The Aesthetic Constitution of Genders

1. Social Roles and Social Practices

The aesthetic model of genders I’m developing here is essentially a social role account of gender, with an emphasis on aesthetic practice. So a first step towards explication of the model is explication of social role accounts of gender in relation to social practice. Social role accounts say that genders are stable positions, defined by and within societies, that people can occupy. These positions consist in sets of norms, constraints, and enablements that describe how a person occupying a role can and should act, be treated, be judged, and so on (Ásta 2018: 2; Jenkins 2023: ch. 3). If you occupy the social role of prime minister, there are things you can do (launch nuclear missiles), things you can’t do (take six-month sabbaticals), and norms that shape your behaviour and others’ expectations and interpretations of it (prime ministers are, or were, expected to be truthful, unlike estate agents).⁶

Social roles such as “prime minister” are occupied and vacated deliberately, if sometimes reluctantly. But many roles are not easily taken on or sloughed off by choice. Social role accounts of gender highlight the ways in which societies impose genders on people by treating them as bound by particular sets of constraints, enablements, and norms (I’ll often just use one of these words synecdochically for the trio). Often, social role accounts include explicit reference to a hierarchical relation between masculine and feminine roles.

Among social role accounts, one can find various proposals for what these constraints compose, and thus for what sort of thing genders are. For example, Ásta (2018) treats genders as properties that are conferred on people. Such properties, on her view, consist in constraints and enablements. Ron Mallon (2016), meanwhile, treats people kinds, including genders, as entrenched representations of social roles, which describe stable clusters of properties. The contents of such representations encode constraints and so on. These approaches aren’t contradictory; Ásta describes them as “largely compatible” (2018: 52).⁷ Perhaps one answers to a certain gender representation because one has a certain conferred property. I wish to remain silent on this ontological question, and instead follow Elizabeth Barnes’ recommendation that “the substantial work of the metaphysics of gender … should be in explaining the nature of [the] constraints” that compose genders (Barnes 2020: 724). Part of explaining their nature is explicating their dependence on the social world.

So where do the norms that constitute social roles originate? We can trace an answer by following Mallon’s account of how social roles come to exist (2016: ch. 2). A social role exists when there is a representation of it whose contents are “common knowledge” (58). Some social roles become entrenched: psychological, social, and environmental mechanisms create loops whereby features of the representation influence the world, and vice versa (Mallon 2016: ch. 3). A homeostatic property cluster arises, corresponding to the representation: properties co-occur non-accidentally and regularly, precisely because they’re commonly thought to co-occur (Mallon 2016: 3.4). Take the social role of prime minister. Suppose that common knowledge about prime ministers says that they’re unusually self-assured. This feature of the representation influences consideration of candidates: self-assured people are (self-)selected. So most prime ministerial candidates really are unusually self-assured. This both makes it non-accidentally true that prime ministers are unusually self-assured, and reinforces the common belief that they are. Not only do we now have part of an entrenched prime ministerial social role; we also now have a norm conditioning prime ministerial behaviour.⁸

On this picture, the norms that constitute social roles depend upon whatever drives the mechanisms that entrench them.⁹ Mallon characterizes these drivers as “human mental states, decisions, culture, and social practices” (5, 88, 95, 207). The generality of this characterisation is deliberate (5) and apposite for his purposes, but I want to concentrate on social practices, and in particular rational social practices.

According to Sally Haslanger, social practices are “patterns of learned behaviour” that allow coordination among group members with respect to management of things the group finds valuable. Coordination depends upon the responsiveness of group members to those valuable things and to each other’s behaviour; that behaviour and those things are “interpreted through shared meanings [or] cultural schemas” (Haslanger 2018: 245).¹⁰

Some social practices are rational, and some aren’t. By a rational practice, I mean one in which the “learned behaviour” consists in patterns of value-guided, reason-based inference issuing in judgements. The most obvious practices of this kind are those that are overtly concerned with argumentation or reasoning, such as scientific inquiry or legal proceedings. A number of other practices are less overtly rational, but are only comprehensible or explicable when considered as rational: think of collective moral judgements, art criticism, news analysis, astrological divination. And some social practices are simply not rational, in that they do not secure their valuable outcomes directly through reasoning: consider sport or religion. There may well be uncertain cases: is voting a rational practice or not? But all I need is the idea that some practices are rational and others not.

I should emphasize that I do not mean that a practice is only rational if there is good reason to do it. All social practices, by Haslanger’s definition, are oriented towards value, so there is at least some prima facie good reason for engaging in each. Conversely, a rational practice may well turn out to be irrational in the “good reason” sense: perhaps the value at its centre turns out to be empty, or perhaps there is an overwhelming prudential reason not to pursue it. A practice is rational in my sense if it secures its valuable outcomes via reasoning, not if pursuing it is wise. I’m not engaged in rationalizing explanations of social practices; I’m engaged in exploration of social practices that are essentially rational(izing).

Many practices that contribute to construction of social roles are rational in this sense: they are concerned with judgements, reasons, and inferences. For example, the “conferral” of social properties at the heart of Ásta’s account of social roles is a rational practice: a society confers a social property on a person by engaging in a coordinated pattern of reason-driven judgement. Meanwhile, Mallon’s representations become entrenched through their repeated use in explanation and prediction of the behaviour of individuals: as he puts it, their involvement in “causal pathways” (2016: 3.5). Treating people as if their behaviour is explained by their gender makes genders real, and does so because of coordinated practices of reasoning.

These descriptions of the mechanisms of conferral and entrenchment illustrate why rational social practices are especially apt for reifying gender kinds. Treating facts or observations as the grounds of inference bestows on them the gravity of reasons; a conclusion derived from those reasons, or an observation apparently explained by them, acquires normative authority. It’s one thing to simply say that women are more emotional than men; it’s another thing to “explain” this by reference to differentiated brain chemistry. The explanation places both the observation and the alleged explanation within the space of epistemic reasons, and so confers upon both special epistemic status. Similarly, it’s one thing to just assert that women should be excluded from leadership positions, and another to say that this is so because they are naturally conciliatory (and, perhaps, to further explain this with speculative reference to inscrutable pre-history). The practical conclusion and its apparent grounds are brought into the space of practical reasoning, and acquire normative weight. This is why rational social practices concerned with reasons and judgements about people are especially apt to generate norms, constraints, and enablements that demarcate gender roles: norms and constraints supported by reasons have force.

Thus far I have suggested that social roles consist in norms, constraints, and enablements, which depend upon social, rational, human-oriented practices (I’ll just call these “social practices”). These are collective, coordinated patterns of behaviour in which judgements are made about people, and reasons given for those judgements (whether implicitly or explicitly). I now want to use this framework to organize social role accounts of genders by interpreting two prominent accounts in its terms. I say “interpret”, because the readings I give move away from the intentions of the accounts’ authors; all the same, I will call them Mallon’s and Haslanger’s accounts.

2. Two Ways to Model Gender Kinds

Mallon investigates how human kinds can be constructed such that the kinds, and correlative categorizations of people, play substantive roles in “induction, explanation, and prediction” (2016: 93). He is concerned with the question of how something socially constituted can be causally efficacious, and hence explanatory of things like behaviour or health outcomes. He doesn’t focus on genders, but his account is intended to encompass them. To apply Mallon’s approach to genders is to say that there are entrenched social roles of genders, with associated common-knowledge representations, consisting of norms and constraints; that features of representations of genders license and underwrite reliable explanations and predictions of relevant behaviour, characteristics, or outcomes, guided as these are by norms and constraints; that those features of the representations accurately reflect homeostatic clusters of properties possessed by kind members; and that the clustering and the possession of properties owe to the practices of explanation and prediction that they underwrite.

Genders, on this account, are fundamentally dependent on epistemic social practices. The account involves inferences premised on epistemic grounds: reasons to believe or predict things about people based on their presumed gender. The model says how such epistemic practices entrench representations of social roles, so that the predictions and explanations continue to be reliable. The constraints, enablements and norms that describe the social roles are to do with what explanations and predictions are appropriate for a person within that social role.

An alternative is to concentrate on how practical reasoning constitutes gender roles. For example, on the simple version of Haslanger’s account:

S functions as a woman in context C iff_df

1. S is observed or imagined in C to have certain bodily features presumed to be evidence of a female’s biological role in reproduction;

2. that S has these features marks S within the background ideology of C as someone who ought to occupy certain kinds of social position that are in fact subordinate (and so motivates and justifies S’s occupying such a position); and

3. the fact that S satisfies (i) and (ii) plays a role in S’s systematic subordination in C, that is, along some dimension, S’s social position in C is oppressive, and S’s satisfying (i) and (ii) plays a role in that dimension of subordination.

   (Haslanger 2012a: 235)

The first condition of the account describes “natural” features generally taken to be criterial of kind membership. The second characterizes the kinds of practical inferences made on the basis of the features adduced in the first condition: kind members ought to be treated in certain ways, and their treatment is motivated and justified by ascription of those “natural” properties. The third links these practical directives to actions and social positions.

The second and third conditions make clear that the account is predicated on rational social practices which involve making inferences about and giving reasons for ways in which people ought to be treated. To function as a woman in a certain social context is to be subject to these inferences, and thus to the kinds of subordinating treatment they license. The social role is demarcated by constraints and enablements related to the appropriate treatment of people occupying it. The gender social roles are what they are owing to the social practice of making practical inferences and working through their consequences.

I sketch these two accounts not to endorse or dispute their particularities, but rather to show similarities and thus exemplify generalities. On both accounts, gender social roles depend on rational social practices. The social roles exist because of social practices of explanation, prediction, and practical reasoning. People are subject to practical and explanatory inferences and judgements on the basis of their ascribed role, judgements which delineate role-related constraints, norms, and enablements. These things constitute the social roles. The roles and the rational practices on which they depend interact in a reinforcing loop: the roles influence the practices, which reinforce and reify the roles.

The fundamental difference between the two accounts concerns the kinds of practice with which each is concerned. According to Mallon’s account, gender roles depend on epistemic social practices, whereas according to Haslanger’s account, it’s practical social practices. So you might think that the two accounts are competitive. However, Mallon suggests that his approach and Haslanger’s are compatible (2016: 209–210). Haslanger, likewise, does not see her account as the only viable one, but rather as a way to make salient certain features of genders and their construction (Haslanger 2012d: 187–189). I now wish to amplify this suggestion, and in so doing clear space for an aesthetic approach.

3. The Space for an Aesthetic Model

The two exemplary accounts are compatible in two related ways. The first concerns the kind of task towards which they’re directed. The second concerns the domains in which each pursues that task.

Both accounts are directed towards modelling gender roles. This is a different task from that of trying to determine the correct application conditions of gender terms, which is how the metaphysics of gender is sometimes approached. Like Barnes (2020), I am dubious that clear, consistent, and extensionally adequate conditions can be found, since the use of gender terms is complicated, confused, labile, and contextually sensitive. It is also a different task from that of providing an exhaustive description of all that genders are; unlike definitions or descriptions, models are explicitly and deliberately selective. The point of a model is to illuminate the target phenomenon by focusing attention on some aspect of it. Haslanger’s model characterizes the conclusions of the practical inferences it involves in terms of subordination and domination. This makes manifest the hierarchical relation between gender social roles. Mallon, however, concentrates on epistemic practice in order to explain how socially constructed kinds can be causally efficacious.¹¹

In a weak sense, this explains already how the two accounts can be compatible: two selective models of a phenomenon can avoid inconsistency by selecting different things. But these two aren’t just weakly consistent; they’re more strongly compatible, even complementary. It’s both possible and valuable to model complex phenomena—weather, consciousness, migration—in different, equally adequate ways that make salient different aspects. It’s possible because the phenomena are complex, because models don’t capture every detail, and because a model’s adequacy is relative to explanatory ends. It’s valuable because complex phenomena are hard to understand, and their aspects are so multifarious that we benefit from examining them from many angles. Genders are among these complex phenomena. They are so complicated, their constitution so convoluted, that they need to be repetitively, selectively modelled in order that we can make sense of them.¹²

The two models of genders illuminate important and distinct aspects of their objects. This is because they are concerned with different domains of reason and practice. A domain of reason consists in a nexus of inference, value, norm, and ground, that demarcates a distinctive form of rationalizing. There’s an epistemic domain of reason, in which epistemic inferences, governed by epistemic norms, which are conditioned by epistemic values, connect epistemic reasons to epistemic outcomes such as belief, knowledge, and understanding. There’s also a distinct practical domain of reason, within which inferences result in such things as goal-directed actions or moral judgements.¹³ Associated with each domain of reason is a domain of practice, a field of human activity. The domain of epistemic practice encompasses, for example, scientific inquiry, crossword puzzles, and criminal investigations; the practical domain encompasses, for example, political protest, funeral rites, and dietary decisions.

These domains of reason and practice are imbricated, not isolated; many activities belong in both, and norms and reasons can be similar or shared. You can do something practical by doing something epistemic: say, pursue justice by investigating a murder. Nonetheless, the distinction is useful. It makes forms of life legible by showing ways in which activities’ rational character makes them cogent and comprehensible. There is room for both an epistemic and a practical approach to genders because the rational practices that make up genders are various. Some practices only make sense given epistemic reasons, norms and values, some only make sense given practical ones. So, again, epistemic and practical models of genders are not directly competitive: they pertain to, make salient, and make comprehensible the constitutive role of practices belonging to different domains.

The story being told here will have the ring of familiarity, and perhaps even truth, to people acquainted with normative talk in theoretical and practical philosophy.¹⁴ Whatever the details, the general idea that there are these two discrete domains of rationality and practice has an ingrained plausibility. I’m just going to trade on that plausibility; I’m not going to argue further that these domains exist, nor that social role accounts of gender are illuminatingly interpreted by reference to them.

Now, if there are these two domains, and if genders can be valuably modelled within each, we might think it equally valuable to model genders in any comparable domain. And so, if there is a comparable aesthetic domain, it might be valuable to model genders within it. To conclude that it is valuable, we would need to be convinced that there is indeed an aesthetic domain of rationality and practice; that it includes rational, social practices centred on humans; and that these aesthetic practices really can delimit gender roles. My aims in the next two sections are to argue that there is such an aesthetic domain, that there are such aesthetic practices, and that they can construct genders.

4. Aesthetic Rational Practice

The aesthetic model of genders I’m pursuing says that the feminine and masculine roles in a social context consist in aesthetic constraints, enablements and norms, which depend on rational, social, human-centred aesthetic practices, in which gender differentiation is essential. In this section and the next, I’ll try to establish the plausibility and cogency of this model by explaining its elements, concentrating here on aesthetic practice, and there on its constitutive relation to gender roles.

I’ll start by noting the confluence between the conception of a social practice outlined above and Dominic McIver Lopes’ “network theory” of aesthetic practice (2018). In Haslanger’s terms, aesthetic social practices are patterns of learned behaviour allowing coordinated activities directed towards aesthetic value. Similarly, according to Lopes, “aesthetic practices are regularities in the performance of aesthetic agents that are explained by cognitive schemas and resources” (2018: 120).¹⁵ Such practices structure people’s pursuit of aesthetic value: they “enable agents to develop specialized competence for [aesthetic] achievement” (127). They do so by instituting norms that establish what is aesthetically valuable within them (120–121). Practices are oriented towards particular objects of value (133), around which norms of appreciation revolve and evolve.

Reading such practices in rational terms is easy enough. As Lopes describes them, aesthetic practices establish norms and standards that say what is aesthetically (dis)valuable about the objects at their centre. These norms and standards provide reasons for appreciation of objects. Aesthetic practices are structured around appreciation, including the elucidation of reasons to appreciate. To engage in a certain aesthetic practice, it’s essential to engage in the patterns of reasoning and appreciation distinctive of it. It’s the fact that appreciation is at the centre of these practices, rather than belief, knowledge, or action, that make them aesthetic.¹⁶

This particular idea that there is a distinctively aesthetic kind of rational practice depends on the general idea that there is a distinctively aesthetic kind of rationality. A growing body of work explores just this idea.¹⁷ The aesthetic domain of reason parallels the epistemic and practical domains. It encompasses distinctive aesthetic values; distinctive aesthetic reasons that ground inferences; distinctive outcomes of such inference, including aesthetic judgements; and distinctive aesthetic norms that regulate inferences, such as norms of aesthetic testimony.¹⁸ This framework for thinking about aesthetic rationality allows us to explicate rational aesthetic social practices.

As an example, consider the community of people who make or appreciate jungle music. They’re engaged in creating works and appreciating them appropriately. What makes a jungle tune aesthetically valuable, what ought to be appreciated (or not) in such a tune, is established by what those creators and appreciators communally take to be aesthetically valuable. The norms and standards of appreciation both guide and follow appreciation and judgement by the community; they evolve as the practice they concern does so. To be a part of this community is to participate in the appreciation at its centre, and the elucidation of reasons for appreciation. In so doing, one contributes to the development of those reasons.¹⁹

This rational approach to the aesthetic domain is not the only one on offer.²⁰ A contrasting approach conceives of the aesthetic as a domain of a special kind of experience. The broadest version of this approach says that any sensual experience is, or can be, an aesthetic experience given the right kind of attentive practice (Irvin 2008). A narrower version says that a distinctive kind of experience, perhaps a heightened version of everyday experiences, marks out the aesthetic (Dewey 1934).

I can’t pursue the argument between these two approaches here, so I’ll just gesture towards two reasons for preferring the rational, practice-based approach. The first is to do with weighting pre-theoretical intuitions. Aesthetics is a field in which balances must be struck between the notion that there are such things as expertise, rightness, critical acumen in aesthetic judgement, and the notion that aesthetic engagement has some definitively subjective or personal character. The more one is tempted to emphasize the personal, the less one will be inclined to value expertise, and vice versa. I am inclined to value expertise, and so to favour the rational approach, because it promises to explain how judgements and values can have normative weight.²¹

The second is a thematic reason that should emerge in what follows: thinking of the aesthetic in terms of rational practice makes sense of the aesthetic constitution of genders in a way that thinking of it in terms of experience cannot. In this respect, the positions adopted in this paper are mutually reinforcing: we should buy the aesthetic model of genders in part because of the plausibility of the rational practice account of the aesthetic, and we should buy the rational practice account in part because of the plausibility of the aesthetic model.

Thus far I have explained the general notion of rational social aesthetic practice. Now it is time to focus on person-centred practices. Lopes’ view most clearly applies to making and appreciating art, but galleries are not the only sites of aesthetic activities: they also occur in tattoo parlours, coffee shops, nightclubs and national parks.²² Lopes’ account is explicitly pluralistic (127–129) about the possible objects of aesthetic practice and the possible values to be found in them. If norms for appreciation of a kind of object have been established, and if people engage in evaluation according to those norms, we have an aesthetic practice, never mind if the objects are biscuits, bicycles, boulders or books.

Aesthetic social practices pertaining to people clearly exist, and clearly fit Lopes’ schema. Widdows (2018) discusses in great detail practices concerned with a narrow conception of bodily “beauty”: grooming, adornment, comportment; the shaping of the body through diet, exercise, and surgery. There are also many person-centred aesthetic practices concerned with different kinds of aesthetic value: fashion, tattooing, bodybuilding, perfume, the cultivation of personal style.²³ In each of these cases, we can understand the practice as one in which aesthetic norms are established and followed, taking humans as the objects of appreciation.

The harder question is whether human-oriented practices are rational in the sense required. One might accept that aesthetic judgement of artworks or landscapes has a rational character, and yet maintain that aesthetic judgement of humans does not. And this matters, because within the framework for thinking about gender roles outlined above, it’s the rationality of practices that lends normative force to the constitutive constraints, norms and enablements that they generate. So I had better meet this challenge.

The challenge begins from the observation that, in genuinely rational thought, the norms that govern the transition from reason to conclusion demand some degree of objectivity. You should not settle what you believe by what you want to believe, but rather by what you ought to believe. As in thought, so in genuinely rational practices: reasons ought to lead to conclusions impersonally.

So far there is no challenge, for this demand of objectivity has its counterpart in the aesthetic domain: the doctrine of disinterest. Disinterest is used to distinguish genuine aesthetic judgements from mere judgements of taste. The latter might be made with personal interests in mind, and might be connected to action: desire for the object under consideration, calculation of its monetary value, acquisitive action towards it. But genuine aesthetic judgements are coloured by no such interests: they are, purely, considerations of the aesthetic merits of the object in question.²⁴ Given this, we can see how aesthetic judgements can respect the demands of objectivity: they must be structured and normed such that they can be made disinterestedly, with only aesthetic merit in mind.

But one might worry that aesthetic judgement of humans cannot be so structured, and so aesthetic practice centred on people can’t be rational. After all, human aesthetic practices plausibly begin from presumptively sexed bodies: from differentiation of male and female, and adoption of different practices for each. This is alike for epistemic and practical practices (compare the first condition of Haslanger’s account). But a signal difference with aesthetic practices, or at least the most prominent ones, is that they also stay focused on bodies, and often on sexualized bodies. And now we do have a challenge. If aesthetic social practices are centred on sexualized bodies, surely they are also centrally concerned with desire: what it is to judge someone else aesthetically pleasing is precisely to express some kind of interested approbation of their appearance. And so there cannot be genuinely rational aesthetic practices focused on people.

Or so the objection goes. But it can be met. Its source is the observation that human aesthetic practices begin from bodies and their appearances. This is incontrovertibly true. But it is not true that all such practices are concerned with bodies, nor even that practices concerned with bodies are always imbued with desire. Further, the aesthetic rationalist can plausibly deny that judgements imbued with desire are genuinely aesthetic in character.

To the first point, note that many aesthetic practices centred on people go way beyond sexualized appearances. We can see this by considering the range of aesthetically evaluable aspects of people. Some are undoubtedly both perceptible and associated with attractiveness. Some are perceptible, but not (obviously) sexual: comportment, tone of voice, style of dress. Some are not perceptible. One can aesthetically evaluate the life choices of another: one can regard a life as well-lived and aesthetically commendable for being so.²⁵ Of course, not every reason to value someone is an aesthetic reason to do so, but the aesthetic reasons and properties predicable of people are various, multiple, and range beyond the perceptible. So human-centred practices need not focus on the body.

To the second point, note that many judgements and practices that are body-based have nothing to do with sexual attractiveness, and so are not subject to the worry about the irrational influence of desire. Perfume is an excellent example. The appreciation of perfume is manifestly a gendered practice, both in the incidental sense that it’s mostly women who do it, and in the essential sense that it’s internally gendered (there are masculine scents, there are feminine ones). It is also manifestly aesthetic. But people do not, in the main, wear perfume to make themselves more desirable: as Larry Shiner says, “much advertising suggests perfumes are about sexual attraction, but the motives for wearing them are complex, including olfactory pleasure and personal identity as well as grooming” (2015: 377). To the extent that the challenge links reasons in aesthetic body practice to sexual desire, this constitutes one among many counterexamples.²⁶

Consider also that it is possible for an object to be the focus of both interested and disinterested judgement. One can judge the same thing to be both desirable and aesthetically pleasing. So the fact that humans are at the centre of aesthetic practices, and the fact that humans can be objects of desire, does not entail that aesthetic judgements within that practice are all judgements of sexual desirability. It might in practice be hard to tell apart a dispassionate judgement of aesthetic merit and a passionate judgement of, well, passion. But this is no mere formal separation: the two can come apart, and one sign of this is that they often don’t when they should. Many aesthetic actions and activities become significant sites of cultural conflict exactly because they are intended by participants to invite genuine aesthetic appreciation within a practice, but are taken by outsiders as invitations to sexualized judgement. The most prominent of these are feminine beauty practices whose aesthetic significance is really for the self and others within the practice, but which are often critiqued in terms predicated on a sexualizing male gaze.

So, much aesthetic practice centred on people is not connected with desire, or at any rate should not be. The judgements involved in those practices can thus be disinterested; the judgements and the practices can thus be appropriately rational. However, for all that, the objection might be pressed. Granted that some practices do not essentially involve interest, it remains obvious that a significant number of human-aesthetic practices are concerned with sex and sexiness; with provoking and admitting desire, with judging desirable. And if these practices form the core of human aesthetics, aesthetic rationality runs out at the crucial point. For all such judgements are essentially personal, subjective, and unresponsive to reason.

The aesthetic rationalist’s response is that the pressed objection begs the question against their view. The rationalist advances a model of the aesthetic domain in which genuine aesthetic appreciation takes place within the strictures and norms established by communities and their practices. Further, the rationalist opposes the idea that aesthetic judgement is necessarily predicated on a particular kind of experience, still less experiences that might be concerned or confused with desire. The objection insists that, in fact, there are genuine aesthetic judgements that are wholly personal and subjective, not subject to reason, and this insistence is implicitly predicated on the notion that the basis of aesthetic judgement is a kind of incorrigible personal experience; specifically, an experience of pleasurable desire, or desirable pleasure. This is precisely what the rationalist denies, and insisting on it is no argument against.²⁷ As Lopes says: “intuitions about aesthetic value are bound to be so heavily infused by aesthetic hedonism that we must not take them at face value” (2018: 137).

Let me sum up. The challenge was this: aesthetic practices centred on people can’t be properly rational, because they are suffused with desire, because they are about bodies. I have countered, first, that many such practices have nothing to do with bodies; second, that even practices to do with bodies are often nothing to do with desire, or shouldn’t be; third, that insisting on the aesthetic character of judgements of sexual attractiveness or desirability is begging the question against the aesthetic rationalist.

All this is aimed towards vindicating the idea that there can be genuinely rational aesthetic social practices centred on people. These practices provide reasons for appreciating people, for valuing their aesthetic achievements. People participating in the practices contribute directly to establishing what is valuable within them. What remains to be shown is that these practices are apt for generating the norms, constraints and enablements that constitute gender roles. So let us turn to that task.

5. Aesthetic Practice and Gender Roles

Of course, the claim cannot be that all aesthetic practices centred on humans make constitutive contributions to gender roles. The claim is that some do. My aim is to provide a programmatic framework for saying which do so, and how. I will first suggest that the constitutive practices are those that essentially differentiate between genders, and exemplify the ways in which such practices generate appropriate norms, constraints, and enablements. I will then suggest that the way the constitutive practices differentiate between genders is by presuming a classificatory distinction and then aestheticizing qualities accordingly.

Gender differentiation is essential to a practice when it is a constitutive norm of that practice that differently gendered people should engage in it differently. A way of telling whether a practice differentiates essentially is to ask whether the same basic features have different appreciative implications depending on gender classification.²⁸ For example, a woman’s red lipstick might be thought of as conventional, or basic, or perhaps more positively classic; the same lipstick worn by a man might instead be provocative, shocking, or inappropriate.

This conception of what it means for practices to essentially differentiate genders is deliberately silent on gendered hierarchies. It is probably typical of practices that differentiate that they do so in a way that reinforces hierarchies. For example, feminine grooming practices are much more demanding of time, effort and money than masculine ones, and to that extent contribute to oppression. However, for present purposes it is enough to have in mind the idea that practices differentiate, without also adducing the oppressive effects of differentiation.²⁹

The line between practices that differentiate essentially and those that do so incidentally is not sharp. There are some human-aesthetic practices whose appreciative standards clearly, constitutively depend on differentiation of genders and differential engagement (even if implicitly, by normative exclusion of one or some genders). Grooming and dress are clear examples. It is constitutive of the practice of formal dress that women and men engage in it differently: ball gowns for one, black tie for the other. Perhaps surprisingly, drag is another example. Granted that one of the expressive effects of drag performance is to undermine and complicate norms of gendered appearance, the practice all the same constitutively depends on there being differential appreciative standards for different genders in the first place, and it is essential to the practice that people of different genders engage in it differently.

There are also some human-aesthetic practices in which it is less clear that gender differentiation is essential. Modern dance performance is one example. Historically at least, dance performance invited differential appreciation of people of different genders (or, more precisely, of their movements). But it is less obvious that performances one might see today of works created recently invite one to appreciate differentially, and so it is less obvious that contemporary dance essentially differentiates among genders. So it is not constitutive of this aesthetic practice that people of different genders engage in it differently.

So some, but not all, human aesthetic practices gender-differentiate essentially.³⁰ I will now start to make the case that such practices generate norms, constraints and enablements that are fundamental to gender roles by distinguishing two kinds of aesthetic practice.

Recall that aesthetic social practices depend on aesthetic networks or communities. People belong to various networks and communities of various sizes and scopes, and varying degrees of voluntarism. Joining the bridge club is entirely voluntary; being born British isn’t; somewhere in between are various networks one might join incidentally (the network of parents of children attending a given school), or as a fait accompli or necessity (the homeowners’ association). As with networks in general, so with aesthetic networks. Some are large, some are small; some we join voluntarily, some we can’t help but be enmeshed in. And as with aesthetic networks in general, so with those that establish the feminine and masculine roles in a given context.³¹

Large, broad, coarse-grain cultures, the kind of things referred to as “Western culture” or “American society”, undoubtedly establish gendered standards and norms of human aesthetics. The plain fact that we can refer to the “Western beauty standard” demonstrates this. Cultures have aesthetic practices that are devoted to making humans aesthetically valuable, such as dressing, grooming, adorning and embellishing (Davies 2020). Reasons to appreciate people and their presentation are established by collective appreciative actions within such practices. The standards are often vague, uncodified, even contradictory. The strictures they impose are loose, because so general, but nonetheless extant: apparent diversity often disguises deeper conformity. And the strictures, I claim, are substantial enough to constitute gender roles.

It’s best to make this case by concentrating on obvious examples. So, take the particular practice of “professional dress” for white-collar work in Northern Europe or the USA. Obviously, within this practice, there are different appreciative standards for men and women. And these standards generate substantive norms, constraints, and enablements. A woman’s professional dress should normatively be demure; she ought not to wear a tie; she may wear a wider range of colours than a man. A man’s dress should be plain; he ought not to expose his legs; he may wear much the same thing day after day. These are aesthetic constraints arising from aesthetic practice. They are aesthetic because the reasons they generate are reasons to appreciate, or not appreciate, people and their appearances. In fact, disapprobation is more telling in this context: the centrality of these aesthetic norms to gender is disclosed by the policing of violations more than by the approval of conformity.

That last point goes some way to answering the pressing question of whether norms such as those of professional dress are really substantial and serious enough to be constitutive of gender roles. Ask yourself: in any context where gender is salient, which are the norms and constraints whose transgression is most routinely punished? Which are the boundaries that are most rigorously enforced? Punishment and enforcement need not, of course, be meted out by authorities: they can consist in stares, frowns, cat calls, shunning. As per the modelling task I’m engaged in, I don’t take the aesthetic norms to be more fundamental than any others. But I do claim that, if we reflect on the questions about enforcement, we will find that the answer is aesthetic as much as it is practical or epistemic. This is the case for thinking that the coarse-grain cultures into which we are born, and from which it is hard to rescind ourselves, form aesthetic networks whose aesthetic practices essentially differentiate genders, and in so doing generate norms that constitute gender roles.

However, as I noted above, there are all sorts of communities and networks to which we belong, some more voluntary than others. Applied to aesthetic networks, this both enriches and challenges the picture being drawn.

The enrichment comes from the observation that many of the more voluntary networks and practices with which we might engage also essentially incorporate gender differentiation on aesthetic grounds. This is most apparent in explicitly aesthetic practices. For example, social media is rife with small, tightly-knit communities focused on looking a certain way, or aestheticizing certain features. These communities and the practices they support are often constitutively gendered, often implicitly by exclusion (try searching Instagram for words like “eyebrows” if this claim doesn’t already strike you as obvious).

But many voluntary networks that are not explicitly aesthetic nonetheless engage in aesthetic practices and establish gendered aesthetic codes. There are gendered aesthetic codes for punks, furries, hippies; for parents, academics, politicians, actors. This means that there are gendered aesthetic practices associated with each of these communities. It is a nice question which of these practices are essentially gender-differentiated, and which are incidentally so. But I take it that there are examples on either side of that divide. And I take it that the voluntary practices that differentiate essentially generate norms, constraints and enablements just as much as the involuntary practices do.

The important enriching point, on which I will lean later, is that people who engage in voluntary aesthetic practices do so deliberately. In so doing, they actively embrace the aesthetic norms, value and standards that the practice encompasses, and they play an active role in establishing those norms. Furthermore, the practices that people engage in within their chosen aesthetic networks frequently influence how they work with the constraints of their involuntary networks. The way in which, say, someone who identifies as a goth inflects their professional dress to reflect that identity manifests an interaction and negotiation between the (gendered) elements of goth presentation and those of professional presentation.

How aesthetic communities and social roles overlap, contradict, and complement, how such relations affect appreciation, how individuals negotiate those relations; these are tricky topics. Being human is a complicated business. But the sheer pervasiveness of aesthetic gender differentiation across communities, large and small, emphasizes the point that gender is, at bottom, an aesthetic phenomenon as much as a practical or epistemic one: when you have gender differentiation, you have aesthetic differentiation of genders, and vice versa.³²

But this pervasiveness and richness also challenges my picture. All these overlapping and distinct social roles, communities, networks and practices present many ways of “doing” gender, many ways of being a man or a woman, many aesthetic norms to which one could conform. And this raises the question of whether such widely heterogeneous practices can really be the foundations of gender roles. If the norms they present are so various, what do they have in common that makes all of them about gender? This is, in a sense, a problem for any social role account of gender. Whatever aspect of social role you focus on, the sheer variability of what gender has meant and does mean across time and space makes it difficult to say much concrete about what unites multifarious social roles as gender-constituting.

This brings us to the “how” question: how is it that the constitutive aesthetic practices differentiate genders? My answer has two parts: classification and aestheticization.

On classification, I return to Haslanger’s account of genders (2012a), which embeds a crucial role for presumptive judgement of biological sex. Roughly, if someone is presumed to have the biological reproductive features typical of females, they are marked out for feminine treatment on the basis of that very judgement. Similarly, in the aesthetic case, presumptive attribution of bodily features leads to classification as suitable for the aesthetic treatment typical of one gender or another. So the first thing that unites the various aesthetic practices as gendering is the fact that they are fundamentally predicated on presumptive classification by body.

In Haslanger’s account, feminine treatment is subordination. I will say something shortly about what feminine- and masculine-typical treatment is on the aesthetic model, but let me first forestall a possible objection to the recruitment of Haslanger’s condition to the aesthetic model. I said above that we should be wary of thinking of human aesthetic practices just in terms of bodies, since they are much broader than that. But I’ve now said that what they have in common is that they’re applied to people based on body-identification. So is recruiting Haslanger’s condition a self-sabotaging move?

I think not. We can distinguish the content of gendered practices from the classificatory work that the practices presume. The classification is predicated on bodies, but that doesn’t mean that the practices predicated on that classification need themselves involve bodies. This is so for practical models: oppression is predicated on body classification, but oppressive practices don’t necessarily, or even often, have anything to do with bodies. As with practical models, so with the aesthetic model.

It might be worth noting, though, that this classification can be thought of in aesthetic terms. In Haslanger’s model, classification involves “observing or imagining” bodily features. However, it is plausibly more often a matter of judgement, an inference from features of personal presentation—clothes, hair, comportment, and so forth—to presumed features of a body.³³ This is usually a subconscious, automatic inference; correspondingly, it is usually lazy, biased, and heuristic. But it’s still an inference, issuing in a judgement.

If classification involves inferential judgements, we can ask about the domain of rationality to which they belong. There is reason to suppose that they are at least sometimes aesthetic, insofar as they involve descriptive-evaluative attribution of sex-typical properties. Consider, for example, judgements concerning people’s height. The claim that women should be between 5’ and 5’8” tall can be normative both epistemically—women do, in fact, typically fall within this range—and aesthetically.³⁴ It may well be that some properties or features are only present in archetypes as aesthetic or epistemic reasons: perhaps not every property does double duty. But the general thought is that gender appearance archetypes are both descriptive and normative, and it is hard to separate clearly two different archetypes, aesthetic and epistemic.

So I suggest that, at the least, the judgements on the basis of which people are classified into gender kinds are importantly influenced by aesthetic norms, standards and practices. This reinforces the suggestion that gender roles are fundamentally constituted by aesthetic reasoning.

Returning to the main line of argument, I promised to complete the “how” of differentiation by saying something general about feminine- and masculine-typical treatment, parallel to Haslanger’s claim that the practices that constitute genders and their differentiation are subordinating and domineering. This is a common thread that ties together a great many gendering practices and gender roles that are otherwise dissimilar. My claim is that the way in which aesthetic practices differentiate by gender is by aestheticizing features taken to be indicative and constitutive of gender membership. These might be the very presumed bodily features that lead to gender classification. For example, professional dress for men aestheticizes broad shoulders. They might also be features more broadly associated with genders. For example, insofar as submissiveness is a normatively feminine characteristic, an aesthetic practice might differentiate between genders by aestheticizing aspects of personal presentation or style that express submissiveness: specifically, by making it a feature of aesthetic approbation in women and disapprobation in men.

There is an obvious circularity to this suggestion. The proposal is that aesthetic practices constitute genders by differentiating between them, which they do by aestheticizing features that are somehow constitutive of genders. I could avoid this circularity by simply claiming that only bodily features are aestheticized. But truth trumps theoretical tidiness, and in truth, any replete descriptive model of what genders are today is going to involve something like this circularity. The gender roles we have now are accretions of practice building upon practice, of epistemic reasoning influencing aesthetic reasoning, of interactions across times and spaces complicating the gender roles prevalent in any given society. It would be surprising if the aesthetic (and other) practices that today constitute genders did not depend in some degree on the results of previous practices, and instead were built only upon the body.

A final point might help the reader to swallow this circular pill. I don’t think that the aesthetic model describes how genders have always been. It’s plausible that the aesthetic aspects of genders have become more salient in their construction over time: that we now treat genders as aesthetic entities more than people used to (cf. Widdows [2018: 1]). Two possible reasons for this are, first, that the idea that gender distinctions have a substantive basis in differences in sexed bodies that lead to justified epistemic and practical conclusions is being steadily undermined. If one thinks that, when challenged, systems of domination find ways to mutate, replicate and consolidate, one might think that an emphasis on the aesthetics of gender is a way to shore up these divisions in the light of their undermining; for aesthetics putatively provides a link to the sexed body, and thus to the supposed primitive basis of division between two genders.³⁵ Second, perhaps the vastly more visual culture prevalent in the last 150 years, in terms of production and especially distribution of images, has contributed to the foregrounding of the aesthetic. One might also add that more people these days enjoy the level of material comfort that allows the “luxury” of consumer attention to appearance (cf., again, Widdows 2018).

In the space available, I can’t do any more than this to prop up the aesthetic model. I’ve claimed that there are rational, human-centred aesthetic practices, which generate reasons to appreciate people. I’ve argued that these practices really are rational in the required sense. I’ve then suggested that some of these practices differentiate essentially among genders, and in so doing generate constraints, norms and enablements sufficiently substantial to constitute gender roles, in just the same way that constraints generated by epistemic or practical considerations do so. I’ve finally suggested that the way in which the constitutive practices differentiate genders involves a presumptive classification based on bodies and aestheticization of gender-typical features.

I’ve made a lot of claims about the aesthetic in my explication. You may disagree with many of them. But the main thing I wanted to establish is that the conceptual space available for an aesthetic model can be filled by a plausible candidate. If you agree that there is a separate aesthetic domain, and you agree that the model I have advanced is cogent and plausible within a credible conception of that domain, then the aesthetic model does the job I want it to do. You could, if you like, advance a model that conforms to your own preferred conception of the aesthetic domain. But we can agree that aesthetic practices somehow constitute gender roles.

However, even given the model’s cogency and plausibility, you’ll still want reasons to think that it’s useful, besides its general function of highlighting the pervasive way in which aesthetic practices inflect gender in so many aspects of life. I’ll now suggest that, by exploring the constitutive role of particular aesthetic practices in individuals’ identities, we can use the aesthetic model to close the gap between social role and self-identity accounts of gender.

6. Social Role and Self-Identity

The gap, recall, concerns inhabitation. Social role accounts of genders say, credibly, that genders are constructed by societies and imposed upon individuals. Self-identity accounts say, credibly, that individuals’ gender identities are nonetheless importantly bound up with how they themselves experience and express their genders. Given this, we might think that gender self-identity involves embrace of aspects of the roles created by societies: inhabiting those roles, not just existing within them. But this should lead us to ask what this inhabitation is. My suggestion is that this question can be fruitfully addressed by construing identification with a certain gender as embrace of the very same aesthetic constraints, enablements and norms that constitute that role: that is, active participation in the practices that are constitutive of the social role. In so participating, one asserts the appreciative norms that one wishes to be applied to oneself qua gendered individual, and also shapes the appreciative norms that apply to oneself and to others gendered similarly. Embracing a gender role involves active participation in, shaping of, and self-valuation within its constitutive aesthetic practices.

This isn’t the only good way to square away social role and self-identity. But the story one can tell in aesthetic terms does have an appealingly fluent narrative about embrace. Many people really do treat genders as central to their identities. At the very least, people want other people to acknowledge their gender aright, and often they want more than that; they want to affirm it, express it. They demand the right to do so.³⁶ Given this, we should want an account of our personal relation to gender roles that does not limit our options to sullen compliance or grim defiance. This is what the aesthetic story provides.

I’ll start telling the story by crystallizing the worry that motivates it. According to Katharine Jenkins, “to say that someone has a female gender identity is to say that she experiences the norms that are associated with women in her social context as relevant to her” (2018: 728). This, she argues, is a better account than a dispositional view (one is a woman if one is disposed to act in ways one’s society sees as womanly) and a mere-identity view (one is a woman if one sincerely identifies as such). I won’t rehash Jenkins’ arguments, but briefly, her convincing case is that the norm relevancy account best meets some explanatory ends that an account of gender identity ought to have if it is to aid and advance trans rights.³⁷

However, one might worry whether talking about “experiencing norms as relevant” is strong enough to capture the relation in which we ideally stand to our gender identity. Perhaps one experiences a norm as relevant if one is aware that compliance or non-compliance with it somehow speaks to one’s gender identity. This certainly seems too weak. Such awareness is too vague to anchor a substantial identity, and it allows that almost any gender norm could be relevant to one’s gender (since presumably one is either compliant or non-compliant with any given norm). We want something stronger: the norms at issue are those that one experiences as importantly relevant to one’s identity, which somehow ground or structure one’s sense of who one is qua gendered being. Further, to capture the relations in which people ideally stand to their gendered selves, we want the norms to be actively, affirmatively embraced.³⁸

So, which kinds of norms, constraints and enablements are suitable for such embrace, for grounding and structuring a gendered self? Within the framework discussed above, the candidates are aesthetic, epistemic, and practical. Now, it’s unlikely that a single kind of norm grounds gendered identity. Genders and identities are too complicated. However, there are prima facie doubts about whether epistemic or practical norms, constraints or enablements are apt for embrace, and a prima facie case for saying that aesthetic norms are.

The doubts about epistemic and practical norms concern their typical contents and normative force. The epistemic model of genders is built around norms that dictate explanations of behaviour in terms of underlying properties. But if one comes to see that gender norms are social, labile, and violable, one might find that epistemic gender norms no longer provide good explanations of one’s actions. There is little here to embrace as one’s own. The practical model, meanwhile, is built around oppressive treatment of women. Why would one reasonably wish to be treated so? Why would a clear-sighted person embrace oppressive norms as central to their identity?³⁹

The case for aesthetic norms also concerns their content and force. To adopt a feminine aesthetic norm is to invite aesthetic approbation for specific reasons. It is to partially structure your identity around a reason for valuing you. This lends such norms a positive personal force even when their contingency is apparent. This makes them apt for embrace.⁴⁰

Let me explain in more detail how this is so, and in so doing relate the explanation to the account of social roles above. People present themselves through a multiplicity of practices. The most obvious are those related directly to the body: how we dress, adorn, groom, comport, perfume; how we use cosmetic surgery and exercise programmes to shape our bodies. A more extended sense of presentation might include the ways in which we shape our personal environments, like our homes, desks, and dashboards. It might also include the ways we express ourselves in text, talk, or TikToks.

These presentational practices are clearly social and clearly aesthetic. They are governed by aesthetic norms, and concerned with aesthetic properties, values, and reasons. The norms, reasons, and values are established and conditioned by the coordinated judgements and beliefs of aesthetic networks. The aesthetic values at which people aim through presentation are many and various. The same person might have different aesthetic aims in different contexts. But whatever the complexities of cases, personal presentation in general is a matter of showing qualities that are intended to meet with the approving aesthetic judgement of a relevant aesthetic network. To embrace an aesthetic norm as relevant to oneself is to (wish to) externalize its relevance; to wish to align one’s presentation with it: to present oneself in a way that makes salient the things one wants to be appreciated for. This is to participate in an aesthetic practice.

Now, we already have the idea in place that embracing an aesthetic identity is a matter of engaging in the very aesthetic practices that make up a social role. But we can emphasize the importance of the aesthetic for gender identity by considering embrace under the auspices of expression.

Expression has often been explicated as the process that underlies artistic creation.⁴¹ But you can find the explications illuminating, and adaptable to the present matter, without committing to that thesis about creation. Consider R.G. Collingwood’s explication in The Principles of Art (1938). In his view, expression is a process in which someone begins with an obscure, partially grasped feeling or idea, and works in a public, social medium to refine and clarify it, with the aim of being able to honestly claim both understanding and ownership of the idea. Importantly, the idea is not internally clarified and then externalized in a medium: working in a medium precisely is the process of clarification.⁴²

We can move from art to genders via Nick Riggle’s work on personal style (2015). According to him, having a personal style is a matter of expressing your ideals. If you have style, your personal presentation embodies or communicates ideals that you genuinely hold. This view suggests that personal presentation and the practices within which it occurs can be media in Collingwood’s sense: public and social means of expression that someone can manipulate in order to externalize an idea, an ideal, or an aspect of personality.

With respect to gender identity, the idea is that to embrace gender—to fully understand and take ownership of it as a part of one’s self—is to express that gender. To express a gender is to externalize one’s sense of self-identity, understood in terms of the norms one embraces as one’s own, in the medium of one’s personal presentation and the particular aesthetic practices in which one engages. The norms that one seeks to express are aesthetic norms, and the process of expression is a process of working out and externalizing, making apparent, the aesthetic norms one wishes one’s presentation to be appreciated under.⁴³

Note that artistic expression is not necessarily deep, profound, or interesting. On Collingwood’s view, most acts of expression are fairly trivial; this is a corollary of his insistence that expression is a commonplace achievement available to all, not only to a coterie of geniuses. Similarly, many people express their genders in ways that conform fairly straightforwardly with archetypical aesthetic norms. Such expression can be both unoriginal and genuine. But conversely, note that some acts of Collingwoodian expression are difficult, painful, experimental, and protracted. The starting point of expression can be inchoate, obscure, and poorly understood; it can take time, effort, and many false attempts to genuinely clarify the idea at issue (Collingwood 1938: 109–11). Similarly, gender expression can begin from a poorly understood sense of the norms that one wants to apply to oneself, and proceed slowly through a process of trying to get those norms straight via one’s expression of them. One can experience some shadowy norms as somehow relevant, and gradually come to understand exactly what they are and how they belong to one’s self. You make legible and salient to yourself the reasons for why you wish to be aesthetically valued, and do so by engaging in practices structured by those reasons.

This idea of expression exploits the resources of the aesthetics model of gender, insofar as it makes reference to personal presentation within socially established practices. But we shouldn’t think that presentation is the only medium of aesthetic gender expression. Besides manipulating how one looks, one can also manipulate the norms, constraints, and enablements of the practices within which presentation happens. One can express one’s gender identity not just by adopting those norms, but by deliberately manipulating them, playing with them, bending them, testing them. One can thereby treat those norms as one’s medium of expression; presentation becomes a means to manipulate the medium, but is not the medium itself.⁴⁴

So expression of gender identity within an aesthetic social practice can involve not just participation in that practice, but also active contribution to the shaping of its norms. This, in turn, provides further reason to think that embracing a gender identity is an aesthetic matter. Engaging in aesthetic practices provides one with reasons for self-valuation; it also provides one with the opportunity to shape those reasons, both for oneself and others. In these respects, aesthetic practices provide opportunities to truly inhabit social roles that are less obviously available in epistemic or practical domains. The reasons to be gendered that aesthetic practices provide are apt to be reasons for you in a way that epistemic and practical reasons are not.

This is obviously not a complete theory of gender expression and its relation to aesthetic practice. Again, the aim is to demonstrate in outline the programmatic viability of the thought that the aesthetic model of gender roles can illuminate the relation of gender self-identity to social role; in particular, it can help to explain what embrace of a gender role might be. Granted that much more could be said, I hope that the potential utility of the aesthetic model has been demonstrated.

Conclusion

In the preface to Anarchy, State, and Utopia, Robert Nozick disarmingly suggests that his philosophical method involves pushing recalcitrant material into an assemblage that holds together just long enough for an orderly-looking snapshot to be taken, then printing the photograph and suppressing the fact that the assemblage has since fallen apart (Nozick 1974 xii-xiii). I feel much the same about this paper. I have tried to assemble social role accounts of gender, an account of the aesthetic domain, and an account of gender identity and expression. There are ragged edges and awkward joints. All the same, the snapshot of the assemblage seems to me to show that there is promise in the idea that genders are fundamentally aesthetic entities.

If so, the work to be done is fulfilling that promise. Besides saying more about genders, one might wonder about other routes to fulfilment. Here are two. First, can the model adapt to other people kinds besides genders? I think so: I think that many people kinds are fundamentally constituted by aesthetic practice as much as they are by other kinds of practice. Which kinds are so constituted, and how exactly are, is a question that deserves consideration.⁴⁵ Second, can the model contribute to amelioration of the deleterious effects of hierarchies of genders? Again, I think so. Charles Mills and Paul C. Taylor both argue that recognizing the constitutive role of aesthetic practice in the constitution of races is a necessary step towards amelioration (Mills 1997: 61–62; Taylor 2015: ch. 4).⁴⁶ So it goes with gender, too; and perhaps, so it goes with many of the social kinds by means of which people are subordinated. The sooner we see clearly the constitutive role of aesthetic practices in people kinds, the sooner we can change them for the better.