What is gender for? Brian D. Earp Yale University & University of Oxford This is the author’s copy of a published paper. It may be cited as follows: Earp, B. D. (2020). What is gender for? The Philosopher, 108(2), 94-99. Abstract There is something of a war going on between, broadly, two groups of people who have very different views about gender. These groups call each other by various names, many of them quite loaded, but let’s just call them Group A and Group B. Group A, on the whole, has a view of gender centred on the connection between physical traits and social positioning. Group B, by contrast, leans more toward a view focused on mental traits and self-understandings. Maybe surprisingly, it turns out that people from these two groups tend to hold very similar beliefs about many, if not most other topics within feminist philosophy and politics: things like abortion rights, say, or the need to address sexual harassment in the workplace. What could be going on? There is a tradition within Western feminism that distinguishes sex and gender. Sex, in this tradition, is a biological classification of our species into males and females, where males produce sperm and females produce eggs (assuming typical development). But, it is usually acknowledged, there is more to sex than just this distinction. Due to gene-level processes involved in sexual differentiation, the sperm- and egg-producing members of the species each have a whole cluster of traits and processes related to reproduction that tend to “hang together,” both statistically and anatomically. So, for example, XY chromosomes tend to go with sperm production, development of a penis, vas deferens, testicles, scrotum, and certain patterns of hormonal activity, fat distribution, and hair, chest, and muscle growth during puberty. And XX chromosomes tend to go with egg production, development of a vulva, vagina, fallopian tubes, uterus, and certain other patterns of hormonal activity – et cetera – during puberty. (I’m not saying anything you didn’t learn in high school.) Now, these clusters of traits and processes are not perfectly dichotomous, as is becoming more widely known. Some of them, like fat and muscle distribution, are highly overlapping between conventional sex categories; and there are all kinds of variations in sex development that lead to a diverse range of outcomes. These include so-called intersex traits in some people, like having a clitoropenis, or various combinations of traits that are characteristic of both sex categories, like having XY chromosomes (characteristically a “male” trait) and a vulva (characteristically a “female” trait), due to a difference in how certain hormones are processed in the body. So sex is not a simple binary. It’s a multi-dimensional constellation of traits that fall into two main groupings, albeit with a good deal of overlap and variation. What about gender, then? In the tradition I am referring to, gender has mostly been understood in terms of a system of social structures, scripts, norms, and expectations that usually correspond to sex, with boy/man as the conventional gender “role” for males, and girl/woman as the conventional gender “role” for females. But there is complexity here, too. Some cultures, for instance, have developed more than two gender roles; and the specific norms that actually govern such roles can be strikingly different across societies. Nevertheless, to oversimplify, what it means to “be” a woman or man, on this view, is to occupy a certain position in social space on the basis of one’s real or perceived reproductive features (that is, the position that has been culturally designated for people who meet the relevant bodily criteria). More recently, within some strands of Western feminism, a view about gender has emerged that shifts the focus from reproductive features and culturally linked social roles to aspects of individual psychology, subjective experience, and self-knowledge. On this view, again to oversimplify, to “be” a woman or man – or an agender or non-binary person, among other potential gender identities – is primarily to experience or understand oneself as such, at least 2 partially independently of one’s bodily attributes or gendered social positioning. So, for example, a person on this view could have all of the reproductive features that are characteristic of the sperm-producing members our species, but nevertheless experience oneself as (or alternatively, know oneself to be), say, a woman. One of way of explaining this example is to say that the social scripts, norms, and so on, that have been culturally designated for females in this person’s society may intuitively feel much more personally relevant, applicable, natural, or fitting than the ones that have been culturally designated for males, even though their body fits the “male” criteria. For someone who is a man on this view, the inverse may be true; and for an agender or non-binary person, there may be no sufficiently relevant social scripts or norms (see the work of philosopher Katharine Jenkins for more on this approach). Now, there are various accounts of how these felt discrepancies between, roughly speaking, one’s “sexed” body and one’s “gendered” mind can come about for a given person. In some cases, for example, a person may experience a strong dislike toward—or visceral dissociation from—their sex-specific body parts. In such a case, the neural representation of those body parts may be coded as, in effect, “not me” (that is, as alien to the brain’s working model of the self). But the specific pathways involved, which may be different for different people, won’t be our focus here. So what’s the big deal? As you may have noticed, there is something of a war going on between, broadly, two groups of people who have very different views about gender. These groups call each other by various names, many of them quite loaded, but let’s just call them Group A and Group B. Group A, on the whole, has a view of gender more like the first one I described: the view centred on the connection between physical traits and social positioning. Group B, by contrast, leans more toward the second view, focusing on mental traits and selfunderstandings. Maybe surprisingly, it turns out that people from these two groups tend to hold very similar beliefs about many, if not most other topics within feminist philosophy and politics: things like abortion rights, say, or the need to address sexual harassment in the workplace. What could be going on? One thing that, on the surface, appears to be going on is a deep-seated disagreement between these groups about ontology – about what there fundamentally is in the world, what exists. Or more accurately, a disagreement about the nature of what exists, in the realm of sex and gender. 3 The apparent disagreement is this: Group A claims that what it is to be a woman or man is to have (or at least, widely be presumed to have) female or male reproductive features; to have certain gender roles or expectations thrust upon one on that basis; and to be socially punished in characteristic ways for deviating too far from the script. A further claim made by Group A is that within patriarchal societies in particular, those who are presumed to have female reproductive features are – by definition – systematically disadvantaged or disempowered along certain dimensions, while those who are presumed to have male reproductive features are systematically advantaged along certain dimensions. Group B, by contrast, claims that what it is to be a woman or man is to experience (or know) oneself as a woman or man, typically based on a strong, inner conviction about the genderrelated aspects of one’s self-concept or biographical identity. And this is so irrespective of the reproductive features one may have, the gender roles or expectations that are thrust upon one, or more generally how one is treated by society. That being said, it is often emphasized by Group B that those whose gendered self-conception does not “match” their sexed body in the culturally prescribed way are generally treated very poorly in patriarchal societies, if not by definition then as a matter of practice. Accordingly, such people may face systematic disadvantages in these societies that are at least as harmful as the ones that are characteristically faced by people who have – or are presumed to have – female reproductive features. Two groups, then, with apparently contrary answers to the question, “What is it to be a woman or man?” Now, sometimes, when two groups of people appear to disagree about the very nature of what there is in the world, there is more to the story. Often, there is something beneath or behind their disagreement that reflects a different set of values or priorities. And instead of hashing out the value-claims directly (by arguing about what is morally or politically more important, for example), the dispute gets couched in the language of ontology (claims about what exists). So, one possibility is that Group A, for example, thinks that dismantling patriarchal social systems – specifically by identifying, helping, and making the world more just for people who are systematically disadvantaged by virtue of having (or being perceived to have) female reproductive features – should be a top priority for feminist politics. For purposes of 4 determining who counts, or who should count, as a member of the category “woman,” then, Group A might put relatively more weight on those dimensions of the concept that correspond to female-coded biological traits and associated social positioning. But this valueladen weighting is not usually made explicit. Instead, the values get subtly baked into claims about what a woman is, as though this were a simple matter of describing reality. Meanwhile, Group B might think that making the world safer for people who are marginalized, abused, and otherwise treated horribly for experiencing themselves as (or on this view, being) a gender other than the one that culturally corresponds to their sexual anatomy should be a top priority for feminist politics (not that the two priorities are necessarily incompatible). A further concern of Group B is that having to constantly fight for recognition as a member of the gender category through which one is best able to make sense of one’s own embeddedness in the world is demoralizing. For purposes of determining who counts, or should count, then, as a member of the category “woman,” Group B might put relatively more weight on those dimensions that correspond to certain psychological properties and gendered self-understandings. But here, too, the apparent moral disagreement about what properties should be conceptually prioritized for certain purposes is occluded by language that purports to be about what a woman fundamentally is. In cases like this – where there is an apparent disagreement about social ontology – you often find that, when you dig a little deeper, there is an even more basic set of facts about which the two sides do not disagree. And you find that it is really the mapping between these facts and some higher-level phenomenon (here, gender categorization) that is at the root of the explicit conflict. If so, it can sometimes be helpful, for the sake of a more productive discourse, to temporarily shift the conversation down to that “lower” level of agreed-upon facts, as a way of finding common ground. What sort of facts do I mean? I mean genetic facts like whether a person has XX or XY chromosomes (or perhaps XO chromosomes in the case of Turner’s syndrome). It includes anatomical facts like whether a person has a uterus, vagina, testicles, clitoris, or penis (or perhaps clitoropenis of a certain length). Functional facts like whether a person can lactate or grow a beard. Psychological facts, like whether a person feels strongly drawn toward – or understands their lived experience most coherently in terms of – a particular gendered social script or role. Social facts, like whether someone with a given cluster of physical or behavioural attributes is likely to be raised in certain ways, treated in certain ways, expected 5 to dress certain ways, and so forth. There is typically little disagreement about these kinds of facts. So, for now, let’s shift the conversation down to this level. I propose that, when someone seeks to know what gender someone is, there is almost always a more specific, metaphysically less controversial, thing (or group of things) they are trying to find out about that person – a thing for which the sought-after gender categorization, such as “man” or “woman,” is merely a proxy. In other words, categorizing people by gender is, in large part, a shorthand tool or heuristic for flagging properties of interest in the world (including, but not limited to, certain statistical regularities related to sexual reproduction), boiling them down, and communicating about them with others. It is, then, a useful and efficient, but also imprecise and imperfect way of dividing up the complex, multi-dimensional, biopsychosocial space in which we all sit, to make navigating this space a bit more manageable. To paraphrase a line from essayist Scott Alexander, “the categories were made for us, not us for the categories.” In other words, what if there is no unified, ontologically basic fact about what gender someone “is”? What if, instead, at the most foundational level, there are only particular facts about the properties a person (or group of people) has, so that the real disagreement is about which of those facts or properties should get more weight when picking out category members? If that is what is going on, the answer to the “should” question will depend on our values, and on the purpose of drawing a line around certain properties, rather than others, in the first place. Consider a (relatively) de-politicized example to illustrate. Why might we want to draw a line around some facts or properties of an organism rather than others? One reason we might want to do this is because it helps us predict or explain something. So, the purpose in drawing a line here would be to help us achieve this aim, and the properties of interest would depend on whatever it is we were trying to predict or explain. Suppose you are running an experiment to figure out the differential behavioural effects of some drug on male versus female rats. If you control for social context and other external factors, what specific properties of the rats are going to be relevant to this experiment? Presumably, it isn’t the abstract “maleness” or “femaleness” of the rats with which the drug itself will be interacting. Rather, the drug will 6 interact with—among other things—certain hormones, such as testosterone or oestrogen, that tend to work differently or to occur in different concentrations in rats that have, say, penises versus those that have vulvas. In such a case, depending on your hypothesis, the most explanatorily robust line to draw might not be between the visibly penis-having “male” rats and the visibly vulva-having “female” rats, but rather between the rats that fall on the higher end of some distribution of hormone levels, say, and those that fall on the lower end of that distribution (among numerous other potentially relevant factors). Assuming that this is the most promising model, the best thing to do might be to actually measure the relevant hormone levels directly, and divide the rats accordingly. And after having made such a division, you might find that some rats with penises are in the so-called “female” arm of the experiment and some rats with vulvas are in the so-called “male” arm of the experiment, but this would be entirely appropriate given the aims of the experiment. Similarly, we humans may have different aims in dividing ourselves along the lines of sex or gender in different contexts. Which underlying features or properties are most appropriate, useful, accurate, or predictive – or perhaps fair, just, respectful, or ethical – for grounding such divisions will depend, at least in part, on the purpose of the division. (It will also depend on how, exactly, the underlying properties are causally related to each other, and on our ability to accurately measure these relations.) And of course, people may disagree about whether a given purpose is itself justified, whether on scientific, pragmatic, moral, or political grounds. For example, why do we have different bathrooms for “women” and “men”? What is the purpose of that division? The historical purpose is complicated, but it seems to have had something to do with the industrial-era emergence of women from the domestic sphere (where they could attend to their business in private) into the public sphere of factory work (where people of all sexes now needed accommodation). Why not one big open restroom for everyone, then? The answer is debated among historians, but one theory is that a kind of benevolent sexism kicked in, where allegedly “weak, vulnerable” women, no longer confinable to the home, needed to have a separate, homelike space even in public so that their feminine “virtues” could be preserved and protected. 7 Whether such a purpose for segregating bathrooms is defensible today is an open question. Some argue that women do need protection in public spaces, if not of their virtue, then of their privacy and safety when it comes to men. Others propose that all bathrooms should be single occupant where architecturally and economically feasible. But let us suppose, purely for the sake of argument, that there is indeed some legitimate reason to have different bathrooms for “women” and “men” in contemporary public spaces. Whatever that reason is, it is unlikely that a person’s chromosomes, for example, will be the most relevant feature for determining who counts as a woman or man in this context, that is, for the purpose of using a public restroom. Rather, something like their gendered appearance, or whether they need to sit down to urinate, or some other feature is likely to be more relevant. Accordingly, which bathroom a person uses, or is entitled to use, should turn on whatever those relevant features are, not their “maleness” or “femaleness” – whatever that might mean – much less their “being” a man or woman. Another example: Why are there different leagues or associations for “women” and “men” who want to play basketball at a professional level? Presumably, it is not the ontological status of a given group of people as being “really” women or men that grounds such a division. Rather, it is likely to be more concrete, descriptive facts that have to do with things like average differences in height, jumping ability, and so forth, that strongly correlate with the two main clusters of reproduction-related features I described at the beginning. Basically, if there were only one professional league – reserved for the members of our species that could score the most points under certain conditions – there would likely be a preponderance of people with male-typical physical features, and not very many sex-matched role models for aspiring basketball players with female-typical features. Having two leagues solves that problem. In short, the mere possession of XX or XY chromosomes, or a vulva or a penis, or a deep sense or knowledge of oneself as a woman or man – or some combination of these – is not always going to be directly relevant to the underlying point of a given social division. Sometimes, only certain of these features will be relevant, in which case, we might have different standards for who “counts” as woman or man for different purposes or in different contexts. For example, someone who counts as a woman for purposes of using a public restroom might not count as a woman for purposes of playing professional basketball. But in neither case is their really “being” a woman in some abstract, metaphysical sense the most 8 relevant thing to determine. And in neither case is “womanhood” or “manhood” (as such) the underlying basis for the original division. Rather, those designations are rough approximations – proxies – for the real set of features that matter in the context. Of course, different stakeholders may disagree about the emphasis that should be placed on different features in different contexts. They may have different conscious or unconscious values, priorities, or political reasons for drawing a line around certain biopsychosocial facts or features rather than others for purposes of grounding category membership, whether generally or in a particular case. But we should remember that the line is just a line. It is doing something for us. It is lumping together certain facts or attributes to simplify a complex reality; it is not itself the reality we are trying to pick out. Brian D. Earp is a philosopher and cognitive scientist who studies, among other things, sex, gender, and identity. Brian is Associate Director of the Yale-Hastings Program in Ethics and Health Policy at Yale University and The Hastings Center and a Research Fellow in the Uehiro Centre for Practical Ethics at the University of Oxford. Parts of this essay have been adapted from an informal lecture entitled, “What is your gender? A friendly guide to the public debate,” which was itself based on an unpublished paper written for a doctoral seminar at Yale led by Robin Dembroff. An edited transcript of the lecture – with references to key works that have informed the ideas presented here – is available at http://blog.practicalethics.ox.ac.uk/2020/03/what-is-your-gender-a-friendly-guide-to-thepublic-debate/. Thank you to Peggy Cadet, Clare Chambers, Zack M. Davis, Robin Dembroff, Anne Fausto-Sterling, David Ley, Moya Mapps, Yuri Munir, Qazi Rahman, Ira Reiss, Adam Safron, John R. Sylla, Richard Wassersug, and Lori Watson for helpful feedback on an earlier draft. Any flaws in the final version are my responsibility. 9