Jump to ContentJump to Main Navigation
Disagreement$

Richard Feldman and Ted A. Warfield

Print publication date: 2010

Print ISBN-13: 9780199226078

Published to Oxford Scholarship Online: September 2010

DOI: 10.1093/acprof:oso/9780199226078.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 19 February 2018

Disputing about Taste

Disputing about Taste

Chapter:
(p.247) 10 Disputing about Taste
Source:
Disagreement
Author(s):

Andy Egan

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780199226078.003.0011

Abstract and Keywords

This chapter discusses disagreement about matters of taste. It contends that these disputes are best explained by saying that what is at stake in them is the self-attribution of certain properties and that resolution of the disputes amounts to the parties making the same self-attributions. It argues that this account explains a special sort of defectiveness to which disputes about taste are subject.

Keywords:   disagreement, matters of taste, self-attribution, disputes

1. Introduction

“There's no disputing about taste.” That's got a nice ring to it, but it's not quite the ring of truth. While there is definitely something right about the aphorism—there is a reason why it is, after all, an aphorism, and why its utterance tends to produce so much nodding of heads and muttering of “just so” and “yes, quite”—it is surprisingly difficult to put one's finger on just what the truth in the neighborhood is, exactly. One thing that is pretty clear is that what is right about the aphorism, that there's no disputing about taste, is not that there's no disputing about taste. There's heaps of disputing about taste. People engage in disputes about which movies, music, paintings, literature, meals, furniture, architectural styles, and so on are good, beautiful, tasty, fun, elegant, ugly, disgusting, and so forth all the time. This is obvious to anyone who has watched dueling‐movie‐critics shows, read theater reviews, or negotiated with a group or partner about which movie or restaurant to go to, or which sofa or painting to put in the living room. It takes great care and good aim to fling a brick without hitting somebody who is engaged in a dispute about taste.

We might suggest instead that what is right about the aphorism is that there is no sensible, worthwhile disputing about taste—that disputes about taste are, across the board, defective in some way that makes them nonsensical, irresolvable, bereft of a genuine subject matter, or otherwise second class (and therefore generally not worth pursuing). This can be a tempting thought. As Hume (1757/1965) notes, it is attractive to think that “a (p.248) thousand different sentiments, excited by the same object, are all right: Because no sentiment represents what is really in the object . . . To seek in the real beauty, or real deformity, is as fruitless an enquiry, as to pretend to ascertain the real sweet or real bitter.” Something has clearly gone wrong when I say that broccoli tastes better than Brussels sprouts, my friend Mira disagrees, and we launch into a deeply committed dispute aimed at getting to the bottom of this question once and for all, and uncovering the real facts of the matter. This sort of disagreement is defective in a way that makes it a mistake to invest it with any great significance, pursue it too deeply, think that anyone who takes a view contrary to mine must thereby be getting something wrong, or devote a lot of resources to resolving our dispute and arriving at a collective view on the question. The same seems to go for disputes about, for example, whether Annie Hall is a more entertaining movie than Indiana Jones and the Raiders of the Lost Ark, or whether (to steal an example from Peter Lasersohn (2005)) a particular roller coaster is or is not fun.

We might think, based on looking at these sorts of cases, that there is something defective about all disputes about taste. There is a project of engaging in a particular sort of argument and discussion, aimed at arriving at a common view about the matter in question, by both parties becoming convinced that one of the two views is the uniquely correct view to hold, that we engage in about ordinary matters of fact. That project is (we might think), across the board, a bad project to go in for in this domain, because the relevant facts in this neighborhood are either absent, observer‐relative, or otherwise incapable of sustaining this sort of collective investigation and debate.

We might think this, but, as Hume points out not long after the passage quoted above, we would be wrong: “whoever would assert an equality of genius and elegance between OGILBY and MILTON, or BUNYAN and ADDISON, would be thought to defend no less an extravagance, than if he had maintained a mole‐hill to be as high as TENERIFFE, or a pond as extensive as the ocean.” There are some disputes about taste that clearly are in order, and are not defective in the way that Mira's and my dispute about broccoli and Brussels sprouts was defective. If I maintain that some tune I have just idly pinged out on a piano is a more beautiful piece of music than Mozart's Requiem, or that the limerick about the guy from Nantucket is better poetry than Shakespeare's sonnet about love being not love which (p.249) alters when it alteration finds,1 I have just got it wrong. There are a lot of cases in which the parties to the dispute should arrive at a common view—in which one of the candidate positions is clearly the one that both parties to the dispute ought to endorse. And disputes about taste often do wind up producing agreement. People do frequently manage to convince each other to revise their views about matters of aesthetics or personal taste, and in such cases the parties to the dispute wind up, at the end, with a common view about the matter, and in agreement about how and whether to apply the contested vocabulary to the items in question.

Think, for example, about the case in which Smith judges that Zingerman's pastrami is not tasty because she ate it with too much mustard, or when she was already over‐full. Or the case where Jones judges that the symphony is not beautiful because the performance he attended was badly executed, he was angry at the conductor, he had an ear infection, his tinnitus was acting up, he was not paying attention, and so on. In these sorts of disputes, Smith's and Jones's interlocutors are likely to be able to convince them to revise their views by drawing their attention either to features of their situation that were liable to generate interference, or to features of the pastrami or the symphony that they had overlooked when forming their initial judgments.

Other cases are, of course, different. If it emerges that Smith's judgment was not the result of some sort of outside interference—if, for example, trying the pastrami with less mustard and on an empty stomach does not change Smith's mind—her interlocutor should probably conclude that they are just differently gustatorily constituted, and give up on the dispute. Continuing the dispute after this emerges would be a mistake. Similarly, if it emerges that Jones's judgment was not the result of interfering circumstances, or failure to attend to all of the relevant features of the symphony, it may well be that his interlocutor should conclude that they, too, are simply differently constituted, and give up on their dispute.

We are not using our time well if we spend a lot of it in deeply committed disputes about whether broccoli tastes better than Brussels sprouts or vice versa. And not just in the way that we are not using our time well if we spend it in deeply committed disputes about whether Caesar had more or less than 1,000 fleas on his person when he crossed the Rubicon. The (p.250) second dispute has got all sorts of things wrong with it—for example, which answer is correct is not very interesting and is not very important, and we are in an extremely bad position to figure it out. But there seems to be something more profoundly wrong with the first dispute—it is not just that it would be a mistake to care a lot about the answer, or that we are badly positioned relative to the evidence we would need in order to establish which answer is right. It is that the idea that there even is a unique correct answer to be discovered, however boring, trivial, and epistemically inaccessible, seems suspect.

Here are two phenomena to notice. (1) The status of a dispute as defective or not does not depend just on subject matter. In the pastrami and symphony examples above, we had pairs of disputes that were alike in their subject matter (the tastiness of the pastrami, the beauty of the symphony), of which one was defective and the other sensible, depending on the circumstances of dispute. (2) Disputes can start off sensible and become defective. Even the defective disputes above were in order (or at least, seemed to be in order) until it emerged that the parties to them were differently constituted with respect to their relevant bits of sensory apparatus (or whatever—I do not want, at this point, to commit myself to very much about just what sorts of differences between the disputants undermine the sensibility of these sorts of disputes).

A third phenomenon that is worth attending to is that there are variations in how robust different disputes about taste are. Disputes about gustatory taste—about what is tasty, for example—are comparatively fragile. It does not take that much to convince us that such a dispute is not worth engaging in. Disputes about the aesthetic value of literary works are much more robust. Others fall somewhere in between. (The robustness of a dispute, it should be noted, seems to vary both with the subject matter and with the circumstances of the dispute. We will—quite reasonably—stick with the project of arguing with one another, attempting to reach agreement, for longer in some cases than in others. I will stick with the project of convincing my friends of the musical virtues of the Ramones for longer than I will stick with the project of convincing my grandparents. And I will stick with the project of convincing my grandparents for longer than I will stick with the project of trying to convince a Martian, or an Earthling who comes from a completely different musical culture.) (p.251)

So here are the central phenomena to be explained. First, that there are some disputes about taste that seem to be perfectly in order, perfectly worth pursuing, and perhaps even liable (if all goes well) to wind up producing agreement at the end of the day, and other disputes about taste that seem to be somehow defective, shallow, not worth pursuing, and/or not subject to any sort of resolution. Second, that disputes which start off in the first category can move into the second when the parties to them get a certain kind of evidence—at a first pass, evidence that they are differently constituted with respect to a certain class of capacities or sensory faculties.

What is right about the aphorism that we started with seems to be not that there is no disputing about taste, or even that there is no sensible disputing about taste, but rather that some disputes about taste are defective, and that they are defective in a distinctive sort of way. Aesthetic disputes are in danger of falling into a certain distinctive sort of defectiveness that many other sorts of disputes are not subject to. My goal in what follows will be to offer a theory that explains the difference between the sensible and the defective disputes about taste, in a way that draws the line in what seems like the right place, and gives us some insight into the sort of defectiveness that is at issue. I will do this by giving an account of what is at stake in aesthetic disputes—of the upshot of these disputes' being resolved in favor of one of the competing views, and of what the parties to an aesthetic dispute are (typically and centrally) aiming to achieve by engaging in it.

Before we move on, it is worth noting a distinction between two different sorts of uses of aesthetic vocabulary, and pointing out that the phenomena I have just been drawing attention to arise for only one of them. One sort of use, which does not behave in the ways I have just been describing, is what we might call the “baby‐food” use of aesthetic vocabulary: when feeding the baby, I might say, “those puréed green beans sure are tasty,” even though I know that they would taste absolutely revolting to me. Similarly, I might, after noting that Fido rolled enthusiastically in the three‐day‐old dead fish while ignoring the lilacs, say, “I guess the dead fish smells better than the lilacs.” 2 Disputes over these sorts of uses of the vocabulary of taste (p.252) do not seem to be subject to the special sort of defectiveness that will be the topic of this chapter. Call these (to have a more dignified and less misleading name than “baby‐food uses” or “dead‐fish uses”) sympathetic uses of aesthetic vocabulary. A natural thing to say about these cases is that I am sympathetically using the aesthetic vocabulary as if from someone else's perspective.

We can contrast these with first‐personally committed (henceforth just committed) uses of aesthetic vocabulary, on which one is subject to accusations of dishonesty or hypocrisy if one first asserts “Vegemite is tasty,” and then balks (without special pleading) at eating it, or asserts “this opera is captivating,” and then yawns (without special explanation) through the performance.

When we are using aesthetic vocabulary sympathetically, our willingness to assert and assent to assertions of “the dead fish smells better than the lilacs” hinges not on our views about our own reactions or dispositions to react to the smells of dead fish and lilacs, but on our opinions about Fido's reactions or dispositions to react to the smells of dead fish and lilacs. I am not subject to any charges of insincerity if I assert, sympathetically, that “the dead fish smells better than the lilacs,” and then steer clear of the dead‐fish‐smelling areas and seek out the lilac‐scented ones. Nor will you be subject to charges of insincerity if you accept my assertion, and then join me in my fish‐avoiding, lilac‐seeking behavior.

When we are using aesthetic vocabulary committedly, our willingness to assert, and to assent to assertions of, “the dead fish smells better than the lilacs” does hinge on (our views about) our own reactions, or dispositions to react to, the objects in question. I will typically be subject to charges of insincerity if I follow up a committed assertion of “the dead fish smells better than the lilacs” with behavior that reveals a clear, stable preference for the smell of lilacs over that of dead fish. And you will typically be subject to charges of insincerity if you accept my committed assertion when you in fact have a robust preference for lilacs over dead fish. The point about acceptance is important. It is not the case that we should accept committed assertions of aesthetic claims whenever we think the speaker has got the relevant sorts of preferences, dispositions, or whatever. The person who accepts an aesthetic assertion, no less than the one who makes it, is liable to charges of insincerity if they later betray that they lack the relevant (p.253) preferences, attitudes, and so on that go along with that sort of aesthetic evaluation.3

At this stage, I want to avoid signing up for any specific proposals about the semantics of either committed or sympathetic uses of aesthetic vocabulary. What I want to do is just note the contrast, and set the sympathetic uses aside. It is only disputes involving committed uses of aesthetic vocabulary that are subject to the distinctive sort of defectiveness that we are investigating here, and that make it attractive to say things like “there is no disputing about taste.” Disputes about sympathetic uses are in general pretty straightforward—once we have figured out how Junior responds to the beans, or how Fido responds to the fish, that is the end of it. Any defectiveness in these sorts of disputes is just of the same kind as we find in the Caesar's fleas‐type cases.

2. Ground Clearing: Contrasts and Desiderata

So far, I have been a bit sloppy about setting up the problem. Let us now be more careful. First, let us get clearer about what a “dispute about taste” is, exactly. In order not to close off any options prematurely, I am going to characterize disputes about taste as a certain sort of conversational exchange involving a certain sort of vocabulary.

I will call the sort of vocabulary in question aesthetic vocabulary, or vocabulary of taste, by which I mean to include predicates such as “fun,” “tasty,” “disgusting,” “beautiful,” “elegant,” and the like. I will not say much more about what it takes to be part of “and the like”—I intend the account that follows to apply broadly to aesthetic predicates and predicates of personal taste, but I want to proceed by identifying some core cases, offering an account that seems to work for them, and then letting the chips fall where they may as far as the exact boundaries of the class of expressions for which such an account is appropriate. (I have in mind something like Sibley's notion (1959, 1965) of an aesthetic expression.) (p.254)

The sort of exchange in question is the sort that happens when the following two things happen:

  1. 1. One party to a conversation assertively utters some sentence S and the other assertively utters a sentence (call it ¬S) that looks, as far as its surface structure goes, like S's negation; call this a case of superficial denial. (I will be concerned, almost exclusively, with pairs of sentences of the form x is F and x is not F.)

  2. 2. The parties take their utterances to be in conflict, and go on to engage in a process of argument, negotiation, bullying, and so on with the aim of arriving at a common position, in which both parties to the dispute are prepared to assert, and to accept assertions of, one of the contested sentences, and both are prepared to deny the other. I will call this sort of process, aimed at this sort of outcome, a dispute, and, if and when the aim is achieved, I will say that the dispute has been resolved in favor of whichever sentence the parties to the conversation wind up collectively adopting as the one to assert and accept, and that the parties are aligned with respect to which of the contested sentences they are inclined to assert.4

Many cases of superficial denial are merely superficial. For example, crucial expressions are sometimes indexical or otherwise context dependent. (Philosopher Phil asserts “philosophy is my field,” and Linguist Larry responds with “philosophy is not my field.”) In these cases, there is typically no occasion for dispute, since the pairs of assertions are not in any tension with each other—one can perfectly well accept both, and each party to the conversation can perfectly well continue to stand by his own assertion while accepting the other's. There is no need to arrive at a common position about which of the sentences is the one for them, collectively, to assert, and indeed the project of getting (p.255) aligned on which sentence to assert would often be deeply misguided and counterproductive. In Phil's and Larry's case, for example, getting both of them into the sort of doxastic state in which they could sincerely assert the same sentence would require one of them to believe something false.

It would be natural to contrast merely superficial denial with semantic denial, in which the proposition asserted by the second party to the exchange is the negation of the proposition expressed by the first. But, since there are other ways for a pair of assertions to be in conflict than by one expressing P and the other expressing ¬P, it will be useful to use a more general, though somewhat harder to pin down, notion of genuine conflict. There is a genuine conflict between an assertion of S by A and an assertion of ¬S by B iff neither party can consistently accept the other's assertion without withdrawing, ceasing to stand by, and ceasing to be prepared to repeat their own. That is, A cannot consistently both stand by her original assertion and remain willing to assert S, while simultaneously accepting B's assertion of ¬S, and B cannot consistently stand by her original assertion and remain willing to assert ¬S, while simultaneously accepting A's assertion of S. When two assertions are in genuine conflict, each party's making, and continuing to endorse, its own assertion commits it to rejecting the other's.

We can think of this, if we like, in terms of the incompatibility of the conversational demands imposed by the two assertions: speech acts (including assertions) have, if the parties to the conversation all cooperate, distinctive sorts of effects on the conversational context in which they occur. We can model this by saying that speech acts impose certain sorts of demands on the context—they demand that the context be changed or updated in a certain way. Since the state of the context depends (at least to a large extent) on the actions and attitudes of the parties to the conversation, speech acts also impose, in the same sense, demands on the other parties to the conversation: demands that they do what they have to do, and adopt the attitudes that they need to adopt, in order for the context to undergo the relevant changes. There are, that is, certain things that the audience members must do, and/or attitudes that they must adopt, in order to go along with a given assertion (or other sort of speech act). If everyone goes along, their going along brings it about that the context has been changed in some distinctive way. We have a genuine conflict (p.256) between two assertions when they impose incompatible conversational demands.5

Disputes in the absence of genuine conflict are pretty clearly defective, and are bad uses of one's time and energy. If there is no genuine conflict (as in the case of Phil and Larry), the thing to do is typically for each party happily to accept the other's assertion, and move on to other topics. (Of course, one of the parties might have some independent reason to reject the other's assertion—Phil might suspect that Larry is lying about what he does for a living, for example. But the mere fact that Phil has gone on record with his own assertion of “philosophy is my field” does not present any obstacle to his accepting Larry's assertion of “philosophy is not my field.”)

The presence of genuine conflict is not all that is required in order to have a dispute that it makes sense to pursue, though. Return to the dispute about Caesar's fleas. Martha says, “Caesar had more than 1,000 fleas when he crossed the Rubicon,” and Elizabeth responds, “Caesar did not have more than 1,000 fleas when he crossed the Rubicon.” It would be a mistake for Martha and Elizabeth to spend a lot of time and energy fighting about this. What is wrong with their dispute is not that they have not succeeded in making conflicting assertions. Of course they have. Here, what is wrong with the dispute is that the project of resolving it just is not going to be a good use of their time. It will not be a good use of their time because (a) there is not much hope of success, and (b) success would not bring much of a payoff, anyway. It is going to be impossible (or nearly so) to get the evidence that they would need in order to resolve the dispute in favor of one claim over the other, and the question is (unless their circumstances are extremely odd) pretty uninteresting and unimportant.

Some genuine conflicts are sufficiently unimportant, given the purposes at hand, that the best thing to do is to pass over them in silence, rather than expending the effort that it would take to resolve them. Some would be extremely difficult to resolve, owing to the inaccessibility of the evidence that would need to be deployed in order to convince both parties to get (p.257) aligned on which sentences to assert and accept. Some would be extremely difficult to resolve, owing to the stubbornness with which one or the other party is likely to cling to the view with which they began. (These considerations interact with one another in all of the obvious ways. If the question is important enough, it is worth fighting against a lot of intractability, or going to a lot of trouble to obtain the relevant evidence, in order to resolve the dispute. The more intractable the disagreement, the more important it has to be in order for the project of resolving it to be worth engaging in, and so on.)

There are, then, two requirements on a dispute's being a sensible one to engage in: first, there has to be a genuine conflict. And, second, the conflict has to be one that it is worth resolving—the project of getting both parties aligned on which of the competing sentences is the one to assert and accept has to be one that it is worthwhile to engage in. Call these the CONFLICT and the WORTHWHILENESS requirements on sensible disputes.

Whatever has gone wrong in defective disputes about taste, it does not seem to be the same thing that has gone wrong in Phil's and Larry's dispute. That is, it does not seem to be a failure to satisfy the CONFLICT requirement. When I say “broccoli tastes better than Brussels sprouts” and Mira says “broccoli does not taste better than Brussels sprouts,” it would (probably) be a mistake for us to invest a lot of time and energy in a dispute. But the reason for this certainly does not seem to be that we have failed to express conflicting views about broccoli and Brussels sprouts, so that we could just as easily simply accept each other's assertions with an interested nod and a polite, “yes, I see.” Given what I have asserted, I cannot accept Mira's assertion without withdrawing my own. And, given what Mira has asserted, she cannot accept my assertion without withdrawing hers. (Similarly, a third party could not happily accept both of our assertions—an onlooker could not sincerely accept my assertion of “broccoli tastes better than Brussels sprouts” and then go on sincerely to accept Mira's assertion of “broccoli does not taste better than Brussels sprouts” without changing his views about the vegetables in question.)6

So it must be that the defective disputes about taste are defective because they fail to satisfy the WORTHWHILENESS requirement. But the failure of (p.258) WORTHWHILENESS in my dispute with Mira seems to be something deeper than the failure of WORTHWHILENESS in Martha's and Elizabeth's dispute about Caesar's degree of flea‐riddenness. It is not just that the evidence that would point us toward the real, objective facts about whether broccoli tastes better than Brussels sprouts is difficult to come by, or that the facts are not very interesting or important. The problem is that the idea that there is any crucial evidence to be found, that there are any objective facts in this domain to be discovered, seems deeply suspect. It is not that it would be hard to figure out which of us is making the error, or that whichever of us has made the error is unlikely to admit it—it is that there is something suspicious about the idea that there ever needs to have been an error here at all. There is a pretty strong inclination to say that the project of resolving this dispute one way or the other, and establishing a common view about which of our assertions is the one to accept, is a bad one to go in for, not just because the facts about who has really got the right view about broccoli and Brussels sprouts are hard to figure out or are not very important, but because both parties have already got the right view about the relative tastiness of those vegetables for them to have.7

So here are some desiderata for our account of the special sort of defectiveness to which disputes about taste are subject:

  • The defect should be a failure of WORTHWHILENESS, not of CONFLICT.

  • The defective disputes about taste are ones in which there is a conflict, but the project of resolving it is a bad one to go in for, rather than ones in which there is no conflict to be resolved.

  • The failure of worthwhileness should be something deeper than what is happening in the dispute about Caesar's fleas, and should be something that explains why we are drawn to the talk of faultless disagreement and mutual correctness.

3. A Solution

Let us consider a sensible dispute about taste for a moment. Though Brett has never tried peanut butter and chocolate together, he has great (though (p.259) misplaced) confidence in his ability to imagine how it would taste, and imagines that it would taste terrible. He sees someone walking by with a peanut butter truffle. Being a philosopher,8 he first names the truffle “Alfred,” and then asserts “Alfred is not tasty.” Yuri, who has extensive experience with peanut butter truffles, and trusts that Alfred is a peanut butter truffle much like any other, responds with “Alfred is tasty.”

One of the things that will happen if the dispute is resolved in Yuri's favor (so that, at the end of the day, Brett accepts Yuri's assertion, and Brett and Yuri are both prepared to sincerely assert “Alfred is tasty”), is that both Brett and Yuri will expect that, were they to eat Alfred (in typical circumstances), pleasant gustatory sensations would result. That is, there is a certain property, of being disposed to derive a certain sort of sensation from certain sorts of interactions with Alfred, that each of them will take themselves to have—that each of them will self‐attribute. Let us call the property being disposed to enjoy Alfred.

Of course, something similar is true about the consequences of the dispute's being resolved in Brett's favor—if this is how things pan out, then, at the end of the dispute, both Brett and Yuri will take themselves to have the property, not being disposed to enjoy Alfred.

The same sort of thing is a quite common feature of aesthetic disputes, and of aesthetic discourse generally. It is a quite general feature of committed uses of the vocabulary of taste that, when sincerely asserted and sincerely accepted, the end result is a pair (or group) of people, each of whom takes themself to have a disposition to respond to some object or type of object in a particular way (and takes that disposition to be shared by the other parties to the conversation). Quite generally, in any (committed) dispute about taste, there will be some property P (a property of being disposed to have a certain sort of response to certain items) such that, if the dispute is resolved one way, both parties to the dispute will wind up self‐attributing P, and, if it is resolved the other way, both will wind up self‐attributing the complement of P (henceforth written ‘¬P’).

This is not just an upshot of disputes—aesthetic assertions that are simply accepted without argument or discussion have the same sort of effect. One cannot sincerely accept a committed assertion of a sentence of the form x is tasty unless one takes oneself to be disposed to have a certain distinctive (p.260) sort of positive response to certain sorts of sensory encounters with x. One cannot sincerely accept a committed assertion of x is fun unless one takes oneself to be disposed to enjoy certain sorts of interactions with x. And so on. Neither will one be inclined sincerely to make a committed assertion of the form x is tasty, x is fun, and so on, unless one takes oneself to have the relevant sorts of dispositions. It seems to be a condition on both the sincere assertion and the sincere acceptance of committed assertions about taste that one self‐attribute the relevant disposition to respond. It looks as if, for each simple taste sentence—something of the form x is F where F is a predicate of taste—there is some property P (a property of being disposed to have certain sorts of responses to certain sorts of objects) such that it is a condition on one's either committedly asserting, or accepting a committed assertion of, the sentence that one self‐attribute P.9

This fact helps to make sense of some important features of our aesthetic practice. One very major role that aesthetic discourse plays is a sort of connection‐building role, in which people discover commonalities in the sorts of things that they enjoy, appreciate, or despise. This can be a substantial part of the process of building and maintaining interpersonal relationships, and in establishing and maintaining ties to communities and groups. Very many groups and subcultures are defined, at least in part, by the common aesthetic sensibilities of their members (and the contrast between their shared aesthetic sensibilities and those of outsiders). Think of, for example, such subcultures as goths, punk rockers, ravers, trekkies, bikers, and so on.

I propose that we should think of this effect of successful aesthetic assertions, and successful resolutions of aesthetic disputes, of inducing mutual self‐attribution of certain dispositions to have a particular sort of response to a particular (kind of) object, as the central business of assertions and disputes about taste, and not as a mere side effect.10 (p.261)

Note that I am not yet trying to offer a theory about why mutual self‐attribution of certain properties is at stake in disputes about taste—I am not yet offering an account of the semantics of aesthetic vocabulary that explains why that vocabulary is well suited for performing this purpose. I will take that task up later, but for now all I want to do is note that this does seem to be what is at stake (or among the things at stake) in these disputes, and use that fact to explain what is defective about the defective disputes about taste. Regardless of what turns out to be the best explanation of why successful resolutions of disputes about taste produce alignment in the self‐attribution of certain sorts of dispositional properties, the fact remains that the successful resolution of such disputes does in fact produce such alignment.

Later in the chapter I will go on to say some things about (a) just what kinds of dispositional properties are at stake, and (b) what sort of semantic theory we ought to provide for predicates of taste, such that we can predict that this is what will be at stake. But one need not accept these further claims in order to accept the explanation of the special defectiveness of certain disputes about taste that I will be offering presently.

The project of disputing about taste is a project of bringing it about that we are alike with respect to self‐attribution of certain properties. That is only a good project to engage in if we are, in fact, alike with respect to the properties whose self‐attribution is at stake in the dispute. If Juan has P and Jim has ¬P, it is a bad idea for them to bring it about that they are alike with respect to their self‐attribution of P or ¬P—either way they align themselves, one of them is going to have to be getting it wrong and self‐attributing a property that he does not have.

Clearly, this project of coordinating our self‐attributions of certain sorts of dispositions is an unproductive project for people who are not alike with respect to those dispositions to engage in. It is also a bad project for people who have sufficiently good evidence that they are not alike in the relevant respects to engage in. This gives us a way to mark off the distinction between the sensible and the defective disputes about taste: the sensible disputes are the ones where the parties are, and reasonably take themselves to be, alike with respect to the dispositional properties that are at stake in the dispute. The defective disputes are the ones where the parties either are not, or do not, reasonably take themselves to be, alike with respect to the dispositional properties that are at stake. (p.262)

Let us distinguish three species of this genus of defectiveness. First, it can happen that the parties to the dispute in which property P is at stake really are different with respect to P, and so bringing it about that either both parties self‐attribute P or both parties self‐attribute ¬P would require one of the parties to the dispute to self‐attribute a property that she lacks. There is clearly something wrong with such a dispute. (Though the parties to the dispute might be blameless in pursuing it, if they do not realize that they are different in the relevant respect.) Call such a dispute factually defective (since what is wrong with it is that the parties to the dispute are in fact different with respect to the property at stake in the dispute).

Secondly, it can happen that, regardless of whether the parties to the dispute are alike with respect to the property P, they are presupposing that they are different, and so they are presupposing that the dispute is factually defective. This is also pretty clearly not an occasion in which it is a good idea for them to go in for a project aimed at making them alike in self‐attributing P or in self‐attributing ¬P, and so this sort of dispute is also defective. Only slightly better is the case where the parties to the dispute do not presuppose that they are different, but also do not presuppose that they are alike. In this case, it is an open possibility, as far as they are collectively concerned, that the dispute is factually defective. This is also not the sort of occasion in which it is a good idea to coordinate on self‐attributing either P or ¬P. Call a dispute that is defective in either of these ways presuppositionally defective, since what is wrong with it is that the required presupposition of similarity is absent.

A final sort of defectiveness, in which we will be particularly interested in what follows, is one in which, whether the parties to the conversation are alike with respect to P or not, and whether they presuppose that they are alike with respect to P or not, it is not reasonable for them to presuppose that they are alike with respect to P. This is, again, going to be a sort of case in which it is not a good idea to coordinate on self‐attributing P or on self‐attributing ¬P, since it is an open possibility, as far as what they are reasonably entitled to presuppose, that the dispute is factually defective. Call such a dispute justificationally defective, since what is wrong with it is that the required presupposition of similarity is not justified.

Consider a dispute in which Smith asserts some aesthetic sentence S, and Jones responds with ¬S. There will be some dispositional property P (p.263) such that accepting Smith's assertion requires one to self‐attribute P and accepting Jones's assertion requires one to self‐attribute ¬P.

If Smith's and Jones's dispute is defective in any of the ways described above, it will have the following features:

  • INCOMPATIBILITY: Smith's assertion of S and Jones's assertion of ¬S cannot be simultaneously accepted—they really are in conflict.

  • POSSIBLE MUTUAL CORRECTNESS: It is either true, or compatible with everything Smith and Jones presuppose, or compatible with everything they are entitled to presuppose, that Smith is correct to self‐attribute the property whose self‐attribution motivates his assertion (since he really does have P), and Jones is correct to self‐attribute the property whose self‐attribution motivates her assertion (since she really does have not‐P).

  • NON‐SUPERFICIAL DEFECTIVENESS: Smith's and Jones's dispute is defective, and not just because they are talking past each other, or because the facts are hard to figure out, or the question is of no consequence. Their dispute is defective because the project of trying to resolve their dispute, so that Smith and Jones either both come to self‐attribute P or both come to self‐attribute not‐P, is a bad one for them to engage in (since either way of resolving the dispute would in fact either require one of them to self‐attribute a property that he lacks, or might, compatibly with everything they presuppose, or compatibly with everything they are entitled to presuppose, require one of them to self‐attribute a property that he lacks).

This combination of features makes, I think, for a pretty attractive story about what's defective about the defective disputes about taste. INCOMPATIBILITY lets us say that, even in the defective disputes, we have really got a disagreement, and really got a conflict between the two parties' assertions—the two parties to the dispute really are attempting to press incompatible views on each other. POSSIBLE MUTUAL CORRECTNESS lets us respect the intuition that, in the case of defective disputes, there is something suspicious about the project of trying to figure out who really is getting it right, and who has made the mistake. There is a pretty widespread inclination to say that these could be cases of disagreement without error, in which both parties to the dispute are getting it right, and that part of what is wrong with the dispute is that the parties to it are not warranted in making the assumption that one or the other of them must be getting (p.264) it wrong, an assumption that they need to make in order to sustain the dispute. This account of what is at stake in these disputes allows us to see what is right about these thoughts. NON‐SUPERFICIAL DEFECTIVENESS lets us say that what has gone wrong here is not just a shallow sort of talking‐past, or a routine sort of unimportance or epistemic inaccessibility, but something more specific to discourse about taste.

Another nice feature of this account of the distinctive sort of defectiveness that disputes about taste are subject to is that it predicts, as seems correct, that we will have many more sensible aesthetic disputes with our friends and neighbors, and with others who share our perceptual apparatus and cultural background, than we will with those who are very different from us, either in the configuration of their sensory organs or in the sort of culture they come from. We will not have a lot of sensible aesthetic disputes with the Martians, but we will be able to have lots of sensible disputes with our fellow humans who were raised just down the block from us. Sensible disputes will become fewer and farther between, the farther our interlocutors get from our close cultural and biological neighbors, and the closer they get to the Martians.

We can also say something about how a dispute in which there is a danger of these sorts of defectiveness might be expected to proceed. Let us start by looking at an example.

Alan and Clare start off with a default presupposition that they are, in general, alike with respect to their dispositions to enjoy various foods. And so they presuppose that they are alike with respect to their disposition to enjoy Vegemite or not. Alan has good reason, based on his experiences with Vegemite, to believe that he is disposed to enjoy Vegemite. Presupposing that he and Clare are alike, and hoping to get her to recognize this similarity, he asserts “Vegemite is tasty.”

In order to accept Alan's assertion, Clare has to self‐attribute being disposed to enjoy Vegemite. (Call this property V from now on.) Clare has good reason, based on her experiences with Vegemite, to believe that she lacks this property. One of two things could happen at this point:

  1. (i) She could take Alan's assertion of “Vegemite is tasty” to be a sign that he has never tried Vegemite, or that his experiences with it have always been in unusual circumstances, and he has mistakenly concluded, based on a few flukey pleasant interactions (p.265) with Vegemite, that he has got a general disposition to enjoy the stuff. In this case, she is likely to respond with “Vegemite is not tasty,” and attempt to get Alan to accept her assertion and self‐attribute not being disposed to enjoy Vegemite (¬V).

  2. (ii) She could take Alan's assertion of “Vegemite is tasty” to be good evidence that they are not, after all, alike with respect to their dispositions to enjoy Vegemite, and stop presupposing that they are. In that case, she will say something like “maybe it's tasty to you—I find it disgusting,” which will, probably, get Alan to accept that they are not alike in the relevant respect, and end the dispute.

Suppose Clare opts for (i). Alan now has a similar pair of options. He can take Clare's resistance to indicate that they are different in the relevant respect, or he can take it to indicate that Clare has had no interactions, or has had flukishly unpleasant interactions, with Vegemite. If he does the second thing, Clare and Alan will be engaged in a dispute.

It is quite likely that they will each, for a while, make efforts to convince the other to self‐attribute the relevant property. Clare might encourage Alan to expose himself to Vegemite in the sorts of situations in which his true dispositions are likely to manifest themselves. Alan might attempt to draw Clare's attention to enjoyable‐making features of her experiences of Vegemite to which he expects she has not attended.

At some point, one of two things will happen. Perhaps one or the other will be convinced by the other's efforts. Clare, for example, might finally try Vegemite spread to the correct thickness on the right sort of cheese, and become convinced that she really is disposed to enjoy Vegemite. She will then withdraw her original assertion and join Alan in asserting “Vegemite is tasty.” Alternatively, they might eventually conclude, based on the persistent failure of their efforts to change each other's views, that they really are different with respect to their disposition to enjoy Vegemite. In that case, they will drop the presupposition of similarity.

They are likely, at that point, to stop insisting on their original assertions and retreat to explicitly relativized assertions. Alan is likely to stop asserting “Vegemite is tasty” (which Clare cannot sincerely accept without self‐attributing V) and retreat to asserting “Vegemite is tasty to me” (which Clare can sincerely accept without self‐attributing V—in order to accept (p.266) this assertion, all she has to accept is that Alan has V). Meanwhile, Clare stops asserting “Vegemite is not tasty” (which Alan cannot sincerely accept without self‐attributing ¬V) and retreats to asserting “Vegemite is not tasty to me” (which Alan can sincerely accept without self‐attributing ¬V—in order to accept this assertion, all he has to accept is that Clare has ¬V). Since these new, explicitly relativized assertions are not in conflict—it is perfectly possible to accept both—this should be the end of the dispute.

In general, the predicted pattern is:

The dispute starts with a presupposition of similarity in place. The parties to the dispute make conflicting assertions. They try for a while to bring each other around. Eventually either one of them succeeds, or they get enough evidence for difference that the presupposition stops being reasonable. If they continue to engage in the dispute at that point, the dispute turns defective. (And, of course, the dispute will be defective right from the start if the presupposition was not reasonable to begin with.)

Some disputes will be more robust than others, because the robustness of the presupposition of similarity will be different in different cases. There will be some disputes about taste such that it is a bad idea to engage in them at all, since the presupposition is just not reasonable from the start. There will be others that we ought to give it up at the first sign of conflict. There will be other disputes such that it is just a bad idea to invest a lot of time in them, since the presupposition will very quickly (though not quite immediately) become implausible. Other disputes will, owing to the robustness of the presupposition of similarity, support a great deal of debate despite persistent disagreement.

When the presupposition of similarity does become implausible, the thing to do is typically to stop insisting on the original assertion (and thereby stop imposing the demand on one's interlocutor to self‐attribute the property) and retreat to talking in explicitly relativized terms. That is, to stop trying to get your interlocutor to accept that he has the property in question, and just aim for the weaker goal of getting him to accept that you have it. (This is what happens when Alan stops asserting “Vegemite is tasty” and starts asserting “Vegemite is tasty to me.”)

So the story is: disputes about taste are aimed at getting all of the parties to the dispute on the same page with respect to the self‐attribution of some property—either everybody taking themselves to have it or everybody (p.267) taking themselves to lack it.11 This explains why the disputes turn defective when the parties to them get enough evidence that they are different from each other in the right kind of way. It explains why there are some disputes about taste that are quite robust, and others that stop being sensible at more or less the first sign of disagreement. It explains why we get the sense that it is possible for both parties to such a dispute to be right—each really does have the property whose self‐attribution they are attempting to push on their interlocutor.

There are still two big questions in need of answers. First, given that self‐attribution of some dispositional properties or other is at stake in disputes about taste, what should we say about which properties, exactly? Secondly, what do we need to say about the semantics of predicates of taste (and perhaps about the pragmatics of the surrounding discourse) in order to make it turn out that disputes about taste aim at getting all of the parties to the conversation to self‐attribute those properties? I will address the first of these in the course of answering a potential objection, the full response to which will also require a bit of a revision to the account of what is required for these sorts of disputes to be sensible. The second I will address, rather sketchily, at the end of the chapter.

4. Which Properties?

A natural worry to have about this sort of proposal is that it will predict too much defectiveness and fragility. The defective disputes about taste are the noteworthy exception rather than the rule, and we might be concerned that sensible, robust aesthetic disputes will be too hard to come by on such an account. Here is a reason to be worried. If, when I try to get you to accept my assertion of “broccoli is tasty,” what I am trying to do is get you to self‐attribute a disposition to enjoy broccoli, why is it that these disputes do not just fizzle at the first sign of resistance? After all, you are in a better position to know what kinds of responses you get from broccoli than I am—should I not just take your word for it when you, by initially refusing to accept my assertion, signal to me that (p.268) you do not take yourself to be disposed to enjoy broccoli? Why think that the presupposition of similarity ever survives even the first sign of conflict? This is a potentially very serious worry—there is a danger that the account according to which what is at stake in disputes about taste is the participants' self‐attribution of certain properties (having to do with one's dispositions to respond in particular ways to the objects in question) will predict that disputes about taste should be a lot less robust than they actually are.

There are two kinds of response to this sort of concern. First, we can look for properties such that the presupposition that the parties to the dispute are alike with respect to them will be reasonable in a lot of cases, and will remain reasonable even in the face of a fair bit of conflict. Secondly, we can expand the range of disputes that are predicted to be worthwhile to engage in by finding some circumstances in which disputes that aim at the participants' self‐attributions of some property P are reasonable to engage in, even in the absence of a reasonable presupposition that the parties to the dispute are alike with respect to P. I will pursue the first strategy in this section, and the second in the next.

In figuring out which properties are at stake in which aesthetic disputes, we are not just out to maximize robustness. What we want is some properties such that the presupposition that we are alike with respect to them has the right degree of robustness. The right degree will be different for different disputes—what we want is, for each dispute, to find a property to be at stake in that dispute such that the presupposition that we are alike with respect to that property has the same degree of robustness as the dispute. We want the plausibility of the presupposition to run out at the same time as the sensibility of the dispute. Following are some ways to make the disputes increasingly robust, by making the presuppositions of similarity with respect to the properties at stake more robust. Which moves to make (or not) in a given case will depend on how much robustness we want in that case. (And, in fact, robustness will be a multidimensional affair—it matters, not just how much evidence it takes to undermine the presupposition of similarity, but also what kind of evidence.)

The first move toward greater robustness is one that we have already made: make sure the properties are dispositional. We can be wrong about our dispositions. For example, as mentioned in several examples above, it (p.269) could be that our past experiences with the items in question have been in deviant circumstances, which have prevented our actual, robust dispositions from manifesting themselves.

The second way to increase the robustness of a dispute about taste is to make the dispositions that are at stake ones that are liable to be widely shared. Dispositions that are products of widely shared features of our sensory apparatus, for example, would be good candidates. The more local the dispositions are, the more rooted in culture rather than biology, or rooted in individually variable biological features rather than species‐wide features, the less robust the presupposition of similarity is likely to be.

This maps pretty nicely onto the sort of “fit with natural capacities” account of taste and aesthetics that we find in, for example, Hume (1757/1965) and Railton (2003, n.d.). Hume finds the foundation for sensible disputes about taste in “a certain conformity or relation between the object and the organs or faculties of the mind.” Railton, following Hume, finds it in the facts about “what matches best and most durably the potentials of our underlying structures” (Railton 2003: 96) or “a particular sort of robust and general match between objects or performances and widespread human sensory capacities and sentiments . . . that permits those objects and events to bring about intrinsically sought, perceptually based experiences in those who become acquainted with them” (Railton 2003: 102). On a Humean account of the sort Railton favors, what it is for something to be beautiful is, approximately, for it to be the sort of thing that is robustly disposed to interact with widespread human sensory capacities and sentiments in a way that produces experiences of an intrinsically desirable sort. The natural extension from beauty to tastiness will say, approximately, that what is tasty is what is robustly disposed to interact with widespread human gustatory capacities and sentiments in a way that produces gustatory experiences of an intrinsically desirable sort. This sort of account of aesthetic qualities is, I think, extremely appealing.

On the Railtonian account, then, “Vegemite is tasty” will express something like the proposition that Vegemite is robustly disposed to interact with widespread human gustatory capacities and sentiments in a way that produces gustatory experiences of an intrinsically desirable sort.12 We can capture much of (p.270) what is appealing about this sort of proposal by modifying it slightly, so that what is at stake is not (or is not merely) acceptance of possible‐worlds propositions about widespread human capacities being thus‐and‐so, but rather (or also) self‐attribution of properties of having capacities that are thus‐and‐so. The property we find in the vicinity of the Railtonian proposal about “Vegemite is tasty,” then, is: having gustatory capacities and sentiments that are robustly disposed to interact with Vegemite in a way that produces gustatory experiences of an intrinsically desirable sort.

The general schema:

The property whose self‐predication is at stake in a dispute about taste will be of the type: having F‐capacities that are robustly disposed to interact with x in a way that produces G experiences. How we substitute for F and G will depend on which predicate is being deployed in the dispute, and how we substitute for x will vary depending on what the predicate is being applied to.

Besides inheriting many of the attractive‐making features of Railton's Humean account, this will also get us quite a bit of robustness for aesthetic disputes. These sorts of dispositional properties are clearly the sorts of things that we can mistakenly take ourselves to have, or mistakenly take ourselves to lack. And so, it will not always be a good idea to defer to our interlocutors about whether or not they have the dispositions in question.

Another way to increase the plausibility and robustness of the presupposition of similarity is to make the properties in question ones having to do not with our current responses or disposition to respond but with the ones we would have in the ideal. For example, maybe what is at stake in a dispute over ⌜x is beautiful⌝ is something like, being someone whose suitably idealized self would be disposed to have experiences of an intrinsically desirable sort as a result of looking at/hearing/contemplating/etc. x.

This is likely to make the properties in question even less luminous than the sorts of dispositional properties we have just been discussing, and allows more room for the parties to the dispute to be mistaken about whether they have the property or not. I could be disposed, after idealization, to enjoy Vegemite without being disposed, as I am now, to enjoy Vegemite. It also makes it more plausible to think that we are alike, even if we do not think that convergence in the ideal is guaranteed. This sort of account also helps us account for the apparent fact that the (p.271) proper appreciation of certain sorts of aesthetic qualities requires a sort of training, experience, and so on, and captures an attractive bunch of intuitions about taste—about the knowledgeable and experienced being more reliable trackers of the relevant features of things, and about the possibility of gustatory, culinary, and aesthetic self‐improvement (rather than mere change).13

There is a species of the idealization move available here that will make the presupposition of similarity even more plausible. We might want to say that, in some cases anyway, congruence with one's neighbors is part of what constitutes being ideal in the relevant respect. Then just the fact that we are interacting with one another would give us reason to think that we will converge in the ideal, because part of what goes into determining where our ideals are is pressure toward convergence. This would be attractive on the sort of metaethical view advocated by Gilbert Harman (in, e.g., Harman and Thomson 1996), where morality is something like a negotiated system of norms whose purpose is to let us interact smoothly with each other. I suspect that this might have some appeal for at least some aesthetic qualities, too.

Relatedly, we might say that the properties in question are sometimes group membership properties—properties such as being a member of a natural biological kind/cultural group/etc., typical members of which are disposed to have G experiences in response to x. This would allow us another sort of flexibility—if I take myself to be an atypical member of my group, I can self‐attribute the relevant group‐membership property, even if I do not take myself to have the relevant dispositions.

There is a lot of room for variation, across different aesthetic predicates, in which sorts of properties one must self‐attribute in order sincerely to apply them to something. There are, I think, a lot of extremely interesting questions here about just which aesthetic predicates to associate with which sorts of properties. But now is probably not the time to pursue the detailed questions about particular predicates. What I hope to have shown is just that the sort of account proposed here has a battery of resources at its disposal that seem well suited to capture the interesting differences between different sorts of aesthetic predicates, and to capture the variety of phenomena that we find in a satisfying way. (p.272)

5. Sensible Disputes without (or with Dubious) Presuppositions of Similarity

One way to expand the range of disputes about taste that your theory predicts will be sensible—the one we have just been exploring—is to be sure to pick the right properties to be at stake in those disputes. Another way—the one we are about to explore—is to draw attention to phenomena that either (a) lower the bar for how plausible the presupposition of similarity has to be in order for it to be reasonable to make, or (b) make room for sensible disputes about taste even in cases where the presupposition is absent.

One bar‐lowering phenomenon that it is worth drawing attention to is the potential pragmatic importance of establishing a common view about some questions of taste and aesthetics. In some cases, a failure to get on the same page with respect to whether we have got the relevant properties is going to make trouble for our capacity to cooperate with one another, and to coordinate our actions. And in some of these cases the sort of coordination of action that failing to coordinate our self‐attributions of the relevant properties would make trouble for will be quite important. Sometimes it is not a big deal whether we agree about the attractiveness of a certain sofa. But sometimes we are trying to decorate a house together, and it is important that we find a sofa that interacts nicely with both of our suites of sensory capacities and sentiments.

In these sorts of cases—where it is going to be a bit of a disaster if we turn out to be different in the relevant respect, and cannot coordinate on how we think that the object in question is liable to interact with our (possibly idealized) sensory capacities, and so on—it might be reasonable to cling to the presupposition that we are, after all, alike in the relevant respect, even after a fair bit of evidence has accumulated that we are not. This will be particularly likely in cases where there is not much to do about it if we are not alike except to give up on any kind of coordinated action in this domain—if diverging opinions on this sofa would scuttle the whole project of living in the same house, or diverging opinions on the tastiness of various foods would scuttle the project of cooking together, for example. (And, of course, the effect will be stronger if this is a domain in which it is particularly important to be able to coordinate our actions.) (p.273) This is, of course, going to be a degreed phenomenon: the more important establishing a common view is to the prospects of coordinating behavior, and the more important it is to coordinate in the relevant domain, the greater the incentive to stick with the presupposition of similarity in hopes that it will be borne out.

(A related way to lower the bar for how plausible the presupposition of similarity has to be in order for aesthetic assertions to be in order is to note the possibility of using assertions about taste as a sort of low‐cost bid to establish similarity—and thereby to find potential partners for coordinated action. One can make the assertion and quickly withdraw it if it meets with resistance, but if somebody jumps up and says “yeah! that is absolutely right!,” you may have found a friend.14)

There are also two kinds of phenomena that make room for sensible disagreement even in the absence of a pre‐existing presupposition of similarity.

The first is the familiar phenomenon of accommodation. In the absence of a pre‐existing presupposition that we are alike with respect to some property F, I could still make an assertion that is felicitous only in the presence of such a presupposition, in hopes that the audience will, recognizing that I have just made an assertion that is felicitous only in the presence of a presupposition that we are alike with respect to F, accommodate my assertion by bringing such a presupposition into effect. This opens up quite a bit more room for sensible disputing about taste.

The sorts of pragmatic considerations canvassed above, about the potential importance of establishing common ground to coordinated action, also point to another kind of case where we could potentially find sensible disagreement without a presupposition of similarity. In these sorts of cases, we are liable to have reasons to try to produce similarity where it was previously absent. In these sorts of cases, we can understand the parties to the disagreement as trying to arrive at a common view not by one party recognizing that she has in fact had the disputed property all along, but by acquiring the property in response to pressure from the other disputant. These are cases in which the aim of the dispute is to change the other party's (p.274) taste, not to get him to correct his previously mistaken views about what his tastes were like at beginning of the dispute.

Suppose you want someone to self‐attribute being green. One strategy for how to bring this about involves a mirror. Another involves a bucket of green paint. If your interlocutor is already green, and just does not know it yet, you can get him to self‐attribute being green by holding up the mirror. If he is not already green, you can still get him to self‐attribute being green by doing the right sorts of things with the paint. (In fact, there are two different paint‐involving strategies. One is to paint him yourself, and ensure that he notices that he has been painted. The other is to convince him to paint himself. In normal circumstances, he will notice that he is doing this.)

There is a conversational precedent for this last, get‐them‐to‐paint‐themselves way of getting people to self‐attribute properties. Think about the sorts of assertive orders discussed by Anscombe (1957) and later put to use by David Velleman (1989). The doctor asserts, in the presence of the orderly, “the orderly will take the patient to the operating room,” or “the orderly is taking the patient to the operating room now,” in order to bring it about that the orderly takes the patient to the operating room, not in order to get her audience (including the orderly) to accept the antecedently well‐supported fact that the orderly was about to take the patient to the operating room. (Depending on one's view of future contingents, there may not have been such a fact prior to the doctor's assertion and the orderly's cooperation.)

Other examples of this are not hard to find—it is easy to construct the sorts of contexts in which “you will bring me a coffee,” “Brian will be at the club at midnight,” and “Juan and Carolina are taking down the front door, Sarah and Liz are covering the back,” could be uttered in order to get the audience to bring it about that things are as the assertion represents them to be, not to get the audience to recognize that things already were (going to be) that way.

We can understand these in terms of the usual sort of conversational demands being complied with in an unusual way. The assertion, “you will bring me a coffee,” even when Daniels uses it to give an order, in the first instance applies pressure on O'Leary to come to accept that O'Leary will bring Daniels a coffee. O'Leary cannot (in the usual case) accept that unless he takes it to be true, and he is in a position to make it true, or to make it false. So, in order to comply with the pressure to accept that he will bring Daniels (p.275) a coffee, he has got to make it true that he brings Daniels a coffee. And so the pressure to accept that the content of Daniels's assertion is true is leveraged into pressure to bring it about that the content of Daniels's assertion is true.

The same sort of thing could happen on the present account of disputes about taste, so long as the properties in dispute are ones such that we, or our interlocutors, have some control over whether we have them. It is, of course, an open question which, if any, of the relevant properties we have such control over. (It is easier to be sympathetic to this sort of picture if one bears in mind that we need not have immediate, instantaneous control—it need not be that we can just, as it were, flip some mental switch and acquire the properties in question. It could be that it requires some long‐term Pascalian process of exposing ourselves to the right influences and so forth.)

There is also, at least in some cases, the prospect of using an analogue of the paint‐them‐yourself method. Consider an idealization property, and a situation in which there are multiple equally good ways for me to idealize from my present position—my present condition requires that I idealize either toward end‐state A or toward end‐state B, but does not determine which one. Once I have moved far enough toward either endpoint, though, I will be committed—further idealization from any point far enough along toward A can only move toward A, and further idealization from any point far enough along toward B can only move toward B. (Think of ideal endpoints as exerting a gravitational pull, and think about the positions in which one is just in between two equally strong attractors. Alternatively, think of movement toward the ideal as climbing a mountain, and think about the people who are presently living in valleys between two equidistant mountains.) In this sort of case, it could happen that the upshot of your arguing with me, emphasizing certain features of the object(s) we are talking about, getting me to undergo some experiences and do some thought experiments, and so on, is not that I come to realize that I was already going to idealize to be an x‐enjoyer, but that you push me far enough down the road toward being an x‐enjoyer that the previously indeterminate idealization‐facts become determinate.15 (Also, on the sort of idealization story where convergence is part of what constitutes the relevant (p.276) sort of idealization, maybe you can push my endpoint of idealization around just by interacting with me, or by taking steps to ensure that you are in the group convergence with which matters.16)

There are also some disputes about taste that are sensible, not because of any prospects of arriving at a resolution, but for some other reason. Many cases of arguments between sports fans are like this—the participants engage in a sort of pretense of attempting to convince each other that their favorite team or player has the most electrifying offense, the most terrifying defense, the filthiest curveball, and so on, but what makes the dispute worth engaging in is not the prospects of a successful resolution. Sometimes the dispute is just enjoyable in itself. Sometimes it is valuable because engaging in the dispute helps one better to appreciate the aesthetic qualities of the items under discussion. Sometimes the value is in the extra appreciation of the merits of one's own view that one acquires in the process of defending it against attack. Sometimes the process of mutual discovery, in which the parties to the dispute come better to understand each other's aesthetic sensibilities, even without coming to share them, makes the dispute worthwhile.17

So while, to a large extent, the status of an aesthetic dispute as sensible or defective will track the presence or absence of a reasonable presupposition that the parties to the dispute are alike with respect to the property whose self‐attribution is at stake in the dispute, these two will not walk perfectly in step.

6. Semantic Proposals

We have got a story about the upshot of disputes about taste—about what a successful resolution requires—that lets us explain the distinctive sort of defectiveness that they are subject to. What we are still missing is a story about the semantics that predicts that those disputes will have that sort of upshot. I will not attempt a detailed spelling‐out of such a semantic theory here—instead, I will give a rough sketch of what seem to me to be the (p.277) two most promising proposals, and say a little bit about which I prefer and why.

The first proposal begins by adopting a Lewisian (1979) account of the propositional attitudes, on which the objects of belief, desire, and so on are properties, and a Stalnakerian (1978) account of assertion, on which accepting an assertion requires that one accept its content. Then one very straightforward way to explain the fact that, in order to resolve a dispute about S and ¬S, both parties need to come to self‐attribute either P or ¬P, is to say that the content of S is P, the content of ¬S is ¬P, and accepting an assertion requires believing (that is, self‐attributing) its content.18 (Officially, we will want to allow for forms of acceptance that are weaker than belief, and so we will want to allow for something like for‐purposes‐of‐this‐conversation self‐attribution. I will continue to abstract away from this.)

Here is some motivation for adopting the Lewisian view about the objects of propositional attitudes: there is a certain doxastic similarity between all of the well‐informed people with burning pants, and a certain conative similarity between all of the kids who want to grow up to be firefighters. One way to capture these similarities is to say that there is some potential object of propositional attitudes that all of the well‐informed people with burning pants believe, and some potential object of propositional attitudes that all of the kids who want to grow up to be firefighters desire.19 We cannot say this if we think that, necessarily, the objects of the propositional attitudes are always possible‐worlds propositions. (The only candidate possible‐worlds propositions in the neighborhood, when both Jane and Carlos want to be firefighters, seem to be the singular proposition about Jane and the one about Carlos, and the existentially quantified proposition. And it is not desiring any of these that marks the relevant conative similarity between Jane and Carlos. Jane and Carlos could both want to be firefighters without sharing a desire that Jane be a firefighter. And everybody, not just the aspiring firefighters, desires that somebody be a firefighter.) We can say it, however, if we think that properties are (or can be) the objects of propositional attitudes. What Jane and Carlos have in common is a desire directed toward the property, being a firefighter, and what all the (p.278) well‐informed people with burning pants have in common is that they all self‐attribute (that is, believe) the property, having burning pants.

Once we have made room for properties as potential objects of belief, we have gone a long way toward making room for properties as the objects of assertion. And if we think that the upshot of a successful assertion is the addition of the assertion's content to the conversation's presuppositions (to the stock of potential objects of belief which all of the parties to the conversation believe, believe that the others believe, etc.), then we have a straightforward explanation why successful resolution of a dispute over some aesthetic sentence S results in the mutual self‐attribution of some property (of one of the sorts discussed above). It's because successful resolution of the dispute requires both parties to accept an assertion of a sentence (either S or its negation), the content of which is a property of the relevant type. Acceptance of an assertion requires believing its content, and to stand in the belief relation to a property is to self‐attribute it.

Given this Stalnakerian view of the relation between assertion and content, and the view that sentences like “Vegemite is tasty” or “the symphony is beautiful” have self‐locating content, we can explain all of the phenomena.

There is genuine conflict between assertions of Alan's assertion of “Vegemite is tasty” and Clare's assertion of “Vegemite is not tasty”, because accepting Alan's assertion would require us to self‐attribute (something like) being disposed to enjoy Vegemite, and accepting Clare's would require us to self‐attribute not being disposed to enjoy Vegemite. No one can simultaneously accept both assertions, and so they are in conflict—they impose incompatible conversational demands on the parties to the conversation.

It is absolutely crucial to making this sort of story work that we take the relation between content and assertion to be the one described above, according to which the essential effect of an assertion with content P is that cooperative and credulous audience members come to accept P. (Which means, in the case of assertions whose content is some property P, that cooperative and credulous audience members come to self‐attribute—that is, take themselves to have—P.) We do not get any sort of conflict if our view of the relation between assertion and content is that, in the case of self‐locating assertions whose content is some property P, cooperative and credulous audience members come to accept that the speaker has P. (p.279)

In the case in which Alan and Clare are just differently gustatorily constituted, their dispute is defective because it is a bad idea for them to add either the property being disposed to enjoy Vegemite or not being disposed to enjoy Vegemite to their conversation's presuppositions. That is, it is a bad idea for them to get aligned on their self‐attributions of being disposed to enjoy Vegemite and not being disposed to enjoy Vegemite. It is a bad idea, because both parties to the dispute are absolutely correct to have the view that they do, and getting aligned on their self‐attributions of these properties would require that one or the other self‐attribute a property that they in fact lack.

This, incidentally, shows why the very first place in which one might be inclined to look for self‐locating content in natural languages—sentences involving first‐person indexicals—is not in fact a very good place to look. (At least, it is not a good place to look so long as you are assuming that the role of an assertion of S is to add the proposition or property that S expresses to the conversation's presuppositions. And that is the only view about assertion on which attributing self‐locating contents to sentences in context makes predictions that are interestingly different from those on which the sentences just express ordinary possible‐worlds propositions about the speaker.)

Suppose that the sentence “I am John Malkovich” expressed the property being John Malkovich. Then introductions would be disastrous. The effect of Mr Malkovich's assertion of “I am John Malkovich” (if it were accepted) would be to add being John Malkovich to the conversation's presuppositions. Part of what would be involved in this would be all of the parties to the conversation self‐attributing being John Malkovich. This is not what happens when people introduce themselves. Sometimes people do mistakenly come to self‐attribute such properties as being John Malkovich, being Hume, or being Napoleon. They do not do this, however, simply by being credulous when Malkovich, Hume, or Napoleon introduces himself. So “I am John Malkovich” does not express the property being John Malkovich. Other attributions of self‐locating content to sentences involving first‐person indexicals are similarly disastrous. So we ought not to believe that sentences involving first‐person indexicals have self‐locating content. We ought instead to believe the usual sort of Kaplanian theory, according to which first‐person indexicals are, well, indexicals—they refer to different individuals on different occasions of use. (p.280)

(How, then, do we learn anything other than the necessary truth that John Malkovich is John Malkovich from Mr Malkovich's introduction? By exploiting the sort of pragmatic mechanism Stalnaker (1978) sets out: we, as competent English speakers, know that, depending on who is speaking, “I am John Malkovich” expresses either a necessary truth or a necessary falsehood. If we take the speaker to be sincere and well informed, we trust that he is among the individuals who is in a position to say something true, rather than something false, with an utterance of “I am John Malkovich.” The only person in such a position is Mr Malkovich. And so we come to accept that the person before us who just uttered “I am John Malkovich” is John Malkovich.)

(A few clarifications for those who are concerned about how the semantic details are going to go. (1) This kind of account of aesthetic sentences does not force us to adopt a weirdly bifurcated theory of content that deals sometimes in possible‐worlds propositions and sometimes in properties or self‐locating propositions. Everything that we can do with a possible‐worlds proposition, we can do with a self‐locating proposition that does not distinguish between different positions within the same world.20 (2) It also does not force us to say that sentences and predicates have the same kinds of semantic values. Whatever the semantic values of sentences are, the semantic values of predicates are functions from objects to sentence‐type semantic values. So, if the semantic values of sentences are properties, the semantic values of predicates are functions from objects to properties. (See Egan 2006a, b.) (3) We also are not forced to add any additional formal apparatus to our formal semantic theories. What this sort of proposal requires is that we say that these sentences have contents that take truth‐values relative to possible situations, positions, or predicaments rather than relative to possible worlds. We can do this by leaving the formal apparatus of our theory of types just the same (we still say that sentences in context have semantic values of type <s,t>, for example), while revising our interpretation of what things of type s are like—we interpret them not as worlds, but as positions or predicaments.)

Another option is to adopt a straightforward contextualist account of the semantics of aesthetic vocabulary, and say that the connection between (p.281) accepting S (¬S) and self‐attributing P (¬P) is pragmatic, rather than semantic.

We can do this by exploiting the possibility of fighting not over the truth or falsity of the propositions that are in fact the contents of our assertions, but over which propositions we are asserting—that is, over the nature of the context that we are in. For example, we can use assertions of, for example, “Joe is rich” and “Joe is not rich,” or “Michael is tall” and “Michael is not tall,” to fight not about truth or falsity of the propositions expressed, but about what the contextually salient standards of wealth or height are or ought to be. If we understand disputes about taste on this model, we can predict that what is at stake in such disputes is the self‐attribution of the relevant sorts of properties.21

A little more detail about the sort of dispute in question. There are two kinds of disputes we might have about “Joe is rich.” In one kind, we are in agreement about how much Joe has to have in order to be rich, and we are fighting about whether he has that much or not. In another kind, we know how much Joe has, and we are fighting about whether that is enough to be rich. The second kind of dispute is naturally thought of as a fight about what sort of context we are in. (Of course there are also disputes—probably most of the ones we actually engage in—in which both questions are open.)

There are three things to notice. First, that sort of fighting‐about‐the‐context use of “Joe is rich” and “Joe is not rich” makes sense only so long as we are presupposing that “Joe is rich” has the same truth value in each of our mouths. (The easiest way for this to be so is for our utterances to express the same proposition, which will happen only if our uses of “rich” have the same semantic values.) In order to have this sort of context‐shifting fight, your utterance of “Joe is rich” has to be in competition with my utterance of “Joe is not rich.” If we thought that our two contexts might be relevantly different, such that we were (or might be) just talking past each other, we could just accept each other's assertions without revising our views about what our own contexts were like at all.

Secondly, notice that what is at stake in these sorts of context‐shifting disputes just is the self‐attribution of a certain sort of property. For any (p.282) context‐sensitive sentence, there is a property that one must have in order to be in a position to use it to utter a truth, and that one must take oneself to have in order sincerely to assert it. In general, when S is context‐sensitive, sincere assertors have to self‐attribute being in a context in which an utterance of S would express a truth, and, to the extent that they are semantically competent with S, they have to self‐attribute whichever properties one must have in order to be in such a context.

For “I am hungry,” the relevant property is being hungry. For “France is hexagonal,” the relevant property is being involved in a conversation whose standards of precision are such that France counts as hexagonal. It is plausible that we sometimes care about whether the context is a certain way, not for its own sake, but because we care about whether we, and the other parties to the conversation, are in the right condition to make the context that way. Whether I am in a context in which “Joe is rich” expresses a truth depends not just on how much wealth Joe has, but also on what sorts of attitudes I, and the other parties to my conversation, have toward wealth. How much do we think that one has to have in order to pursue the sorts of projects and life plans that we take to be important? Which sorts of things are we taking to be necessities, and which luxuries? The answers to these, and surely other, questions about our attitudes will be relevant to determining which property “rich” picks out in a given context. We can apply pressure on each other to change those semantic‐value‐affecting attitudes by applying pressure on each other to accept that, for example, “Joe is rich” expresses a truth in our context. (Note that this actually divides into two different kinds of fight: one in which we are attempting to change each other's views about the antecedent facts about the context, and one in which we are attempting to shift the context in one direction or another.)

In cases where we are presupposing that S has the same truth‐value in your mouth as in mine, I will not be able to accept your assertion of S without taking myself to be in a position to truly assert it myself. And so my sincere acceptance of your assertion of S will require that I self‐attribute the same property that your sincere assertion of S requires you to self‐attribute. So, the upshot of a resolution of a dispute about some context‐sensitive sentence will be the mutual self‐attribution of whatever property (or properties) one must have in order to be in a position to use it to assert a truth. And, on plausible contextualist accounts of aesthetic vocabulary, these are liable to be precisely the sorts of properties discussed in Section 3. (p.283)

Finally, notice that this sort of dispute will be subject to the right sort of defectiveness. If aesthetic vocabulary is context‐sensitive, then these sorts of disputes over aesthetic sentences, in which the aim of the disputants is to change the other's view about the context they inhabit (and therefore about which proposition is expressed), rather than to change their view about the truth or falsity of the particular proposition expressed, will be sensible to engage in only so long as the parties to the dispute take themselves to be alike with respect to the property (or properties) that one must have in order to be in a position to assert something true by use of the sentences in question.

So, at least at first glance, this sort of contextualist account looks pretty good. Let me say briefly before closing why I think that the previous, self‐locating proposal is preferable.

My main concern is that it makes the wrong predictions about what we should do when the presupposition of similarity fails. On the sort of contextualist proposal we are now considering, once the presupposition of similarity fails, it should be clear that we are in a situation where the parties to the dispute are simply talking past one another, and their assertions are not really in conflict. Alan's assertion of “Vegemite is tasty” means something like Vegemite is tasty to Alan, Vegemite is tasty to Australians, or Vegemite is tasty to Alan‐type subjects. Clare's assertion of “Vegemite is not tasty” means something like Vegemite is not tasty to Clare, Vegemite is not tasty to North Americans, or Vegemite is not tasty to Clare‐type subjects. Once it is clear to the parties to the dispute that this is the case, the thing for them to do should be just to accept each other's assertions, and the thing for a third party to do should be just to accept both assertions. (In the same way that, once it becomes clear that we are using “here” to talk about different places, I should happily accept your assertion of “the Rock and Roll hall of fame is here,” and you should happily accept my assertion of “the Rock and Roll hall of fame is not here.”)

But this seems wrong. The right reaction to the failure of the presupposition of similarity is not for each party just happily to accept the other's assertion. The right reaction is to stop asserting those sentences. It would be very strange if, once Alan and Clare stopped presupposing that they were gustatorily similar, Alan continued to assert “Vegemite is tasty,” and Clare responded by nodding acceptance and saying “that's interesting.” It would also continue to be weird and infelicitous for a third party to accept both (p.284) Alan's and Clare's assertions. In short: the contextualist theory predicts that, once the presupposition of similarity fails, the dispute should become defective because the CONFLICT condition is no longer satisfied (since there is no longer any difficulty in accepting both parties' assertions). But this seems mistaken. Even after the presupposition has given way, the assertions are still in conflict. It is just that the goals one would be pursuing by continuing the dispute are not good ones to pursue—what gives out is WORTHWHILENESS, not CONFLICT. The self‐locating account predicts this; the contextualist account does not.

7. Conclusion

We can explain the puzzling phenomena about disputes about taste with which we began with the chapter by saying that what is at stake in such disputes is the self‐attribution of certain properties. The project of seeking to resolve such a dispute is the project of trying to bring it about that the parties to the dispute all self‐attribute either the relevant property or its complement. If we say this, and take note of the conditions under which it makes sense to engage in that sort of project, we can explain the special sort of defectiveness to which disputes about taste are subject, and we can make the right kinds of predictions about the sorts of circumstances under which that defectiveness will manifest itself.

That is what I take to be the main point of this chapter. I have also argued more tentatively for some proposals about just which kinds of properties might be at issue, and about what sort of semantic theory we ought to endorse such that we can predict that the self‐attribution of those sorts of properties is what is at stake in disputes about taste. I think that these are likely to be right, or at least on the right track, but they are detachable from the central point.

References

Bibliography references:

Anscombe, G. E. M. (1957). Intention. Cambridge, MA: Harvard University Press.

Barker, Chris (2002). “The Dynamics of Vagueness,” Linguistics and Philosophy, 25: 1–36.

DeRose, Keith (2004). “Single Scoreboard Semantics,” Philosophical Studies, 119/1–2: 1–21. (p.285)

Egan, Andy (2006a). “Secondary Qualities and Self‐Location,” Philosophy and Phenomenological Research, 72/1: 97–119.

—— (2006b). “Appearance Properties?,” Noûs, 40/3: 495–521.

—— (2007). “Epistemic Modals, Relativism, and Assertion,” Philosophical Studies, 133/1: 1–22.

Harman, Gilbert, and Thomson, Judith Jarvis (1996). Moral Relativism and Moral Objectivity. Cambridge, MA: Blackwell.

Hume, David (1757/1965). “Of the Standard of Taste,” in Of the Standard of Taste and Other Essays, ed. John W. Lenz. Indianapolis, Bobbs‐Merrill.

Lasersohn, Peter (2005). “Context Dependence, Disagreement, and Predicates of Personal Taste, Linguistics and Philosophy, 28/6: 643–86.

Lewis, David (1979). “Attitudes De Dicto and De Se,” Philosophical Review, 88: 513–43;

repr. with postscripts in David Lewis, Philosphical Papers, i. Oxford: Oxford University Press, 1983.

MacFarlane, John (2007). “Relativism and Disagreement,” Philosophical Studies, 132: 17–31.

Mill, John Stuart (1859/1978). On Liberty. Indianapolis: Hackett Publishing Co.

Nolan, Daniel (2006). “Selfless Desires,” Philosophy and Phenomenological Research, 73/3: 665–79.

Perry, John (1979). “The Problem of the Essential Indexical.” Noûs 13: 3‐21.

Railton, Peter (2003). “Aesthetic Value, Moral Value, and the Ambitions of Naturalism,” in J. Levinson (ed.), Aesthetics and Ethics: Essays at the Intersection. Cambridge: Cambridge University Press, 1998; repr. in Peter Railton, Facts, Values and Norms. Cambridge: Cambridge University Press, 85–130. Page references are to this reprinting.

—— (n.d.). “Objectivity, Realism and Aesthetic Value.” Manuscript.

Richard, Mark (2004). “Contextualism and Relativism,” Philosophical Studies, 119: 215–42.

Sibley, Frank (1959). “Aesthetic Concepts,” Philosophical Review, 68/4: 421–50.

—— (1965). “Aesthetic and Non‐Aesthetic,” Philosophical Review, 74/2: 135–59.

Stalnaker, Robert (1978). “Assertion,” Syntax and Semantics, 9: 315–32; repr. in Stalnaker, Context and Content: Essays on Intentionality in Speech and Thought. Oxford: Oxford University Press, 1999, 78–95.

Stephenson, Tamina (2007a). “A Parallel Account of Epistemic Modals and Predicates of Personal Taste,” in E. Puig‐Waldmüller (ed.), Proceedings of Sinn und Bedeutung, 11: 583–97.

—— (2007b). “Judge Dependence, Epistemic Modals, and Predicates of Personal Taste,” Linguistics and Philosophy, 30: 487–525. (p.286)

Velleman, J. David (1989). “Epistemic Freedom,” Pacific Philosophical Quarterly, 70: 73–97;

repr. in J. David Velleman, The Possibility of Practical Reason. Oxford: Clarendon Press, 2000, 32–55.

Williams, Bernard (1995). “Internal Reasons and the Obscurity of Blame,” in Bernard Williams, Making Sense of Humanity. Cambridge: Cambridge University Press.

Wright, Crispin (1992). Truth and Objectivity. Cambridge, MA: Harvard University Press.

Notes:

(1) Sonnett 116.

(2) These are slight modifications of examples from Stephenson (2007b). She attributes them to Kai von Fintel and Danny Fox.

(3) This is one of the reasons why the first contextualist theory of taste that we might try out—the one according to which “tasty” always means tasty to x, where x is the speaker—does not actually look very attractive.

(4) We will be looking at only a subset of all of the very many kinds of disputes there are. There are all kinds of ways to have disputes—and perfectly sensible ones—where the sentences in question are not of this sentence and its negation form. (For starters, “This mug is hot”/“That mug is cold,” “I think it is raining”/“No it is not,” “None of the philosophers is in his office”/“Bob is in his office,” “Everybody is at home”/“Bob is in his office,” and so on). I will be attempting to draw the line between the sensible and the defective disputes only with this very specific sentence‐and‐negation form—the presentation would otherwise be a bit of a nightmare, and the extension to the rest of the cases should be pretty straightforward.

(5) This means that the presence or absence of a genuine conflict between a particular assertion of S and a particular assertion of ¬S will not always go along with genuine conflict between later assertions of the same sentences. In the case of, for example, sentences containing temporal indexicals, like “now” (or context‐sensitive expressions designed to refer back to previously uttered sentences, like “your last sentence”), subsequent assertions of the same sentence are likely to impose different conversational demands than those imposed by the original assertion.

(6) These facts about our actual practices of disputing about taste are another reason why the first pass, ‘tasty’‐means‐‘tasty‐to‐me’ version of contextualism about aesthetic predicates does not look very attractive.

(7) This is a feature that has loomed large in discussions of aesthetic predicates and claims involving them. (See, e.g., Wright (1992) on cognitive command.) It has also loomed large in discussions of secondary qualities more generally.

(8) And wanting to avoid complications about generics.

(9) I will not, at the end of the day, quite want to endorse this. I want to allow that the properties I need to self‐attribute in order to accept some aesthetic assertions will be idealization properties—properties of not having a certain disposition D just now, but of being such that a suitably idealized version of me would have D. I also want to allow that some of the relevant properties are group‐membership properties—properties of being a member of some group typical members of which have D. This maps onto Railton's discussion (2003) of the vertical and horizontal features of aesthetic properties. More on this in Section 4.

(10) Tamina Stephenson (2007a, b) gives a similar account when she explains how we ought to understand the formal apparatus of her (independent, and somewhat differently motivated) semantics for the vocabulary of personal taste.

(11) More generally, everybody taking themselves to have F or everybody taking themselves to have some incompatible G.

(12) Or perhaps of kind K, where K is some intrinsically desirable sort of experience.

(13) We can also use this to draw a distinction between cases that are mere change and those that are genuine improvements.

(14) Both of these are akin to phenomena about the offering of reasons that Bernard Williams (1995) discusses.

(15) Here again, Williams (1995) says some similar things about reason‐talk. It is also possible to read the later bits of MacFarlane (2007) as proposing that we understand most, or all, disputing about taste in this sort of way, though I am not completely confident that this is what is intended.

(16) Again, it is easier to be sympathetic to the idea that some disputes about taste work this way if one thinks about long‐term, running arguments over the course of days, weeks, months, or years.

(17) Many of these positive functions of various sorts of disputes will be familiar from Mill (1859/1978).

(18) I argue for a similar account of epistemic modals in Egan (2007).

(19) Though not the only way. See Perry (1979) for an alternative. I think that Lewis's way (1979) is preferable, but that is a topic for another paper.

(20) See Lewis (1979). See Nolan (2006) for dissent.

(21) See, e.g., Stalnaker (1978), Lewis (1979), Barker (2002), DeRose (2004), and Richard (2004) for discussions of this sort of phenomenon.