Skip to main content
Open Access

Moral Realism, Aesthetic Realism, and the Asymmetry Claim*

Abstract

Many people accept, at least implicitly, what I call the asymmetry claim: the view that moral realism is more defensible than aesthetic realism. This article challenges the asymmetry claim. I argue that it is surprisingly hard to find points of contrast between the two domains that could justify their very different treatment with respect to realism. I consider five potentially promising ways to do this, and I argue that all of them fail. If I am right, those who accept the asymmetry claim have a significant burden of proof.

I. Introduction

A growing number of philosophers are sympathetic to moral realism (sometimes called robust moral realism). This is the view, roughly, that there are moral facts, and these facts are robustly mind-independent, in the sense that they are not constituted by people’s attitudes, say, of approval and disapproval, nor are they constituted by the attitudes of hypothetical ideal observers. The wrongness of slavery, for example, according to moral realism, does not consist in the fact that anybody disapproves of it, nor does it consist in the fact that maximally rational, well-informed observers would disapprove of it. According to moral realism, actual and ideal observers alike play the role, at best, of trackers of moral facts, not determiners of them. So slavery would be wrong even if actual people approved of it, and even if ideal observers (specified in nonmoral terms) approved of it.

While realism is an increasingly popular position in metaethics,1 most people are less sympathetic to its counterpart in aesthetics. At first blush, it might look as though this is false: many aestheticians, after all, describe themselves as aesthetic realists.2 But when they do so, they are using the term ‘realism’ differently from how I am using it here, and differently from how many in metaethics use the term.3

As I am understanding ‘realism’ here, aesthetic realism would be the view that there are aesthetic facts and that these facts are robustly mind-independent, in the sense that they are not constituted by people’s attitudes, say, of admiration or pleasure, nor are they constituted by the attitudes of hypothetical ideal observers. According to aesthetic realism, the beauty of Venice doesn’t consist in our admiration of it, or in the pleasure it gives us; nor, further, does it consist in the fact that hypothetical ideal observers would admire or take pleasure in it. According to aesthetic realism, actual and ideal observers alike play the role, at best, of trackers of aesthetic facts, not determiners of them. So Venice would be beautiful even if nobody enjoyed or admired it, and even if ideal observers (specified in nonaesthetic terms) would not enjoy or admire it. Aesthetic realism, in contrast with its metaethical counterpart, has few, if any, supporters, and in fact it is considered to be such a nonstarter that any expression of support for it is likely to meet with what David Lewis called the incredulous stare.4

This is a striking disparity. For it to be justified, there would need to be some difference between beauty and morality that renders the former less susceptible to realist treatment. Call the view that aesthetic realism is less defensible than moral realism the “asymmetry claim.” The asymmetry claim has not before been challenged. And despite, or perhaps because of, its widespread acceptance, it has rarely been explicitly defended. Here it is subjected to detailed scrutiny. I show that it is significantly harder to justify than has been appreciated. I discuss five ways to argue for the asymmetry claim. These are arguments that score relatively highly on one or both of the following desiderata: match with actual motivation (they are arguments made in the literature or that otherwise plausibly capture reasons why people have found the asymmetry claim attractive), and chance of success (they are arguments that look potentially promising). All five arguments take the form of identifying some apparent point of contrast between the moral and the aesthetic and arguing that it entails that aesthetic realism is vulnerable to an obstacle that moral realism does not face. I argue that each of these arguments fails.

While this does not conclusively refute the asymmetry claim, it does raise a serious challenge for those who are sympathetic to it. If I am right that the differences traditionally appealed to aren’t up to the task, those who accept the asymmetry claim can no longer simply take for granted that it holds, but must provide some argument for thinking that it does. If this challenge cannot be met, then certain very common combinations of views might need to be reexamined. The vast majority of aestheticians stop short of aesthetic realism; in the absence of any good arguments for the asymmetry claim, they should also stop short of moral realism. Equally, moral realists, if no compelling arguments for asymmetry are forthcoming, should take aesthetic realism more seriously than they have.

II. Preliminaries

A. Aesthetic Value

In aesthetics there are a number of different issues with respect to which the question of realism arises.5 The comparison I’m interested in here is between morality and beauty. The widespread tendency I have in mind, then, is the tendency to think that realism about what is morally good and bad is more defensible than realism about what is beautiful. In what follows I use the terms ‘aesthetic value’ and ‘beauty’ interchangeably. Some philosophers include more than beauty under the category of aesthetic value,6 but most at least take beauty to be an example of aesthetic value. I will be adopting the narrower usage.

B. Realism

I introduced the relevant sense of realism in terms of a commitment to a particularly robust form of mind independence. It is now time to set out a full definition of realism. For the purpose of this article, realism, with respect to a domain D, is defined as the following view:

R1:

D-statements are in the business of stating facts.

R2:

There really are such facts.

R3:

These facts are (relevantly) robustly mind-independent, in the sense that they are (relevantly7) independent of

(i) the actual attitudes of actual observers;

(ii) the attitudes actual observers would have under ideal conditions; and

(iii) the attitudes ideal observers would have under ideal conditions.8

This usage is common in metaethics,9 but it is virtually unheard of in aesthetics. In aesthetics it is common to define realism in terms of just R1 and R2, omitting any kind of mind-independence clause.10 And those who do take some kind of mind independence to be criterial of aesthetic realism tend to have in mind a less robust form of mind-independence than that given in R3.11 Both kinds of approach classify at least some response-dependence accounts as realist. Response-dependence accounts of beauty hold that for something to be beautiful is for it to be disposed to produce a certain kind of response in (actual or ideal) observers, under certain conditions. These accounts don’t qualify as realist by R1–R3, since they don’t take facts about beauty to be robustly mind independent. They take facts about beauty to be constituted by facts about the actual or hypothetical attitudes of actual or ideal observers, and this is exactly what R3 rules out.

Needless to say, no fruitful comparison of two domains with respect to the tenability of realism can be conducted if the term ‘realism’ is understood in a more demanding sense when discussing one domain and a less demanding sense when discussing the other. My interest here is in whether realism in the more demanding sense is more defensible for ethics than for aesthetics, so I shall be using the term ‘realism’ to refer to the position given in R1–R3.

C. The Asymmetry Claim

Some philosophers express support for the asymmetry claim explicitly. David Enoch asserts that “there is a sense in which taking morality seriously seems appropriate, but taking aesthetics seriously does not.”12 Marcia Eaton argues that realism about beauty is untenable in a way in which realism in other domains is not: “Scientific realists believe that there are states of the world, and hence facts about the world and statements describing those states, that are true independently of whether anyone believes them. … Theorists have described moral realism analogously. There are moral truths or facts, and these truths or facts are independent of the evidence for them. … An analogous interpretation of aesthetic realism would require that there be aesthetic facts or truths independent of the evidence for them. Such a strong interpretation has few adherents, for the very meaning of aesthetic terms carries a subjective element.”13

But acceptance of the asymmetry claim is frequently more implicit. First, moral antirealists often claim that their theories transfer unproblematically to aesthetics (as do J. L. Mackie regarding his moral error theory and A. J. Ayer about his moral noncognitivism), while moral realists don’t tend to be quite so cavalier about their claims transferring straightforwardly to aesthetics.14 This tendency suggests that people are assuming that aesthetic realism faces all the potential obstacles that moral realism does, and more besides: if moral realism is shown to face insurmountable problems, then these problems would also block aesthetic realism; but if moral realism turns out to be right, aesthetic realism might still be false.

Second, as noted above, moral realism is a popular position and is taken seriously by its opponents, while its aesthetic counterpart is neither popular nor taken seriously. The orthodoxy in aesthetics appears to be that a response-dependence account—coupled with some story about how this is compatible with there being standards of correctness, or objectivity in some meaningful sense—is the closest to realism that one can reasonably get with respect to beauty.15 Response-dependence approaches not only are popular in aesthetics but also (at least to my knowledge) virtually exhaust the category of accounts that seek to accommodate standards of correctness with respect to aesthetic value.16 People rarely go further, in other words, and say that beauty is an entirely mind-independent matter in the sense articulated in R3. This is in striking contrast with the landscape in metaethics.17

Are there any good reasons to be squeamish about aesthetic mind-independence if one is not worried about moral mind-independence? In what follows I discuss five potential reasons. In each case, we have an argument that some feature of the aesthetic (i) is a point of contrast with the moral and (ii) strongly suggests that aesthetic value is mind-dependent. Each argument, then, is an argument that aesthetic realism is implausible in a way that moral realism is not:

Dependence:

Beauty depends on sensuous properties, such as colors and sounds, which are themselves mind-dependent. Moral goodness, on the other hand, does not depend on mind-dependent properties.18

Pleasure:

Beauty depends constitutively on pleasure in a way that is incompatible with realism, but this is not the case with moral goodness.

Descriptions:

You can make a moral judgment on the basis of an accurate, nonmoral description. This is not the case for aesthetic judgments—no amount of nonaesthetic, or nonevaluative, descriptions of something will put you in a position to judge whether it is beautiful.19

Acquaintance:

To judge something beautiful, you need to see it for yourself, but to judge something morally good or bad, you don’t.20

Principles:

One way to judge the moral status of the action is to apply principles, but you can never come to a view on whether something is beautiful by applying principles.

I argue that none of these arguments succeed. In each case, either the feature of the aesthetic that it identifies is not a genuine point of contrast with the moral, or it doesn’t in fact lend any support to the claim that aesthetic value is mind-dependent.

III. Dependence

Dependence:

Beauty depends on sensuous properties, such as colors and sounds, which are themselves mind-dependent. Moral goodness, on the other hand, does not depend on mind-dependent properties.

The Dependence point has sometimes been taken to support the asymmetry claim. The thought is that here we have an argument against realism about beauty that lacks a moral analogue:

D1:

Beauty depends on sensuous properties, such as colors and sounds (premise).

D2:

Beauty can only be mind-independent if colors and sounds are mind-independent (from D1, transitivity of dependence).

D3:

Colors and sounds are not mind-independent (premise).

D4:

So beauty can’t be mind-independent (from D2, D3).

D5:

So aesthetic realism is false (from D4, definition of aesthetic realism).

I will not take issue here with D3. It is beyond the scope of this article to enter into the debate on the mind-dependence or otherwise of colors and sounds, so I will assume for argument’s sake that D3 is correct. I shall have more to say on D1 later, but for now, note that D1 should not be allowed to stand without further argument. Many people hold that things can be appropriately judged beautiful which lack sensuous properties entirely. Proofs, chess moves, and theories are often held to be capable of being beautiful.21 But even setting aside these worries, the argument is unconvincing as an argument for the asymmetry claim.

The problem is the move from D4 to D5 and the availability of a parallel argument against moral realism. As noted in Section II, it’s not mind-dependence per se that’s at issue in the debate between moral realists and antirealists, but only a certain kind of mind-dependence. It’s widely acknowledged, even by moral realists, that moral goodness and badness can depend on mind-dependent things in the following kinds of ways. If I make a remark that upsets you, your upsetness looks like it bears on the moral status of my act. If, in addition, I believe that you’ll be upset, and I make the remark anyway—or if I actually intend to upset you in making the remark—these features (my beliefs and intentions, your hurt feelings) all look highly relevant to the moral status of my action. Given this, an analogous argument can be constructed for the moral case:

D1m:

Moral goodness and badness depend on intentions, hurt feelings, and beliefs (premise).

D2m:

Moral goodness and badness can only be mind-independent if intentions, hurt feelings, and beliefs are mind-independent (from D1m, transitivity of dependence).

D3m:

Intentions, hurt feelings, and beliefs are not mind-independent (premise).

D4m:

So moral goodness and badness can’t be mind-independent (from D2m, D3m).

D5m:

So moral realism is false (from D4m, definition of moral realism).

There are two upshots of this for the argument from Dependence. First, the moral case looks to be in the same boat as the aesthetic case with respect to the Dependence point, so they stand or fall together: if Dependence shows aesthetic realism to be implausible, its counterpart would show moral realism to be implausible. So the argument from Dependence would be useless as an argument for the asymmetry claim even if D1–D5 succeeded as an argument against aesthetic realism.

A second upshot is that the parallel with the moral case should make us doubtful that D1–D5 do succeed as an argument against aesthetic realism. If moral realism doesn’t commit to mind independence per se, but rather only to a specific kind of mind-independence, then D5 doesn’t follow from D4. And if D5 doesn’t follow from D4 in the moral case, it shouldn’t follow in the aesthetic case either. In other words, if mind-dependence per se isn’t enough to preclude realism in the moral case, it shouldn’t be enough to preclude realism in the aesthetic case.

Can the Dependence argument be amended to target the more specific kind of mind-independence that is criterial of realism? This is doubtful. Call the kind of mind-dependence realists are concerned to deny the bad kind of mind-dependence, and call the kind of mind-dependence they are not concerned to deny the okay kind of mind-dependence. A canonical example of the bad kind would be “kicking puppies is wrong, and what makes it wrong is that people would disapprove of it,” and a canonical example of the okay kind would be “my sharp remark to you was wrong, and what makes it wrong is that it hurt your feelings.” Some have rested content with there being an intuitive distinction here. But others have tried to pinpoint what principle underlies the distinction. Philosophers have, variously, taken the bad kind of mind-dependence to be stance-dependence (Milo, Shafer-Landau),22 dependence of the moral truths on our evidence for them (Brink, Sturgeon),23 and dependence on mental or mind-dependent properties that is not a simple consequence of their being normatively relevant (Enoch).24

What the asymmetry theorist needs is an argument that (i) shows that beauty is mind-independent in the bad way and (ii) lacks a moral counterpart. But it’s far from clear that the dependence of beauty on colors and sounds could be made into such an argument. Consider, for example, Enoch’s normative relevance criterion. Colors and sounds look normatively relevant to beauty, just as hurt feelings and nefarious intentions look normatively relevant to moral badness. And this normative relevance looks to be what’s making D1 look plausible. In this case, the kind of mind-dependence that the argument from Dependence can get us is not the bad kind but the okay kind.

At the very least, then, someone who wants to use the Dependence point to argue for the asymmetry claim will need to argue for some way of drawing the line between the bad kind and the okay kind of mind-dependence, that has the following features: (i) classifies dependence on feelings, beliefs, and intentions as mind-dependence of the okay kind; and (ii) classifies dependence on colors and sounds as mind-dependence of the bad kind. This looks like a tall order.

IV. Pleasure

Pleasure:

Beauty depends constitutively on pleasure in a way that is incompatible with realism, but this is not the case with moral goodness.

It is striking that not only does nearly everyone accept aesthetic antirealism but also there is one specific kind of aesthetic antirealism that is extremely popular. This is a response-dependence view according to which what it is for something to be beautiful is for it to occasion pleasure in observers of a certain kind.25 Call this view hedonic response-dependence (HRD) about beauty.

HRD about beauty comes in different varieties, depending on what the relevant kind of pleasure is and how the relevant kind of observer is characterized. Most aesthetic HRD theorists hold that there is some distinctively aesthetic kind of pleasure and that beauty is a disposition in objects to occasion that kind of pleasure in observers who meet certain conditions. According to this kind of view, what it is for Venice to be beautiful is just for it to have the capacity to give rise to this distinctive kind of pleasure in observers who meet certain requirements and who are observing the object under normal, optimal, or ideal conditions. On this view, it’s not that these kinds of people take pleasure in it because it’s beautiful, but rather, the fact that these kinds of people take pleasure in it is what it is for it to be beautiful.

In contrast with HRD about beauty, HRD about morality is nowhere near as popular. But this point requires a bit of clarification. The comparison is not about just any kind of constitutive dependence on pleasure. As discussed in Section III, not all kinds of mind-dependence are incompatible with realism. And this point is especially salient here because there is a well-known family of views that hold that moral goodness depends constitutively on pleasure that are not antirealist, namely, hedonic utilitarianism.

As noted in Section III, there are a number of proposals in the literature of how to distinguish between the antirealist kind of mind-dependence and the other kinds. It’s an interesting question whether all of these do an equally good job of capturing and explaining the intuitive distinction.26 What’s important here, however, is this: assuming that the distinction can be drawn in the right kind of way, you might think that Pleasure could be a basis for an argument for the asymmetry claim. HRD about beauty entails aesthetic antirealism, and if HRD about morality doesn’t hold, we have an argument against aesthetic realism that lacks a counterpart in the moral case:

H1:

HRD about beauty holds: what it is for something to be beautiful is for it to occasion pleasure in certain kinds of observers under certain conditions.

H2:

If HRD about beauty holds, then realism about beauty must be false.

H3:

So realism about beauty is false.

The argument is certainly valid. Despite its validity, however, the argument shouldn’t worry the aesthetic realist. HRD about beauty entails aesthetic antirealism because HRD is just a species of aesthetic antirealism. So the argument from HRD to antirealism is not dialectically useful—you would already need to be an aesthetic antirealist to be an HRD theorist about beauty. The obvious way out for the aesthetic realist, then, is to reject the first premise.

There are two ways the argument could be rescued. One is if it could be supplemented by a compelling argument for HRD about beauty; the other is if HRD about beauty were intuitively obvious enough to not require argument.

Consider the latter first. Could it be that HRD about beauty is intuitively obvious? It’s not clear that there is strong intuitive support for it. It may be highly intuitive that beauty is (often) pleasing to experience. But that’s a different (and weaker) claim than the HRD claim that the occasioning of pleasure in a certain kind of observer is what beauty consists in. The realist can accept that weaker claim—she can say that beauty is something we take pleasure in, or something that does occasion pleasure in appropriately situated observers. She just denies that beauty is constituted by such pleasure—just as the moral realist can say that we tend to approve of moral goodness and disapprove of moral badness—but these responses aren’t what moral goodness and badness consist in.

If HRD about beauty is right, then, it’s not obviously the case. The aesthetic HRD theorist would need an argument for her view. Are there any good arguments for it? The first thing to note is that there is a conspicuous absence of arguments for HRD about beauty in the literature.

The arguments I will go on to discuss in Sections VIVIII could each—if successful—be taken as an argument for HRD about beauty. But even if this is so, each could equally be taken as an argument for the less demanding thesis of antirealism about beauty. Since their success as arguments for asymmetry requires only that they support aesthetic antirealism, my focus will be on whether they succeed in that objective. I will argue that they fail to establish aesthetic antirealism. If I am right, then a fortiori they fail as arguments for HRD.

V. Affective Responses

Not only is it commonly held that beauty is constitutively dependent on pleasure in the way discussed in Section IV, but it is also commonly held that pleasure has an essential role in aesthetic judgment.

Note that the latter isn’t the same view as the HRD view discussed in Section IV, though the two views may go well together. It’s not the same view because you could think that beauty is constitutively dependent on pleasure in the way HRD says it is, without holding that the only possible way to judge something beautiful is by having a feeling of pleasure. (For example, you could think that beauty is a disposition to occasion pleasure in observers who meet certain conditions. This view lends itself naturally to the thought that, at least for observers who don’t meet the relevant conditions, finding out that something is beautiful is a matter of finding out whether the relevant kinds of observers take pleasure in it—and this is something you could do without you yourself having any kind of feeling of pleasure, or indeed any other affective response.)

Call the view that affective responses are essential to judgment in a given domain the affective model of the relevant kind of judgment.27 While many take the affective model of aesthetic judgment to be obviously right, moral judgment is not usually taken to be an obvious candidate for the affective model: a lot of people want to say that while you might have certain affective responses when you make moral judgments (moral outrage, for example), these aren’t essential to moral judgment. I think that many have found the asymmetry claim plausible because they take this contrast to obtain, and they take realism to be incompatible with the affective model. The argument would be this:

P1.

Affective Asymmetry Premise: Aesthetic judgment differs from moral judgment in essentially involving affective responses.

P2.

Incompatibilist Premise: If judgments in a particular domain essentially involve affective responses, then realism with respect to that domain is false.

C.

So aesthetic realism is untenable in a way that moral realism is not.

Is this argument any good? One way to challenge it would be to put pressure on the incompatibilist premise, the claim that the affective model is an obstacle to realism. But this is not the challenge I will pose here. Instead, I will focus on the affective asymmetry premise, the claim that the involvement of affective responses is a point of contrast between moral and aesthetic judgments. This is a very widely held claim, but, I will argue, it is actually rather hard to motivate.

Note that the affective asymmetry premise doesn’t look to be obviously true. First, if judging something to be beautiful essentially involves having a response of aesthetic pleasure, then we should expect that there should be something wrong with asserting (1) in something like the way that there is something wrong with asserting the Moore-paradoxical sentence (2):

(1)

It’s beautiful but I don’t like it.

(2)

It’s true but I don’t believe it.

But it’s not clear that there is. Asserting (1) doesn’t sound contradictory in the way that asserting (2) does.

Second, if affective asymmetry holds, we should also expect that (1) would sound worse than (3):

(3)

It’s morally good but I don’t approve of it.

And it’s not clear that it does. In fact, you might think not only that (1) doesn’t sound worse than (3) but also that in some ways it sounds much better than (3).

If the affective asymmetry premise is to be justified, then we need some other kind of argument for it. In Sections VIVIII, I consider three related differences that have been taken to hold between the moral and the aesthetic domains and generate arguments for affective asymmetry. These are the arguments from Descriptions, Acquaintance, and Principles. Each of these arguments pinpoints an apparent restriction on when aesthetic judgments can be made, which, it is claimed, (i) lacks a counterpart in the moral case and (ii) strongly suggests that aesthetic judgments essentially involve affective responses.

VI. Descriptions

Descriptions:

You can make a moral judgment on the basis of an accurate, nonmoral description. This is not the case for aesthetic judgments—no amount of nonaesthetic, or nonevaluative, descriptions of something will put you in a position to judge whether something is beautiful.

On the face of it, it sounds like the Descriptions claim picks out a genuine contrast. If you tell me, entirely in nonmoral terms, that your friend Betty did such-and-such, I can, in principle, be in a position thereby to make a moral judgment about whether what Betty did was morally wrong or not (provided that you give me accurate and sufficiently detailed information). Matters look different for aesthetic judgment: no amount of nonaesthetic, or nonevaluative, descriptions of something will put me in a position to judge whether it is beautiful.

Why would the Descriptions claim support the affective model of aesthetic judgment? One argument would be the following inference-to-the-best-explanation argument:

Ds1.

Aesthetic judgments can’t be based on descriptions of their objects.

Ds2.

The best explanation for this is that aesthetic judgments essentially involve affective responses.

Ds3.

So aesthetic judgments essentially involve affective responses.

This argument is unconvincing. The problem is Ds2. Why should we think that the involvement of affective responses would even be an explanation, let alone the best one? In order for the involvement of affective responses to be an explanation for the inadequacy of descriptions, it would have to be the case that descriptions don’t put you in a position to have affective responses to the object described. But this is not in general the case. Some people hold that moral judgments require emotional responses, and these people don’t have to deny that moral judgments can be made on the basis of descriptions. Many of the situations we emotionally respond to, we do so having heard them described, rather than actually watching them unfold. People are often morally upset, outraged, and so on, about events they have heard about via descriptions in news stories. So given that these emotional responses can be had on the basis of descriptions in the moral case, and since emotional responses are a kind of affective response, it certainly can’t be that in general the impossibility of forming a particular kind of judgment on the basis of a description could be explained by the hypothesis that judgments of that kind are based on affective responses. And if that kind of argument doesn’t work in general, those who make the argument in the aesthetic case need, at the very least, to say what is special about this case.

The asymmetry theorist can respond with a refined version of the argument. As noted above, those who hold that aesthetic judgments must be based on affective responses usually have a particular kind of affective response in mind: standardly pleasure of some specific-but-hard-to-pin-down kind. Call this “aesthetic pleasure.” And it’s this specific response of aesthetic pleasure, they will argue, that can’t be had on the basis of a description, even if other affective responses can be. The best explanation, they will then argue, for why aesthetic judgments can’t be based on descriptions is that aesthetic judgments are based on that particular kind of affective response.

Ds1.

Aesthetic judgments can’t be based on descriptions of their objects.

Ds2*.

The best explanation for this is that aesthetic judgment essentially involves having the specific response of aesthetic pleasure.

Ds3.

So aesthetic judgments essentially involve affective responses.

This argument is an improvement on the previous one. It does look plausible that descriptions of an object don’t put you in a position to have this particular kind of response to it. Suppose you described to someone a painting: it’s in the impressionist style, there’s lots of blue, it’s an outdoor scene with three figures. It would probably strike you as odd or otherwise infelicitous if they said “Ah, I’m at a loss for words!” and claimed to be experiencing this response of aesthetic pleasure (just as it would strike you as odd if they had said “It’s beautiful” just on the basis of that description). The involvement of aesthetic pleasure, then, is a more promising candidate for an explanation of the inadequacy of descriptions. But is it the best explanation? This looks doubtful, I will argue, since there is a promising alternative explanation.

A. An Alternative Explanation

By and large, independently of whether they are evaluative realists or antirealists, most people accept that evaluative judgments of an object should be responsive to certain nonevaluative features of that object. Without being aware of certain details of what Betty did, you’re not in a position to judge that what Betty did was wrong. Similarly, without any awareness of what Ryoanji looks like, you’re not in a position to judge that it is beautiful.28 Call this the minimal epistemic requirement (MER). MER should strike you as quite a banal point. But it can do a surprising amount of explanatory work.

Let’s consider three quick points about MER. First, MER imposes different requirements on judges depending on the kind of evaluative judgment they are making. The kinds of features of something that you need to be aware of in order to be in a position to judge it morally are very likely different from the kinds of features you need to be aware of to judge it aesthetically.

Second, MER doesn’t presuppose any particular metaevaluative view. In particular, it’s worth noting that it is compatible with an affective model of the judgments in question. Consider a view that takes moral judgments to be affective responses of certain kinds of approval and disapproval. This kind of picture can accommodate MER because it can hold that if you don’t know what Betty did, then the relevant kind of disapproval is not available to you, and you aren’t able to even count as judging her action morally wrong. Or consider a view that takes aesthetic judgments to be responses of certain kinds of pleasure and displeasure. People who hold this view can accommodate MER because they can hold that without any awareness of what Ryoanji looks like, the relevant kind of pleasure is not available to you, and so you aren’t able to even count as judging it beautiful.

Third, whether the responses in question are cognitive or affective, MER is a genuinely epistemic requirement, in the sense that it requires the judger to be aware (at least at the time of making the judgment) of certain features of reality. Perhaps the awareness doesn’t need to be knowledge; perhaps it doesn’t even need to be belief (in the sense that you can be aware of something without having any conscious beliefs about it). But in any case, awareness is still an epistemic state (even if it turns out to be a rather minimal one), and so the requirement that one be aware of certain features is a genuinely epistemic one.

MER suggests a plausible explanation of why you can’t judge an object beautiful on the basis of a description of it. You can’t come to know what it looks like (or indeed, sounds like) from a description of it. Nor can you even become aware of what an object looks like or sounds like from a description of it. And so (at least if judgments of beauty must always be responsive to an object’s perceptual appearance), descriptions won’t be an adequate basis for such judgments. Descriptions don’t put you in a position to make aesthetic judgments because they don’t enable you to meet MER. Descriptions of objects don’t put you in touch with the features that aesthetic judgments should be responsive to.

Call this the epistemic explanation, since it takes the problem to be that descriptions don’t allow you to meet an epistemic requirement for aesthetic judgment. Interestingly, the epistemic explanation of the Descriptions claim doesn’t take it to indicate any particularly deep contrast between moral and aesthetic judgments. The two different kinds of judgments are responsive to different kinds of features of their objects. The kinds of features that are relevant to moral judgment are different from the kinds of features relevant to aesthetic judgment. And as you might expect from this, different ways of apprehending their objects are appropriate as bases for the different judgments. Aesthetic judgments can be made on the basis of photographs, but can moral judgments? Perhaps they can given a certain amount of context, but can a photo alone be enough? Probably not.29

Both photos and descriptions are what we might call proxies, things we can use to become aware of certain features of something else. It happens to be the case that the features that are relevant to the moral status of an action can in general be captured quite well in a description and not very well in a photo. And the features that are relevant to the aesthetic status of an object can be captured quite well in a photo and not very well in a description. This shouldn’t be surprising. The nature of proxies is to capture some features well and others less well. Depending on the kind of evaluative judgment one is making, different nonevaluative features will be important, so different kinds of proxy will be suited to different kinds of evaluative judgment.

B. Why Prefer the Epistemic Explanation?

We have two explanations to choose between: the epistemic explanation (descriptions don’t put you in touch with all the features to which an aesthetic judgment should be responsive) and the affective explanation (you can’t have a response of aesthetic pleasure to something on the basis only of a description of it). For the argument from descriptions to work, the affective explanation needs to be the best explanation. But there is good reason to think that the epistemic explanation is a better explanation. Here is why.

Proponents of the affective explanation need it to be the case that the particular affective response of aesthetic pleasure can’t be had on the basis of a description. We might ask why it can’t be. What can they say in response to this question? Given that affective responses in general are not impossible to have from a description, it must be that the reason this particular kind of affective response can’t be had on the basis of a description has something to do with the particular kind of affective response it is, rather than with the fact that it is an affective response.

What is it about this particular kind of affective response? The most plausible thing to say is that it’s the fact that it is a response to certain features of an object. Descriptions don’t put us in touch with these features, and that’s why the response isn’t available. This seems sensible. If there is a distinctive response of aesthetic pleasure, then like aesthetic judgment, it should be a response to the perceptual appearance of objects. And descriptions can’t put us in touch with objects’ perceptual appearances.

If the proponent of the affective explanation says this, however, she is committing to everything that the epistemic explanation is committed to. She’s accepting that descriptions can’t put us in touch with the perceptual appearance of objects. And she’s accepting that aesthetic judgments (derivatively on aesthetic responses) must be responsive to objects’ appearances. But she is also committing to one extra thing: that aesthetic judgments involve affective responses. This extra commitment is not doing any explanatory work. The epistemic explanation, after all, works without it. So the epistemic explanation is to be preferred.

The argument for the involvement of affective responses was precisely that it does important explanatory work. I’ve shown that this is false; all the same explanatory work can be done without it. All you need is the claim that aesthetic judgments are responsive to perceptual appearances of objects and the claim that descriptions can’t put us in touch with these appearances. The inference-to-the-best-explanation argument from the Descriptions claim fails to motivate the affective model of aesthetic judgment.

C. Nonperceptual Beauty

There is potentially a further problem for the argument from Descriptions. In my objections to it so far I have assumed that beauty depends on perceptual appearances and that judging something beautiful requires awareness of its perceptual appearance. But is this assumption correct? Many people hold that the things that can be appropriately judged beautiful are not exhausted by perceptible objects. As noted in Section III, proofs, chess moves, and theories are often held to be capable of being beautiful.

If there are cases of beauty that are not perceptual, this would cause further problems for the argument from Descriptions. Let me explain.

If there is nonperceptual beauty, this may spell trouble for the Descriptions claim itself. At the very least, it introduces a need for clarification on what a description is. Does a statement explaining a theory in enough detail for someone to fully grasp it count as a description of it? Does one count as describing a proof if one spells it out very precisely, line by line? If these count as descriptions, then presumably one can, after all, judge the beauty of a proof or a theory on the basis of a description of it. If this is right, then the Descriptions point would be false: the inadequacy of descriptions would not, after all, be a point of contrast between moral and aesthetic judgments. The inadequacy of descriptions would not characterize the aesthetic case as a whole; rather, it would hold only for a subset of aesthetic judgments: judgments of perceptual beauty. And it would be a feature of these aesthetic cases simply by virtue of the fact that they are also perceptual. The inadequacy of descriptions would be a hallmark not of the aesthetic but of the perceptual.

If, on the other hand, the asymmetry theorist insists that spelling out a proof line by line and explaining a theory thoroughly enough for someone to grasp it don’t count as giving descriptions of the proof and the theory, respectively, she owes an explanation of why. Perhaps the suggestion will be that it is criterial of descriptions that they leave out some information about the thing described. Not just any verbal recounting qualifies as a description: these examples don’t count as descriptions because they are too thorough—they don’t leave out any information.

But if this is what is meant by ‘description’, is Ds1 plausible? Is it plausible that one can’t make aesthetic judgments, even of nonperceptual beauty, if any information is left out? I think the best argument the asymmetry theorist can give is to appeal to what I’ll call the extreme Sibleyan picture: the idea that every characteristic of an object is aesthetically relevant. Frank Sibley, in an influential article, makes a milder version of this point:

First, the particular aesthetic character of something may be said to result from the totality of its relevant nonaesthetic characteristics. It is always conceivable that, by some relatively small change in line or color in a picture, a note in music, or a word in a poem, the aesthetic character may be lost or quite transformed (though possible also that by some considerable changes it may not be significantly altered). A somewhat different totality might result in an aesthetic difference. Features one would hardly think of singling out as notably contributing to its aesthetic character—say, background colors, hardly noticed brush strokes, and so on—nevertheless do contribute because, being as they are, they at least allow it to have the character it has, a character it conceivably might not have if they were altered.30

Sibley’s point is more moderate than the kind of picture the asymmetry theorist would need here, since he allows that it’s possible that some changes would fail to impact aesthetic character. But a more extreme version of Sibley’s thought would hold that every characteristic of an object is aesthetically relevant. If this extreme Sibleyan picture is right, then descriptions, understood as verbal recountings that leave something out, will be inadequate bases for aesthetic judgment, since they always leave out something aesthetically relevant.

However, while this line of thought supplies a motivation for Ds1, it does so at the cost of putting Ds2* (and indeed the original Ds2) on shaky ground. The way the extreme Sibleyan picture motivates Ds1 is by supplying an explanation for it, namely, the reason that descriptions fail to be adequate bases for aesthetic judgment is that they fail to put us in touch with all of the aesthetically relevant features. But this just is the epistemic explanation again, and since the epistemic explanation is not committed to the affective model, it entirely undermines Ds2*.

So the extreme Sibleyan picture can’t motivate Ds1 without undermining Ds2*. Unfortunately for the asymmetry theorist, it’s not clear what, besides the extreme Sibleyan picture, could motivate Ds1. If you tell me quite a lot about a theory, for example, and you leave out some details (so, by the criterion under discussion, you count as describing it), but you leave out nothing that is relevant to its beauty, it’s hard to see why we should deny that I am in a position to judge it beautiful. The extreme Sibleyan picture amounted to denying that this was possible—if you leave something out, then a fortiori you leave something out that is relevant to beauty. But if the extreme Sibleyan picture is false, this situation is possible, and it’s hard to see the motivation for denying that, in such a situation, I could be in a position to judge the theory beautiful.

To summarize, the Descriptions point was raised as a way to argue for affective asymmetry. For this to work, there needs to be an argumentative route from the inadequacy of descriptions for aesthetic judgment to the affective model of aesthetic judgment. The best candidate looked to be the following inference-to-the-best-explanation argument:

Ds1.

You can’t judge something aesthetically on the basis of a description of it.

Ds2*.

The best explanation for this is that aesthetic judgment essentially involves having the specific response of aesthetic pleasure.

Ds3.

So aesthetic judgment essentially involves having the specific response of aesthetic pleasure.

I noted that there may be a question of how to interpret Ds1, depending on whether there is nonperceptual beauty. I argued that on any plausibly true interpretation of Ds1, Ds2* looks false. On any plausible interpretation of Ds1, what does the explanatory work is the claim that descriptions can’t put you in touch with the relevant features of an object (either because they can’t put you in touch with the object’s perceptual appearance, and judgments of beauty are responsive to perceptual appearance, or because descriptions always leave out some information, and that information is relevant to beauty).

VII. Acquaintance

Acquaintance:

A constraint known as the Acquaintance Principle holds in aesthetics, but not in ethics. In order to judge, say, that your friend Betty acted wrongly, I don’t need to have been there, seen her action unfold, observed the consequences of it for myself, and so on, but in the case of my judgment that, say, the Torre del Oro in Seville is beautiful, things are different: I do need to see it for myself.

Some philosophers have argued that the Acquaintance Principle (AP) holding in the aesthetic case is best explained by the affective model of aesthetic judgment.31 The reason you need to see the Torre del Oro for yourself to judge it beautiful is that judging something beautiful is a matter of having a certain kind of affective response to it. Sean McKeever and Michael Ridge, for example, suggest that it is the role of affective responses in aesthetic judgments that “holds the key to explaining the special role of direct engagement in aesthetic judgment.”32

If McKeever and Ridge are right that (i) AP applies to all aesthetic judgments, (ii) the best explanation for this is that aesthetic judgments essentially involve affective responses, and (iii) AP doesn’t hold in ethics, then this is an argument for thinking that aesthetic judgments essentially involve affective responses, without any corresponding commitment to thinking that moral judgments involve them.

The argument from Acquaintance is in fact underspecified. There are various different ways of understanding AP and what exactly it amounts to. The crucial question is whether there is any reading of AP on which both (i) and (ii) are plausible. I will argue that there is not.

The usual way to understand AP is as a requirement that one’s judgment be based on a perceptual encounter with its object. You have to see it, hear it, and so on. Call this AP1.

AP1:

Aesthetic judgment must be based on a perceptual encounter with its object.

AP1 doesn’t look to be true. Most people accept that you can, at least in some cases, judge the visual beauty of something on the basis of a photograph of it. It is also plausible that you can judge the beauty of a musical work on the basis of reading its score. If this is right, and if AP is to hold for all aesthetic judgments, acquaintance needs to be construed in a way that allows that looking at a photo of an object can acquaint you with that object and reading a score of a musical work can acquaint you with that work.

Photos and scores can put you in touch with the perceptual appearance of their objects, even though they are not ways of encountering those objects. Through looking at a photo, you can come to know what an object looks like, and through reading a score, you can come to know what a piece of music sounds like. So perhaps AP should be understood not in terms of perceptual encounters but in terms of awareness of perceptual properties. Call this AP2.

AP2:

Aesthetic judgment must be based on an awareness of the object’s perceptual properties.

Would understanding acquaintance in terms of AP2 help the argument from Acquaintance? No; if acquaintance is understood in terms of AP2, the asymmetry theorist faces a dilemma. Either there is nonperceptual beauty, or there isn’t. If there is nonperceptual beauty, AP2 is false. If there isn’t, and all beauty is perceptual, then AP2 just falls out of MER: it just amounts to the banal claim that aesthetic judgments must be responsive to the aesthetically relevant features of their objects. This is paralleled by the fact that moral judgments must be responsive to the morally relevant features of their objects. MER’s applying to aesthetic judgments is no more in need of explanation than its applying to moral judgments. So the claim that the affective model of aesthetic judgment must be brought in to do explanatory work is entirely unconvincing.

Perhaps the asymmetry theorist should resort to a reading of AP that has nothing to do with the perceptual. Perhaps she should say that all that is required for acquaintance is awareness of all of the object’s properties, perceptual or not. Call this AP3.

AP3:

Aesthetic judgment must be based on awareness of all of the object’s properties.

AP3 is far too demanding. We make aesthetic judgments all the time without knowing everything about the object we’re judging. But even putting that worry on hold, it’s entirely unconvincing that the best explanation for AP3—were it true—would be the affective model of aesthetic judgment. Again, the epistemic explanation looks more plausible: if aesthetic judgment of something requires awareness of all of its properties, the natural explanation for this would be that all of its properties are aesthetically relevant, and aesthetic judgments must be responsive to the aesthetically relevant features of their objects.

In summary, the problem with the argument from Acquaintance is similar to the problem faced by the argument from Descriptions. If there is a sense of “acquaintance” in which it is plausible that acquaintance is required for all judgments of beauty, this looks to be easily accommodated as a consequence of the epistemic requirement that judgments of beauty must be based on awareness of that object’s aesthetically relevant features.

VIII. Principles

Principles:

Aesthetic judgments do not involve the application of principles, but moral judgments do. At least one way to judge the moral status of the action is to apply principles, but you can never come to a view on whether something is beautiful by applying principles.33

Principles is the third alleged point of contrast between moral and aesthetic judgments that might be used to generate an argument for affective asymmetry. For the argument to work, two things would need to be the case. First, Principles needs to be a genuine point of contrast between the moral and the aesthetic. Second, it also needs to be the case that the lack of a role for principles in aesthetic judgment would put pressure on us to adopt the affective model of aesthetic judgment. There may well be good reasons to accept the former, but—I will argue—there are no good reasons to accept the latter.

The contrast in Principles is both widely accepted and plausible. At least since Kant,34 the contention that aesthetic judgment is not principled has been orthodoxy in philosophical aesthetics.35 Those who defend the view that aesthetic principles even exist are often described ironically as “heroic”!36 But in the moral case, the mainstream view is that moral judgment is principled. There are people—particularists, of whom Jonathan Dancy is the best-known example—who deny that moral judgment is principled, but this is a minority view, and particularists tend to take themselves to have the burden of proof. Tellingly, in his arguments for moral particularism, Dancy takes aesthetics to be a domain where particularism is uncontroversially right, and he tries to support moral particularism by drawing an analogy with aesthetics.37

The mainstream view, then, is that Principles marks a genuine contrast between moral and aesthetic judgments. Moreover, there are good reasons to think that the mainstream has it right. In order to see this, we need to distinguish between two kinds of principle: decisive principles and contributory principles. Decisive principles of moral goodness would be principles that say which nonmoral properties would guarantee that an action is morally good; contributory, or pro tanto, principles would be principles that identify nonmoral features as counting toward or counting against an action’s being morally good.38

Here are two reasons to think that it is plausible that moral judgment involves at least contributory principles and less plausible that aesthetic judgment does. First, there has been some degree of convergence on what contributory moral principles would be, but much less on contributory aesthetic principles. Most people will accept that keeping promises and occasioning pleasure count toward an action’s goodness, while lying and stealing count against it, but it is hard to find any counterparts that enjoy such widespread acceptance regarding beauty. There has been very little progress in identifying even plausible candidates for aesthetic principles. Second, the phenomenology of the respective judgments supports the contrast too. Aesthetic judgment just doesn’t typically feel as though it involves reasoning: it has what some have called “phenomenological immediacy,”39 in that it feels more like one is just struck by something’s beauty than that one reasons to a conclusion that it is beautiful. Moral judgments, on the other hand, do often feel as though they involve reasoning.

There may be room to contest this picture, but in any case this won’t be my focus here. Let’s assume, then, that Principles marks a genuine contrast between moral and aesthetic judgments. What I’m interested in is the argument from Principles to the affective model of aesthetic judgment. As with Descriptions and Acquaintance, the argument is an inference-to-the-best-explanation argument:

Pr1.

Aesthetic judgment does not involve applying principles.

Pr2.

The best explanation for this is that aesthetic judgment essentially involves affective responses.

Pr3.

So aesthetic judgment must essentially involve affective responses.

Pr2 is highly doubtful. In fact, you might think it’s distinctly odd. After all, moral particularists are often moral realists, so isn’t there something strange about using aesthetic particularism to argue ultimately for aesthetic antirealism? The asymmetry theorist would certainly need to say something to this worry. But perhaps she can. Perhaps she will be able to show that moral particularists are wrong to be moral realists. In any case, I will pursue a different line of argument against Pr2 here.

The affective model of aesthetic judgment would be a possible explanation for the lack of a role for principles. If judging something beautiful were always a matter of feeling a certain way about it, then we should expect that the application of principles would play no role in such a judgment. However, there is another possible explanation for why judgments of beauty never take this inferential form. Some properties we discern noninferentially, rather than as the outcome of a process of reasoning, from the presence of other properties.

Many hold, for example, that perceptual judgments fall into this category: we don’t infer from other propositions (e.g., about our sensory experience) that certain objects with certain properties are before us. Rather, we just notice the objects.

Further, some perceptual noninferentialists point out that if there are properties of visual experience that are linked in a principled way to the presence of particular objects, these are not properties that we ordinarily notice. As Strawson puts it, “An observer, gazing through his window, may perhaps, by an effort of will, bring himself to see, or even will-lessly find himself seeing, what he knows to be the branches of the trees no longer as branches at all, but as an intricate pattern of dark lines of complex directions and shapes and various sizes against a background of varying shades of grey. The frame of mind in which we enjoy, if we ever do enjoy, this kind of experience is a rare and sophisticated, not a standard or normal, frame of mind.”40

It’s highly plausible that analogous points hold concerning beauty:

(i) Standardly, we discern beauty immediately, not via first discerning properties other than beauty and then inferring from the presence of these properties, together with a principle linking them to beauty, that the object is beautiful.

(ii) If there are other properties that are linked in a principled way to beauty, these are not properties that strike us with the immediacy that beauty does, and we’re not ordinarily in a position to characterize these properties.

If this noninferentialist picture is right, it would be a natural explanation for the lack of role for principles in judgments of beauty. If beauty were a property that we could discern noninferentially, we wouldn’t have any need, in aesthetic judgment, for principles linking other properties to beauty. Further, if other properties that are linked in a principled way to beauty are not easy for us to consciously pick out, or at least are less easy to pick out than beauty itself, it would be entirely expectable that we would not use principles in judgments of beauty. More generally, principles linking certain nontarget properties to some target property are only useful if these nontarget properties are easier for us to pick out than the target property itself. If we have immediate awareness of the target property and are only able to identify the nontarget properties through considerable effort, these principles are of no use to us in judgments concerning the target property. It should be unsurprising that judgments of the target property proceed without the use of such principles.

Moreover, this explanation doesn’t rely on characterizing judgments of beauty as affective, just as the noninferential account of perception doesn’t rely on characterizing perceptual judgments as affective. So we have here an alternative explanation to the affective model. Unless the asymmetry theorist can supply some reason to think that the affective model of aesthetic judgment would be a better explanation than the noninferential explanation, the argument from Principles fails to establish the asymmetry claim.

A. An Objection

There appears to be a tension between the view I’ve argued for here and what I said in Sections VI and VII. I argued in Sections VI and VII that aesthetic judgments must be based on awareness of certain nonaesthetic features. But here I’ve argued that judgments of beauty don’t involve reasoning. Are these two claims in tension with each other? How can judgments of beauty be based on awareness of the object’s aesthetically relevant features, unless they involve reasoning from these features? If aesthetic judgments are noninferential, shouldn’t you be able to judge something beautiful without being aware of anything else about it?

The objection assumes that the only form the “based-on” relation can take is inferential. But there are reasons to reject this assumption; there are good reasons to think that there is noninferential based-on-ness. Robert Audi argues, for example, that facial recognition judgments must be based on awareness of certain visual patterns but nonetheless don’t typically involve inference:

Granted, facial recognition depends on seeing the features of the face, as is evidenced by the impossibility of recognition where a number of features are blocked, say the eyebrows. But such recognition is not dependent on inference from the relevant features, as is evidenced by the possibility of recognition even where the perceiver has no belief corresponding to those features. I need not believe Karl’s brows have the look they do until I focus on the matter. Indeed, the look they have that is important for my recognition may be so distinctive—or so subtly related to other features, such as the nose and hairline—that it would be difficult, or perhaps impossible, to capture the content of beliefs of properties or, correspondingly, in a set of premises for inference.41

It is worth looking closely at the arguments here, both for facial recognition’s being noninferential and for its being based on awareness of certain features. Both arguments, I will suggest, can be made just as easily in the aesthetic case. If I’m right, and if Audi is right about facial recognition, the noninferential nature of aesthetic judgment is not in tension with its being based on awareness of nonaesthetic properties.

Why should we take facial recognition judgments to be noninferential? Audi argues that in typical cases of facial recognition we just don’t have the kinds of beliefs that would be needed as premises if the judgment were inferential. While you need to see certain features in order to recognize a face, you don’t need to have any beliefs about those features. Audi offers two considerations in support of this. First, the relevant kinds of beliefs often don’t happen without conscious reflection: “I need not believe Karl’s brows have the look they do until I focus on the matter.” Second, the kinds of beliefs that would be needed as premises for such an inference would be very difficult to come by: “The look [Karl’s eyebrows] have that is important for my recognition may be so distinctive—or so subtly related to other features, such as the nose and hairline—that it would be difficult, or perhaps impossible, to capture the content of beliefs of properties or, correspondingly, in a set of premises for inference.”42

It’s worth adding that these arguments apply to two kinds of belief: what we might call specific visual beliefs, and principles of facial recognition. If facial recognition involved inference, we’d need beliefs of both kinds. We’d need beliefs such as “the brows have this particular look, which is related in such-and-such a way to the nose, and so on,” but we’d also need beliefs such as “if the brows have this particular look, which is related in such-and-such a way to the nose, and so on, then it’s Karl,” or “then this counts defeasibly toward its being Karl.” Not only is it plausible that we can lack beliefs of the first kind and still form facial recognition judgments, but it is also highly plausible that we can do so while lacking beliefs of the second kind.

Why should we take facial recognition to be based on awareness of certain visual features? Audi argues that awareness of certain visual features must be necessary for recognition, because recognition can become impossible if certain features are blocked (e.g., if you can’t see someone’s eyebrows). To this we can add that facial recognition is definitely going to be impossible if you can’t see any of the features. So some awareness of visual features has got to be a requirement for facial recognition to happen.

If Audi is right about facial recognition, then this shows that a judgment can be based on awareness of something without being an inference from it. And if these two features aren’t in tension in the facial recognition case, then it’s hard to see why they should be in tension in the case of aesthetic judgment.

Of course, there is potentially room to contest the picture of facial recognition as noninferential. Some philosophers stress that inference can be unconscious, as can the beliefs that form its premises. Those who are sympathetic to that view will have some resources to resist Audi’s arguments about facial recognition. But note that to the extent that this can be made to fly in the recognition case, it will also be a plausible way to resist the thought that aesthetic judgment proceeds without the use of principles, and with it the thought that Principles constitutes a genuine contrast between aesthetic and moral judgments.

To summarize, in the previous three sections I considered three kinds of argument for affective asymmetry. These took the form of three alleged features of aesthetic judgment that are claimed to (i) support the affective model of aesthetic judgment and (ii) be a point of contrast with moral judgments. In each case, while there is plausibly a moral-aesthetic contrast in the vicinity, there is no compelling argument from it to the affective model of aesthetic judgment. This might be surprising. The affective model is often taken to be not only well suited to aesthetic judgments but also better suited to them than to moral judgments. But in the absence of any good arguments, we should take seriously the possibility that the affective model is no better a fit for aesthetic judgments than it is for moral ones. Many philosophers are sympathetic to the thought that, while we may feel strongly about moral matters, these strong feelings are importantly distinct from moral judgments. If the arguments of Sections VIVIII are right, we should be equally open to an analogous line of thought with aesthetics: finding things beautiful is enjoyable, but that doesn’t mean that all there is to finding things beautiful is enjoying them. Either that, or we should be equally resistant to both.

IX. Conclusion

This article has challenged a particularly pervasive and obstinate feature of current philosophical orthodoxy: the view that moral realism is more defensible than aesthetic realism. I discussed five broad ways in which this view might be supported, and I argued that none of these are successful. This is not a knockdown refutation; at best the arguments here play the role of undercutting defeaters: showing that the asymmetry claim is unmotivated. But at the very least it places the burden of proof squarely on the side of those who are sympathetic to the asymmetry claim. Unless some arguments are given for the asymmetry claim, we should conclude that aesthetic realism should be put on the map of serious philosophical positions worth engaging with, or that moral realism is less worthy of being taken seriously than is currently supposed.

If what I’ve argued here is correct, we might wonder why people have found the asymmetry claim so appealing. If I’m right, its appeal outstrips its argumentative support, so we might wonder if there are other factors that have contributed to its attraction. One such factor might be the appeal of another thesis from which the asymmetry claim is not often clearly distinguished. It is a much-repeated platitude that there is a sense in which the stakes are higher in ethics than aesthetics. While this claim would need some fleshing out, there is no doubt something right about it. The thought that, in some sense, it matters more to get things right morally than aesthetically is a sensible one. But, of course, this has no bearing on the viability of realism in the two domains. Getting things right morally matters more—in pretty much any sense of mattering worth considering—than getting right how many paper clips are in the drawer of my desk. And yet we don’t take realism about the latter to be less tenable than realism about the former: on the contrary, realism about the number of paper clips in my desk is largely taken to be more defensible than moral realism. Nonetheless, it strikes me as broadly right that the intuitive plausibility of the claim that the stakes are in some sense higher with ethics than aesthetics goes some way to explaining the widespread acceptance of the asymmetry claim.

Notes

*. I would like to thank Chris Cowie, Guy Fletcher, and Bob Stecker for helpful comments on earlier drafts. I am also grateful to audiences at the University of Leeds in 2014, the Humane Philosophy and the Arts Conference in Oxford in 2014, and the Realism and Antirealism in Metaethics and Aesthetics Conference in Cambridge in 2014, as well as to two anonymous referees and various associate editors at Ethics.

1. See, e.g., Russ Shafer-Landau, Moral Realism: A Defence (Oxford: Oxford University Press, 2003); David Enoch, Taking Morality Seriously (Oxford: Oxford University Press, 2011); William FitzPatrick, “Robust Ethical Realism, Non-naturalism, and Normativity,” in Oxford Studies in Metaethics, ed. Russ Shafer-Landau (Oxford: Oxford University Press, 2008), 3:159–205; Michael Huemer, Ethical Intuitionism (New York: Palgrave MacMillan, 2005); Matthew Kramer, Moral Realism as a Moral Doctrine (Chichester: Wiley-Blackwell, 2009).

2. See, e.g., Jerrold Levinson, “Aesthetic Properties, Evaluative Force, and Differences of Sensibility,” in Contemplating Art (Oxford: Oxford University Press, 2006), 315–35; Marcia Eaton, Merit: Ethical and Aesthetic (Oxford: Oxford University Press, 2000); María José Alcaraz León, “The Rational Justification of Aesthetic Judgments,” Journal of Aesthetics and Art Criticism 66 (2008): 291–300; Philip Pettit, “The Possibility of Aesthetic Realism,” in Pleasure, Preference and Value, ed. E. Schaper (Cambridge: Cambridge University Press, 1983), 17–39.

3. According to the usage of ‘realism’ common in aesthetics, response-dependence accounts are classified as realist. Accounts that take beauty to be a matter of the responses actual agents in fact have, the responses they would have under ideal conditions, or the responses hypothetical ideal observers would have all count as realist under the definition favored in aesthetics. But none of these kinds of accounts take beauty to be mind-independent in the robust sense I am discussing here, and so none of them count as realist in the sense relevant to this article.

4. David Lewis, On the Plurality of Worlds (Oxford: Blackwell, 1986), 133.

5. To name a few: beauty, fictional truth, pictorial representation, musical expression of emotion. These questions are largely independent: it is perfectly intelligible that someone could take one position with respect to fictional truth (say, relativism), for example, and quite another with respect to beauty (say, realism). Further, many philosophers distinguish between beauty and artistic merit: a beautiful artwork is not necessarily a good one, and a good artwork is not necessarily beautiful. See, e.g., Arthur Danto, “Works of Art and Mere Real Things,” in The Transfiguration of the Commonplace (Cambridge, MA: Harvard University Press, 1981), 1–33; Robert Stecker, “Artistic Value Defended,” Journal of Aesthetics and Art Criticism 70 (2012): 355–62; Carolyn Korsmeyer, “On Distinguishing ‘Aesthetic’ from ‘Artistic,’” Journal of Aesthetic Education 11 (1977): 45–57. Depending on how beauty and artistic merit are related, there may also be room for taking different meta-aesthetic stances with respect to these two notions. See Louise Hanson, “Artistic Value Is Attributive Goodness,” Journal of Aesthetics and Art Criticism 75 (2017): 415–27, esp. 422–24.

6. James Shelley, “The Problem of Non-perceptual Art,” British Journal of Aesthetics 43 (2003): 363–78; Noël Carroll, “Non-perceptual Aesthetic Properties,” British Journal of Aesthetics 44 (2004): 413–23; Gary Iseminger, The Aesthetic Function of Art (Ithaca, NY: Cornell University Press, 2004); Dominic McIver Lopes, “The Myth of (Non-aesthetic) Artistic Value,” Philosophical Quarterly 61 (2011): 518–36.

7. Most realists who include a mind-independence clause don’t want to deny all kinds of dependence of normative facts on human attitudes. They will want to allow that the wrongness of a sharp remark or an ill-judged joke can depend on the fact that it caused someone to feel hurt or embarrassed. For discussion of this, see Shafer-Landau, Moral Realism, 15; David Brink, Moral Realism and the Foundation of Ethics (Cambridge: Cambridge University Press, 1989), 15–16; Enoch, Taking Morality Seriously, 4. See also Sec. III of the present article for further discussion.

8. There is perhaps some scope for debate about whether Fitting Attitude accounts of value can qualify as realist. Fitting Attitude accounts of moral goodness hold that it is necessarily true that “an object S has a value V (e.g., moral goodness, beauty) if and only if it is fitting to have a certain attitude (e.g., approval, admiration) toward S.” See Justin D’Arms and Daniel Jacobson, “Sentiment and Value,” Ethics 110 (2000): 722–48; Graham Oddie, Value, Reality and Desire (Oxford: Oxford University Press, 2005); Mark Schroeder, “Value and the Right Kind of Reason,” Oxford Studies in Metaethics, ed. Russ Shafer-Landau (Oxford: Oxford University Press, 2010), 5:25–55; Conor McHugh and Jonathan Way, “Fittingness First,” Ethics 126 (2016): 575–606.

Certainly reductive forms of FA, according to which something is morally good/beautiful in virtue of the fact that certain attitudes toward it are fitting, will not qualify as realist. However, if all it means to adopt an FA account is to hold that the above biconditional is necessarily true, then FA looks compatible with realism, since you could accept the biconditional and hold that the right-hand side holds in virtue of the left-hand side, not the other way around. For a discussion of this point, see Graham Oddie, “Fitting Attitudes, Finkish Goods, and Value Appearances,” in Oxford Studies in Metaethics, ed. Russ Shafer-Landau (Oxford: Oxford University Press, 2016), 11:74–101.

9. See Ronald Milo, “Contractarian Constructivism,” Journal of Philosophy 92 (1995): 181–204; Shafer-Landau, Moral Realism, 13–18; Oddie, Value, Reality and Desire; Enoch, Taking Morality Seriously; Huemer, Ethical Intuitionism; Matthew Evans and Nishi Shah, “Mental Agency and Metaethics,” in Oxford Studies in Metaethics, ed. Russ Shafer-Landau (Oxford: Oxford University Press, 2012), 7:80–109; Kramer, Moral Realism as a Moral Doctrine.

10. See Pettit, “Possibility of Aesthetic Realism”; Alcaraz León, “Rational Justification of Aesthetic Judgments”; James Young, “Aesthetic Antirealism,” Southern Journal of Philosophy 35 (1997): 119–34.

11. See Levinson, “Aesthetic Properties”; John Bender, “Sensitivity, Sensibility, and Aesthetic Realism,” Journal of Aesthetics and Art Criticism 59 (2001): 73–83; Eaton, Merit, 51–53; Stephanie Ross, “Ideal Observer Theories in Aesthetics,” Philosophy Compass 6 (2011): 513–22.

12. Enoch, Taking Morality Seriously, 268.

13. Eaton, Merit, 51–52.

14. See A. J. Ayer, “Critique of Ethics and Theology,” in Language, Truth and Logic, 2nd ed. (London: Victor Gollancz, 1948), 102–20; J. L. Mackie, Ethics: Inventing Right and Wrong (London: Penguin, 1977), 15.

15. For examples of such accounts, see Frank Sibley, “Objectivity and Aesthetics,” Supplement to the Proceedings of The Aristotelian Society 42 (1968): 31–72, esp. 32–33; John McDowell, “Aesthetic Value, Objectivity, and the Fabric of the World,” in Pleasure, Preference and Value, ed. E. Schaper (Cambridge: Cambridge University Press, 1983), 1–16; David Wiggins, “A Sensible Subjectivism?,” in Needs, Values, and Truth, 3rd ed. (Oxford: Clarendon, 1987), 185–211; Richard Miller, “Three Versions of Objectivity,” in Aesthetics and Ethics: Essays at the Intersection, ed. Jerrold Levinson (Cambridge: Cambridge University Press, 1997), 26–58; Pettit, “Possibility of Aesthetic Realism”; and, at least on some readings, David Hume, “Of the Standard of Taste,” in Essays, Moral, Political, Literary, ed. Eugene F. Miller (Indianapolis: Liberty Classics, 1985). See, e.g., James Shelley, “Hume and the Nature of Taste,” Journal of Aesthetics and Art Criticism 56 (1998): 29–38.

16. Shelley and Watkins might be an exception. See Michael Watkins and James Shelley, “Response-Dependence about Aesthetic Value,” Pacific Philosophical Quarterly 93 (2012): 338–52. Eddy Zemach’s Real Beauty (University Park: Pennsylvania State University Press, 1997) might also be an exception, but, as others have remarked, it is not easy to tell whether his preferred view is one that takes aesthetic value to be genuinely mind-independent. See Glenn Parsons and Allen Carlson, review of Real Beauty, by Eddy M. Zemach, Canadian Journal of Philosophy 29 (1999): 635–53.

17. The following all argue for moral mind-independence: Shafer-Landau, Moral Realism; Huemer, Ethical Intuitionism; Enoch, Taking Morality Seriously; FitzPatrick, “Robust Ethical Realism”; Kramer, Moral Realism as a Moral Doctrine.

18. For this argument, see W. D. Ross, The Right and the Good (Oxford: Oxford University Press, 2002), 126–27; and Nick Zangwill, “Skin Deep or in the Eye of the Beholder: The Metaphysics of Aesthetic and Sensory Properties,” Philosophy and Phenomenological Research 61 (2000): 595–618.

19. The claim that aesthetic judgments cannot be made on the basis of descriptions is frequently made. See, e.g., Peter Strawson, “Aesthetic Appraisal and Works of Art,” in Freedom and Resentment, 2nd ed. (Oxford: Routledge, 2008), 196–207, esp. 204; Philip Pettit, “Substantive Moral Theory,” in Objectivism, Subjectivism, and Relativism in Ethics (Cambridge: Cambridge University Press, 2008), 1–27, 9 n. 10.

20. The claim that the Acquaintance Principle holds in aesthetics has a long history. See Immanuel Kant, Critique of Judgement, ed. Nicholas Walker, trans. James Creed Meredith (Oxford: Oxford University Press, 2009), sec. 8. Those who explicitly draw a contrast along these lines with ethics include Richard Wollheim, Art and Its Objects (New York: Harper & Row, 1971), 233; Pettit, “Substantive Moral Theory,” 9; Miller, “Three Versions of Objectivity,” 35; Pettit, “Possibility of Aesthetic Realism,” esp. 25.

This claim is used by Sean McKeever and Michael Ridge in a rare explicit argument for the asymmetry claim. See Sean McKeever and Michael Ridge, “Aesthetics and Particularism,” in New Waves in Metaethics, ed. M. Brady (New York: Palgrave Macmillan, 2011), 264–85. McKeever and Ridge are not alone, however, in suggesting that the Acquaintance Principle’s holding in aesthetics supports aesthetic antirealism: Cain Todd argues that “no advocate of the Acquaintance Principle ought to hold that aesthetic judgments are beliefs or genuine assertions at all.” Cain Samuel Todd, “Quasi-realism, Acquaintance, and the Norms of Aesthetic Judgement,” British Journal of Aesthetics 44 (2004): 277–96, 290. And Roger Scruton argues that “in aesthetics you have to see for yourself precisely because what you have to ‘see’ is not a property.” Roger Scruton, Art and Imagination (London: Methuen, 1974), 54 (emphasis mine).

21. Further, it’s not clear, even for things that have sensuous properties, that they can never be beautiful by virtue of qualities other than those sensuous properties. If one holds, for example, that not only a novel but also a copy of a novel can be beautiful by virtue of features such as its plot progression and the nuance of its use of language, then this would be a case of nonsensuously dependent beauty.

22. See Milo, “Contractarian Constructivism,” 182; Shafer-Landau, Moral Realism, 15.

23. See Brink, Moral Realism and the Foundations of Ethics, 15–17; Nicholas Sturgeon, “What Difference Does It Make Whether Moral Realism Is True?,” Southern Journal of Philosophy 24 (1986): 115–41, 117.

24. See Enoch, Taking Morality Seriously, 3–4.

25. For arguments against, see James Shelley, “Against Value Empiricism in Aesthetics,” Australasian Journal of Philosophy 88 (2010): 707–20; Dominic McIver Lopes, “Aesthetic Experts, Guides to Value,” Journal of Aesthetics and Art Criticism 73 (2015): 235–46; Watkins and Shelley, “Response-Dependence about Aesthetic Value.”

26. It is an interesting question why HRD is antirealist while hedonic utilitarianism is realist (or at least not automatically antirealist). My own view on this is that Enoch’s normative relevance criterion pinpoints the relevant difference between the kind of mind-dependence the realist is denying and the kind she is not. So the debate on mind-dependence between moral realists and antirealists is about whether anything mental or mind-dependent bears on the moral facts in any way other than by being normatively relevant. If this is right, then it suggests the following rationale for classifying HRD as antirealist and hedonic utilitarianism as realist. HRD theorists are not saying that pleasure is normatively relevant to beauty; they are saying that pleasure bears on beauty in some way other than being normatively relevant. Hedonic utilitarians, on the other hand, say that pleasure bears on moral status by being normatively relevant.

27. I borrow this terminology from Terence Cuneo, “Signs of Value,” British Journal for the History of Philosophy 14 (2006): 69–91.

28. One possible exception to this might be judgments formed on the basis of testimony, but this is highly contentious. Many deny that testimony can yield knowledge on moral or aesthetic matters. See, e.g., Roger Crisp, “Moral Testimony Pessimism: A Defence,” Supplement to the Proceedings of The Aristotelian Society 88 (2014): 129–43; Daniel Whiting, “The Glass Is Half Empty: A New Argument for Pessimism about Aesthetic Testimony,” British Journal of Aesthetics 55 (2015): 91–107. But some argue that testimony can give you aesthetic knowledge. See Malcolm Budd, “The Acquaintance Principle,” British Journal of Aesthetics 43 (2003): 386–92. In any case, the spirit—if not the letter—of the above is highly plausible. If testimony can give you knowledge on moral and aesthetic matters, then the claim above has an exception—but it is still plausible that firsthand (i.e., nontestimonially acquired) evaluative judgments need to be responsive to certain features of the object. If testimony can’t give you evaluative knowledge, on the other hand, then the formulation above appears to be in good shape.

29. It’s hard to think of a case where the moral features of an action would be apparent from a photograph. Perhaps the most promising kind of case would be something like showing you a photograph of Betty tripping someone up. But even in this kind of case, it’s unclear that the photo alone could put you in a position to make a moral judgment. Unless I also tell you that what you see in the photo really is what it looks like, and that it wasn’t an accident, nor was it a joke, and that she really intended to trip this person up, it’s not plausible that you’ll have enough information to be in a position to pronounce on the moral status of her action.

30. Frank Sibley, “Aesthetic and Nonaesthetic,” Philosophical Review 74 (1965): 135–59, 138.

31. See McKeever and Ridge, “Aesthetics and Particularism”; Scruton, Art and Imagination, 54; Todd, “Quasi-realism, Acquaintance, and the Norms of Aesthetic Judgement.”

32. McKeever and Ridge, “Aesthetics and Particularism,” 274.

33. The PRINCIPLES point might be thought to be very closely related to the Descriptions point, but it is worth noting that they can come apart. If you can judge on the basis of a description that something is morally wrong or right, it doesn’t obviously follow that if you are doing so, you are using moral principles. Again, even those who argue for moral particularism can acknowledge that moral judgments can be made on the basis of descriptions. I might judge that Betty acted wrongly, after hearing a faithful nonmoral description of what she did, but the question of whether my judgment involved applying principles is still open.

34. See Kant, Critique of Judgement, sec. 8.6: “If we judge objects merely in terms of concepts, then we lose all presentation of beauty. This is why there can be no rule by which someone could be compelled to acknowledge that something is beautiful. No one can use reasons or principles to talk us into a judgement on whether some garment, house, or flower is beautiful.” See also Christopher Janaway, “Kant’s Aesthetics and the ‘Empty Cognitive Stock,’” in Kant’s Critique of the Power of Judgment: Critical Essays, ed. Paul Guyer (Lanham, MD: Rowman & Littlefield, 2003), 67–86, 74: “There are no principles of taste; again because my judgements of taste require as ground this ‘direct’ inspection of an object and this testing of it against my own feelings, there could be no intersubjective rules anyone could use to prove the things’ beauty to me.”

35. Mary Mothersill, in what appears essentially to be an argument from authority, writes, “It does not matter whether you speak of criterial features or of principles of taste: there are none. It is Kant’s single greatest contribution to the subject to have made the point and to have insisted upon it.” See Mary Mothersill, “Aesthetic Laws, Principles and Properties,” Journal of Aesthetics and Art Criticism 47 (1989): 77–82, 78.

36. The oft-repeated dig at Monroe Beardsley’s attempt to argue for aesthetic principles originates from Frank Sibley. See Monroe Beardsley, “On the Generality of Critical Reasons,” Journal of Philosophy 59 (1962): 477–86; Frank Sibley, “General Criteria and Reasons in Aesthetics,” in Essays on Aesthetics, ed. J. Fisher (Philadelphia: Temple University Press, 1983), 3–20.

37. See Jonathan Dancy, “Ethical Particularism and Morally Relevant Properties,” Mind 92 (1983): 530–47, esp. 546.

38. See Jonathan Dancy, Ethics without Principles (Oxford: Clarendon), 5–6.

39. Fabian Dorsch, “Noninferentialism about Justification—the Case of Aesthetic Judgements,” Philosophical Quarterly 63 (2013): 660–82.

40. Peter Strawson, “Perception and Its Objects,” in Philosophical Writings (Oxford: Oxford University Press, 1979), 125–45, 131.

41. Robert Audi, Moral Perception (Princeton, NJ: Princeton University Press, 2013), 53.

42. Ibid.