2.1 An alleged asymmetry

In his insightful discussion of NA Robert Hopkins writes

An obvious worry is that the picture I am presenting in this paper will not explain the crucial asymmetry Hopkins highlights; or rather – since I am explicitly not endorsing the rationality of our attitudes towards belief revision in aesthetics – that it will not explain the descriptive version of Hopkins proposed normative asymmetry. I am claiming that while DA correctly describes our practice of belief revision (or the lack of same) in aesthetics this is, in large part, a result of our overestimating our abilities as aesthetic judges. Yet I have claimed above that the tendency to overestimate our own abilities is a general one and, as such, it surely cannot be appealed to as an explanation of the psychological difference between our attitudes to belief revision in aesthetics and elsewhere. In the next section I will respond to this worry by presenting my own account of DA and highlighting what I take to be the relevant differences between aesthetic and mundane cases; differences which are perfectly compatible with our self-aggrandisement being general rather than restricted to the aesthetic domain. First, though, it is important to clarify exactly what is at stake in the autonomy debate.

2.2 Responding to disagreement

To begin let us consider the following pair of disagreements; one aesthetic, the other more mundane

How would I respond to disagreement in these cases? And how should I respond? These are not easy questions to answer for a number of reasons; most obviously because these examples are underspecified in a number of respects and leave a range of potentially relevant questions unanswered. Do my friends have significantly better vision, or aesthetic judgment, than I? What are our track records in these kinds of disagreement? How honest are my friends (and how honest do I take them to be)? How confident was I in my initial belief? How much is at stake for each of us in getting the right answer? And so forth. To prevent such issues complicating matters unduly let us stipulate that in the cases above (and, unless otherwise stated, in any further cases discussed below) all those involved in the disputes are epistemic peers. Peers in the general sense that they possess equivalent ‘intelligence, perspicacity, honesty, thoroughness, and other relevant epistemic virtues’Footnote ¹² and the specific sense that they are alike ‘with respect to their exposure to evidence and arguments which bear on the question at issue’Footnote ¹³ as well as with respect to the reliability of the relevant faculties (none of us are colour blind, tone deaf etc.).Footnote ¹⁴ Furthermore an impartial assessment of the evidence available to each of the disputants would warrant their believing that their interlocutors were their epistemic peers in all these respects.Footnote ¹⁵ Finally we should assume – in order to focus on the relevance of the ‘higher order’ evidence arising from the disagreement itself rather than on any ‘first order’ evidence –Footnote ¹⁶ that none of the parties puts forward any independent evidence (concerning typical bird populations in the area, the flautist's credentials etc.) in favour of their judgements. The extent to which such first order evidence is (and ought to be) significant in aesthetic disputes is, of course, a controversial issue in its own right but not one I will pursue here.Footnote ¹⁷

Even given all these stipulations (and ignoring any peculiarities introduced by considering specifically aesthetic judgements) there is still significant controversy regarding how best to respond in the face of disagreement. One thing that is evident from the growing literature on this subject of peer disagreement is that there is no clear consensus as to how we ought to respond to disagreement even in straightforward cases like Colourful Birds.Footnote ¹⁸ As such it would be implausible to think that any straightforward comparison of how we ought to act in aesthetic and non-aesthetic cases will be possible. Crudely speaking there are three stances which we could take with respect to disagreements of either kind, we could (i) retain our original belief; (ii) withhold belief endorsing neither our original belief nor that of our opponents; (iii) reject our initial belief in favour of the incompatible belief put forward by our interlocutors.Footnote ¹⁹

Although NA (as stated above) does not tell us anything about how we should respond in non-aesthetic cases it is clear that if the proponent of NA intends their doctrine to reveal anything interesting about aesthetics then they must propose, as does, a marked contrast between the correct response in aesthetic cases and those concerning more mundane disagreements.Footnote ²⁰ It is not immediately clear, though, from the quotes from Kant and Hopkins above – nor from other defences of NA – exactly what the contrast is supposed to be. Certainly that response (iii) is often correct in mundane disputes and never (or rarely ever) with respect to aesthetic disagreements, even if true, though, this doesn't tell us what our response in aesthetic cases should be. Should I retain my initial belief or merely refrain from adopting that of my fellow disputants? While the quotes I have marshalled above are perhaps consistent with either interpretation – suspending judgement could, for instance, be construed as one form of placing ‘less confidence in’ a view – it is clear that advocates of NA standardly take their doctrine to entail the claim that I should retain my initial belief and not merely suspend judgement.Footnote ²¹ From now on, then, I will take talk of revising aesthetic beliefs (or failing to do so) on the basis of disagreement to include not only cases where I adopt my opponent's belief but also those in which I come to suspend judgment on the issue in question. Of these two options I will focus primarily on the latter and argue that the best explanation for our peculiar unwillingness to abandon our aesthetic beliefs on the basis of disagreement is one which does not provide any normative vindication for this refusal. Of course if we do abandon our original belief in cases such as Flute Solo this raises the further issue of whether we should also adopt the views of our opponents. However, at this point the discussion becomes primarily concerned with issues surrounding whether (and under what circumstances) one should form aesthetic beliefs on the basis of the word of others; that is with the debate over aesthetic testimony. As mentioned above, though, I do not intend to address issues of testimony directly here and as such will focus almost exclusively on the question of whether we should abandon our previously held aesthetic beliefs on the basis of disagreement. It is important, especially considering the ease with which they are conflated, to stress the differences between these two topics particularly since several prominent arguments for rejecting aesthetic testimony would be clear non-starters if proposed as reasons not to abandon our current aesthetic beliefs. The accounts offered by Robert Hopkins, Philip Pettit, Roger Scruton and Cain Todd,Footnote ²² for instance all place additional constraints on the legitimate formation of aesthetic beliefs (or their expressivist equivalents).Footnote ²³ Constraints which, they allege, prevent beliefs formed on the basis of testimony from being legitimate. Even if such constraints were in place, though, they would only add to the conditions which an aesthetic belief must meet in order to be legitimate and as such they could never justify retaining aesthetic beliefs in circumstances where mundane beliefs would be abandoned.

3.1 Unreliability

I have suggested above that when it comes to making certain colour judgements most of us are extremely reliable. In contrast I maintain that aesthetic judgements – by which I will hereafter, unless otherwise stated, mean the first-hand judgements of non-experts regarding complex artworks – are typically unreliable. That is to say that our judgements are frequently incorrect, far more often than they are with respect to e.g. straightforward colour judgements, to the extent that such judgements (even when correct) frequently fail to achieve the status of knowledge. Further even in those cases where we do attain aesthetic knowledge through first-hand judgement such knowledge is likely to be far more precarious – subject to potential defeaters etc. – than our knowledge concerning ordinary empirical matters.Footnote ²⁹

There is a large and rapidly expanding literature showing that our aesthetic judgements are heavily influenced by a number of biases and distortions; ordering effects, exposure effects, snobbery, belief polarisation, echo chambers and many more besides.Footnote ³⁰ Such studies have led some to make rather extreme claims; for example that mere exposure or order effects frequently play a larger role in determining preference than does actual aesthetic quality.Footnote ³¹ I do not, however, propose to argue that the empirical results support such ambitious assertions - nor do I intend to deny this – but merely that they support my earlier claim that aesthetic judgements are unreliable in the sense outlined above. This general unreliability picture, or something very like it, has already been argued for at length elsewhereFootnote ³² and I will not rehearse these arguments here. Instead I will focus on some specific aspects of this unreliability picture which will be crucial for my argument.

Consider Aaron Meskin's attempt to show that aesthetic testimony is unreliable as a source of knowledge and, in particular, that it is less reliable than first-hand aesthetic judgement.Footnote ³³ Meskin argues that very few of us meet, or even approximate, Hume's standards for being a true judge; ‘strong sense, united to delicate sentiment, improved by practice, perfected by comparison, and cleared of all prejudice’Footnote ³⁴ and that an individual is ‘much more likely to know the extent to which he or she meets the requirements for being a true judge than the extent to which another person meets those requirements’.Footnote ³⁵ As such she would often have a defeater for aesthetic beliefs based on testimony (the high proportion of testifiers who are not true judges) which is not itself defeated (as it would be if she were able to e.g. establish that a particular testifier were a true judge). By contrast she is able to trust her own judgement provided she has taken the appropriate steps to check that she possesses delicate sentiment, is free from prejudice and so forth. A similar story could be told with respect to disagreement in aesthetics. If I am able to tell, through introspection say, that I at least approximate Hume's standards with regards to a particular aesthetic judgement but am unable to make such a determination with respect to my opponents – and indeed possess good statistical grounds to doubt it – then this may explain, and justify, my unwillingness to revise my aesthetic beliefs in many (though doubtless not all) cases of aesthetic peer disagreement.Footnote ³⁶

Yet the situation with respect to such matters is, in fact, very different to that which this argument suggests. As I have stressed at length above we are generally remarkably poor at assessing our own abilities in a whole range of areas and, in particular, tend to overestimate our abilities in comparison to those of our peers. As such we should be very much inclined to doubt – in the absence of any compelling independent evidence – any claim which rests on our ability to successfully gauge our own abilities in some area. Beyond this general picture, though, there are specific reasons to worry about our self-assessment of our fitness to be Humean judges. As indicated above, I am in agreement with Meskin that the source of much of our unreliability concerning aesthetic matters lies in biases and distorting factors of various kinds – order effects, snobbery etc. – which would clearly disqualify us from the role of true judges. The situation with respect to our ability to detect such biases, however, seems to be the exact reverse of that which Meskin's argument requires.Footnote ³⁷ Research has repeatedly shown that individuals greatly underestimate their own susceptibility to various biases and distorting factors – including many relevant to unreliability in aesthetics – even in the face of powerful evidence that they are greatly influenced by such biases.Footnote ³⁸ They readily accept that the evidence shows such biases to be widespread in others (and even that people typically underestimate their susceptibility to such biases) but resist applying these conclusions in their own case. Such ‘bias blind spots’ – combined with the impressive evidence that such biases play a uniquely (or at least atypically) prominent role in aesthetics - surely give us even more reason to defer to the judgements of others.

3.2 Irrepressibility and Aesthetic Self-aggrandisement

That aesthetic unreliability is largely rooted in such biases is also extremely important for establishing my second claim; that we are curiously irrepressible in our tendency to make aesthetic judgements. Those who suffer from significant maladies which impair their ability to judge colour are, for the most part, well aware of their conditions and as such refrain from making confident first-hand judgements about such matters. Similarly most of us are well aware that our knowledge of various areas – pre-Columbian South America languages, subatomic physics and so forth – is not what it could be and so refrain from confident pronouncements in these areas.Footnote ³⁹ The problem in the aesthetic case, by contrast, is that it combines bad judgement with frequent judgement. To see why we exhibit such irrepressibility with respect to aesthetic judgement we need to turn to my third claim concerning self-aggrandisement in aesthetics.

The claim that we standardly overestimate our own capacities as aesthetic judges is, in several respects, central to my overall argument. I have already alluded to the wealth of empirical evidence supporting the claim that individuals typically overestimate their abilities in a range of areas but what evidence is there that we overestimate our ability as aesthetic judges? Firstly, it is not unreasonable to maintain, given our ubiquitous tendency to self-aggrandisement, that the burden of proof is on the objector to show that the aesthetic domain is an exception rather than on me to show that it is not. Secondly, there is direct evidence of self-aggrandisement in areas at least analogous to aesthetic judgement; for example judgements of our ability to ‘get’ jokes.Footnote ⁴⁰ Finally, if - as I have claimed - our aesthetic judgements are heavily influenced by a number of biases and we are disposed to underestimate the negative effects of these biases in our own case then this alone makes it extremely likely that we would overestimate the reliability of our aesthetic judgements compared to those of our peers.

There seems to be good reason, then, to accept that we self-aggrandise with respect to our capacities as aesthetic judges. As with many of the biases discussed above, though, self-aggrandisement effects show a remarkable resistance to evidence. Those who are well informed about the prevalence of such biases seem no less susceptible to them and even those (myself for example) who spend a great deal of time writing about such biases - explicitly acknowledging that they are as much disposed to be influenced by them as the next person – still find themselves believing in accordance with such biases most of the time.Footnote ⁴¹ And it is easy to see why such beliefs would develop so as to be insensitive to evidence. Research indicates that self-aggrandising beliefs are often highly conducive to psychological wellbeingFootnote ⁴² and, pre-theoretically, it is difficult to imagine that believing yourself to be below average at your profession or as a conversational or romantic partner would give you much motivation to pursue the development of various personal and professional projects necessary for human flourishing; ‘if a job's worth doing, it's worth leaving for someone else who'll do it better’ is not the mantra of someone likely to survive and thrive. As such there would be good (non-epistemic) reason to have these self-aggrandizing beliefs hard-wired into us in such a way that they are insensitive to any evidence we have to the contrary. And, luckily for us, this again appears to be exactly the state we are in.Footnote ⁴³

Further, given that our self-aggrandising tendencies (along with, I suspect, many of the other biases discussed above) appear to be (at least to some extent) hard-wired, our irrepressibility as aesthetic judges also becomes a simple matter to explain. Since we are incapable of responding appropriately to our own limitations as aesthetic judges we continue to confidently make aesthetic judgements despite significant evidence of their unreliability.

3.4 Explaining the asymmetry

Having presented and defended the three crucial claims on which my account of DA depends it is now time to lay out the account itself. To begin, consider again our hypothetical individual Bobby observing my disagreements but let us now assume that Bobby has fallen into error while attempting to assess the capacities of individuals as systems for producing true beliefs. Specifically Bobby now shares my self-aggrandising beliefs with respect to my own abilities as well as my resistance to revising those beliefs on the basis of evidence. Bobby believes I am every bit as good as I (in my unreflective moments) think I am.

How would this alter things with respect to Colourful Birds? Not, I suspect, very much. Let us assume that Bobby judges my friends to be mistaken twice as often as I am and that while he still (correctly) judges them to be 98% accurate with respect to colour judgement he now (mistakenly) believes me to be 99% accurate. He will then judge the situation as follows: One system with 99% accuracy judges that P, ten systems with 98% accuracy judge that Q (where Q is incompatible with P). The odds of the single system judging P when ¬P is the case are one in a hundred. The odds of the ten systems independently judging that Q when P is the case are – as in the original case – too low to be worth considering. As such Bobby will, again, judge that I should abandon my colour belief.

Turning next to Flute Solo is there a significant difference here? I think there is but it is one which will require a little unpacking. First, recall that our aesthetic judgements are nowhere near as reliable as our colour judgements; let us assume that they are only a little better than chance (say 60% reliable).Footnote ⁴⁴ Further, recall that a great deal of this unreliability is the result of various biases and distorting factors which we are quick to ascribe to others but incredibly reticent to recognise in ourselves. As such it would not be unreasonable to suppose that I (and by extension Bobby) overestimate my superiority in aesthetic judgement to a greater extent than with respect to colour judgement. Let us say that Bobby correctly judges my friends to be 60% accurate but mistakenly judges me to be 90% accurate (that is he mistakenly judges them to be wrong four times more often rather than twice as often as in the colour case). With numbers like this it is easy to see why I would refuse to revise my belief in one on one cases of disagreement – as well as more complex cases where opinion is, more or less, equally divided - but they are not enough to explain why I continue to do so when faced with Kant's hundred voices or even the ten friends in Flute Solo. The odds of ten systems with a 60% accuracy all independently being mistaken are, once again, close to zero.Footnote ⁴⁵ Yet in this case, unlike Colourful Birds, the odds of all ten of my friends independently being right – given their low level of reliability would also be extremely remote. And here lies the crucial difference between the two cases.

In my discussion above I have been tacitly assuming that Bobby is disposed to treat my friends and I as akin to something like veritistic slot machines. When asked to provide a judgement with respect to a particular subject we will pay out a certain percentage of the time by producing true beliefs and correspondingly true assertions. Crucially, though, there will be no systematic patterns to the pay outs.Footnote ⁴⁶ Yet in a case such as Flute Solo this assumption - given the extreme unlikelihood of all ten of my friends returning the same result (whether mistaken or not) on such a model – does not appear to be legitimate. A better response would be to hold that there are certain instances in which the ten systems are very reliable and others in which they are prone to systematic error. Further, given that they all disagree with a system which is 90% reliable in this case, we have some reason to think that we are in the latter category but hardly – without some explanation as to the nature or the source of the error – an overwhelming one. Yet again, though, an appeal to the kind of biases and distorting factors prevalent in aesthetic cases can shore up Bobby's reasoning here. Bobby already thinks (correctly) that my friends are susceptible to all manner of biases to which he believes (incorrectly) I am immune. As such he can readily explain our, initially puzzling, disagreement by assuming that the disputed case is one where such biases play a particularly significant role. My friends are all lead astray by snobbery (or by exposure effects or ordering effects or…), as they so frequently are, whereas I remain immune. Such cases, Bobby reasons, are very common and certainly more common than circumstances in which I am mistaken and my friends are unanimously correct. And, of course, the situation in which the fictional Bobby finds himself (given his stipulated beliefs) is intended to be closely analogous to my actual position with regards to the question of revising my own aesthetic judgements.

We have seen, then, that anyone who holds the kind of self-aggrandising beliefs that I have argued we typically hold with respect to our own aesthetic judgement will – in combination with the other factors highlighted above – take themselves to have good reason to refrain from revising their aesthetic belief on the basis of disagreement. Further, their taking themselves to have such a reason provides an explanation for the existences of DA. Given that these beliefs are mistaken, though, the account I have given of DA provides no justification for adopting NA. Of course merely having an alternative explanation of DA – even one which I hope I have shown to be independently plausible – on the table does not in itself demonstrate that NA is mistaken. In the remainder of this paper, then, I will turn to criticise what I take to be the strongest NA account currently available.

4.1 McGonigal's sensibilities

In my view by far the most promising defence of NA is the relativistic one recently proposed by Andrew McGonigal. Consider again McGonigal's argument from §2 and, in particular, premise (4) that ‘The best explanation of the disagreement in this scenario is that I am at fault’.Footnote ⁴⁷ McGonigal introduces (4) since he thinks that previous arguments concerning belief revision in aesthetics (in particular those offered in Hopkins 2001) are ‘crucially enthymematic’ as they do not include this vital premise.Footnote ⁴⁸ He contends, further, that in the aesthetic case we can reject this premise since we have an alternate explanation for our disagreements which enables us to avoid the humiliating concession in (5) that our own judgement is most likely at fault.

According to McGonigal a promising rival explanation for our divergences in these cases is that I have a ‘different [aesthetic] sensibility’ from that of my interlocutors.Footnote ⁴⁹ A sensibility in McGonigal's sense is, at least partially, constituted by dispositions to have certain experiences in certain conditions. Here McGonigal draws an analogy with judgements of deliciousness; someone finds strawberries delicious if they are disposed to take certain kinds of pleasure in eating them and they should not let disagreement from those who lack such dispositions sway them into thinking they are not delicious (though it may perhaps motivate them to check their judgements more carefully).Footnote ⁵⁰ Similarly my judging an object beautiful, on McGonigal's view, is at least partially constituted by my being disposed to have particular positive experiences when viewing that object in relevant conditions and I should not let disagreement from those who lack such dispositions pressure me into retracting my judgement that it is beautiful.

In essence we can take McGonigal's proposal as introducing a further option into the kinds of aesthetic disagreement discussed above. As well as the standard possibilities – I am mistaken or they are – I can now consider another explanation; that both my interlocutors and I judge correctly according to our own sensibilities. I can therefore accept that their judgement is right for them (or similar) without having to concede that my own judgement is in error.Footnote ⁵² This new option certainly makes the situation in aesthetics different from that which we find in standards cases of disagreement; however, I think that it is important not to overestimate the difference this third option will make to the debate.

4.2 Relativism and Belief Revision

Let us for the time being assume that something like the relativistic sensibility picture which McGonigal proposes is correct – those who find such relativistic accounts implausible will, of course, have an additional reason to prefer my debunking account of NA to the vindicating account McGonigal proposes - and ask what consequences this will have for belief revision in aesthetics. Consider first some experiments on aesthetic judgment carried out by Chris McManus, and others, which revealed that subjects displayed ‘significant and temporally stable, but somewhat idiosyncratic, individual preferences’ between simple figures (squares, triangles and so forth).Footnote ⁵³ Let us suppose, for the sake of argument at least, that these results carry over to other kinds of indicator besides mere preference and, in particular, that the same pattern applies to judgements concerning beauty. If in such conditions I find squares, on the whole, more attractive than triangles and you have the opposite disposition then it strikes me as extremely plausible that such differences will be accounted for primarily (if not entirely) by appeal to something like McGonigal's sensibilities. This is not to say, however, that we could not be mistaken with respect to such judgements. Any credible relativistic picture will allow that we can be mistaken with regards to our own aesthetic standard just as we can be mistaken about what is delicious according to our own standards. I could mistakenly believe, for example, that strawberries are not delicious according to my standard if I have only ever eaten them immediately after brushing my teeth. Similarly I could be mistaken – even according to my own standards – if my judgements are too heavily influenced by order effects and the like (I assume that few relativists will want to defend a view according to which the order in which I view objects in anyway determines how beautiful they are according to my own sensibilitiesFootnote ⁵⁴).

Still, with respect to aesthetic judgements of this kind the new option which McGonigal introduces will frequently be a live option and under certain circumstances we may well be entitled to accept it and thus avoid the need for belief revision. How often this occurs will depend on a number of factors; the reliability of our aesthetic judgements concerning basic shapes, our evidence that others share our sensibilities and so forth. But it does – again assuming the relativistic picture is correct – legitimise some differences between our belief revision practice in aesthetics and our practice with respect to ordinary empirical matters.Footnote ⁵⁵ Yet this result does not appear to fully capture the alleged phenomena. Kant, Hopkins and McGonigal all seem to endorse an unconditional version of NA according to which disagreement is never reason enough to revise one's aesthetic beliefs. Surely, though, there will be cases where such revisions would still be justified on the picture under consideration. Perhaps my belief only barely counts as knowledge such that even the slightest piece of counterevidence would render the belief illegitimate or perhaps I have very strong evidence based on our past judgements that my interlocutors and I share the same aesthetic sensibilities. As such McGonigal's picture cannot be taken to justify an unconditional version of NA. Perhaps, though, the NA advocate would be willing to make their position less ambitious and merely claim that (i) disagreement is much more rarely a legitimate basis for belief revision in aesthetics than with respect to ordinary empirical matters (ii) this disparity is explained by some principled difference (in this case whether the beliefs in question are relativistic) between the two domains. Such a position – particularly when combined with a sceptical attitude with respect to testimony in aesthetics - may well be enough to capture the spirit of NA with respect to aesthetic judgements of the kind under discussion here. Such judgements have not, however, been the focus of this paper nor of most extant discussions of NA. Rather these discussions have tended to focus on aesthetic judgements with respect to artworks, and it is to these which I will now turn.

4.3 Artworks and Experts

I have looked above at the prospects for NA with respect to certain simple cases such as aesthetic judgements concerning basic shapes and in many ways this sounds like the kind of case with which McGonigal is primarily concerned; focusing on a certain kind of beauty, talking about ‘My being capable of taking a distinctive kind of pleasure in the perceptual appreciation of a given aesthetic object’Footnote ⁵⁶ and drawing frequent analogies with judgements of deliciousness. Yet, many of the aesthetic judgements we make – particularly those concerning artworks on which I have focused above – are of a rather different kind and it is clear from some of the examples he uses that McGonigal intends his conclusions to apply to judgements of this kind also.Footnote ⁵⁷ When it comes to considering such judgements, though, NA becomes significantly less plausible.

A film's being exquisitely plotted, well-acted or original has very little to do with dispositions or, at least, with dispositions of the straightforward kind which would make such judgements interestingly analogous to judgements of deliciousness. They may perhaps correspond to, or even be constituted by, the dispositions of idealised observers in rarefied circumstances, such as Hume's ‘true judges’ but seem, at most, loosely related to any distinctive kind of pleasure. Of course I am assuming here that we accept, as I think we should,Footnote ⁵⁸ a rather liberal view of aesthetic properties according to which judgements of well-plotedness, originality and so forth are genuinely aesthetic but my argument in no way depends on making this assumption. A similar line of argument can be constructed with respect to those aesthetic properties of artworks which are often treated (incorrectly I believe) as being, in some sense, essentially perceptual; elegance, gracefulness, dumpiness, gaudiness and so forth. In many cases judging that a line of music is elegant, a painting gaudy or a dance graceful is not very closely analogous to judging a cheesecake to be delicious or a square to be attractive. To see why, consider some of the criteria highlighted by, amongst others, Kendall Walton concerning the proper conditions for perceiving aesthetic properties in a work. In order to adequately judge the aesthetic properties of an artwork (and even to perceive it correctly) we have to properly orientate the work with respects to the category to which it belongs, other artworks in that genre, its historical relations to events in the art world and so on.Footnote ⁵⁹ How often do we non-experts really do this though? Not, I suspect, very often. In part because there are so very few art forms for which most of us have anything like the knowledge required to situate works in this way.Footnote ⁶⁰ Given how rarely we are able to do this then – and even more rarely that we actually do do this – we will (even assuming a thoroughgoing relativism) rarely, if ever, be justified in retaining our aesthetic beliefs given disagreement from genuine experts who have employed such knowledge in making their aesthetic judgements.

Notice here the shift in focus. I have been assuming above that our interlocutors are our peers with respect to the judgements in question. Focusing on such cases simplified matters somewhat and allows for fruitful comparisons between the autonomy debate and the epistemological literature on peer disagreement but, of course, many of our disagreements in aesthetics (and elsewhere) are not with our peers. And with respect to such cases it is important to notice that we have an interesting tendency to maintain our aesthetic beliefs even in the face of disagreement from genuine experts (consider that favourite film or beloved author derided by the critics), how are we to explain this?

Consider two kinds of factor relevant to being an aesthetic expert with respect to particular complex artworks or genres of artwork.Footnote ⁶¹ First, an expert must have an impressive range of aesthetically relevant knowledge to draw from; knowledge of genres, of art history, of the contemporary artworld and of much more besides. Second, there are those elements of aesthetic character which contribute to an expert's ability to make appropriate judgements with the knowledge she possess; for instance those Hume attributes to his ideal critics such as strong sense, delicate sentiment and an absence of prejudice.Footnote ⁶² Minimally aesthetic judgements – such as those in the shape cases discussed above – in contrast seem to require only the second kind of expertise. As such persistent disagreements between us in these cases will very often be explained either by flaws in at least one of our aesthetic characters (yours of course), by some blameless divergence in sensibilities or some combination of the two. When it comes to the appreciation of artworks, though, the first kind of factor is often extremely important and far more so than we are pre-theoretically accustomed to thinking. Our reluctance to recognise the importance of such factors leads us to mistakenly treat all aesthetic debates as if they were essentially ‘taste offs’ which – coupled with our ubiquitous mistaken beliefs in the superiority of our own faculties – leads us to be unwilling to revise our aesthetic beliefs on the basis of expert disagreement. In combination these two errors account for our maintaining DA even in the face of disagreement with genuine experts concerning complex artworks; something which the sensibility account (at least in isolation) is unable to do.

This is not to claim, of course, that no aesthetic disagreement with respect to complex artworks can ever be explained by reference to differing sensibilities and the like (nor do I deny that this is the case, I find the relativistic sensibility picture quite attractive but I am not willing to unequivocally endorse it). It may be, for instance, that I regard a play as well-acted when an expert denies this not due to our considerable differences in theatrical knowledge, or even owing to their superior capabilities as an aesthetic judge, but because of difference in our sensibilities. Perhaps I have a sensibility according to which a certain style of acting should be praised as vibrant and expressive whereas she has one which condemns it as histrionic and unrealistic.Footnote ⁶³ Still, to return to an earlier point, it seems unlikely that in any given case I will be able to establish that our disagreement finds its root in such factors rather than as a result of her superior aesthetic expertise. As such, disagreements of this kind will frequently serve as a defeater for my aesthetic beliefs even if they do not go so far as establishing that I ought to adopt the beliefs of my interlocutor.

It may be objected that this appeal to aesthetic experts is problematic given that there are no reliable (non-circular) methods for identifying such experts. The only way, the objector maintains, to establish whether someone is an expert in aesthetics would be to look at their track record of getting the right results concerning disputed aesthetic cases and the only way to do this is to use my own judgement. This need not, however, be the case. Footnote ⁶⁴ There may be problems with establishing just who the genuine aesthetic experts are but there is no reason to think that this will not be possible without appealing to our own aesthetic judgements; just as I can (at least to some extent) establish who the experts in physics are despite my very limited knowledge of physics.Footnote ⁶⁵ Of course there may be worries which need to be dealt with in identifying aesthetic experts which do not arise with respect to experts in physics; checking that the alleged expert possesses an appropriate Humean character for example. However, while assuaging such worries might complicate our search for aesthetic experts somewhat I see no reason to think that they create any in principle barrier to our identifying such experts.