Is Sex Socially Constructed?
Examining the arguments
“Sex is not gender,” the wildly popular YouTuber and sex educator Laci Green tweeted last year. On one standard way of explaining the difference, sex is biologically given, natural, the raw material on which culture and society can do their work. Gender is the result of that work, the social significance that we invest in sex. Nature gives us our sexed bodies, and “gender is the cultural meanings that the sexed body assumes.”
That last line is from Judith Butler’s Gender Trouble, the book at the pinnacle of the gender studies canon. According to the European Graduate School’s brief biography of Butler, she is “one of the most challenging thinkers of our time.” Since the previous paragraph seems quite sensible and not especially challenging, you might guess that Butler disagrees with it.
Indeed she does. According to Butler, nature gives us something, a kind of clay that can be molded by cultural norms and power structures, but it doesn’t give us sexed bodies any more than it gives us police officers, people who identify as genderqueer, or professors of philosophy. Sex is not a purely biological matter, it is in part a social or cultural one; it is socially (or culturally) “constructed.” As she puts it in Gender Trouble: “perhaps this construct called ‘sex’ is as culturally constructed as gender.” This is a Butlerian “perhaps” — she evidently means to endorse it. Green, for her part, does not — she also tweeted “sex is not a social construct!”
If Butler is right, then radical possibilities open up. If we construct sex, we might be able to demolish it, or construct it differently. The idea that sex is socially constructed did not originate with Butler, but her influence has been enormous, both in academia and popular culture. You can find the idea endorsed in sociology, gender studies, and philosophy — it sometimes even makes an appearance in psychology. (Biology is an exception — the biologists are apparently not up to speed with the latest discoveries.) Whenever sex and gender are the topic online, “sex is a social construction” is likely to be a bone of contention sooner or later.
But what does this slogan mean, precisely? Why does it matter? And is it true?
Let’s take our three questions in order.
What Does It Mean To Say That Sex Is A Social Construction?
Start with sex, which is shorthand for the categories female and male. So the thesis that sex is socially constructed can be better put like this: the categories female and male are socially constructed. Categories are basically the same as the properties (or features, or attributes) of things. My cat Maisie is furry — equivalently, she has the property of being furry. In other words, furriness is one of her features or attributes. Said another way, Maisie belongs to the category furry.
But what does it mean to say that a category is socially constructed? The terminology of “social construction” is often thrown around so carelessly that its intended meaning is impossible to divine, but there is one clear answer that corresponds closely with what Butler has in mind. A socially constructed category is a category that meets this condition: If an object belongs to the category, the object must exist (or have existed) within a society or social organization.
As the philosopher (and my colleague) Sally Haslanger puts it in her book Resisting Reality: Social Construction and Social Critique, in order for an object to belong to a socially constructed category it “must exist within a social matrix.”¹ The relevant kind of “social matrix,” or social organization, needs to be of a human-like level of complexity — ants are social insects, but do not exist within a social matrix in Haslanger’s sense.
That’s a bit abstract; some examples will help. Consider the categories engagement ring, planet, and queen. Engagement ring is a socially constructed category: if something is an engagement ring, then it must exist (or have existed) within a society with the institution of marriage — a band of silver could be found in pre-social nature, but not an engagement ring.
Planet, on the other hand, is not a socially constructed category: the Earth currently hosts many societies, but this is irrelevant to its status as a planet. The planets in our solar system were around for billions of years before life even evolved. In 2006, the International Astronomical Union proposed a definition of “planet” that excluded Pluto. The IAU’s decision to define “planet” in this way was a social matter — the IAU itself is a social organization. But that does not mean that the astronomical category the IAU decided should be labeled by the word “planet” is socially constructed, and it isn’t.
Finally, queen (in the monarchy sense) is a socially constructed category. Elizabeth Windsor, for instance, is only the Queen of the United Kingdom because she is embedded in a society with certain political practices. In a world with no societies, there are no queens.
The terminology of “social construction” is unfortunately very misleading. Engagement rings are literally and straightforwardly constructed: they are manufactured by jewelers. Queens are constructed only in a loose and extended sense: the present British queen wasn’t built by anyone (your parents don’t exactly build you), but Elizabeth only became a queen because certain social groups did certain things, like place a crown on her head in Westminster Abbey. However, the categories themselves — engagement ring and queen — are not “constructed” in any helpful sense. Societies produce engagement rings, and also various phrases to label them (“engagement ring” in English, “anillo de boda” in Spanish), but not the category engagement ring. If there hadn’t been any engagement rings, the category engagement ring would not have disappeared — rather, the category would have had nothing belonging to it. Similarly, the category existed before the ancient Romans started the tradition of giving engagement rings.
More cautiously put: It is highly controversial whether categories are the sorts of things that can be “constructed.” A much better term than “socially constructed category” would simply be “social category,” but let’s keep the misleading terminology for uniformity.
We should guard against another potential confusion. It is crucial to distinguish socially constructed categories from socially significant ones — categories that social groups regard as important in one way or another. But not all socially constructed categories are socially significant, and not all socially significant categories are socially constructed. The category Pepsi drinker is a socially constructed category, but is pretty insignificant: apart from the Pepsi company, no one really cares whether you drink Pepsi, Coke, or something else. On the other hand, despite the fact that the category gold is not socially constructed (gold is just a chemical element), it is highly socially significant: King Midas personifies the human obsession with this metal. It is not in dispute that the categories female and male are socially significant — nothing could be more obvious. What is very much in dispute is whether those categories are socially constructed. And that leads to our second question.
Why Is It Important Whether Sex Is Socially Constructed?
What exactly rides on whether female and male are socially constructed categories? We have already gestured at the answer. At least for many socially constructed categories, if we change society, or the position of people within a society, we can change what belongs to the category. Legalizing same-sex marriage might have increased the number of things belonging to the category engagement ring. We can ensure that there will be no more queens (or kings) by abolishing the monarchy and establishing a republic. If sex is socially constructed, someone might be able to change his sex by changing his social position — a truly spectacular feat of self-determination. When the transgender woman and reality star Jazz Jennings says “I am female,” she could be speaking the plain truth. She is socially female, in almost all the ways that matter, and this might be sufficient — or might be sufficient with some surgery or hormone treatment — for simply being female. More ambitiously, perhaps if we drastically reorganize society we can look forward to an androgynous future, without females and males, or one with a number of liberating new sexes. To the revolutionary minded, social construction presents an opportunity.
Of course, what belongs to categories that are not socially constructed can also change. If you put on a few pounds you are changing your weight: you cease to belong (say) to the category weighs 164 pounds and now belong to the category weighs 166 pounds. Neither of those categories is socially constructed. If sex is not socially constructed, someone might still be able to change his sex solely by medical means. (However, current medical means do not reproduce the processes that allow some animals to change sex.) Humans could perhaps be genetically engineered to be sexless, with reproduction taking place by cloning. But all this is science fictional speculation; at least in the case of female and male, if those categories are not socially constructed then the revolution will have to wait.
Is Sex Socially Constructed?
So, is sex socially constructed? Since a yes answer is hardly obvious, we need an argument. Unfortunately, explicit arguments are hard to find, let alone ones that are carefully laid out and worth taking seriously. Still, we can extract two arguments from the literature. The first can be found in the writings of Butler, and the second is close to the surface in many discussions of so-called intersex conditions (for instance, in Anne Fausto-Sterling’s Sexing the Body). There is a third argument which—unlike the first two—is carefully laid out and needs no extraction; we will get to that later. Let us start with Butler.
The Performative Argument
Butler has an argument for the social construction of sex that turns on a kind of sentence called a “performative.” (You might have heard of Butler’s “performative” theory of gender, which is related; luckily, we do not need to examine it here.) The term “performative” is due to the British philosopher J. L. Austin, who wrote and lectured about the topic starting in the 1940s. What are performatives?
Sometimes we use language simply to report the facts, as when we say “It’s noon,” or “The Earth is a planet.” But sometimes we bring about what we report, just by reporting it, as when we say “I apologize,” or “I name this ship Queen Elizabeth 2,” or “I sentence you to three months in prison.” As if by magic, saying it makes it so: if I say “I apologize” in suitable circumstances then I will have made it the case that I apologize. The qualification about “suitable circumstances” is important. If I say “I apologize” as an actor on the stage, then I am only pretending to apologize, not really apologizing. The need for the qualification is even clearer with the second and third examples. Saying “I name this ship Queen Elizabeth 2” will only succeed in naming a ship if the speaker has some social position that confers authority in ship-naming. (The QE2 was christened by Elizabeth herself.) If I spot the USS Constitution in Boston Harbor and exclaim “I name this ship Queen Elizabeth 2!” I will not have renamed Old Ironsides. Similarly, without any legal authority, saying “I sentence you to three months” will not succeed in sentencing anyone.
Sentences like “I apologize” — which can be used, given the appropriate social setting, to bring about what they report — are called explicit performatives. Note that there are other ways to apologize than to utter an explicit performative. “I’m sorry” will usually do just as well, but is not itself an explicit performative. Saying “I’m sorry” does not report the speaker as apologizing, and does not bring about what it does report, namely a feeling of regret; it is, rather, an implicit performative. Similarly, one can name a ship or sentence someone without uttering an explicit performative. In the appropriate social setting one can name a ship simply by keeping quiet and cracking champagne on the bow. And a judge can sentence someone to three months by saying “You will go to prison for three months.”
What could this have to do with whether sex is socially constructed? One more preliminary observation and all will be revealed. By uttering a performative, the speaker can make it the case that she apologizes, or that a ship is named, or that a defendant is sentenced to three months. Put in the jargon of categories, the speaker can make it the case that a thing belongs to a category — that the speaker belongs to the category apologizer; that this ship belongs to the category named “Queen Elizabeth 2”; that the defendant belongs to the category sentenced to three months. Now notice that all three categories are socially constructed categories: without some kind of society, there would be no apologizers, no things with names in a public language like English, and no one would be sentenced to jail terms. Consideration of these and other examples should make the following principle plausible: if performatives can be used to bring it about that a thing belongs to a category, that category is socially constructed.
In Bodies That Matter, Butler writes:
Consider the medical interpellation which (the recent emergence of the sonogram notwithstanding) shifts an infant from an ‘it’ to a ‘she’ or a ‘he’.
Never mind what “interpellation” means — the example is simply one in which a medical authority examines a fetus or a neonate and says “It’s a girl,” or “It’s a boy.” Butler claims that those sentences are “initiatory performatives.” As Sara Salih puts the idea in her useful book on Butler,
When the doctor or nurse declares “It’s a girl/boy!,” they are not simply reporting on what they see…they are actually assigning a sex…to a body that can have no existence outside discourse. In other words, the statement “It’s a girl/boy!” is performative.
Using the principle mentioned at the end of the second paragraph above, we can now set out the Performative Argument.
- Premise 1: “It’s a girl” as uttered by an appropriate medical authority is a performative; specifically, such performatives can be used to bring it about that the baby is a girl, and so belongs to the category female. Similarly with “It’s a boy.”
- Premise 2: If performatives can be used to bring it about that a baby belongs to the categories female or male, the categories female and male are socially constructed.
- Conclusion: The categories female and male are socially constructed.
Since the conclusion follows logically from the premises, the only way of resisting the argument is to reject one of the premises. And premise 1 seems very easy to reject: it is not plausible at all that “It’s a girl” is a performative. Explicit performatives contain an action-verb like “apologize,” “thank” (“I thank you”), or “bet” (“I bet $100”), and “It’s a girl” does not. That sentence is therefore not an explicit performative. And if it is an implicit one, then we should be able to make what is implicit, explicit. Just as we can roughly paraphrase the implicit performative “I’m sorry” by the explicit performative “I apologize,” we should be able to find an explicit performative that corresponds to “It’s a girl.” Explicit performatives are typically in the first-person present indicative: “I apologize,” “I thank you,” and so on; their performative character can be further highlighted with the adverb “hereby” — “Hereby, I apologize.” So is the doctor or other medical authority in effect saying “Hereby, I make the baby a girl”?
That doesn’t sound right! If it were right, then provided the doctor had the appropriate authority and sincerely proclaimed “It’s a girl,” then the baby would be a girl, with no possibility of error. Similarly, if I say “I apologize” in suitable circumstances, then I really did apologize, even if I later realize that I had nothing to apologize for. I will say “I shouldn’t have apologized,” not “I didn’t apologize.” But in the case of sexing a baby there is a possibility of error in “suitable circumstances.” If the authoritative doctor later realized that she had (for instance) misinterpreted the ultrasound image, her previous announcement that the baby was a girl was simply false. She won’t say “The baby shouldn’t have been a girl,” but rather “The baby wasn’t a girl.”
The Performative Argument is a complete failure, then. Charity might lead us to suspect this was not quite what Butler intended. Rather than enter the labyrinth of Butler interpretation, let us try something different.
The Assignment Argument
Cloacal exstrophy is a particularly serious (and, fortunately, exceptionally rare) birth defect which affects lower abdominal development, with part of the large intestine appearing outside the body; in infants with male sex chromosomes (XY), the penis is either absent or malformed. (It is not an intersex condition, but bringing those in needlessly complicates matters.) In these cases a practical decision needs to be taken about the assignment of sex. Given the available surgical techniques, and previous evidence about patient outcomes, should we raise the infant as a girl, a boy, or — an option never considered — neither? The focus is not on whether the infant is really a female or a male, but on which option will lead to the best quality of life. And that means that social considerations are important to the assignment.
This is a dramatic example, but actually the point generalizes. With very few exceptions, we assign babies with penises to the category male and raise them as boys; similarly, we assign babies with vaginas to the category female and raise them as girls. But that practice is not forced on us by the biology. There usually seems little harm in doing it this way, but if a baby with a penis would be utterly miserable being raised as a boy and would flourish personally and socially being raised as a girl, why not go for it? Social considerations, in other words, are always lurking in the background of sex assignment, even in perfectly normal cases. If someone says (in the case of a seemingly unexceptional birth of a child with a penis) “But the child will be much better off assigned female,” the right response is “Why do you think that?” rather than “Who cares?”
When Elizabeth became a queen, that had something to do with biology, in particular her family structure. But, just like sex assignment, the austere biological facts about Elizabeth’s parentage and siblings did not dictate that she would be a member of the category queen — social considerations also mattered. In general, it seems plausible that if belonging to a category depends on social considerations, the category is socially constructed. Putting all this together gives us the Assignment Argument:
- Premise 1: A baby’s assignment to the categories female or male at birth depends on social considerations — biology does not dictate how the baby should be assigned.
- Premise 2: If a baby’s belonging to the categories female or male depends on social considerations, the categories female and male are socially constructed.
- Conclusion: The categories female and male are socially constructed.
The Performative Argument had a false premise. The Assignment Argument has a different defect: the premises do not logically imply the conclusion. The first premise is about being assigned to a category; the second premise is about belonging to a category. If we tried to repair the argument by replacing “assignment” in one premise by “belonging” (or vice versa), the repaired argument would logically imply the conclusion. But then it would have a false premise, or at least one without any support. (If we replaced “assignment” by “belonging” the first premise, the claim would be false or unsupported; vice versa, and the culprit would be the second premise.) Assigning and belonging — although connected — are quite different.
To assign an item to a category (in the sense of premise 1) is to perform a public classificatory act, to publicly label or mark the item as belonging to a category. Assignment can be used to get people to treat the item (in certain respects) as if it belonged to the category. Assigning an infant to the category female at birth almost invariably results in the child being socially treated as if it is female — the child is given a girl’s name, perhaps, or dressed in girl’s clothes, or is expected to have a preference for dolls over toy cars. In the vast majority of such cases, the child also belongs to the category female, but assigning and belonging can come apart.
This point should be perfectly familiar. Consider the category fool. It often isn’t nice to label fools as such, and to treat them publicly as if they are fools. Often it’s best to suffer them gladly. So someone can belong to the category fool even if he is not assigned to that category. Conversely, you might insult your Machiavellian enemy by calling her “a fool,” and treating her with the contempt befitting fools; we can imagine that this catches on, and everyone else gets in on the act, to your enemy’s chagrin. Your enemy is no fool, which is precisely why your insults are so effective. She is assigned to the category fool, but does not belong to it.
The moral is that sometimes someone can be assigned to a category even if the person does not belong to the category. No reason has been given for thinking that this moral fails to apply in the case of sex assignment at birth. Assigning depends on social considerations, but this is quite compatible with belonging not depending on social considerations at all.
The two arguments examined so far have been duds. Is the third one any better?
The Explanatory Argument
The philosopher Ásta’s recent book Categories We Live By contains a clearly stated argument for the social construction of sex. (She also takes the time to explain what “social construction” means; her explanation is close enough to the one given earlier.) Ásta concedes that sex appears to be “biologically given,” but she argues that it is not. “Being of a certain sex,” she thinks, is to have a “conferred legal status,” a property that someone has only because she or he is embedded in a society with certain legal institutions.
Why does Ásta think that sex is socially constructed? This is her key reason:
if a property [i.e. a category] chiefly figures in explanations of social facts, and not natural facts, then that suggests that the property is a social property [i.e. is socially constructed].
There is something to this. Go back to our earlier examples of socially constructed categories: engagement ring and queen. One can explain various social facts by citing facts about engagement rings: for instance, Romeo had to borrow money because engagement rings are expensive. Similarly, one can explain various social facts by citing facts about queens: there is a parade occurring in London because it’s the Queen’s official birthday. On the other hand, facts about engagement rings or queens are not very well suited for explaining “natural” facts — facts in the non-social world, for example, about the weather or the tides. That is not to say that they can never explain non-social or natural facts. Why does this silver band fit perfectly on this human’s fourth finger? (A natural fact.) “Because she just got engaged, and it’s her engagement ring” might give a reasonably adequate explanation. The point is just that a socially constructed category does its main explanatory work in the social world, not outside it.
Ásta thinks that the categories female and male fit this pattern too. All sorts of social facts are explained (at least in part) by citing facts about these categories. There are doctors specializing in obstetrics and gynecology because there are many human females; the app Grindr has been successful because many human males are same-sex attracted; many children in Afghanistan do not get an education because they are females, and so on. It is not immediately obvious that all sorts of non-social or natural facts are explained by citing facts about females and males. (What about the natural fact that many humans can bear children? Ásta argues that this is better explained by specific physiological facts, rather than by the fact that many humans are female.) Putting the points of this and the previous paragraph together gives us the Explanatory Argument.
- Premise 1: The categories female and male chiefly figure in explanations of social facts, and not natural facts.
- Premise 2: If the categories female and male chiefly figure in explanations of social facts, and not natural facts, then the categories female and male are socially constructed.
- Conclusion: The categories female and male are socially constructed.
Like the Performative Argument, the conclusion logically follows from the premises. The problem is that, on closer examination, there is no reason to believe the second premise. (The first premise is also debatable, but the problem with the second is easier to explain.²) We noted earlier that the category gold is socially significant, but not socially constructed. Now consider another category that is like gold in these two respects: the category red diamond. Red diamonds are very rare, and the most expensive kind of diamonds. The category chiefly figures in explanations of social facts: the crowds flocked to the museum because a red diamond was on display; the collector paid a high price because the item was a red diamond, and so on. The category figures in few, if any, explanations of natural facts: red diamonds can scratch anything because they are diamonds, not because they are red diamonds. And yet the category red diamond, although a socially significant one, is not socially constructed. So even if female and male do their main explanatory work within the social world, this does not indicate that those categories are socially constructed.
None of these three arguments come close to establishing that sex is socially constructed. In fact, no argument could establish that, because sex is not socially constructed. In the contest between Judith Butler and Laci Green, we can declare the vivacious YouTuber the winner.
Recall my cat Maisie. The ordinary English word “female” has no special sense in which it applies only to humans: Maisie is female, just like Elizabeth II. Females are to be found across the animal (and vegetable) kingdoms: Elizabeth, Maisie, and numerous fish, beetles, and earthworms all belong to the category female. (Every earthworm enjoys the privilege of belonging to both categories, female and male.)
Now return to the definition of “social construction.” If a category is socially constructed, then in order for an object to belong to the category, the object must exist (or have existed) within a society or social organization. Clearly many animals have belonged to the category female (or male) without existing within a society of any kind. Indeed, there would have been females and males even if life on Earth had been destroyed by an asteroid half a billion years ago and humans had never evolved. Female and male are therefore not socially constructed categories; that is, sex is not socially constructed.
It might be replied that the slogan “Sex is socially constructed” was supposed to be restricted to humans. Human sex is socially constructed, not feline sex or beetle sex. But this reply is confused. The categories of “human sex” are presumably human female and human male. To belong to the category human female is simply to have the property being human and the property being female; in other words, it is to belong both to the category human and the category female. But the category human is no more socially constructed than the category female — and if it were, then the thesis should have been that the category human is socially constructed, not that female and male are socially constructed. Admittedly, human children and adults almost invariably exist within social groups, but the occasional human who has always been completely isolated from any society is doubtless part of our history too. (Even if not, there could easily have been such an isolated human.) By the same token, there have surely been socially isolated human females and human males. Perhaps an early hunter-gatherer couple left their group and fended for themselves; she became pregnant, he was eaten by a saber-toothed cat, and she died shortly after giving birth. Their short-lived but developmentally typical child did not exist within any society, but was a human being, and was either female or male.
Why are these elementary points almost always overlooked? If the gender theorist’s focus is on the human condition, and especially how existing social arrangements can be improved, the rest of the natural world (and long-dead hunter-gatherers) can seem irrelevant. Seeking (in Butler’s words) “to undermine any and all efforts to wield a discourse of truth to delegitimate minority gendered and sexual practices,” it is easy to ignore my cat Maisie, not to mention beetles.
Those who press the claim that sex is socially constructed are likely to sympathize with Marx’s adage that “Philosophers have hitherto only interpreted the world in various ways; the point is to change it.” That may be, but if you want to change the world you need to understand it first.
Notes:
¹For a very similar account, see “constitutive construction” in section 1.3 of Ron Mallon’s Stanford Encyclopedia of Philosophy entry, “Naturalistic Approaches to Social Construction.” That entry also explains some of the many bewildering ways in which “social construction” is used.
²Biologists are frequently concerned to explain facts about females and males, but the first premise is about what facts about sex explain, not what explains facts about sex. Here is one observation relevant to the assessment of premise 1. Once it is understood what the categories female and male amount to (roughly, large gamete producers and small gamete producers), they figure in an explanation of why males across the animal kingdom are usually more sexually competitive. See J. Lehtonen, et al., “Why Anisogamy Drives Ancestral Sex Roles.”
