Unfashionable but True: Sex Is Binary
“Fashion” and “Style” typically refer to products and practices that are valued in specific cultures and time periods. In the contemporary U.S., vintage clothing, athleisure, and oversize water bottles are in style, whereas cigarettes, neckties, fur coats, wood paneling, shag carpets, and white wall tires are out.
Ideas also vary in their popularity and prestige across cultures and times. These days, mindfulness, sustainability, plant-based diets, and pet parenting are fashionable, while eugenics, homophobia, hierarchical work environments, and colonialism are not.
Perhaps the most fashionable current idea is that the binary distinction of females and males—and girls and boys, and women and men—is scientifically incorrect and harmful. Instead, leading social scientists, activists, and even professional journals and organizations, have adopted the view that sex should be considered a nonbinary variable, either a continuous spectrum or something with more than two categories.
Yet, the traditional, binary view of sex, despite being unpopular, is basically correct. Crucially, I am confident that holding this view is in no way at odds with being fully respectful to individuals who are transgender or intersex. Here are eight reasons for affirming that biological sex is binary.
Let’s review each of them in more detail.
1. Evolutionary biologists define sex based on gametes, a binary framework. Although scientifically fruitful, this doesn’t work for many individuals.
Reproduction is fundamental to life. Although there are many ways of reproducing, one basic distinction is that of asexual reproduction versus sexual reproduction. In asexual reproduction, sometimes called cloning, an organism reproduces without the participation of another organism. Although asexual reproduction occurs in some vertebrate animals, including reptiles as large as Komodo dragons, it is rare. Instead, most vertebrates, including humans, only reproduce sexually, meaning that two organisms combine their genetic material to produce offspring.
In all vertebrate species, including humans, there are two kinds of sexual reproducers: females and males. Females contribute an egg, which is a gamete (i.e., sex cell) package with a large nutrient bundle and no ability to move. Males contribute sperm, which is a gamete package with no nutrients, but excellent mobility, usually because of a flagellum, a miniature tail. In humans, a single egg is roughly 100,000 times larger than a single sperm.
In all vertebrate species, including humans, there are two kinds of sexual reproducers: females and males.
Most evolutionary biologists, including Richard Dawkins and Jerry Coyne, define sex according to gametes, whether the individual produces eggs (female) or sperm (male).1, 2, 3, 4, 5 This gametes framework undergirds our understanding of reproduction across all sexually reproducing animal species. For instance, in terms of parental responsibility, producing eggs is a high investment strategy whereas producing sperm is a low investment strategy. Building on this initial pattern, in all placental mammals, females, but not males, gestate offspring and provide milk once they are born.
Although the gametes framework has proven highly fruitful for evolutionary biology, it has two key drawbacks when applied to humans. One is that gametes in mammals can’t be observed easily—we don’t lay eggs as do all birds and most reptiles. A second drawback is that, although this framework works well for species or populations, it doesn’t work for many individuals. Postmenopausal women generally don’t produce eggs and neither do women who have had their ovaries removed. Boys don’t produce sperm, and neither do men who have had their testes removed.
2. People in all societies define sex based on reproductive traits, another binary framework. This framework builds on the gametes definition and is far more practical.
Because of the limitations of the gametes framework, I propose the following two “new definitions” of sex based on reproductive traits that are related to gametes.
In humans, a female should be defined as an individual who possesses, or is on a trajectory to possess, or previously possessed, the traits necessary for reproduction as an egg producer; these traits include ovaries, fallopian tubes, uterus, cervix, and vagina. These traits are often referred to as the primary sex characteristics.
In humans, a male should be defined as an individual who possesses, or is on a trajectory to possess, or previously possessed, the traits necessary for reproduction as a sperm producer; these traits include the testes, penis, and scrotum.
I put “new definitions” in quotes to make the point that, although I am arguing we should use this pair of definitions, it is more accurate to say that we should acknowledge our traditional use of them, rather than thinking of them as new. This is because people in all human societies distinguish between males and females, and, historically, they must have done this by observing a newborn’s external genitals. What other option could they have had?
Even in modern industrial societies, where blood and tissue tests are often available, genital inspection is still the main method of sex categorization, although that often now occurs by looking at ultrasound images months prior to birth.6
Of course, outside of the birthing room, most of us rarely examine anyone’s genitals to learn their sex. Instead, we use various traits that typically, but not always, co-occur with the primary reproductive traits. These include secondary sex characteristics, which are physical traits that are not necessary for reproduction but that usually accompany the respective set of reproductive traits, particularly after puberty. In women, these include enlarged breasts and widened hips; in men, they include facial hair and Adam’s apples. There are other relevant traits, including men’s generally lower vocal pitch and their greater height and upper body musculature. Even facial shape is extremely helpful: adults can, with high accuracy, correctly determine another adult’s sex from pictures alone.7, 8
In addition, many components of fashion and style serve, in part, to signal the wearer’s sex. For example, women typically have fuller lips than men, and the current popularity of lip filler procedures can be understood as means of amplifying femininity. Many other products and practices work similarly, including jewelry, piercings, tattoos, hairstyles, makeup, and clothing. Thanks to all this biological and cultural information, most of us interact with many people each day and rarely have doubts about whether a person is female or male.
Even in modern industrial societies, genital inspection is still the main method of sex categorization.
Nonetheless, it bears stressing that the initial assignment of a newborn’s sex, apparently in all societies, is based on external genitalia. In fact, this assignment can be considered part of our universal folk knowledge, “common sense” that sometimes turns out to be correct. Moreover, this universal folk knowledge about sex isn’t limited to knowing that sex is assigned based on external genitalia; it also includes recognizing that there are two, and only two, kinds of human reproducers, that only females bear children, and that females can only reproduce if they have had sexual intercourse with males.
Is this folk knowledge about sex universal, meaning that it has occurred in all human societies that ever existed? Logically, of course, nobody can prove that any belief or behavior is universal; there could always be some exceptional society that has never been studied.9 Nonetheless, we can be highly confident that these patterns are universal or nearly so. The reason is that, for the past few hundred years, cultural anthropologists, ethnographers, missionaries, and explorers have traveled the world documenting exotic beliefs and practices, such as taboos against eating nutritious foods or body modification that involves painful procedures. This attention to describe cultural beliefs extends to sexuality. We know, for example, that in some traditional societies, people believe (incorrectly) that a woman can only get pregnant if she is inseminated repeatedly;10 in other societies, people believe (incorrectly) in partible paternity, meaning that a child can have two biological fathers;11 in others, people believe (incorrectly) that a woman becomes impregnated by a spirit but that a penis must first open the vagina to allow the spirit’s entrance.12 If there were a society where people believed in a third manner or mode of reproduction or were unaware that penile-vaginal intercourse is, with very few exceptions, necessary for conception, we can be confident anthropologists would have studied it and described it by now.
3. Sex is consequential for many reasons, including that it is women who have babies, not men.
In addition to the ones just noted, humans hold other universal beliefs and practices. A fundamental practice in all societies is that there are distinct words for girls and boys, women and men, mother and father, and daughter and son.13 These words exist everywhere because they indicate reproductive roles, and reproduction is always potentially consequential. Even in contemporary countries with sub-replacement level fertility, a substantial portion of women bear children, an event with major consequences for education, work, recreation, family life, and, indeed, survival.
Healthcare is a particularly notable sphere where sex, based on reproductive traits, is consequential. Practitioners aiming to provide the best care frequently must consider if their patient is female or male. Diseases that differ substantially in prevalence, manifestation, or treatment, include Alzheimer’s, COVID-19, depression, diabetes, influenza, pneumonia, and several kinds of cancer.14
Some have asked whether our society would be better off if we simply ignored sex. Although this seems desirable in some situations, it is not practical in many others.15
4. Third genders are nonbinary, but they do not challenge the biological sex binary, reproductive traits framework.
Another frequent question is whether societies with so-called third genders pose a problem for the binary, reproductive traits framework. For example, in Samoa, there is a third gender kind of person called Fa’afafine, which roughly translates to “in the manner of a woman.” Fa’afafine are biological males who often dress in female-typical clothing, adopt feminine names, and do female-typical, people-oriented jobs such as teaching and nursing; they are exclusively attracted to male-typical, masculine men as sexual partners.16
Third gender individuals either do not reproduce or they reproduce in the typical male or female manner.
Fa’afafine are accepted in Samoa as a third kind of person, neither a typical male nor a typical female. However, everyone there recognizes that Fa’afafine do not reproduce in a third way; they rarely or never reproduce, and, if they do, they do so as men. In other societies, there are other kinds of third gender individuals, including biological females who adopt male-typical roles and, in other societies, individuals who do not conform to either a male-typical or female-typical role.17 However, the story is essentially the same everywhere: everyone knows that third gender individuals either do not reproduce or they reproduce in the typical male or female manner.
Many people, even educators and policymakers, express confusion about third gender individuals and claim they challenge the idea that sex is binary. However, this is attributable to many scholars and activists using the terms “sex” and “gender” interchangeably. They will say things such as, “The Fa’afafine embody a third gender and sex.”
Gender has a rich and broad web of meanings, but we don’t need to unpack them all to address the confusion. We only need to remember that what we can call gender roles associated with biological sex—that is, doing male-typical or female-typical things—can be nonbinary. That is, there can be three or more genders or gender roles, and there can be intermediate genders. Further, there are many other traits that are associated with biological sex that are nonbinary; these include skeletal traits, hormones, and personality. None of this, however, contradicts the binary reproductive traits framework discussed above. Again, there are exactly two kinds of sexual reproducers, male and female, and people in all societies—even societies with third genders—recognize this binary distinction, and they recognize it as consequential.
5. Intersex individuals challenge the binary, reproductive traits framework, but they don’t invalidate it because they do not reproduce in a third way.
A more substantial challenge to the binary, reproductive traits framework comes from intersex individuals, who are sometimes described as having DSDs (Disorders or Differences of Sexual Development). These individuals are born with genetic, hormonal, or physical characteristics that are unusual for males or females. For example, a person might have male-typical chromosomes (e.g., XY) yet their phenotype or appearance may be female. A critical point is that, unlike most third gender individuals, intersex individuals typically do not possess a full set of traits necessary for reproduction, and often they are unable to reproduce. Frequently discussed intersex conditions include complete androgen insensitivity syndrome (CAIS), congenital adrenal hyperplasia (CAH), 5-alpha reductase deficiency (5-ARD), and ovotesticular disorder (also called true hermaphroditism). In some communities, some intersex conditions occur fairly commonly (e.g., 5-ARD), and people with this condition may be described as embodying a third gender.18
There is debate about the frequency of intersex conditions with some writers claiming that 1 in 60 live births is intersex and others suggesting the true frequency is roughly 1 in 5,000. The debate largely centers on what counts as a true intersex condition. Using a very broad definition,19, 20 anyone who doesn’t fit their “exacting criteria” for being a typical male or female should be considered intersex, making the prevalence relatively high. In this perspective, a man with a short but functional penis could be called intersex, as could a boy with hypospadias (i.e., their urethra opens on the underside of their penis instead of at the tip) or a woman who bore three children but learned later in life that her androgenic hormones were unusually high.
The binary nature of human reproduction is about as complete a binary distinction as one can find in the natural world.
However, if we reserve the label intersex for individuals who do not possess all traits necessary for successful reproduction, whose chromosomal sex does not match their phenotypic sex (e.g., XY female), or for whom there was genuine uncertainty about their birth sex—all of which might be called classic intersex conditions—the more accurate estimate is roughly 1 in 5,000.21 This estimate is not particularly controversial; it was cited without challenge in 2020 in a progressive editorial in the New England Journal of Medicine advocating for the removal of sex designations on birth certificates.22
Regardless of whether they are fairly common (1 in 60) or truly rare (1 in 5,000), some intersex individuals do not fit neatly into the binary, reproductive traits framework. Nevertheless, they don’t invalidate the framework for two reasons.
First, many definitions of real-world phenomena— including of species, vegetables, or games—become fuzzy or must admit exceptions if scrutinized.23 This is true even of binaries that we usually take for granted:
- Nearly all of the earth’s water can be classified as freshwater or saltwater, yet there are transitional, intermediate areas, which are sometimes called brackish.
- When coins are flipped, they land on one side or the other, heads or tails, however, coins occasionally land on their edge.
- In all societies, people are considered dead or alive, yet, with modern medicine, some individuals may exist in intermediate states, including comas and brain death.
A second, and more crucial, reason that intersex individuals don’t invalidate this framework is that this framework is based on female and male modes of reproduction, and intersex individuals do not reproduce in a third way. All intersex individuals who reproduce, reproduce in either the male or the female manner. To summarize: the binary framework accommodates all modes of sexual reproduction in humans, but it does not (quite) accommodate all humans.
It’s worth stressing that, as Dawkins has eloquently explained,24 the binary nature of human reproduction is about as complete a binary distinction as one can find in the natural world.
6. Hyde and colleagues’ (2019) popular nonbinary definition of sex is indefensible.
What about the nonbinary definitions of sex? Because these have become influential, one might assume that they are better than the binary definitions. I want to note that there is apparently no nonbinary definition of sex that has been recognized as being the best. Nevertheless, a good place to start is with a review article by Hyde and colleagues entitled, “The Future of Sex and Gender in Psychology: Five Challenges to the Gender Binary.”25 This article was published in 2019 in the American Psychologist, a leading journal of the American Psychological Association, the largest psychological society in the world. Hyde is one of the most influential sex and gender scholars, and this article has already been cited more than 1,100 times—a very high number for any academic article, particularly one published so recently. Hyde et al. offer this definition of sex: “The term sex is used here to refer to biological systems involving the X and Y chromosomes, pre- and postnatal sexual differentiation, and hormones that influence sexual differentiation of the external genitals, which, in turn, serve as the basis for sex assignment at birth.”
This is a terrible definition, on several counts. The first issue is that it fails to acknowledge the unifying or organizing feature of the properties included in the definition—that unifying feature being, of course, reproduction. Imagine someone defined a door as: “A human-created object, situated in a house, dwelling, or vehicle, that can be comprised of various materials, that can be decorative, and that can have other objects affixed to it.” These stated properties of doors are all true, yet this is an inadequate definition because it doesn’t state that the chief purpose of a door is to serve as a movable barrier. Similarly, it would be a poor definition of an eye if someone listed many of its components (e.g., cornea, lens, retina) but failed to mention that the eye’s function is to see (that is, to transduce some property of some portion of the electromagnetic spectrum into neural impulses).
A second issue is that Hyde et al. have not explained why a new definition of sex is needed. A basic principle of communication is that if there is an established definition, one should not alter it, or introduce a new one, without justification. Of relevance here is that the gametes definitions of sex developed by biologists is well established—it has been used for more than 100 years and has proven extremely fruitful. (Yes, it’s true that earlier in this article I proposed defining sex based on reproductive traits, however, I explained why this is a sensible extension of the gametes definitions and that it is best viewed as an acknowledgment of our intuitive, universal folk knowledge.)
A third problem is that Hyde et al. have not provided a constructive, practical definition of sex. To make this clear, recall that according to the reproductive traits framework, an individual is female if they possess (or did possess or will possess) the traits necessary for reproduction as an egg producer, and they are male if they possess (or did possess or will possess) the traits necessary for reproduction as a sperm producer, or they are outside the binary if they don’t possess traits for either of these modes of reproduction. There will always be a few edge cases requiring further careful consideration. That said, the vast majority of people can be categorized easily by the reproductive traits framework.
Hyde et al. don’t offer anything similarly useful. They do not state what, according to their definition, the various kinds or categories of sex are—whether it be three, four, five, or more categories. Or, if they consider sex a gradient or spectrum, rather than a categorical variable, they do not specify what concepts or variables that gradient represents. This is worth emphasizing: Based on those very definitions, one cannot categorize a single person as female, male, a third sex, or some intermediate sex. Practically speaking, a 99.9 percent easy classification (i.e., the reproductive traits framework) is preferable to a 0 percent classification!
What issues do Hyde et al. address in their article? One emphasis is defining additional terms, such as “gender,” “transgender,” “cisgender,” “nonbinary,” “agender,” and “genderfluid.” This creates the impression that sex is mixed up with these other terms. The other approach Hyde et al. take—and this comprises most of the article—is to argue that most traits that typically differ between women and men are affected by many factors that can be placed on a gradient or spectrum. These traits include the amounts of various hormones, the size of various brain structures, and the amount or frequency of cognitive abilities, social behaviors, romantic attraction, and feeling like a man or a woman. Although it’s true that these traits are nonbinary, that is not at odds with the reproductive traits framework. The take-home message of that framework is not that all sex-differentiated traits are binary; it’s that the two modes of reproduction—and the reproductive traits that support them—are binary.
So, how does the review by Hyde et al. deal with the fact that reproduction is binary? It doesn’t—it simply ignores the issue. In particular, although it is more than 20 journal pages and over 14,000 words, the words “gamete,” “baby,” “child,” “pregnancy,” “mother,” or “parent,” do not appear, except in parentheses or in the references section. There are a few mentions of “reproductive phases” and “reproductive structures,” but there is no acknowledgment that people bear and raise children, that these events are important, or that reproduction should be considered when defining sex.
Of course, Hyde et al. might argue that ignoring reproduction is a feature, not a bug—indeed, one that justifies their new definition of sex. Perhaps so, but this then raises the question of what trait should serve as the foundation for defining sex. Since they do not provide any answer, I submit the reason is that no nonbinary definition is defensible.
Without a theoretically defensible definition of sex to provide the endpoints of a gradient or spectrum, no individual could be placed on any female-male spectrum for any trait.
Imagine that someone proposed that the amount of testosterone a person is exposed to prior to birth should be the foundation for defining sex, an idea that seems reasonable given that such exposure is known to correlate with variation in several typically sex-differentiated important areas, including play preferences, work preferences, cognitive abilities, and sexual orientation. Although apparently reasonable, one might argue that it is in fact an adult’s current level of circulating testosterone, not their prenatal exposure, that is the more salient measure; another might suggest that prenatal estrogen exposure as the key measure; a fourth might posit that yet another hormone is critical.
The resulting situation is actually more nebulous because other scholars might just as reasonably claim that hormones are no more important than brain shape, skeletal anatomy, personality, or other areas, and each area has many possible measures. There is simply no single area or measure that is more defensible than the others.
In addition, a more fundamental problem is that any scoring system requires a reference point, and scholars who do not accept a binary definition of sex have committed themselves to abandoning the reference point of binary sex. This issue is relevant to every kind of trait—behavioral, anatomical, physiological—that might be considered a candidate for characterizing sex along a spectrum. Without a theoretically defensible definition of sex to provide the endpoints of a gradient or spectrum, no individual could be placed on any female-male spectrum for any trait.
7. No better nonbinary definition has been offered.
If the definitions and ideas proposed by Hyde et al. in 2019 are unworkable, then perhaps there is some other framework that works better? Though I follow the scholarship and discussions in this area closely, I’ve yet to encounter one.
In debates and essays, scholars arguing against the “sex is binary” position, including Alice Dreger26 and Steven Novella,27 invariably echo the points made by Hyde et al., including that some intersex individuals do not fit the binary and that most traits related to sex are not binary. However, these scholars never cite nor develop a viable nonbinary alternative framework, and they fail to acknowledge the fact that all humans who reproduce do so in either the male or female mode, never in a third or intermediate way.
I’ve long taught the course, “Psychology of Sex Differences,” and so am familiar with the leading textbooks in this field, and have participated in a content analysis of the leading textbooks.28 These books also deny that sex is binary yet fail to offer any constructive nonbinary alternative. Furthermore, although they invariably address reproduction and childrearing in later chapters, when they first introduce and define the terms sex and gender in their opening chapters, they do not acknowledge either that reproduction exists or that it should be incorporated into a definition of sex.
8. Nonbinary definitions of sex are popular because they are viewed as being progressive. However, one can hold progressive political views while still retaining the traditional binary view of sex.
If the binary framework of sex based on reproduction is solid and the nonbinary alternatives are so weak, why have the nonbinary alternatives gained so much traction? Why is the nonbinary view presented not only in texts but especially in the media?
The short answer is that scholars, especially evolutionary biologists, haven’t done an adequate job in making the binary basics accessible. While concentrating on the logical and heuristic power of the gametes framework, including its applicability to species where individuals can alter their sexual strategies or reproduce asexually, evolutionary biologists have neglected to provide intuitive and practical definitions for nonspecialists. I hope this article has made the case for how and why a reproductive traits framework—which builds on the gametes framework and makes explicit humans’ universal folk knowledge—remedies this problem.
Those with intersex conditions don’t fully fit the traditional binary pattern, and neither do gay people, or third gender individuals.
A second reason nonbinary definitions of sex have become popular is that they seem to accommodate better the diverse and varied biologies, psychologies, and experiences of individuals who do not conform to the traditional binary pattern, i.e., that most people have male or female reproductive traits, and this difference is accompanied by a corresponding package of female-typical or male-typical nonreproductive traits.29 Among many others, these include secondary sex traits (e.g., enlarged breasts, facial hair), normative social roles and interests (e.g., homemaker, breadwinner), heterosexual orientation, and identifying with one’s biological sex (cisgender). Those with intersex conditions don’t fully fit the traditional binary pattern, and neither do gay people, or third gender individuals. Further, in the U.S. and other Western societies, traditionally there has not been any third gender category or categories, but this is changing as increasing numbers, especially of younger people, identify as transgender.
Although there have always been people who don’t fit the traditional binary pattern, in recent years the visibility of these individuals—and the respect and legal rights afforded to them—has increased tremendously. A major contributor to this change has involved rejecting the presumed practical and moral superiority of the traditional binary pattern. Some scholars call this the cisgender heteronormative pattern. Whatever the nomenclature, the critical point is that more and more individuals have recognized that it is fine and safe to be, for example, a biological woman who is sexually attracted to women, a biological man who identifies as a woman, a biological woman who prefers fixing cars rather than crocheting baby blankets, or an intersex person whose reproductive traits do not allow categorization as being strictly male or female.
For many, myself included, such rejection of the superiority and dominance of the traditional binary pattern represents true ethical and political progress. However, doing this does not require abandoning the scientifically accurate view that sex is binary.30, 31