Gender is both strange and hard to think about. Anyone who tells you otherwise probably hasn’t given it much thought. In this essay I will describe two common theories of gender. I will criticize them both. Then I will use the more interesting bits of each to build something that I don’t think I’ve seen before: a theory of gender that is neither essentialist nor constructivist, but something else entirely – a liberal and individualist account of gender.

This is just a sketch. I’ll be vague in places and hand-wavy in others. I’m not trying to rewrite the entire field of gender studies. I’m just trying to think about it clearly, in words, and for as much space as my editors have given me.


One common way to approach the topic of gender is to think of it as an essence: Gender is a special attribute associated with, or inherent to, a thing or a person. This attribute causes the thing or person to be either masculine or feminine in a way that is both real and (purportedly) easy to understand.

Once a person’s gender essence has been apprehended, it is often held that this essence entails ethical obligations: If you are a girl, you should act like one. This means bringing non-essential attributes into conformity: Girl toys should be for girls; boy toys should be for boys. Essence gives commands.

Essence is usually held to correlate to an observable (or an assumed) attribute: maybe to chromosomes, or to genitalia, or to the determination made by a doctor at the child’s birth. Whatever the case may be, the essentialist will insist that this attribute reveals the real gender. This attribute shows the essence, and as we have already observed, essence gives commands.

Like left and right, masculine and feminine are generally held to be opposites; a single thing can’t be both truly masculine and truly feminine. Now, androgyny is rather obviously possible in the real world, but an essentialist will often hold it to be suspect, because only one essence can hold true for any one person. Indeed, androgyny can even raise questions about a someone’s humanity: In pop culture and in science fiction, the androgynous always has a touch of the alien to it, and it is a short step from there to much uglier forms of dehumanization.

Thus it is usually thought best for gender-ambiguous people to become more conforming with their real self. Deviations are likened to deception, and they may be subject to remarkably severe punitive violence. In the language of essentialism, being true to one’s real gender can be a project much like telling the truth itself, albeit with penalties that far exceed those of ordinary lying.

The assertion that a transgender man is still really a woman – still really female deep down, any transformations notwithstanding – is one example of an essentialist claim. A gender essence has been claimed to exist in the person; it is commonly though, in this case, to spring from either chromosomes or from the medical determination made at birth. All the rest follows from this, and undergoing gender transition is then held to be improper.

My next claim may be uncomfortable: The assertion that a transgender man is really a man can also be a type of essentialism.

Consider that to make this claim, one must identify a gender essence. The essence corresponds to an observable attribute, which in this case is the self-image or self-understanding of the individual in question. Yet again, this way of thinking holds that non-essential attributes should be brought into conformity. Essence gives commands, only here the process begins with a different identifying trait and reaches a different gender result.

These two positions – about which I personally have passed no normative judgments so far – are congruent when we look at the narrow question of how they build their moral claims: A part of the whole is taken to be more real, and more true, and more demanding of our moral allegiance, and thus the matter is settled.

Now, if I were forced to choose between these two forms of essentialism, I would not hesitate: I would side, always and strongly, with the trans-affirming variety of essentialism. I do this not because I accept essentialism as a valid moral method, but because among the various essentialist views, I find trans-affirming essentialism to be the most humane and the one that best conduces to dignity for all. I believe that there is something deeply human and deeply wonderful about individuals becoming the authors of their own destiny. If for some people self-authorship includes “I am a woman,” then I say: More power to you. Be your real self, and don’t let anything stand in your way.

In short, if we must be essentialists, then let us err in the direction of too much deference to the individual. It’s the only decent thing to do.


Ultimately, though, I do not believe that essentialism is correct. Constructivism offers one of the more powerful critiques of essentialism.

In the previous section, I made heavy use of the words “real” and “really.” They are especially important words for what comes next, where I will describe how constructivism looks at what’s basically the same territory.

Whenever you see the word “real” in gender talk, you’re almost certainly looking at a piece of essentialism. The reality of one’s real gender is more real than whatever other things might be taking place adjacent to it. It doesn’t matter that some other things are also “real” in the more prosaic sense of the word. The beard, the muscles, the cowboy boots, the denim – that stuff’s just plain old real. The chromosomes are really real.

There is a lot that’s intuitively plausible to the idea of a really-real gender. After all, we appear pretty good at tracking something that looks a lot like it. Indeed, just about all of us can identify gender in nearly everyone and nearly everything else, without even trying. Ask a six-year-old whether dolls are meant for boys or girls, and they’ll immediately tell you: Dolls are girl toys. Should you question them, you’ll be told – They really really are. (I’ve tried this.) Dolls’ feminine nature is presented as an irreducible reality. And people, too, have a real gender. They really really do, as almost every child can tell you.

But we mustn’t be content with explanations that reduce to “it really is.” Such explanations amount to question begging. They reutter an assertion without so much as the pretense an argument in support. This gets us nowhere, most noticeably when two different essentialist accounts disagree with one another.

Rather than asking a gender essentialist, young or old, to re-re-affirm his or her belief that gender is really real, we should ask ourselves instead: What work is this assertion doing?

It appears to me, and I believe to constructivists as well, that the more accurate term here might not be that a given attribute is more real, gender-wise, but rather that it is more controlling: The speaker asserts that the individual’s gender is A, and not B, because the choice between A and B is controlled by observable attribute C, which supplies the only correct answer.

But which attribute controls gender, and how does it do it? How do we recognize a controlling attribute when we see it, and how do we differentiate the controlling attribute from all the other, non-controlling attributes, which may also be gendered male or female, but which do not control? How do we know that the consequences that allegedly flow from a controlling attribute are themselves correctly inferred?

This matters a lot, because whenever a non-controlling attribute changes, essentialist theories will say that the real gender hasn’t changed. Indeed, all of the non-controlling attributes might change, but if – for example – the chromosomes do not, then a chromosome-essentialist will say that the person in question has not really changed gender.

The immutability of chromosomes makes an appealing target for essentialism – in part, I think, because western philosophy since Plato has always held that essences are unchanging, and chromosomes are a thing we don’t know how to change yet. But defending the assertion that chromosomes control gender turns out to be difficult: “Gender just is chromosomes” begs the question yet again. Asserting a claim extra hard, with feeling, and vigor, and italics, is never enough. We want to know why, and knowing why takes more than that.

The constructivist critique of chromosome-essentialism begins by noting that it puts a remarkably heavy metaphysical burden on chromosomes. Particularly when it turns out, as we now know, that chromosomes, genitalia, and gender expression are all pretty variable, and none of them always corresponds to any of the others.

We all know that humans are usually either XX (female) or XY (male), but there are also humans that are XYY, XYYY, XO, XXX, XXY, and XXYY. None of these is strictly speaking “normal,” but usually they are found in individuals who present an unambiguous set of genitalia, and who are easily classified visually as biologically male or female. Their gender expression and identity, meanwhile, can be as mutable as anyone else’s.

It doesn’t end there. Other animals have a variety of sex-determination systems. Meanwhile, and even within the human species, some people have XY chromosomes but physically present as women. This phenomenon is known as androgen insensitivity syndrome; it’s only one part of a larger complex of intersex conditions, in which, even at the biological level, one’s chromosomes don’t actually control much of anything. It has even been credibly suggested that at least five human sexual phenotypes exist, and possibly more.

One may always dismiss these as anomalous cases, but doing so seems to ignore an observed reality in favor of an ideology. It also leaves real intersex individuals in an awkward spot: How are they supposed to act in our gender-suffused world? Essence may give commands, but what can it say to them?

And this is where constructivists put the knife in: Essence isn’t giving these commands at all, and it never has. The commands of gender moralism arise from systems of all-too-human authority. Gender is a sort of nefarious trick that these forms of authority have been playing on us, largely without our noticing it. Gender rules our lives, often in an oppressive way. Gender makes us conform, and it tramples those of us who don’t or can’t. There is clearly some validity to these observations.

Now, constructivist accounts of gender are commonly and grossly misunderstood. Thus it is often heard in conservative circles that the constructivist position amounts to the following: “Gender is arbitrary; nothing about it is real. So do whatever you want. You can gender-identify as a rutabaga if that’s what makes you happy.”

Yet one could hardly find a less accurate characterization of the constructivist view as its proponents usually advance it. If you read Michel Foucault or Judith Butler you will discover that this is almost precisely the opposite of what they believe. Gender for them is a system of knowledge – and of power, which amounts to the same thing – and it is exceptionally difficult to escape. The institutions and practices of gender are not to be trifled with. Gender may be socially constructed, and of course any constructivist is committed to saying so, but this does not mean that gender is arbitrary or freely chosen: Gender is constructed, but so are prisons.

Many constructivists, and Butler in particular, recommend the parodic subversion of gender norms as a way to find breathing room in a stifling system of gender conformity, or at least to make the authorities as uneasy as possible: If you can’t beat ‘em, mock ‘em. Yet they are pessimistic about the prospects for authentic long-term change. In their own way, many constructivists make deeply conservative predictions about the future. Not only is gender a prison today, Butler seems to say, but gender will be a prison forever. It’s for exactly this pessimism – and not for a permissive, anything-goes approach to gender – that Martha Nussbaum famously excoriated her.

It is odd, from this standpoint, to consider how some forms of constructivism can end up looking a lot like essentialism: the essence of your gender lies in outside social forces. Only it’s an essentialism that exists for no good reason, and it hurts us rather than purportedly fulfilling us. I do think Michel Foucault is somewhat less guilty here than most, but he’s still far indeed from the usual conservative misconceptions about gender constructivism. Contrary to its loudest critics, much constructivism professes a tragic lack of options in the face of gender oppression, and this, rather than the louder critiques, is why I do not believe it to be quite correct.


I believe that there is something entirely real about gender, but my idea of it is far removed from both essentialism and constructivism. I find it difficult to describe without resorting to a set of extended analogies.

The first analogy I would draw is to pareidolia, the phenomenon by which humans see faces in settings where no likeness of a face was intended. We see faces in clouds, rocks, and trees, and their effect on us is impossible to deny. We even respond to them in socially appropriate ways: We laugh when the faces laugh, and we recoil when they horrify us. Evolutionary psychologists believe that pareidolia likely exists because identifying and reacting to faces has helped our ancestors (and us) to survive.

In like manner it would appear that within most of us something is continuously looking for sex, and thus for sexually identifying traits. Yet it also seems clear that this search system can light up in some unlikely places. Many languages universally apply two genders even to biologically sexless nouns (“fire” in French is masculine; “justice” is feminine). Traditional Hawai’ian culture reserved certain foods for men and denied them to women, even to the detriment of women’s health. We in the modern West view skirts as inherently feminine, and we may even think that a real man who dons a skirt has wrongly disobeyed the commands of his gender essence, although apart from the social sanctions that this action may bring, it’s anyone’s guess how he has been harmed.

Finding sex in objectively sexless things may be a hyperexpression of an inborn identification system a pareidolia of the sexual part of the mind. The biological function of this sexual identification system is obvious, at least when it identifies the sex of a potential mate in a way that allows an organism to reproduce. As with the pareidolia of faces, it makes good evolutionary sense that a system like this should exist, and also that it should be hypertrophic: It is much better from an evolutionary standpoint to over-identify a face, or a potential sex partner, than to do the opposite.

Humans have a powerful tendency to categorize in this and many other ways, and some degree of categorization seems necessary to rational thought. But we also tend to regard our categories as if they arose from the natural world, and as if we only passively observed them. Gender essentialism arises not because it answers to some external Platonic form that we have apprehended, but because we carry within us a subsystem that instantly and rightly or wrongly classifies people and things by gender. This system tells us both (1) gender is a proxy for sex and (2) it can be found in things that lack any biological sex at all, like dresses or toy guns. Gender appears to be real in these items, and we react to them as if they were gendered, but we do so because the gender classification system is real in us.

Humans also have a long history of turning our evolutionary quirks into realms of meaningful personal expression and meaningful choice. Thus our tendency to pareidolia in the ordinary sense has grown into figurative art: Because we see faces in rocks, we can also put faces into rocks. Our tendency to signal using voices a trait we share with other animals we have turned into language, which we have developed far beyond mere animal signaling. Alone among animals, we have built musical instruments; these stimulate our evolved tendency to read emotions into sounds. If I’m right, cultural expressions of gender also began with our evolved biology, which we have turned into a complex, symbolically laden expressive outlet.

There’s a word for this: Gender is an art.

Like the visual arts, language, or music, we use gender expression to communicate in ways that are deadly serious, or that are playful, or that can track the whole range of human emotions. Like other forms of expression, gender expression is often moralized. It can be downright upsetting. It can be inconvenient, or dangerous, or yes oppressive. Clearly we can and ought to take steps to minimize these harmful tendencies all around, and endeavor to use and respond to gender signaling in a responsible manner. But doing so requires a fuller understanding of what gender is neither an innate natural essence nor an artificial tyranny. When we realize that a lot of what we call gender is probably a form of social signaling that’s built atop an evolved substrate like much of the rest of our culture it stops looking like spooky metaphysics, and it starts looking like something wonderful that we can do with the mere matter of our selves. Just as we bring expressivity and fulfillment to the evolutionary imperatives of sex, eating, and vocal signaling, we do the same with gender.

From this perspective, the essentialist account of gender can be somewhat recast as follows: “Persons with observable attribute C will do best to conform to gender B.” This, however, is an empirical claim, not a metaphysical one, and in part it necessarily references the individual’s own self-reported happiness. Understood as an empirical claim, it appears at least in a significant number of cases to fail. In others it may succeed. But either way an outsider observer cannot serve as the final arbitrator. Only the individual can.

Gender is also a way that we communicate information about ourselves. A desire toward gender communication is likely to some degree inborn, but as with communication in language or the visual arts, gender communication is a stepping stone toward culture. Gender uses a medium our bodies, clothing, language, and the everyday choices we make and an inner classification system that we all have, so as to signal to others what we think and feel about who we are, sometimes in creative ways that begin with a strict gender duality but that need not end with it. And thus humans make meanings out of everything, including clouds and trees and rocks. And skirts and beards and vaginas and penises.

Self-identification is by far the most contentious aspect of gender expression. Just as we do with objects and behaviors, we usually classify ourselves as belonging to one gender or the other, and as participating in various threads of expression within that gender. Our self-classifications typically feel inevitable and natural to us, even if others do not agree. This, though, is no less an expression of gender than any other. And within such expressions there are many subtypes that can be more or less voluntarily chosen.

All this was made clear to me several years ago when I met a non-op transman who described his gender as “Boy Scout.” To which I replied: “Hey! That’s my gender too!” His gender expression had been the product of a conscious transformation; mine had been more or less where I’d ended up, but we immediately became friends. That he was attracted to women, and I was attracted to men, mattered hardly at all, apart from an added subversion: the actual Boy Scouts would have rejected both of us.

Whenever an individual asserts their gender identity, they are inviting you to participate in a creative social process that they experience to be a core part of their being. This is equally true of trans and cisgender identities. In both cases, the assertion of a gender identity is an invitation to participate in the creative, pro-social deployment of a pareidolia that we all experience.

This pro-social use of our evolved psychology may appear to be severely or even fatally constrained by technology. To my mind, though, this is so much the worse for our technology, which ought to be improved. Of course some gender expressions are easy; others are difficult; still others we don’t know how to do at all, at least not yet. I could sit in a more feminine way right now if I so chose. Dressing in a more feminine way would take an afternoon shopping trip. If I wanted breast implants, I’d need a few months and a large amount of money, but I could probably get it done. If I wanted a uterus, or two X chromosomes, forget it at least for now though in fairness many women don’t have those either. In the future, though, we may be able to acquire even these, and perhaps we should try. (Then where would the essentialists be?)

Against my characterization of gender, I can imagine someone objecting as follows: You, and not the constructivists, are the one who has made gender into a matter of whim and unreality. I would answer that my account remains grounded in reality at two points: First, this expansive, creative sense of gender begins with biological sex; without affiliating nonsexual things to biological sex, our tendency to categorize here would lack all meaning. Second, the fact that we all experience gender pareidolia, and that we constantly deploy and contest the meanings that develop from it, gives even the culturally contingent and shifting parts of gender the same ontological status as the meanings of words: They are real because we have made them so.

To say as I do that gender is a social construct is therefore not to say that gender is a matter of arbitrary choice. Nor would I say that gender is an inescapable social prison. Like language, gender is externally constrained as to many of its features, and yet it’s also full of potential for nuanced and authentic self-expression. At its best, gender expression is neither arbitrary nor totalizing, but it occupies if we allow it a happy middle ground.

I believe that a free society would allow individuals a wide permission to make their own choices about these matters, and it would likewise permit others to honor the choices so made, insofar as possible. This permission would take two forms: a set of substantially gender permissive laws and a concomitant ethos that held that wherever possible, gender expression was a matter to be settled by the individual doing the expressing. We are not there yet, but I think we should be.

Situating the origin of gender in human evolutionary psychology but not doing the same for all of its cultural implications also holds out the possibility that a liberated, non-oppressive future need not be genderless. The idea that gender itself must be overthrown in order to overthrow gender oppression arises from an analogy that has often been made between gender theory and Marxist theories of class struggle, in which the proletariat ultimately abolishes class distinctions forever. But if the account of gender explored here is at least roughly accurate, then Marxist class struggle, whether plausible in itself or not, is no longer a satisfying analogy to gender. We may say with somewhat greater confidence that a better world is possible, and that it will still largely be populated by men and women.

The exact methods to achieve this goal cannot be discussed in detail at the moment, and much in this area is probably still unclear to nearly everyone, if we are being honest. But when it’s a question of how one thinks of oneself, and how one expresses oneself, the appropriate analogy is perhaps not to telling a truth that is read in the chromosomes, or read anywhere else, but rather to the censorship, or not, of a deeply held artistic vision. We do not condemn Bach because he failed to tell the truth of the songbirds.

On this view there is something not merely off the mark about warning a trans person that they’re not really of the gender that they assert, or that their attempts to achieve a desired gender expression are unlikely to succeed. No: there is something preposterous about these warnings. They are rather like walking around a museum and warning people that the paintings they stare at are not really people, and that they would therefore do better not to stare. In the realm of gender we are all paintings. Gender is a museum of signifiers, one where we are simultaneously the artists, the canvasses, and also the viewers.

And yet we must admit that this analogy is not entirely correct. Gender inferences also help us to do things like find mates, signal affection, and communicate our personal identity to one another. Flights of fancy notwithstanding, sometimes we just want gender to achieve its evolutionary purpose as well. (Or a near approximation of it: As a cisgender gay man, I don’t depend on gender to find romance … with women.)

It may appear at those times that the art of gender interferes with its more practical uses. Nonconformities of this type are perhaps akin to ambiguities in the use of language, that is, to equivocation or amphiboly. (Note that these too are often a part of a deeply held authorial vision.)  And yet in matters of gender the response seems horrifically out of proportion. I cannot recall the last time that someone was stabbed over an amphiboly, and whereupon thousands of defenders rushed to the stabber’s side say that anyone who uses an ambiguous comma deserves to bleed to death. Without wanting to deny the coordinating purposes of gender expression, I would urge more latitude in how we police the gender system, if we must police it at all. To the considerable degree that gender is an art, gender must also be free.
Featured image is the “face on Mars” formation, from NASA.


Jason Kuznicki

Jason Kuznicki is a research fellow at the Cato Institute, the editor of Cato Unbound, and the author of Technology and the End of Authority.

Leave a Comment