Agency and aesthetic identity

Abstract

Schiller says that “it is only through beauty that man makes his way to freedom.” Here I attempt to defend a claim in the same spirit as Schiller’s but by some different means. My thesis is that a person’s autonomous agency depends on their adopting an aesthetic identity. To act, we need to don contingent features of agency, things that structure our practical thought and explain what we do in very general terms but are neither universal nor necessary features of agency as such. Without these things, the question of what to do for any individual would be underdetermined. The problem is that adopting such a contingent form of agency amounts to a restriction on what we can do, and so it is a prima facie threat to our autonomy. I will argue that one way, and indeed the only way, of meeting this challenge lies in aesthetic experience. Granting our capacity for aesthetic pleasure the authority to determine the particularities of our agency is compatible with autonomy because doing so means identifying with one’s capacity for pleasure in free and creative activity. Doing this allows us to be both particular sorts of agents and creatures regulated by the universalizing demands of reason.

“It is only through beauty that man makes his way to freedom.” So Friedrich Schiller says.¹ For him agency confronts us with a problem of rapprochement. Even if we are free and rational wills, we are also corporeal, receptive creatures. And for this reason, our freedom requires something more than the subjugation of our sensible nature to its rational sibling. It requires the harmony of our natures. And here it is natural to think, as Schiller does, that our aesthetic powers—to appreciate beauty, to invent, to play—will have a pivotal role in this reconciliation because they combine features that characterize each half of our nature. Aesthetic experience is at once spontaneous and receptive.

Nowadays this proposition is not so much rejected as ignored. Most interesting forms of freedom, contemporary philosophers suppose, can be had in ways that make no recourse whatsoever to our aesthetic capacities. The aesthetic is a non sequitur when it comes to questions about agency and freedom. My goal in this paper is to show that Schiller was right, more or less. We really do face something like his problem, and he is right to think that our aesthetic powers are the key to solving it. That said, my argument is different from Schiller’s, and I will not spend much time addressing his efforts.  Above all, my argument will differ by being rooted in a more contemporary picture of moral psychology.² The argument in a nutshell goes like this. To act we need to don contingent features of agency: things that structure our practical thought and explain what we do in very general terms but are neither universal nor necessary features of agency. Without these things, the question of what to do for any individual will be woefully underdetermined. But this need brings a challenge in tow. Agency also requires that these contingent features be justified by the right sorts of reasons. For any given agent, I will argue, we should expect these reasons to include ones having to do with what pleases and displeases them because these are the only kind of reasons that are plentiful enough to justify relatively mundane choices about how to configure our agency. Unfortunately, many kinds of pleasure are too mechanical and too pathological to shoulder the burden of agent-constitution, and so identifying with them would threaten our autonomy. The exception, I suggest, is aesthetic pleasure. It is suited to this role because it is pleasure taken in free, creative, and self-directed activity. In short, agents must identify with their capacity for aesthetic pleasure because doing so is the only way to justify the particular and contingent features of their agency consonant with the general demands of autonomy.

1 Contingent features of agency

We do not encounter the world as formless agents. We have commitments that distinguish us as individuals and make us particular kinds of agents. These include our projects, cares, allegiances, skills, statuses, identities, and ideals. Naturally, some features of my agency do arise from the nature of agency “as such”, just as some features of The Dying Gaul come from the nature of sculpture “as such”, but much of the assemblage of powers that make up my agency come from elsewhere.

Culture is an especially obvious source of these features. The monk will develop powers and predilections that enable him to live seamlessly in a highly ordered religious community. The frontiersman will do the same, but for powers and predilections tailored to rough living and self-sufficiency. Likewise for the Roman slave, the medieval vassal, the British tar, the office bureaucrat, the cyberpunk, and the professional philosopher. One example of this kind of adaptation involves what Sally Haslanger calls psychological schemata.

Schemas consist in clusters of culturally shared concepts, beliefs, and other attitudes that enable us to interpret and organize information and coordinate action, thought, and affect. Schemas are public—think of them as social meanings conventionally associated with things in our social world, including language—but are also internalized and guide behavior.³

The monk has internalized one set of psychological schemata, and these allow him to fluently participate in his abbey. The frontiersman has internalized a different set of schemata that allow him to thrive in the wilderness.

These differences will affect each agent’s practical psychology by structuring what we might call the background framework of their practical thought. For example, they may exclude otherwise possible actions from consideration, while elevating others to default status. There are myriad actions that the monk wouldn’t even entertain during mass, and, generally, one specific action that occurs to him as a default. They may also link certain kinds of choices to characteristic ways of making them. The frontiersman and the Roman slave both frequently find themselves deliberating about whether to kill an animal, but those deliberations will unfold according to very different templates. They will also have perceptual effects. The monk will see his ritual objects as possessing a normative status that others do not naturally observe.

In many respects, schemata function like practical versions of Thomas Kuhn’s famous paradigms. For Kuhn, paradigms offer a conception of which problems a scientific enterprise needs to solve and of what a solution to those problems would look like. They suggest a methodology, in the form of explicit rules or exemplary experiments. They promote particular experimental apparatuses and instruments. They help articulate otherwise vague scientific values and ideals. And they fix common ground assumptions about the structure of part of the world being studied. Paradigms function as background frameworks for scientific thought; schemata do something similar for practical thought.

Once we recognize this background structuring, we can see that many other kinds of mental attitudes are capable of the same kind of work. These include all the familiar types of commitment: projects, cares, allegiances, skills, statuses, identities, and ideals. My love for my daughter structures my practical thinking in particular ways, and in doing so, it gives my agency a particular shape. Ground projects, like becoming a famous accordion player or living out the ideal of Christian compassion do the same. Going forward, it will be useful to have a name for the things I am talking about, for things whose adoption structures the background framework of our practical thought and in doing so modifies our generic power of agency. I will call these accidents of agency. (In the scholastic sense of “accident” that contrasts with “essence”.)

Accidents of agency should be distinguished from things like beliefs and desires. An agent might possess these things, but we should not, under normal circumstances, say that they modify their agency because they do not ordinarily affect the framework of our practical thought.⁴ And they should also be distinguished from essential features of agency, like a disposition to adhere to the enkratic principle or maintain means-end coherence. These things are features of agency but they are not accidents.

My claim is that we need to take on accidents of agency to be agents at all. I cannot be a generic agent; I must be a particular kind of agent with a particular contingent practical nature. My reason for believing this is, in effect, an underdetermination argument. Without these accidents of agency, our practical thought would be too unstructured for the enterprise of practical reasoning to get off the ground. The point parallels one Kuhn makes about paradigms:

In the absence of a paradigm or some candidate for paradigm, all of the facts that could possibly pertain to the development of a given science are likely to seem equally relevant. As a result, early fact-gathering is a far more nearly random activity than the one that subsequent scientific development makes familiar. … But though this sort of fact-collecting has been essential to the origin of many significant sciences, anyone who examines, for example, Pliny’s encyclopedic writings or the Baconian natural histories of the seventeenth century will discover that it produces a morass. One somehow hesitates to call the literature that results scientific. The Baconian “histories” of heat, color, wind, mining, and so on, are filled with information, some of it recondite. But they juxtapose facts that will later prove revealing (e.g., heating by mixture) with others (e.g., the warmth of dung heaps) that will for some time remain too complex to be integrated with theory at all.⁵

We must adopt accidents of agency for much the same reason. Without schemata, projects, skills, statuses, identities, ideals, and other accidents of agency, practical agents would find the world too diffuse, too disorganized, and too much of a morass of undistinguished facts for them to profitably engage with. It would, for example, be difficult for such a creature to set even the most basic ends or pose the most rudimentary practical questions without a framework that provides the language for those ends and questions. In other words, we need to articulate our agency in contingent ways because that is the only way to operationalize the vague problem put to us by the possibility of action.⁶

2 Autonomy and constraint

This necessity creates a problem. The same background framework that enables our agency also restrains it. If my practical thought is structured by certain commitments, then I cannot help but be moved by those commitments. The is a very general point. A sculpture needs a contingent nature, a form, if it’s going to be a sculpture instead of a lump of clay. But this contingent nature is also a restriction on the nature of the sculpture. A bust of Helen cannot also be a bust of Dido.

At an extreme this constraint will make certain conduct impossible. If love for my daughter is built into the very structure of my practical thought, then I cannot help but see her welfare as mattering, and I could never seriously entertain an intention to harm her. Harry Frankfurt calls these effects “volitional necessities,” things we have no choice but to do, not because we lack the power not to do them but because the “configuration” of our will makes it impossible for us to exert that power.⁷ If certain commitments create volitional necessities, then it seems they can also stymie our autonomy. There will be things, like neglecting my daughter, that I cannot elect to do. There will be others, like caring about her, that I cannot help but do. And there will be a vast number of subtle inflections to our agency that don’t quite amount to volitional necessities but still reflect the influence of this structure. I can suppress my partiality for my daughter, but it takes an effort that I am not always prepared to exert.

Another way to put the problem comes from Kant. If the will, Kant says, “seeks the law in the constitution of any of its objects,” then “heteronomy” follows. This is because in such a case “the will does not give itself the law but the object through its relation to the will gives the law to it.”⁸ By “object” of the will, Kant seems to mean the inputs of our practical reasoning rather than the apparatus that carries out the reasoning. Inasmuch as my commitment to my daughter is something that I consider in practical reasoning, rather than part of the thing that does the reasoning, it is an object of the will. And so if this commitment “gives the law” to the will—if it is the ultimate decider, we might say—then I am being controlled by something external to my own agency. And that is a failure of autonomy.

This would seem to represent a conflict between the demands of agency. We need to adopt a contingent practical nature to constitute ourselves as particular agents but doing this also seems to constrain us. Frankfurt suggests that this conflict is an illusion, at least some of the time. “Even though a person’s interests are contingent,” he says, “they can belong to the essential nature of his will,” and when a person is governed by their will’s “essential nature”, they are governed by themselves.⁹ The idea is that a person can be fully autonomous despite the sorts of constraints we saw above if those constraints reflect their personal volitional “essence”—if the commitments, projects, and ideals that impose those constraints also constitute the person’s nature as a practical creature.

Frankfurt’s suggestion is on the right track, but it stumbles over a critical feature of human agency. We have the ability to regulate ourselves through the application of reflective scrutiny. We can reflect on our beliefs, desires, intentions, commitments, and ideals to ask whether we should “stand behind” them—whether we should endorse or reject them. We can do the same for what I have called accidents of agency. If we couldn’t reflectively scrutinize ourselves in this way, then it seems unlikely that genuine self-determination would be possible. This ability suggests a difficulty for Frankfurt’s notion of “volitional essence”. Imagine a person who has internalized a monkish schema and so cannot help but relate to the world as a Benedictine does, but who also thinks that there are no good reasons to abide by The Rule of St. Benedict. This is a person alienated from their particular agential capacities in a way that plainly diminishes their autonomy. There is, however, an obvious amendment to Frankfurt’s suggestion that will allow it to accommodate this case. For a contingent feature of a person’s practical nature to be part of a person’s volitional essence in a sense that enables autonomous action, they must endorse that feature. More specifically, they must take themselves to have adequate reasons to adopt or maintain that feature.¹⁰

What I have suggested, then, is that there is a prima facie conflict between autonomy and the indispensable accidents of agency. This conflict can be resolved, in something along the lines of Frankfurt’s proposal, but only if we can justify our accidents of agency in the face of reflective scrutiny. The upshot of this is that agents need good reasons to adopt new accidents of agency or carry on with ones they already have.

3 Hedonic identity

What kind of reasons will we turn to in justifying our contingent nature as a practical creature? Many sorts, naturally. But I am interested in arguing that a particular sort of reason must show up in these justifications. At least some of these reasons, I want to argue, will be grounded in hedonic facts. That is, part of the justification of every agent’s contingent practical nature will ultimately rely on facts about what they like or dislike, what they relish or despise, what pleases or displeases them.

I will call this the Hedonic Justification Thesis. It comes with an important corollary. If I do as the Hedonic Justification Thesis suggests and justify my status as a certain kind of agent by citing a corresponding form of enjoyment, then there is an obvious sense in which I am identifying with that enjoyment. I am, in effect, granting a particular capacity for pleasure the authority to settle questions about which accidents of agency to adopt and how to structure my practical thought. I am granting it the authority to determine what kind of practical creature I am. To do this is to adopt a hedonic identity. The corollary, then, is that agency requires the adoption of a hedonic identity in this sense.¹¹

Before defending these two claims, I need to offer one piece of clarification. Both claims depend on a very broad conception of the hedonic. Pleasure, as I understand it, will not be a distinctive kind of sensation. It is a class of experiences unified by our normal attitude toward them. It will consist, that is, in the experiences we are inclined to love, relish, or enjoy. Hedonic facts, on this view, are facts about our dispositions to these attitudes and their opposites. This is a more liberal conception of the hedonic than, for example, one that takes particular sensory pleasures as paradigms.¹² Understanding pleasure in this way allows us to countenance sophisticated and unorthodox forms of pleasure—to say, for example, say that people go to horror movies because they enjoy them, even if that enjoyment is partially constituted by feelings of terror.¹³ It also allows us to see how even demanding vocations can be justified by a person’s hedonic identity. Parenthood involves taking on certain accidents of agency. And whatever reasons I had for becoming a parent, my agency is now very much structured by these accidents. So I face the question: what reasons do I have for carrying on with parent-agency? Part of my answer is that I enjoy the kinds of activity and perception that it enables, that I enjoy my children in a distinctively parental fashion. This is probably not my only answer, and it by no means suggests that parenthood is free of pain and sacrifice. Nonetheless, for many parents the unique forms of enjoyment found in the exercise of parent-agency will be an important part of the justification for persisting in that form of agency.

Now let’s turn to the case for the Hedonic Justification Thesis. Importantly, this thesis does not just claim that we can justify accidents of agency with hedonic facts but that we must. My argument here, like my earlier one for the necessity of accidents of agency, is an argument from underdetermination. I grant, naturally, that many elements of our contingent practical nature will be justified by non-hedonic facts. For example, we may take some to be demanded by morality. (I endorse my powers of empathy because I think they make me a better person.) Some we may think are non-moral but nonetheless objective excellences. (I endorse my cat-like reflexes because I think they make me a better human animal.) Others will be grounded in non-hedonic facts about ourselves. (I may have a neurological condition that makes it prudent to cultivate certain habits of attention.) But I think these kinds of consideration are only going to do part of the job of specifying our contingent practical nature because some elements of that nature are simply too trivial or too personal to be justified by such considerations.

This is best shown with examples. Suppose you decide that it would be appropriate to form social attachments and in doing so develop the accidents of agency characteristic of love. But this raises the thorny question of to whom you should attach yourself. Some people may be ruled out by morality and prudence. Maybe others will be ruled out by points about compatibility, feasibility, and so on. But this still leaves a large class of individuals who are all reasonable candidates for your love. You can’t love all of them, and here it seems only natural that you will turn to hedonic facts to decide the question. Thus you might think to yourself, “I so enjoy spending time with Phoebe”, or “I find Tim simply charming”, or “Arnold is a gentle person, but he gets on my nerves.” This kind of recourse, I think, seems almost inevitable.

Now suppose you elect to fashion yourself as a great athlete with all the ensuant agential powers. But once again, you have choices. What kind of athlete? Should you do gymnastics or softball or horseshoes? Should you play shortstop or pitch? Do the balance beam or the vault? Some of these questions may be settled by facts about what you have a better knack for. And perhaps morality rules out some of the rougher sports. But once more, these facts will eventually run out, and we will have to turn to turn to other, less lofty considerations. These, I submit, will likely include hedonic facts: “I enjoy softball more than gymnastics, but playing shortstop is nerve-wracking. I think I would most like being a left fielder.” This is a perfectly reasonable way to justify one’s deciding to be a left fielder rather than a shortstop or a vaulter. And insofar as playing left field will involve taking on distinctive kinds of agential powers, those powers will be justified by the same hedonic considerations. Once again, this seems unavoidable. I cannot imagine justifying my choice to be a left fielder from the point of view of the universe.

In these examples, we have to justify the adoption of rather personal, particular, and, from a certain point of view, mundane features of ourselves. These examples suggest four interlocking reasons for why we should expect to always find hedonic facts amidst these justifications. First, there is an immense amount of work to be done in determining all the details of our contingent practical nature, the full breadth of which only emerges in the light of contrast questions like, “why do I want to be a softball-style athlete instead of a gymnastics-style athlete?” and, “why am I attached to Phoebe instead of Arnold?” Second, the classes of reason we have mentioned—morality, prudence, the perfection of the human organism—are generally too blunt to answer these questions. They are not going to tell me anything about the relative merits of softball over and against gymnastics. Third, hedonic facts can do this work because pleasure and pain are ubiquitous and reflect fine discriminations. It is not at all surprising to hear that someone enjoys spending time with Tim more than Phoebe, even if they are, in the grand scheme of things, very similar. (Medium-sized bipeds with decent manners.) This is because our enjoyment can vary significantly even in response to relatively similar stimuli. Our capacity for pleasure is, in this sense, capable of quite fine discrimination. Fourth, hedonic facts have a kind of default, if defeasible, normative force. They are prima facie reasons because, as Anthony Kenny says, “it is always silly to ask a man why he wants pleasure.”¹⁴

Is there anything else that can do similar work? Perhaps. Desire can be fine-grained in the same way hedonic facts are, and they also support prima facie justification. I desire Phoebe’s company but not Arnold’s, and that may be a reason to be friends with her rather than him. This convergence is hardly surprising since desire and pleasure usually go together. We want what we like, and we like what we want. But sometimes the two do come apart. The addict, for example, may desire a drug without liking it.¹⁵ And a person may enjoy an experience while having the corresponding desire repressed by a shadowy psychological mechanism. Importantly, in these cases the normative force seems to follow pleasure, not desire. The addict’s desire does not give them a reason to take the drug precisely because they get nothing out of it, while the repressed person’s potential for enjoyment does give them a reason, even absent the corresponding desire. For this reason, I think pleasure is a better candidate for the work I have identified than desire.

4 Identifying with agreeableness

Agency requires us to identify with some aspect of our capacity for pleasure and pain in the sense of granting it authority to determine features of our contingent practical nature. That is what the Hedonic Justification Thesis says, and it is what I have argued so far. This brings us to our final and most important question. Are all kinds of pleasure equally fit for this work? Or are some compatible with agency while others are not?

I will argue that one kind of pleasure is fit for identification, but another is not. The unfit kind of pleasure involves what I will call, borrowing Kant’s word, the agreeable. Here are some paradigms: the taste of sweet food, the sound of a E-flat major triad, the sight of International Klein Blue, the texture of silk, and the smell of an expensive new car. These examples have a few things in common. First, they are pleasures of the senses, and it is natural to talk about them as aroused by the interaction of a class of object with our sense organs. (“The agreeable,” Kant says, “is that which pleases the senses in sensation.”¹⁶) Second, agreeable pleasure is “interested” insofar as “through sensation it excites a desire for objects of the same sort.”¹⁷ Another way to put this point is that the agreeable is intrinsically conative. There is a constitutive connection between the pleasure of the agreeable and a motivation to engage something suitably related to it. “What directly (through sense) urges me to leave my state (to go out of it) is disagreeable to me,” Kant says: “it causes me pain; just as what drives me to maintain my state (to remain in it) is agreeable to me, I enjoy it.”¹⁸

These features are certainly symptoms of the agreeable, but I don’t think they characterize it. That role falls to a different property that Kant also identifies. The agreeable, he says, is a “pathologically conditioned satisfaction.”¹⁹ For my purposes, this “pathological conditioning” is the essence of the agreeable. What Kant means by this phrase, I think, is that these pleasures are the result of a mechanism. Some stimulus s acts on me, and through a pathway p a corresponding mental state m arises. The state m is at once a phenomenal state (a qualitative feel) and an intentional attitude toward this feeling—liking, disliking, enjoying. If m involves a pro-attitude such as liking, then we say that s is pleasant. A pleasure is agreeable just in case the relationship between s, m, and p is sufficiently rigid—if qualitatively similar stimuli s always produce the same states m and do so according to a single, invariable pathway p. For example, imagine that there is a simple machine fitted to my palate that detects whether a substance has pH between 6 and 7. If I ingest such a substance, the same machinery sends electrical signals to my brain that put me in a state that is constituted by a very particular sensation (“mild acidity”) and a pro-attitude toward that sensation. This arrangement is a model of agreeableness because it involves a strong correspondence between a particular stimulus s and a particular kind of enjoyment m, and this correspondence is grounded in a rigid mechanism.

The paradigms of agreeableness I mentioned work the same way. When I hear an E-flat major triad, the sound waves strike my ear drum and initiate a process that extends up my auditory nerve and eventuates in a particular state of enjoyment. As long as this mechanism is working, the major triad will provide me with agreeable sonority. If someone asks me, “why did you respond to that stimulus with that state?” or “why were you pleased by what you heard?” I would say that they didn’t understand how this kind of pleasure works. “It’s not up to me; it’s just how things work for me.”

The important question for our purposes is whether we can identify with these agreeableness facts, whether it is possible to grant them authority over certain durable features of our agency. I think we can. I find certain experiences especially agreeable: the pleasures of working through an abstract problem, the pleasures of winning an argument, the pleasures of an audience’s rapt attention, the pleasures of intellectual prestige. I come to believe that a certain life—maybe a philosophy professor, maybe a lawyer—is likely to produce those pleasures. This justifies me in doing certain overt things—maybe going to law school, maybe reading more Kant. But it also justifies me in fashioning, or at least endorsing, certain accidents of agency on the grounds that they are conducive to that sort of life. If I am going to be a philosopher, I will need to have certain habits of attention and reflection, certain forms of patience, and a tolerance for uncertainty. The right hedonic identity can justify the cultivation of all these things.

So this kind of justification is certainly possible. Indeed, it is probably quite common. Nonetheless, I want to argue that someone who identifies with what they find agreeable in this way is less autonomous for it. Three features of agreeableness are relevant to my indictment.

(i) We are passive before our agreeable experiences. Even if I identify with my disposition to find some stimulus agreeable, I do not experience the production of this pleasure as something I do. When I taste something sweet, the pleasure of that experience is something that washes over me, not something I produce through my activity. I am the patient of the experience, not the agent. Of course, I may do things to induce a stimulus—focusing my attention on a major triad, looking more closely at the painting, cleansing my palate so I can taste the notes of honeysuckle. But after I have made myself receptive in these ways, there is nothing more for me to do. The pathway that mediates stimulus and experience will unfold quite independently of my will.

(ii) The course of agreeable experiences is not something I can control. This follows from the determinism of the underlying mechanism. A relatively narrow range of stimuli s will reliably produce a correspondingly narrow range of pleasure states m, and it will do so according to the same pathway p. There is no possibility that the same s will produce a different mental state n, nor that it could produce it according to some substantially different pathway q. If it did, it would be a different kind of agreeableness. This means, among other things, that there is scarcely any opportunity for me to control how my pleasure unfolds.

(iii) Our finding a particular experience agreeable in a particular way is not open to critical scrutiny. I like the sweetness of sugar. But if someone challenges this reaction or asks me to justify it, I will think that they misunderstand the nature of my pleasure. Why do I like it? I just do; it’s a brute fact about me. I cannot justify my enjoyment, and that enjoyment would not diminish in the face of effective criticism. (I could have reasons to modify my sense organs in some way to change which things I find agreeable, but that’s a different story.) For these reasons, I would not conceive of a divergent response to sugar as a disagreement but a difference.

What do these three points have to do with autonomy? Autonomy is more than just independence from external forces. It is a capacity that entails non-trivial responsibilities, the most obvious of which is responsibility for oneself. Autonomy requires I be the ultimate ground for my own activity and be capable of justifying important aspects of it, and for this reason, my autonomy can be diminished just as readily by laziness, diffidence, or inappropriate deference as by interference. I think Andrea Westlund has this conception in mind when she says that “to treat someone as autonomous is to treat her as her own representative,” and “treating someone as her own representative means treating her and no one else as the one from whom we are ultimately to seek answers to questions that arise about the commitments that guide her conduct.”²⁰ Westlund uses this conception of autonomy to explain some familiar cases. For example, Thomas E. Hill’s “deferential wife”:

She buys the clothes he prefers, invites the guests he wants to entertain, and makes love whenever he is in the mood. She willingly moves to a new city in order for him to have a more attractive job, counting her own friendships and geographical preferences insignificant by comparison. She loves her husband, but her conduct is not simply an expression of love. She is happy, but she does not subordinate herself as a means to happiness. She does not simply defer to her husband in certain spheres as a trade-off for his deference in other spheres. On the contrary, she tends not to form her own interests, values, and ideals, and when she does, she counts them as less important than her husband’s.²¹

This person seems less than fully autonomous despite identifying with her servility, and Westlund’s conception can help us see why. In deferring authority for such central features of her life to her husband, she fails to exercise ultimate responsibility for essential characteristics of herself.

I want to suggest that the person who identifies with their disposition to agreeableness will suffer from a similar problem. Imagine I have cultivated certain accidents of agency because doing so will be agreeable to me. I have, for example, become a lawyerly sort because beating people at arguments makes me feel big, and that pleases me. In doing this, I have given a disposition to feel pleasure in winning arguments authority to determine what kind of person I am. There are two reasons to worry that I have shirked responsibility for myself in doing this. First, because I am passive before agreeable pleasures and these pleasures unfold deterministically, I cannot claim that I exercise authority over myself through them. I cannot claim that I am controlling myself by allowing my disposition to pleasure to configure my contingent nature as an agent because this disposition is not at all responsive to me or my will. It has its own nature and tendencies, quite independently of my opinions. This, of course, parallels Hill’s case: the deferential wife cannot claim that her deference is a way of her exercising control over herself because her husband makes up his own mind. Second, because pleasures of agreeableness are not responsive to critical scrutiny, the justification I can offer for this identification will be fatally cut short. I can say that I have fashioned myself as a lawyerly creature because my dispositions to pleasure recommend it—recommend it in the sense of producing pleasure in some contexts and not others. But if I ask why this recommendation has been made, why certain things please me but others don’t, I am stymied. The question either doesn’t make sense or it leads us to a pathological answer that in no way justifies my choice. This is an effect of the agreeable’s immunity to critical scrutiny. This reason also parallels Hill’s case. The deferential wife can ask why her husband has recommended a certain course of action, but any answer she comes up with will not be the right sort of reason. It will not be a reason for her to do something but for her husband to do something.²² Thus the person we are imagining is limited in both the self-control they exercise and in their ability to justify themselves, and this is why I think they are unable to fully take responsibility for the particular sort of agent they have made themselves.

To be clear, the problem I am alleging is not with the act of identification and the motives that precipitate it. Nor is it with the intrinsic character of the agreeable. The problem, as I see it, is with the relationship that the agent bears to her dispositions to agreeableness. She does not exercise the right kinds of control over the experience of agreeableness, and she does not stand in the appropriate relationship to facts that explain that agreeableness for those facts to be justifications that she can offer. And that makes it difficult for her to take responsibility for the things that follow from this identification. In making this identification, an agent is giving herself over to something that cannot really be hers.

5 Aesthetic pleasure

Looking at this argument, we might be tempted to conclude that all hedonic identification will involve some shirking of responsibility because all pleasure has the problematic features I identified. But this would be a mistake, for there is another species of pleasure, aesthetic pleasure, and it differs from agreeableness on all three fronts.

(i) Aesthetic pleasure is pleasure in an activity. Mohan Mathen says exactly this and a bit more when he argues that:

Aesthetic pleasure comes from contemplating something intellectually and, in the case of visual and performing arts, perceptually as well—focusing on the object and its properties. To get aesthetic pleasure from a painting is to enjoy looking at it and thinking about it. A novel gives pleasure by engaging the mind: we enjoy its narrative in the context of its themes, its moral stance, its form, the style and articulateness of its presentation.²³

These are just two examples, but we can multiply them endlessly. We enjoy art by looking for connections, patterns, and contrasts, by rehearsing the artist’s choices and wondering why they did as they did, by homing in on design and significant form, by trying to understand, contextualize, and challenge the work. The pleasure of art, and aesthetic pleasure more generally, is not pleasure in the passive reception of experience, but in mental activity.

(ii) The course of this activity is not determined “pathologically”—by some identifiable causal mechanism. Pathological determination was one of the hallmarks of the agreeable: the course of one’s enjoyment was determined in advance by the causal properties of a stimulus and a corresponding pathway. Our aesthetic activities are very different. We cannot, in particular, say in advance how they will go, or even how they must not go. We can make some crude generalizations like, “it’d be very unusual to be amused by the subtle wit of Tosca”, but we cannot completely rule out the possibility of such a response, nor say with absolute certainty that it would be unmerited. There is, then, a certain indeterminacy in our experience of aesthetic pleasure. This indeterminacy need not entail that all the features of my aesthetic pleasure will be free from pathological mechanisms. After all, a painting can include Klein Blue and an opera can have lots of E-flat major triads. The point is that aesthetic contemplation of these objects will be more than an agglomeration of these qualities, and so the pleasure of that contemplation will not be wholly explicable in terms of the operation of these mechanisms: there will be room for something more spontaneous to intervene.²⁴ Nor does this mean aesthetic pleasure somehow stands outside the causal order of the universe: the absence of a rigid causal regularity between two classes of events does not mean that there is no causal relationship between them, just that that relationship does not rise to the level of a mechanism.

(iii) Aesthetic pleasure is open to critical scrutiny. The essence of critical scrutiny is the ability to stand back from one’s attitudes for the sake of considering whether to continue holding them.²⁵ I can scrutinize an attitude just insofar as I can take up this critical distance on it. This is something I can do for my aesthetic pleasures. I once took great pleasure in the poetry of William McGonagall. But you showed me that it is actually quite bad. The rhymes are primitive, the imagery is clichéd, and the meter feels like the gallop of a dying horse. Your comments helped me see that my pleasure in McGonagall’s poetry was unmerited, and, as a result, the pleasure evaporated. I once thought that a Pillsbury commercial was a great work of cinema. I took tremendous pleasure in watching piping hot rolls pulled from the oven. But you showed me that my enjoyment came from the thought of eating the rolls, not the commercial itself. And so I came to realize that my pleasure was not aesthetic, but something else. I never liked Arvo Pärt’s music, but then you showed me how to listen to it, and I now take great pleasure in doing so. In each of these examples the usual forms of criticism we can subject our aesthetic judgments to offers a means by which we can reflect on and scrutinize our putative aesthetic pleasures.²⁶

Here one might wonder how the transformation of judgment we see in these examples differs from similar transitions involving the agreeable. Isn’t my discovery of the beauty of Arvo Pärt’s music rather like my discovery that bourbon is delicious or that Klein Blue is appealing (assuming these are examples of the agreeable)? Isn’t it an example of my opinion changing because I gain a perceptual capacity that sensitizes me to the pleasure-producing properties of these objects? In fact, this marks a crucial difference between the aesthetic and the agreeable. Because pleasure in the aesthetic is pleasure in mental activities, my change in view about Pärt need not involve a change in my perceptual state. (Though of course it can.) It can instead reflect a change in what I do with that state, in the reflective and contemplative activities I undertake in response to it. This means that the control effected by critical scrutiny on aesthetic pleasure is unmediated in an important sense. It is not just the result of manipulating our perceptual faculties to better align with an opportunity for pleasure, but a direct response to my judgments about how to respond to a stimulus. And this constitutes an important form of control over my ability to experience aesthetic pleasure. It also means that the explanation for why I take pleasure in a given object will amount to rather more than brute facts about which perceptual states happen to produce pleasure in me and my being in those states. It will include, for example, familiar sorts of aesthetic reasons: facts about the quality of an object that serve to justify me in taking aesthetic pleasure in it.

These features of aesthetic pleasure immunize it against the arguments I made against identification with agreeableness. It would seem that we can exercise control some control over our capacity for aesthetic pleasure, and we can justify ourselves in doing so. This gives us a prima facie argument that we can indeed take responsibility for the self created in response to our capacity for aesthetic pleasure.

But only a prima facie case. Just because aesthetic pleasure slips through the objections I put to identifying with the agreeable doesn’t mean that we can autonomously constitute ourselves by identifying with it. And indeed, aesthetic pleasure does come with some features that should give us pause. Do we really have enough control over the course of aesthetic experience to exercise this kind of responsibility? On the one hand, I exercise certain familiar forms of control over my aesthetic engagement. I can attend to certain features and ignore others. I can approach an object with certain questions. I can deliberately emphasize certain connections and try to suppress others. I can choose one interpretive framework rather than another. On the other hand, this control is limited in equally obvious ways. I certainly cannot decide to take pleasure in an aesthetic object the way I can decide to wiggle my fingers. So even if there is daylight between our relationship to aesthetic pleasure and agreeableness, I owe a positive account of why aesthetic pleasure fitness for the role I have described.

To do this, let’s consider the limitations on our control of our aesthetic responses. These come in at least three varieties. First, there are limitations related to the object. Even if an aesthetic object does not determine my response in the rigid way that it does for agreeableness, it will still vastly limit what kind of response I am capable of. I cannot simply decide to take pleasure in the subtle wit of Tosca. Second, there are limitations of skill. Certain forms of aesthetic engagement depend on the possession of particular skills, and whether I have those skills is not a question of will. I can appreciate the wordplay and historical allusions in Dickens, but the same appreciation of Tolstoy is impossible for me because I don’t read Russian. Third, there are limitations related to the inherent creativity of aesthetic experience. In distinguishing aesthetic pleasure from agreeableness, I have emphasized the former’s irregularity: aesthetic pleasure is neither rigid nor mechanical. One reason for this, the principal reason, is that aesthetic appreciation is a creative endeavor. Aesthetic experience is spontaneous and unrehearsed—it is not something we undertake by following rules, forming the right kind of intention, or carrying out a plan.²⁷ Insofar as these things are amongst our most reliable devices of agential control, the creativity of aesthetic experience will entail limits to that control.

I think we can take responsibility for dispositions to aesthetic pleasures despite all this. My argument is by analogy. The production of art displays exactly the same limitations of control, but we readily accept that artists can take responsibility for their work in a way that supports the autonomy of that production. Take portrait painters: they are constrained in innumerable ways: by fidelity to their sitter, by fashion, by the demands of their audience. They are also limited by their own skills in obvious ways. There are things the painter can do and things they simply cannot. Finally, if the portrait is any good, there will be some element of creativity to it—some qualities that do not reflect anything like an antecedent plan or design. Despite these limitations, a portraitist can take responsibility for their work in the ordinary way we take responsibility for any of our actions. They can say I did this; I am answerable for it; it is part of who I am as a painter.²⁸

This analogy gives us one final wedge to drive between aesthetic pleasure and the agreeable. Aesthetic pleasure is, in these respects, rather like artmaking, and artmaking is something we can take responsibility for. It is constrained in myriad ways, by subject matter, by skill, by the vicissitudes of creativity, but the artist’s relationship to it is an active one, one which has a meaningful effect on the product and is guided by a distinctive kind of critical scrutiny. The same cannot be said for the agreeable. Identifying with your capacity for aesthetic nature in the way I have suggested is not a way of shirking responsibility for oneself, as identification with the agreeable is. It is a distinctive way of taking responsibility for oneself, one much like the way the artist takes responsibility for their art.

6 Aesthetic identity

Taking on an aesthetic identity will mean granting one’s capacity for aesthetic pleasure authority in choosing fundamental and durable features of agency, including things that affect the structure of one’s practical thought. This kind of justification is not unfamiliar. Aesthetic experience is frequently thought to involve a distinctive kind of authority—to make demands on us on us that other sorts of experience don’t. As Richard Moran notes, “we often speak of the beautiful in terms of something appealing to us or demanding of our attention … and the pleasure we may experience in something beautiful raises the issue of its calling for that pleasure or that attention, in a way that does not apply to other things that may arrest our attention or gratify our senses.”²⁹ As an example, Moran offers a passage from Proust:

That year my family fixed the day of their return to Paris rather earlier than usual. On the morning of our departure … my mother, after searching everywhere for me, found me standing in tears on the steep little path close to Tansonville, bidding farewell to my hawthorns, clasping their sharp branches in my arms. … “Oh, my poor little hawthorns,” I was assuring them through my sobs, “it isn’t you who want to make me unhappy, to force me to leave you. You, you’ve never done me any harm. So I shall always love you.” And, drying my eyes, I promised them … I would never copy the foolish example of other men, but that even in Paris, on fine spring days, instead of paying calls and listening to silly talk, I would set off for the country to see the first hawthorn-trees in bloom.³⁰

In this passage, the narrator—call him Marcel—finds beauty in the hawthorns, and this discovery leads him, in Moran’s words, to place “himself under an obligation, something like the vow to remain responsive to this beauty.” Marcel is “promising that he will be true to [the hawthorns] in spite of the pleasures and distractions of ‘paying calls and listening to silly talk,’” and in doing this “he is both affirming the superior value of the hawthorns and attempting to bind himself to that value.”

But how does this self-binding work? Moran suggests that it comes about through Marcel’s identification with the experience of the hawthorns. The magnetism he experiences when beholding the hawthorns comes to “define” Marcel in such a way that “were he to lose his responsiveness to their appeal, it would count as a loss of the self he presently is and cares about.”³¹ And this, Moran says, is what distinguishes beauty from the merely agreeable. Losing one’s preference for tea or coffee typically means finding “something else more agreeable or equally so and mov[ing] on. There need be no experience of the loss of some part of oneself, nothing to mourn or regret in this change of tastes.” But losing one’s interest in beauty means losing part of oneself.³²

To understand Moran’s suggestion, we need a better idea of what it means to say that were Marcel “to lose his responsiveness to [the hawthorns’] appeal, it would count as a loss of the self he presently is and cares about.” How should we understand this “loss of the self”? One gloss has it that Marcel “loses his self” whenever he ceases to possess some capacity or disposition that he previously valued. Presently Marcel approves of his responsiveness to the hawthorns, but if later he becomes unresponsive to them, then he has lost his self. This proposal is too weak: we shed these sorts of properties far more frequently than we can be said to really “lose” ourselves. Another interpretation understands the “loss of self” in terms of personal identity. Marcel’s responsiveness to the hawthorns is a constitutive part of who he is, so if he loses it, he becomes literally a different person. This proposal seems much too strong: we lapse from aesthetic infatuations more frequently than we cease being one person and become—literally—someone else.

In between these two extreme ways of understanding the “loss of the self” there is a better conception grounded in the agential machinery I have been relying on. Marcel’s responsiveness to the hawthorns is part of his self in the sense that it reflects accidents of his agency that shape his practical thought. For Marcel, perceptions of hawthorns come with a certain halo. Hawthorns are to-be-admired and to-be-contemplated. He is inclined to suspend the more mundanely prudential or scientific way he approaches the world when he comes to the hawthorn tree. He looks on it “disinterestedly”. He is attentive to the formal details of the hawthorn tree in a way that he is not to other things. He is inclined to think in a more open-ended and exploratory way about the hawthorn tree than about other things. He is inclined to see the hawthorn as having unconditional, or at least non-instrumental value, and he is emotionally vulnerable to it. For Marcel to “lose” his self in this sense would be for his practical thought to lose these characteristics—for him to become a different kind of agent who is responsive in different ways.

On this suggestion, the “demand” made on Marcel by the beauty of the hawthorns is grounded not in his higher-order attitudes about this responsiveness, nor in the conditions of his identity as a person. It is grounded in his integrity as a particular kind of agent, an agent who has oriented themselves toward the world as a lover of hawthorns. Marcel’s “vow” bestows authority on one part of himself—the part that orients him to the aesthetic potential of the hawthorns—to make him this particular kind of agent, to configure his agency in ways that will further this potential. Once he adopts this configuration, Marcel will have new reasons, ones grounded in the integrity of his agency, to keep faith with what moves him aesthetically.

I think this is a serviceable way to understand Moran’s seductive description of Marcel’s attachment to hawthorns. But even if I am wrong about this, it does provide us with a model of what it would mean to identify with one’s capacity for aesthetic pleasure.

7 Aesthetic justification

Marcel’s example shows how the notion of an aesthetic identity can help us solve the central problem of this paper, the problem of justifying our contingent practical nature. Let’s develop this thought a little further by considering a slightly different example, one involving the accidents of agency characteristic of professional philosophers. These will include a host of familiar habits of mind: to formulate ideas and arguments as crisply as possible, to anticipate objections, to formulate simple, illustrative examples, to connect a proposition to other ideas that it may justify. They might also include sensitivities to logical structure, ambiguity, conflation, equivocation, and the fruitfulness of certain distinctions. And they may very well entail articulations of intellectual ideals: clarity, rigor, creativity, interest, fecundity. These psychological features will tend to overlap and mutually support each other in ways that warrant our calling them a complex. I have argued that autonomy requires some justification for our adopting or carrying on with this philosophers’ complex. So what reasons might I offer?

One reason I can cite is that I think this configuration of agency is the best one I could adopt. Perhaps I think the philosopher’s way of thinking is the best way to gain knowledge or understanding, and these are things worth having for their own sake. This isn’t a bad justification, exactly, but it is a little chauvinistic, especially if I cite it in defense of being a philosopher instead of a physicist, an anthropologist, or an artist. Another kind of justification would cite the agreeableness of the things this complex can help me secure. Solving a difficult philosophical problem gives me a feeling of heady pride; giving a lecture on a difficult topic to a rapt audience soothes my ego. The problem with these reasons, I have argued, is that giving my disposition to agreeableness this kind of authority threatens my autonomy.

In between these two kinds of justification for the philosophers’ complex is something more promising. It is the sort of justification a philosophically curious version of Marcel might offer: an aesthetic justification. Suppose I take profound pleasure in the contemplation of hawthorns. It seems reasonable to suppose that I will take similar pleasure in similar forms of contemplation. And here my limited experiences of philosophy may then suggest that philosophical contemplation, when done competently, will be relevantly like my contemplation of the hawthorns. In particular, this contemplation may promise similar aesthetic pleasures. But because competent philosophical contemplation requires the complex of agential accidents characteristic of philosophers, I have a good reason to adopt that complex—a good reason to cultivate myself as a philosophical kind of agent. If this is my justification, then I am giving my capacity for taking pleasure in aesthetic activity authority over the shape of my agency. We could imagine similar justifications for many other contingent forms that agency can take: the ones employed by mathematicians, mass transit systems engineers, detectives, intelligence analysts, or healthcare consultants, the ones that allow a person to move seamlessly through a new culture, the ones that allow a person to be a perceptive conversationalist. All of these are activities that some people, in some circumstances, may reasonably believe will resemble the activities that produce aesthetic pleasure, and they can be justified by that similarity.

These are all relatively indirect examples, in the sense that the agent comes to an aesthetic justification for their accidents of agency through a comparison with another aesthetic experience. But if we accept the thesis that we can have aesthetic experiences of particular forms of agency, then there is a more direct path.³³ In this case, we can imagine individuals endorsing certain forms of agency—the agency of the monk, of the frontiersman, of the sea captain—simply because of aesthetic merits of that form that are most visible from the point of view of someone who has taken on that form of agency. This too is a way of justifying our accidents of agency in light of our aesthetic identity.

8 Neglected alternatives

Let me summarize my argument once more. Agency requires us to fix contingent but durable features of our practical nature. Among other things, these features will give our practical thought its structure. But we must justify the adoption of these features. Given that we are justifying something particular and individual—a feature of us but not everyone—this justification will cite, among other things, features of our particular, sensible nature. In particular, I think we cannot help but cite features about what pleases us. This is what the Hedonic Justification Thesis says. Now, one kind of pleasure is unsuitable for this work. Identifying with our dispositions to agreeableness threatens to diminish our autonomy. But another kind of identification works well. Identification with aesthetic pleasure can be a way of taking responsibility for ourselves through the free creativity of aesthetic activity.

This argument does not yet show what I claimed at the outset,that agency requires us to adopt an aesthetic identity. It would show this, save for a neglected possibility. There could be other forms of pleasure—neither aesthetic nor agreeable—that could do the job.

There are two ways to approach this hole. One is to weaken our thesis. A person does not have to adopt an aesthetic identity to be an agent. It may be strictly possible to cobble together a contingent practical nature out of hedonic identities that are neither aesthetic nor agreeable. But it is unclear what these identities would be or how this fashioning would proceed, and anyway this project seems quite difficult. If took this route, we would have to rest content with the thesis that nearly everyone will have to adopt an aesthetic identity to realize their agency.

The second option is to try to close the hole. The only way to do this decisively is by arguing that the features that make aesthetic pleasure suitable for autonomous identification are also sufficient conditions for being an aesthetic form of pleasure. In other words, we would need to claim that any form of pleasure that is active, scrutinizable, and a product of our creative powers is aesthetic. Thankfully, this strikes me as a plausible thesis. It would make aesthetic pleasure a large tent, but that seems appropriate. The pleasure we take in art and natural beauty would be important paradigms, but not exhaustive of the class. For example, many forms of contemplation—scientific thought experiments, mathematical reverie, thinking through the consequences of a philosophical theory—may produce aesthetic pleasure on this conception.³⁴ And on some views, the pleasures of food and drink are sufficiently rich to qualify as aesthetic.³⁵ More cautiously, pleasures of skill—throwing a curveball, doing engine maintenance, playing Czerny exercises—may barely qualify as aesthetic if they reflect some measure of creativity over and above mechanical facility, and erotic pleasures may present an entirely unique relationship between feeling and self.

If we take this route, the distinction between the agreeable and the aesthetic will look less like a dichotomy and more like a gradient. Pleasure can be more or less aesthetic depending on how free, creative, and susceptible to reflective scrutiny it is. Even some of my paradigms of the agreeable may have some aesthetic elements in certain circumstances. My thesis would then have to be amended to also admit of degrees. We would not say that a person’s agency depends on their identifying with their capacity for a definite kind of pleasure, aesthetic pleasure. We would say that their agency will vary (ceteris paribus) depending on how aesthetic—how free, creative, and subject to scrutiny—the pleasures they identify with for the sake of configuring their contingent practical nature are.³⁶

Before concluding my argument I must address two salient objections to this thesis.

9 First objection: aesthetic reasons

I have suggested that we can identify with our capacity for aesthetic enjoyment. What this means is that we can take suitable facts about what we enjoy as reasons for configuring our agency in particular ways. But one could object that reasons grounded in aesthetic pleasure are inadequate to this proposal. There are a few ways to press this objection.

First, we could say that because the pleasure of aesthetic experience is “disinterested”, it cannot provide us with anything like a “practical” reason. What “disinterested” means here is a point of great darkness, but here is one way to put the idea. The pleasure we take in eating a sandwich might be a reason to get a sandwich, but that is because gustatory pleasure produces a corresponding desire. Disinterested pleasure produces no such desire, so it does not produce reasons. My response is that this is an objection to my claim only if we think that the reason aesthetic pleasure gives us for cultivating a certain contingent practical nature must be grounded in a desire. But we do not have to make that assumption. Aesthetic pleasure can give us other reasons that support the kind of cultivation I have suggested. For example, it could give us reasons to appreciate an aesthetic object, which would justify activities that support appreciation.³⁷ That I enjoy listening to Beethoven is plainly a good reason for me to appreciate Beethoven, which then justifies all sorts of appreciation-enabling activities: taking the subway to the concert hall, checking out scores from the library, and cleaning my ears. This is the kind of reason my argument relies on. I claim that our aesthetic enjoyment of something gives us a reason not only to go to the concert hall or listen attentively, but to fashion ourselves in a way that deepens, broadens, and perpetuates that enjoyment.

The second version of this objection concedes that my aesthetic pleasure gives me a reason, but it insists that this reason is far too weak to justify the kind of self-fashioning I claim it can. Identifying with my capacity for aesthetic pleasure therefore represents a mistake about what kinds of considerations are appropriate for determining my nature as an agent. (This is one way of expressing familiar worries about overexuberant aestheticism.) The basic shape of our will should be determined by cold, solemn things—things like our moral convictions, our religious values, our needs and wants, our place in the social order, and what our father expects of us. Aesthetic pleasure can justify spending a few hours at the museum, but not durable features of my agency. The problem with this objection is that it doesn’t take our problem seriously. I agree that the solemn considerations mentioned can and do justify our becoming particular sorts of agents. But our problem was that they don’t do the whole job. The fact that a man is utilitarian, a Unitarian, and from Utah will not provide enough solemn reasons for him to fully configure a contingent practical nature for himself.³⁸ Something else is needed to fill the gap. This kind of objection would have more merit if I claimed that aesthetic pleasure can outweigh other reasons. But I only claim that it can fill a gap left over by those reasons. Whether aesthetic reasons are “weighty” enough for this work is entirely a question of how well they serve the purpose, and I have been arguing that they serve it quite well.

10 Second objection: individuality and universality

I suggested that aesthetic pleasure may help justify our contingent nature as practical creatures because those pleasures reflect our individuality. Each of us takes on a distinctive form of agency that accords with our distinctive aesthetic preferences. But how individual are these preferences? The problem isn’t with how much diversity of aesthetic opinions we find in the actual world; there is plenty of that. The problem is with the legitimacy of this diversity. Some philosophers claim that aesthetic judgments include an implicit demand for agreement.³⁹ If this is true, then interpersonal convergence may be a built-in ideal for aesthetic judgment. And that could mean that our aesthetic preferences are not, or should not be, as individual as we thought.

One way to deal with this concern is to deny that convergence is an aesthetic ideal. An alternative view has it that our aesthetic judgments are not under this normative pressure and that the ideal is one that encourages aesthetic individuality.⁴⁰ That’s a fine way to proceed, but I don’t think we have to accept this view because I don’t think the alleged tension is real. Imagine a disagreement: I prefer Fats Navarro solos, but you prefer Dizzie Gillespie solos. On a certain view, you and I disagree and are under pressure to resolve our differences. We would have a problem if this fact alone meant that at least one of us could not or should not identify with the aesthetic pleasure we take in our preferred trumpeter. For this would mean that much of the aesthetic individuality we find in the world would be unsuitable for identification. But I see no reason to accept this premise. It can be the case that in the long run we ought to converge in our aesthetic preferences, but right now it is perfectly legitimate or even advisable for each of us to identify with our own idiosyncratic likes and dislikes. We can be governed by the demands of both personal integrity and the ideal of interpersonal agreement.

True enough, this pair of demands can create sticky dilemmas. Mary loves Strauss waltzes and forms her aesthetic identity around them. This means that she has a reason—one of integrity—to tend to Strauss waltzes the way Marcel tends to his hawthorns. But there is much better music than Strauss waltzes. Mary also has reason to explore all this music, and she may even have reason to fashion a new identity around some of it. This is a challenge for Mary. But it not a problem for my view because it reflects a completely ordinary kind of dilemma. Mary has a decent enough relationship with her boyfriend, which generates reasons of fidelity. But Mary is also a better fit for a different partner, and that gives her reasons to “trade up”. This puts Mary in a pickle, but it’s no argument against the existence of competing demands.

11 Conclusion

I have tried to defend a version of Schiller’s stirring claim about the connection between beauty and freedom. The argument, in a nutshell, is that agency requires certain work that beauty is uniquely suited to perform.⁴¹ The problem we face is one of fashioning ourselves into particular agents without sacrificing our autonomy. I have suggested that the solution to this problem lies in our capacity to take pleasure in an activity that is at once free, creative, and amenable to critical scrutiny—in aesthetic activities. If we identify with this capacity, then we can justify the adoption of accidents of agency on contingent and highly personal grounds: because those accidents conduce to something that pleases us. And because these activities are free, creative, and amenable to critical scrutiny, we can grant them this authority not as a way of relinquishing responsibility for ourselves but of taking up that responsibility.