It’s nearing the end of paper writing season and I have no clue what I’m going to do for the final section of one of my papers, so now is as good a time as any to revisit a perennial concern of mine: the possibility for objectivity in discourse about art. I wrote a post on this subject a couple months ago, in which I stated that my goal was to move between two equally unpalatable poles: “pure” objectivism and “pure” subjectivism. My goal here is to introduce an apparatus (drawing on the work of Rudolf Carnap and Huw Price) in which to think about these poles, in order to make it clearer that there indeed can be a middle ground between the two.
One vice of that earlier post was that I did not give a very explicit, very clear definition of objectivity. Since the term is used in so many different ways, that is a serious flaw, though I tried to mitigate it through context clues. However, while reading parts of Carnap’s Der Logische Aufbau der Welt recently, I came across a superb definition. (As it happens, reading Carnap has been quite eye opening, stripping me of a great many of my prejudices about logical empiricism. Carnap was a genius and a philosopher quite congenial to my general way of thinking, despite my many points of disagreement with him.) Carnap defines objectivity along two poles (The Logical Structure of the World, §66, George translation):
The requirement that knowledge be objective can be understood in two senses. It could mean objectivity in contrast to arbitrariness: if a judgment is said to reflect knowledge, then this means that it does not depend on my whims. […]
Secondly, by objectivity is sometimes meant independence from the judging subject, validity which holds also for other subjects. It is precisely this intersubjectivity which is an essential feature of “reality”; it serves to distinguish reality from dream and deception.
In just a couple sentences, Carnap has cut through a vast swath of issues. First off, it is useful to understand just how these two concepts are separate, by seeing how you can have either without the other:
Intersubjectivity without non-arbitrariness: This position is exemplified by cultural relativist views (in any domain)—or at least by certain extreme forms of cultural relativism. What is right or acceptable is just what is deemed right or acceptable by the lights of a particular culture. Thus, within that culture, there is intersubjective validity because there is an intersubjective standard by which to evaluate the relevant statements: just check what is deemed appropriate by that culture. However (and this is where the restriction to extreme forms of cultural relativism comes into play), what the culture deems acceptable or appropriate is entirely arbitrary, a result of what we might call communal “whims”. That is, there are no normative restrictions on what the culture may deem acceptable.
Non-arbitrariness without intersubjectivity: This possibility is more difficult to see, but it is possible. A view that would fit this model is one that suggested that e.g. questions about the value of a work of art that say that a person cannot give reasons for her view that have any force for other people, but that her opinion is nonetheless not arbitrary. For instance, she might base her appreciation for a work of art on a rigorous attempt to relate the work to her experiences, ignoring those experiences that don’t matter. Her judgment is then justified in a way that not just any judgment would be, but because her experiences are her own and not shared, there nevertheless is not intersubjective validity.
What is most useful about Carnap’s two axes of objectivity is that there is no mention of ‘truth’—presumably in part because he wished to avoid the metaphysical excess of the correspondence theory of truth, in which a statement is true just in case it corresponds to some fact about the world. If one holds such a theory of truth, then there is an immediate temptation to think that objectivity is a matter of grasping truth in this sense. Science then becomes the paradigm of objectivity: the claims of science correspond to the facts. Moreover, once this view of truth is in place, it becomes very difficult to even see how there could be objectivity about normative matters such as ethics and aesthetics, because it becomes very difficult to see just what an evaluative fact might be. This is the jumping off point for J.L. Mackie’s error theory (specifically, his argument from queerness). Mackie in effect asks what an evaluative fact might be, finds the very idea weird (rightly so, I think), and diagnoses our moral discourse as a thoroughgoing error: we think we are making true claims, but we are not, because there are no evaluative facts to correspond to our claims.
Carnap is having none of this, and in fact I think diagnosed Mackie’s error nearly 50 years in advance: “However, we must here clearly distinguish between a certain kind of language usage and the assertion of a thesis” (The Logical Structure of the World, §178). It is not at all obvious, and in fact I think it manifestly incorrect, to see our moral discourse as implicitly committed to the thesis that there is a special sort of evaluative fact. Of course, many people certainly do believe in such facts, and this belief certainly shapes how they talk. But this is not something inherent to moral discourse. While Carnap would disagree with me about the possibility for some degree of objectivity about evaluative judgments, it is not on the grounds that there is no special class of moral facts. (When I say Carnap would disagree, I mean the later Carnap. The Carnap who wrote the Aufbau in fact made explicit room for the construction of values within his system.)
When I said above, “science then becomes the paradigm of objectivity,” I was not being dismissive. Science is the paradigm of objectivity, and Carnap lets us see why. Good scientific practice fights against arbitrariness and attempts to maximize intersubjective validity (I will say more about what this amounts to shortly). The sciences are rife with disagreement (which is essential to their flourishing), but they have mechanisms to eventually resolve this disagreement. This allows them to enjoy a degree of objectivity that is unparalleled in any other domain.
Recognizing scientific inquiry as a paradigm of objectivity in this way does not make us lose sight of the possibility for objectivity (in different degrees) in other domains. This is because Carnap’s conception of objectivity allows us to see both axes as axes, i.e. as defining a continuum. Once we have this in view, we can see how there is room for objectivity in other domains, even if they are not as paradigmatically objective as scientific inquiry.
Why bother with all of this? One big motivation—for all sides, I would wager—is the phenomenon of disagreement. In my last post on the subject I discussed disagreement at some length, and I’ll discuss it more here. Another, related, aspect of the issue is the need to hold other people accountable. How we think about objectivity in various domains will change how we go about holding one another accountable for their claims, and how we react to disagreements. In the rest of this post, I want to relate these issues to an essay I recently read by Huw Price, “Truth as Convenient Friction” (published in his recent collection of papers entitled Naturalism Without Mirrors). In that paper, he discusses three norms of discourse, which I want to relate to the phenomena of disagreement. My end goal is to motivate a view in which we recognize the need for a norm of truth in aesthetic discourse, and to connect this to the discussion of objectivity above.
When considering the phenomenon of disagreement, it is useful to start by considering the marvelous closing sentences of Nelson Goodman’s delightful book, Ways of Worldmaking (emphasis added):
The vaunted claim of community of opinion among scientists is mocked by fundamental controversies raging in almost every science from psychology to astrophysics. And judgments of the Parthenon and the Book of Kells have hardly been more variable than judgments of the laws of gravitation. I am not claiming that rightness in the arts is less subjective, or even no more subjective, than truth in the sciences, but only suggesting that the line between artistic and scientific judgment does not coincide with the line between subjective and objective, and that any approach to universal accord on anything significant is exceptional.
My readers could weaken that latter conviction by agreeing unanimously with the foregoing somewhat tortuous and in a double sense trying course of thought.
What Goodman so rightly emphasizes, particularly in the italicized portion, is that disagreement is rampant in all walks of life. Science, as I said, is a paradigm of objectivity, but not because it lacks disagreement. Nevertheless, the fact of widespread disagreement in ethical and aesthetic matters is routinely used as an objection to views that think that such discourse can be objective. The reason why such arguments are not simply mistaken on their face is that disagreement in the sciences looks different from disagreement in other areas. In at least certain respects, disagreements in the sciences are more tractable than those in the arts. (Goodman, I should note, is sensitive to this. In a footnote to the passage just quoted, he writes, “The reason that earlier theories but not older works may be rendered obsolete by later ones is often, I think, that the earlier theories insofar as sound, are absorbed into and are rederivable from the later while works of art, functioning differently as symbols, cannot be absorbed into or derived from others.” I have added emphasis to an important and astute point—the point that gives the lie to Mackie’s error theory.)
One reason why scientific disagreements tend to be tractable in a way that aesthetic disagreements are not is that, when a scientific question is sufficiently well-formulated, it tends to have a single correct answer. Much of the most interesting scientific work lies not in determining the answer to such questions, but in learning just which questions to ask. Questions in aesthetics, on the other hand, don’t seem to admit of such unitary answers. Compare: (a) What is the charge of an electron? (b) Are Bach’s Brandenburg Concertos better or worse than Beethoven’s fifth symphony? The former question has a determinate answer, which at this point we know to a fairly high degree of precision. The latter question, on the other hand, does not obviously have a correct answer, and if it does we sure as hell don’t know how to find it. Nor is there any real prospect of us somehow “discovering” the method for finding it at some point in the future.
Part of this is because the comparison is unfair. The scientific question is just the right question to ask, part of a very clear theoretical framework established through a great deal of conceptual and empirical work. The aesthetic question, on the other hand, is precisely the wrong sort of question to ask (at least, not without a very clear context given). It is an argument for another day why it is the wrong question, but see my “screed” to the right for a brief indication of my thoughts. Here are some better questions: (c) What is the mood of the piece? (d) Does it establish this mood effectively? (e) What is the piece saying? (f) Should one listen? — These questions also may not have unitary answers, but they are at least more worthwhile questions to ask, and the reasons they don’t have unitary answers are more interesting. (I have more to say about what questions are interesting in my post Why write about art?)
I think part of why people take disagreement in normative discourse to cut against objectivity is that they take questions like (b)—the bad sort—to be the paradigm sort of question in those areas. But that is hardly the full explanation. It is also because disagreement about even the better sorts of questions do not seem tractable to anything like the same degree as those in science. I think that these arguments can be overcome. I argued in my last post on the subject that certain forms of subjectivism lose sight of the very possibility of disagreement, so I won’t rehash that here. However, I do want to look at Huw Price’s discussion of norms of discourse, and relate them to the phenomenon of disagreement more generally.
Price discusses three norms of discourse, each more stringent than the last: (1) sincerity, (2) warranted assertibility, (3) truth. These are norms of discourse in the sense that we can and sometimes do hold people accountable for failing to uphold them.
The first norm, sincerity, requires that when people make some claim, they advance it sincerely, i.e. it requires that people honestly express their opinions. This norm may be suspended for rhetorical or other purposes, e.g. when one person plays devil’s advocate, but in most contexts it holds (and even in the case of playing devil’s advocate, the norm may hold insofar as the person is expected to make it clear what she doing). One way of distinguishing a lie from just any false claim is by noting that a lie violates the norm of sincerity.
People who hold hardline subjectivist views about aesthetics generally hold aesthetic discourse only to norms of sincerity. It is possible to explore people’s motivations for their views on works of art, but ultimately you cannot fault someone for the opinion they hold. You can only fault them for being insincere. For just this reason, it is inappropriate on such views to talk about reasons for evaluating works of art one way or another, taking the notion of a reason to involve at least some intersubjective rational force. You can only speak of motivations. This norm of discourse thus does not require objectivity in either sense: a person’s judgment may be totally arbitrary (“I just like it”) and it has no claim to intersubjective validity whatsoever. On both axes, an aesthetic discourse that involved only a norm of sincerity would be a paradigm of subjectivity. As Huw Price notes, there is no room for disagreement if such a norm is all that is available. If you sincerely express your opinion and I sincerely express mine, we may have different opinions, but we cannot disagree. As I argued in my previous post on this topic, I think this fact is enough to make us wish to avoid such a form of discourse at all costs if we can, so I will not belabor the point here.
The second norm, warranted assertibility, comes in both personal and communal forms. The personal form says that a speaker is incorrect to make an assertion, not only if she does not sincerely believe it, but also if she does not have adequate grounds for believing what she asserts. This norm basically amounts to requiring that one be able to justify one’s claims and opinions. We can take one another to task for expressing opinions that are not well-founded, where ‘well-founded’ here means well-founded by the speaker’s own lights. (Price gives the example of beliefs that are not well-founded “on the grounds that they [do] not cohere with the speaker’s other preferences and desires.”)
Allowing this norm makes possible a more moderate subjectivism about aesthetics. While opinions may be quite arbitrary and subjective, they cannot be totally arbitrary—they are held at least to standards of minimal coherence. Moreover, it makes possible what Price calls ‘no-fault’ disagreements—disagreements where neither party is at fault (where nobody is wrong). However, there is still a very real sense in which parties to such disagreements “slide past one another,” as Price puts it. Because there is no norm beyond minimal personal justification, there is no disagreement in a fully robust sense. There is no real prospect of resolving our disagreements, and the reason is because there is no real intersubjectivity.
This can be overcome by considering a communal form of warranted assertibility, where one is justified in making a claim only if that claim is warranted by the standards of her community. On one view, the community imagined is some ideally rational future community (when we reach “the end of inquiry”). But, from my perspective, it is not clear what such a community would look like, it is clear that no such community actually exists, and it obvious that we have no way of judging one another by the standards of such a community, seeing as we are not one. (Price makes all of these points, but I arrived at them independently of him.) Thus I will focus on a more modest norm of warranted assertibility, in which one is warranted in asserting something just in case one is justified in believing it by the lights of one’s actual community.
This needs to be handled with care. If taken too crudely, it lapses into a strong cultural relativism, with all its flaws. (How, for instance, could an individual justifiably dissent from the cultural norms, if those norms determine what is justified?) We might, however, imagine that both the personal and communal forms of warranted assertibility are used, which allows for a picture of critique in both directions. The risk this runs, however, is that disagreements still slide past another. When personal and communal warranted assertibility disagree, what arbitrates the disagreement? The problem is that personal warranted assertibility has no intersubjectivity, which means it cannot really engage in an intersubjectively compelling critique of communal norms. Moreover, while communal norms are intersubjective, one can still say, with Price: “My manners are not those of the palace, but so what?” There must be some account of why communal norms should apply to me, why I should be beholden to them. So the picture of mutual critique I suggested is, in fact, a ruse.
This argument, which I have adopted almost wholesale from Price’s paper, leads to the need for a third norm, the norm of truth. The title of the paper is “Truth as Convenient Friction”, which indicates the role the norm of truth plays: it introduces the friction needed to get disagreements to stop sliding past one another. Price: “We could be aware that we have different opinions about what is warrantedly assertible, without that difference of opinion seeming to matter. What makes it matter is the fact that we subscribe to a practice according to which disagreement is an indication of culpable error, on one side or other; in other words, that we take ourselves to be subject to the norms of truth and falsity.”
When we hold other people to the norm of truth, we take them to task for saying false things. We take there to be “facts about the matter” (please do not read this metaphysically) that transcend what any particular person is warranted in asserting at a particular time, and when people disagree, they are culpable in the sense that at least one of the parties to the disagreement is at least partially wrong.
It is easy to see how the norm of truth plays a key role in scientific inquiry, and helps us to make sense of scientific disagreement. Let me illustrate by example. In the 1910s, Alfred Wegener proposed his theory of continental drift. For several decades, there was sharp disagreement among scientists, with many (even most) still holding to immobilist theories. Wegener’s theory eventually won out, of course, but that did not fully happen until the 1960s. We can say about this case that, in those intervening decades, both sides satisfied the norms of warranted assertibility (and sincerity, I presume) in that both sides could offer compelling, empirically acceptable reasons for preferring one theory over the other. In hindsight, we nevertheless praise those who supported Wegener’s theory, because Wegener’s theory was the true one. (This is really very crude, and it sidesteps the myriad issues of scientific realism. But I can only fit so much in a single post.) Scientists can agree to disagree, as the saying goes, but only temporarily. Agreeing to disagree, for a scientist, is an uneasy state to be in.
If my reasons for wanting to save disagreement in the aesthetic realm are compelling, then we have good reason for wanting a norm of truth for aesthetic discourse as well. Nevertheless, it is just this fact of widespread, seemingly intractable disagreement in such discourse that makes such a norm seem impossible to have. I want to offer a diagnosis of this apparent impossibility, and point the way toward seeing how we can have such a norm.
One reason why aesthetic disagreement seems to rule out the possibility of such a norm is that our concept of ‘truth’ is, I think, strongly connected to the existence of a single right answer. Regarding empirical questions, such as, what is the charge of an electron?, the existence of a single correct answer is straightforwardly comprehensible. Price never makes this point explicitly, but I suggest it is implicit when he says, “Assertoric dialogue requires an intolerance of disagreement.”
But in the aesthetic domain, even when we move away from crude questions like (b) earlier and toward more sophisticated interpretive and evaluative questions, it is still difficult if not impossible to get a grasp on there being a single right answer. The reason for this, I suggest, is that disagreements in aesthetics are (or ought to be) something between pure fault and pure no-fault disagreements. They are fault disagreements in that we think (or should think) that it is possible to be wrong, to have bad opinions. But they are no-fault disagreements in that the resolution of such disagreements can indeed be a stable agreement to disagree. Such agreement to disagree is possible precisely because, while we recognize a norm of truth that extends beyond the norms of sincerity and warranted assertibility, we do not think that there exists a single correct answer, for any deviance from which one is culpable.
This creates problems because it stands in tension with our sense that we can only speak of truth where there is a single right answer. There are two ways to resolve this tension. One way is simply to get over that prejudice, and to allow talk of ‘truth’ and ‘knowledge’ even in evaluative discourse. I tend to favor this strategy on the grounds that I think banishing such terminology from aesthetic and moral domains is dangerous, as it creates a temptation to fall into the sort of subjectivist positions I find intolerable. (Of course, keeping the terminology is dangerous, too, as it leads to the opposite temptation of falling into an equally intolerable and implausible objectivism. But I think this temptation is easier to resist, as it looks much less attractive in today’s philosophical landscape.)
In general, then, I am happy to speak of aesthetic truth, aesthetic knowledge, and of art as a “way of knowing” (to use a troubled but lovely phrase). But it is also useful to have a terminological distinction at hand for when it is necessary. For this, I return to the Goodman quote I introduced a long way back, specifically this sentence: “I am not claiming that rightness in the arts is less subjective, or even no more subjective, than truth in the sciences…” Here Goodman draws a distinction between truth and rightness, leaving “truth” to the sciences while speaking about “rightness” for the arts. As I understand Goodman, truth is a species that falls under the genus of rightness. Crudely characterized, we can imagine truth as the special case of rightness where a well-formulated question has a single correct answer. (This probably will not do on closer inspection, but it’s good enough for now.)
The reason this distinction is apt is that while our sense of ‘truth’ seems to involve the existence of a unitary correct answer, the same is not true of our sense of ‘right’, for I am sure we are all familiar with the phrase, “there are multiple right answers” (which is of course distinct from “there are no wrong answers”). Having a notion of rightness allows for a healthy pluralism that has room for agreeing to disagree in the long run (i.e. not just temporarily, with the hope that in the future the disagreement will be resolved).
This should let us see how aesthetic discourse can be objective. It is not arbitrary because we cannot take just any position that suits our whims—we must take due care to ensure, as best we can, that we are right. And it is substantially intersubjective because we must be able to offer reasons that can justify our position to another, showing why we are justified in taking ourselves to be right. It is not purely intersubjective because we cannot take consensus even as an ideal (as we might in science), but it is nevertheless much more than merely subjective.
I hope I have gone some small way to showing how we might save the objectivity of aesthetic discourse.