than gamete size is not part of sex but is part of gender:

Beyond gamete size, everything else — including secondary sex characteristics, body size, shape, color, behavior and social roles — is gender.

Gender in nature is also extraordinarily varied and fluid across plants and animals, including humans. Beyond gamete size, no general binary describes how living things look, act or relate to others. Across species, gender difference ranges from penguins with near identical male and female genders to the extreme dimorphism of lions. Human gender diversity is in the middle, showing some gender difference that varies within and across cultures.

Thus whether or not you have a penis or vagina (secondary sex traits) are, assert R&V, part of your gender, even though their presence, and the other traits mentioned, are highly correlated with biological sex.  The idea that physical traits are part of gender seemed wrong to me, but the notion of what “gender” really is has eluded me for a long time.  So I thought about why it seemed wrong to call physical traits parts of gender instead of sex (behavior, as I note below, is a partial exception). This led me to come to my own tentative definition of gender.

The first thing I did, of course, was look up “gender” in the Oxford English Dictionary.  Virtually all the definitions had to do with the classes of nouns in languages in which words have genders, like French and German. But there was one related definition that did correspond pretty closely to what I see as gender: gender identity.

Oxford English Dictionary’s definition of gender identity:

An individual’s personal sense of being or belonging to a particular gender or genders, or of not having a gender.

Gender identity is generally regarded as distinct from biological sex, or sex as registered at birth. In later use it is also often (and for some commentators controversially) distinguished from gender as a socially or culturally constructed state (cf. gender n. 3b), and from its manifestation in gender expression or presentation (see gender expression n., gender presentation n.).

Thinking further, it seemed to me that gender identity (what people mean when they self-describe their gender) is a psychological rather than a physical trait: it is how someone feels vis-à-vis where they lie the spectrum between being masculine or being feminine. Or perhaps they feel they aren’t on that spectrum at all, or are somewhere in the middle.  Now of course in this sense gender can be described as “biological,” but only in the sense that all human thoughts and feelings are biological because they reside in the neurons of the brain.  But otherwise, you can’t tell someone’s gender by their behavior, genitalia, body size, etc.  You can’t tell what self-conception a person has who possesses a penis and a beard, because you can’t see inside their brain.  Most such people, of course, are of male sex and feel themselves to be pretty close to male on the psychological spectrum, but traits besides gametes are not dispositive of gender. You  would find out someone’s gender not by observing them, but by asking them.

Behavior is a slightly different issue since behavior issues from one’s self=conception. So R&V are correct in saying that historically, different genders have characterized many societies—though I’m using gender in my sense and not theirs:

Across cultures and through time, societies included people corresponding to what the West now calls trans or gender diverse. Anthropologists and historians documented these people across North America, South America, Polynesia, India, Southeast Asia, ancient Rome and other parts of the world. Many cultures accepted these individuals as simply part of everyday life, often holding respected social and spiritual roles unique to their cultures.

I agree with them so long as you consider gender a psychological and not physical feature, and one that can but not necessarily is expressed visibly through behavior. In other words, gender is your self-identification in terms of how you fit on the sex spectrum (or off it), and gender roles are how those self-identifies are acted out in society.

But I disagree with the authors on two issues. First, on their claim that gender identity should somehow be recognized by the courts:

In matters of law and policy, “sex” actually refers to elements of gender because the criteria that have historically determined one’s “legal sex” (typically genitals, chromosomes, appearance and/or behavior) are properties of gender and not sex. As such, the courts should recognize that legal sex encompasses gender diversity.

Second, the authors seem to imply that “affirmative care” (they call it “appropriate care”, though I’m not sure that’s what they mean) is mandated for all young people suffering from gender dysphoria:

Today, every major U.S. medical association — the American Medical Association, American Psychiatric Association, American Psychological Association, Endocrine Society, Pediatric Endocrine Society, American Academy of Pediatrics, American Academy of Child and Adolescent Psychiatry — supports transgender people and their right to appropriate care. Still, if the medical establishment can switch from pathologizing trans people to supporting them, it could switch back under political pressure.

If “appropriate care” means “psychological care up to a certain age—the age when those with dysphoria can decide whether to take hormones or undergo surgery”—then I agree. But if they mean “affirmative care,” in which the wishes of children or adolescents are accepted and acted on therapeutically and medically, then I disagree.  I favor “objective” therapy, not “affirmative” therapy, but therapy done with empathy. But I do not sign onto the use puberty blockers or surgery until a patient is of an age of consent (18 or 21; I waver).

I find the article confusing, both in R&V’s definition of gender and in how they want it to be used in law. There are some people—I believe Alex Byrne is among them—who say that gender is really a word that has no real meaning (it’s analogized to a “soul”).  But there is still the phenomenon of people who don’t feel they adhere to concepts of masculine or feminine psychology, and I’m groping to find a definition of “gender” that can describe such people. (Most of them are, of course, biologically male or female using the gametic definition.)

Feel free to give your own take on gender in the comments. But remember, be civil and don’t call anybody names!

In a recent post on his site “Reality’s Last Stand,” Colin Wright reprints an article he published a few months ago in Archives of Sexual behavior, outlining why there are exactly two sexes and dismantling five common arguments that biological sexes actually comprise either more than two types or a spectrum. Click below to read it or find the free pdf here.

If you’re already familiar with the rebuttals of the more-than-two-sexes arguments, you may want to skip this, but it’s a very short piece and worth refreshing yourself. And it was, of course, peer-reviewed.

I’ll give a few excerpts, listing the five arguments supposedly fatal to the only-two-biological sex view and making a few of my own comments. Wright’s excerpts are indented, and my own commentary is flush left.

He begins with what we all know is true: the two-sex definition (really a “recognition of reality”), based on differences in gamete size, was the accepted view until recently, when gender transformation became common, making people want to redefine biology to conform to their own views or identity.

In recent years, however, this previously uncontroversial fact has been challenged in popular discourse (Fuentes, 2023; Kralick, 2018; Viloria & Nieto, 2020) and now increasingly in scholarly scientific publications (Ainsworth, 2015; Fuentes, 2025; McLaughlin et al., 2023; Velocci, 2024), seemingly driven by cultural and political debates surrounding the concept of “gender identity” and transgender rights. Popular outlets now routinely publish articles asserting that there are more than two sexes or that sex is a nonbinary “spectrum” conceived as a continuum or as a multivariate cluster of traits. Scholarly articles have amplified this framing by characterizing the sex binary as overly simplistic, outdated, and even oppressive, urging its replacement with broader and putatively more nuanced models (Ainsworth, 2015).

Here I synthesize evolutionary and developmental evidence to demonstrate that sex is binary (i.e., there are only two sexes) in all anisogamous species and that males and females are defined universally by the type of gamete they have the biological function to produce—not by karyotypes, secondary sexual characteristics, or other correlates.

“Anisogamous” species are those having different sizes of gametes, and comprise all animals and vascular plants. That of course includes humans. And again, although the sperm vs. egg dichotomy is called a “definition” of biological sex, it recally should be called a concept because, like biological species, it simply recognizes an existing dichotomy and does not impose arbitrary human views onto nature.

Wright goes on to define biological sex, which has evolved several times independently. But isn’t it curious that each time it does—no matter what determines sex—there are only two classes that result? That’s an insight that has led to the creation of good theories for why the sexes are always two.  Here’s what Wright sees as the most common attempts to refute the sex dichotomy, and why they fail. Bold headlines are my characterizations.

a.) There are more than two “sexes” in organisms that have gametes of equal size (“isogamous species”), including some fungi and slime molds

. . . . sexes in anisogamous taxa are defined by gametic dimorphism—the production of small gametes (sperm) versus large gametes (ova). Some anisogamous species may also possess mating-type systems layered on top of male and female functions, but isogamous species, by definition, lack sexes.

Claims of hundreds or thousands of sexes thus refer to many mating types in isogamous systems, not to sexes. Where reproduction is anisogamous, the number of sexes remains two—male and female—defined by gamete type (Lehtonen, 2021).

This may seem like a slippery definitional ploy, but in fact biological sexes were recognized as being of only two types in anisogamous species, not isogamous species. Still, if someone insists on saying that there are many sees in isogamous species, I am not going to argue with them too vehemently. The claim of more than two sexes, of course, is invariably used to apply to animals and plants, especially humans. And there we see only two mating types.

b.) If you define sex by chromosome types, there could be more than two sexes. (Note, though, that many species with two sexes do not have them determined by chromosomes: in turtles sex can be determined by temperature, and in some fish by social hierarchies. ) In humans, for example, the typical XX (female) and XY (male karyotype) are supplemented by rare karuyotypes like XO, XYY, XXY, and others.

Colin:

The fundamental flaw is conflating how sex is determined with how it is defined (Capel, 2017; Griffiths, 2021; Hilton & Wright, 2023). In developmental biology, sex determination refers to the mechanisms that trigger and regulate sexual development. These mechanisms vary widely across taxa (Bachtrog, 2014). Examples include chromosomal (e.g., SRY gene on Y chromosome in mammals), temperature-dependent (e.g., higher temperatures produce males in many reptiles), haplodiploidy (e.g., unfertilized haploid eggs yield males in most Hymenoptera insects), or environmental (e.g., chemical cues in Bonellia viridis).

Yet, regardless of the mechanism by which sex is determined, an individual’s sex—male or female—is universally defined by the type of gamete (sperm or ova) their reproductive system has the biological function to produce (Goymann et al., 2023). Sex chromosome aneuploidies therefore represent variations within the two sexes, not additional sexes.

c.) Sex is a spectrum because individuals have a continuity of male and female traits, as exemplified by individuals having DSDs (differences/disorders of sex development.  As I’ve noted before, the frequency of “intersex” individuals, which supposedly cause the spectrum, is quite low: about 1/5600 individuals—close to the probability that if you toss a nickel in the air, it will land on its edge. Yet we don’t see people flipping coins saying, “Call it: heads, tails, or edge.”

. . .The primary evidence invoked to support the spectrum model is the existence of disorders/differences in sex development (DSDs) (Sax, 2002), including forms of genital or gonadal atypicality, often presented visually along a continuum from “typical female” to “typical male.”

However, the existence of such conditions does not undermine the binary nature of sex, because the sex binary does not entail that every individual can be unambiguously categorized as male or female. Rather, the claim is that in anisogamous organisms there are only two gamete types, sperm and ova, and thus only two sexes. Sexual ambiguity is not a third or intermediate sex because developmental variation does not correspond to producing new gamete types.

These next two objections are those I see most often in the literature, and they both have the problem that they don’t set out criteria for defining or recognizing someone as male or female.

d.) In reality, sex is a “polythetic” category, which Colin defines as “one in which members share overlapping characteristics, with no single feature necessary or sufficient for membership. Inclusion is based on “family resemblance”. This is the objection raised, for example, by people like Steve Novella and Agustín Fuentes.

Proponents of a polythetic sex model draw on this idea to portray sex as multivariate (rather than univariate, as in a simple “spectrum”). On this view, “sex” is an aggregate of traits—chromosomes, gonads, gametes, hormones, neuroanatomy, secondary sex characteristics, and other sexually dimorphic traits—and individuals are assigned degrees of maleness or femaleness according to how their overall profile aligns with what is considered male-typical or female-typical (Dreger, 2000; Fausto-Sterling, 2000).

However, male and female are not polythetic categories. They are reproductive classes defined by a single criterion: The type of gamete (sperm or ova) an organism’s reproductive system has the biological function tomproduce. All other traits—karyotype, genital morphology, hormone profiles, neurological and somatic dimorphisms—are typically causes, proxies, or consequences of that functional distinction. Treating those correlates as jointly definitional blurs the determinants and downstream effects of sex with sex itself.

e.) You can be a member of different sexes depending on which trait you’re looking at (chromosomes, genitalia, hormones, and so on).

As articulated by McLaughlin et al. (2023), sex is framed as “a constructed category operating at multiple biological levels,” with four focal levels: genetic, endocrine, morphological, and behavioral. This framing conflates the determinants and correlates of sex with sex itself (Bachtrog, 2014; Capel, 2017). Genes and gene networks initiate and regulate sexual differentiation; hormones mediate downstream development and phenotypic dimorphisms; morphology and many behaviors are influenced by an organism’s sex. Yet none of these traits defines sex. Sex is an organism-level reproductive class anchored to the type of gamete that organism has the biological function to produce. Treating upstream regulators (e.g., SRY activity, hormonal milieu) or downstream outcomes (e.g., dimorphic morphology, behavior) as coequal “levels” of sex is a level-of-analysis error.

And the kickerm which shows the fact that critics really do recognize two sexes (and use them in their own scientific papers!):

Moreover, the multilevel account inherits the same circularity as the polythetic model. Traits are labeled “male-typical” or “female-typical” only because they correlate with organisms already identified as male or female—an identification that, in anisogamous species, is made ultimately by reference to gametes. Once that reference is removed, the typology loses its interpretive footing. As a descriptive framework to integrate genetic, endocrine, and morphological findings in clinical differential diagnosis, the multilevel schema has pragmatic value; as a definition of sex, it is incoherent.

Why is this important biologically? Colin explains:

The scientific value of clear and precise definitions is enormous (Dawkins, 2025). A gamete-based definition prevents error propagation across comparative biology, physiology, ecology, and medicine. It preserves the interpretability of sex-linked phenomena—sexual selection, dimorphism, and life-history trade-offs—and maintains conceptual discipline by keeping determination mechanisms (e.g., SRY pathways, ZW systems, temperature-dependent determination, social cues) in their proper explanatory lane. It also secures cross-taxon coherence: Whether a species is gonochoric or hermaphroditic, and whether determination is chromosomal, environmental, or social, “male” and “female” remain meaningfully comparable because those terms are anchored to reproductive function rather than to a bundle of traits that shift widely from taxa to taxa.

I like to summarize this by saying that the biological sex definition/concept is both universal and explanatory. No other concept of sex, for example, can explain sexual selection and the differences in behavior and phenotype that appear in animals.

It’s important to recognize that the recent reframing of the two sexes as needing revision did not result from any new discoveries about biology. All the things about sex determination and differentiation have been known for a long time. What has changed is not biology but ideology. It is perfectly clear that arguing that there are more than two sexes is derived from the desire to give solace to those who don’t feel or identify as male or female,  But there’s no need to  change your view of nature to bring such solace. As Wright says:

The societal and ethical stakes are also significant. Accurate biology is distinct from questions of dignity, rights, and how we treat one another. Policy disputes should not be adjudicated by redefining—or defining away—the reproductive realities that make sex a useful scientific concept in the first place. When categories are blurred for nonscientific reasons, we invite downstream harms: muddled clinical protocols, compromised epidemiology, eroding and/or conflicting legal protections, and diminished public trust in science.

It is not transphobic to recognize the two sexes that biologists have known for decades, but, unfortunately, we are dealing with ideologues who are largely impervious to both facts and reason, and so the five points above are aimed largely at those who don’t know a lot about the way biologists conceive of sex.