Art and achievement

An increasingly popular view in the philosophy of art is that some artworks are good artworks at least partly because they are achievements. This view was introduced to explain why two works that look the same, such as an original painting and a perfect copy, can differ in artistic merit. An achievement theory can say that the original is better because it is a greater achievement. Achievement theories have since been used to answer other questions, and they are now a serious alternative to traditional theories of artistic merit. This paper has three aims. The first is to articulate the achievement theory more fully and explicitly than its advocates have. The second is to show that the achievement theory should be rejected, by raising five problems for it. The third is to show that appealing solely to the excellence, aptitude, and ineptitude a work manifests yields a better theory of artistic merit.

knowledge-how (Carter and Pritchard 2015), or understanding (Pritchard 2010;Carter and Pritchard 2015, 188-190) as achievements (for criticism, see Lackey 2007Lackey , 2009Pritchard 2010;Bradford 2015b). 1 Some say that this explains why knowledge or understanding is non-instrumentally valuable, since achievements have final value (Riggs 2002(Riggs , 95, 2007Greco 2010, Chap. 6;Pritchard 2010; for criticism, see Whiting 2012). And some (e.g. Greco 2010, Chap. 5) argue that this explains why the subject in Gettier cases lacks knowledge, since her believing truly is not related to the exercise of her cognitive abilities in the way success and ability relate in achievement. Recently, the first monograph on achievement was published by Gwen Bradford (2015a).
In aesthetics, an increasingly popular view is that some artworks are good artworks at least partly because they are achievements (e.g. Dutton 1979;Currie 1989;Walton 1993;Budd 2007Budd , 359-364, 2014Turri 2016, 126; for discussion, see Goldman 2006, 341;McGonigal 2010, 556-557, 565-566;Lopes 2011Lopes , 530-532, 2014Huddleston 2012;Levinson 2016). This view was introduced to explain how two works that look, sound, or read the same can differ in artistic merit. For example, an original painting can be better than a perfect copy of it. Traditional views that tie artistic goodness only to the work's appearance, or only to what can be discovered by viewing, hearing, or reading it, are thought to struggle to explain this. An achievement theory does not struggle: it can say that the original is a greater achievement, and that is why it is a better artwork. Achievement has since been invoked to answer other questions: for instance, to argue that a poet's insincerity is relevant to criticism of her work (Neill 1999), to establish that a work's moral flaws can be aesthetic merits (Eaton 2012), and to explain why aesthetic value enhances artistic value (Huddleston 2012). An achievement theory is also attractive because, assuming achievements have final value, it implies that good artworks have final value, which many find plausible.
So achievement theories are emerging as an alternative, or at least a complement, to more established views about artistic merit, such as aestheticism, which explains a work's merit by reference to its aesthetic value, and experientialism, which does so by reference to the value of the experience the work offers. Let us call ''achievementism'' any view that explains why something has (a certain degree of) artistic merit by appeal to the fact that it is an achievement (of a certain kind), or by appeal to facts about how great an achievement (of some kind) it is. 2 Achievementism might claim that being an achievement of a certain kind or magnitude gives a thing artistic merit, or that the thing's being such an achievement explains why some other property gives it merit, or that having artistic merit is identical to being an achievement of a certain kind or magnitude. 1 Ernest Sosa's work is a major influence on achievement theories; see Sosa (2007). 2 I also count as an achievementist someone who explains a thing's artistic merit by appeal to the fact that creating it is an achievement (of a certain kind or magnitude). For brevity's sake, I shall usually not mention this view, but my criticisms will apply to it, mutatis mutandis. One motivation for it is the belief that only acts are achievements (Lopes 2011, 530-531 believes this). I think that is false. It seems to me that David Davies (2004, 204) is right in saying that ''achievement'' displays a familiar process-product ambiguity: we can correctly apply the word to acts and to their products.
Here I will argue against achievementism. In Sect. 1, I articulate achievementism more fully and explicitly than has been done elsewhere. I distinguish it from similar views, and set out different forms it can take. This discussion provides a new way of categorizing theories of merit, and should give greater clarity to our thinking about theories of merit generally.
In Sect. 2, I identify four challenges achievementism must overcome. The first is to specify the kind of achievement whose magnitude is correlated with artistic merit. This is more difficult than it appears. The second is to establish that artistic merit is not a determinant of the magnitude of this kind of achievement. The third is to account for apparent examples of things that have artistic merit by luck. The fourth is to answer the objection that achievementism offers one explanation too many of certain facts about merit. I outline strategies for solving the first two problems. These may or may not prove successful. But I do not see a way of overcoming the third or fourth.
In Sects. 3-4, I argue that a different theory of merit, which appeals only to the excellence and aptitude of the artist's acts and perceptions, gives a better explanation of everything achievementism explains, and explains more besides. This is a fifth reason to reject achievementism. I then argue that this theory can retain the attractive features of achievementism. I conclude by identifying challenges and questions it would face. The moral is that, although achievementism reveals the shortcomings of some views it was introduced to correct, the notion of achievement draws our focus away from what is actually doing the explanatory work.

Articulating achievementism
Achievementism explains why something has artistic merit by appeal to the fact that it is an achievement (of a certain kind), or to facts about how great an achievement (of some kind) it is. That it is explanatory, and that it appeals to the thing's being an achievement, are crucial. These features distinguish achievementism from several similar views.
First, achievementism is not the view that one artwork's being a greater achievement than another is simply evidence that it is better. You could believe that the greater achievement is probably the better work without believing that this explains why it is the better work. The achievementist tries to explain why it is better.
Second, achievementism is not the uncontroversial claim that works can have artistic merit in virtue of what is achieved by the artist. An aestheticist can agree that, when aesthetic value is achieved, the work has merit in virtue of what is achieved. This does not make her an achievementist.
Third, achievementism is not even the view that works have artistic merit in virtue of the things achieved being achieved. At least, it courts confusion to express the view this way. This is because there is a distinction between constituting an achievement and constituting something achieved. ''To achieve'' often means, roughly, to attain some aim by your own efforts or abilities. That you achieved an aim does not imply that your achieving it was an achievement. Simon Keller (2009, 670-671) gives the example of having the aim of adding milk to your coffee, and achieving it (compare Pritchard 2010, 68;Bradford 2015a, 4). This is not normally an achievement. But it is the achievement of an aim. So not every achievement of an aim is an achievement.
The basic idea, then, is that a thing's being an achievement explains its artistic merit. There are three ways in which one might develop this thought.
First, one might think that being an achievement (of some kind or magnitude) gives a work (a certain degree of) artistic merit. The fact that it is such an achievement is a reason why it has merit or a certain degree of merit. Let us call this view ''determination achievementism''.
Second, one might claim that the work's being an achievement enables some other property to give it merit. Some (e.g. Plato 1975, 99b;Dancy 2004, 45-49) distinguish the reason why p from what enables the reason why p to be the reason why p. One might argue that, for instance, a work's aesthetic value gives it artistic merit, but only when, and because, its aesthetic value constitutes an achievement. Andrew Huddleston (2012) discusses this view. The work's being an achievement still explains why it has a certain degree of merit. It explains, for example, why the work has more artistic merit than something else with the same aesthetic value. But it is aesthetic value, not being an achievement per se, that gives the work artistic merit. Let us call such a view ''enabler achievementism''.
Third, the fact that the work is an achievement (of a certain kind or magnitude) might be identical to the fact that it has (a certain degree of) artistic merit. Alternatively, one might accept a property identity claim, and hold that being such an achievement just is having (a certain degree of) artistic merit. Let us call such a view ''identity achievementism''. Identity achievementism can still explain why a work has a certain degree of merit, even though it does not say what gives it merit. One might claim that originals have more merit than forgeries because originals are greater achievements, and having more merit just is being a greater achievement. This could be a perfectly good explanation. We might not have realized that having more merit just is being a greater achievement.
So defined, achievementism is compatible with other theories of merit. For example, we can call ''determination experientialism'' the widespread view that offering a valuable experience gives a work merit, and ''determination aestheticism'' the view that the work's aesthetic value gives it merit. Clearly, identity achievementism and enabler achievementism are compatible with these views. What may be less obvious is that determination achievementism is compatible with them. Malcolm Budd is both a determination experientialist and a determination achievementist. In his view, a work's artistic value is determined by the intrinsic value of the experience it offers, and ''the intrinsic value of the experience offered by a work of art is determined by the nature and magnitude of the artist's achievement in creating the work'' (Budd 2014, 19). On this view, being a great achievement of the right kind gives a work merit. But it does so only by giving it another property that gives it merit, namely, the property of offering an experience with intrinsic value. Being a great achievement of the right kind gives it merit indirectly-by giving it another property that gives it merit. There must also be properties that give things merit directly, that is, not only by giving them other properties that give them merit. Accordingly, determination achievementism can be direct or indirect, depending on whether, according to it, being an achievement (of some kind or magnitude) gives the work merit directly or indirectly. Budd is an indirect achievementist and a direct experientialist.
Determination and enabler achievementism can take strong and weak forms. A determination achievementist might say that anything with artistic merit has all of its artistic merit in virtue of being an achievement of a certain kind or magnitude. Alternatively, she might claim that a thing can have some merit not in virtue of this. She might allow that a work could have great overall merit, even if it has only a little merit in virtue of how great an achievement it is. She believes that there can be, as I shall put it, ''achievement-independent merit''. A work's overall merit is a function of its achievement-independent merit and its achievement-dependent merit. Corresponding forms of enabler achievementism could also be formulated. Let us call the strong form of these views ''strong achievementism'', and the weak form ''weak achievementism''. The strong-weak distinction reveals another way different theories of merit can be compatible. You are a weak direct achievementist and a weak direct experientialist if you think that being a great achievement and affording an intrinsically valuable experience both give works merit directly.
In the literature, Denis Dutton is a weak achievementist. Dutton's view is multifaceted, but the best reading of him, as I argue below, takes him to hold that one work can be better than another in virtue of being a greater achievement. 3 However, he criticizes those who emphasize a work's ''status as human achievement, at the expense of attention to the purely formal properties'' (Dutton 1979, 310). It seems that, on his view, purely formal properties make independent contributions to the work's merit. By contrast, Gregory Currie appears to be an identity or strong achievementist. He holds that all judgements of artistic merit whatsoever are ''in part, judgements about the artist's achievement in producing the work'' (Currie 1989, 38-39;cf 19). 4 Huddleston (2012, 707, 709, 712, 713) and Dominic Lopes (2011Lopes ( , 530, 2014 discuss the view that a work's artistic value is identical to its value as an artistic achievement. Lopes (2011Lopes ( , 530, 532, 2014 also discusses the view that a work's artistic value is its value as the product of an artistic achievement. It is unclear to me what value as (the product of) 3 Dutton says (1) forgeries misrepresent their creator's achievement, and (2) a forgery is an achievement of a different kind than an original, because the forger performs a different task. One might think he accepts only these points, not weak achievementism, which he does not explicitly discuss. But if this were so, he would not have provided anything close to a solution to the problem he sets himself. He writes: ''The problem may be stated quite simply thus: if an aesthetic object has been widely admired and is discovered to be a forgery, a copy, or a misattribution, why reject it?' ' (1979, 303). Points (1) and (2) do not answer this question. Misrepresenting achievement is ethically bad. But why is it artistically bad? Dutton does not say. Nor does the fact that a forger performs a different task explain why her work has less merit. A comedian performs a different task than a flautist, but that does not explain why (or imply that) one of their works has less merit. So Dutton must be read as holding that the forgery is a different kind of achievement, not only because a different task is performed, but because it is a lesser achievement. Otherwise, we cannot ascribe to him anything that looks like a solution to his problem. 4 Classifying Currie is complicated by his remark that these judgements are ''in part'' about the artist's achievement. It is not clear to me how he understands this qualification, or that it precludes his being a strong or identity achievementist. an achievement is. It might be the greatness of the achievement. If so, these views are forms of identity achievementism. Or it might be the value (e.g. the final value) the achievement has in virtue of its greatness. If so, these views are forms of strong achievementism.
It has not been recognized that the achievement view comes in these different forms. Those who appeal to achievement to explain merit are rarely fully explicit about the content of the underlying theory. Appeal to achievement might explain facts about merit because achievement magnitude determines merit, because it enables another property to determine merit, or because it is identical to merit. The theory might be the strong claim that all of a thing's artistic merit is due to how great an achievement it is, or only the weaker claim that being an achievement of a certain magnitude confers a certain degree of merit. And the greatness of the achievement might make a difference to merit directly, or do so only because it makes a difference to some other factor that affects merit, such as the value of the experience. Making these distinctions is therefore necessary to get a clear view of what the achievement theory of merit is.
This will enable us to assess the theory fairly. The discussion above shows, for instance, that we cannot object to achievementism simply by providing an argument for an alternative theory, such as aestheticism: only certain forms of aestheticism are incompatible with certain forms of achievementism. And as we shall see below, some forms of achievementism have more resources for dealing with certain objections than others do. The achievementist's first challenge is to specify the kind of achievement whose magnitude makes a difference to merit. An artwork can be an achievement of more than one kind. Any plausible achievementist theory will need to place restrictions on the kind of achievement it appeals to. 5 First, there must be some restriction on what is achieved. A novel could contain a brilliant proof of a mathematical theorem; some books of the Bible contain elaborate legal codes and genealogies. These works, or parts of them, might be mathematical, jurisprudential, or antiquarian achievements. But this would not normally make any difference to the artistic merit of the work or part. 6 The right kind of thing is not achieved. Achievementists can differ over what the right kind of 5 Many acknowledge this, at least implicitly: Dutton (1979, 308), Currie (1989, 40), Budd (2007Budd ( , 360, 2014, Huddleston (2012, 706), Lopes (2014, 97). 6 Perhaps a work of conceptual art could wittily satirize our culture's preoccupation with achievement partly in virtue of being a great jurisprudential achievement. I take it that an achievement-based theory requires more than this. A determination theory of merit must not just identify properties that can sometimes confer merit, but properties that (for instance) normally or by default confer merit. thing is. For example, some might think aesthetic value must be achieved, others that offering an intrinsically valuable experience is what counts. 7 However, specifying the right kind of thing to achieve is not enough. In addition, the magnitude of the kind of achievement appealed to must have the right determinants. Sometimes, an artist's achievement is enhanced by factors that do not make the work better. The literary historian Pat Rogers describes the circumstances in which Boswell wrote his biography of Johnson as follows: The Life came out when he was just fifty; he had recently lost his wife, his long-established drinking habits had started to catch up with him, and he was undergoing bitter failure at the bar and in politics. These facts do not augment the merits of the book, either in terms of artistry or documentary value. They serve rather to emphasize the human achievement involved in creating the Life…. (1980, vii) The problem here is not that the wrong kind of thing is achieved. Rather, the wrong kind of thing makes the achievement greater. Boswell's difficulties make his biography a greater achievement. But they do not make it better literature or a better work of art.
One might think this problem is solved simply by saying that it is artistic achievement, where an achievement is artistic at least partly in virtue of what is achieved, that makes a difference to merit. We can also count as instances of this view theories that connect merit to kinds of artistic achievement, such as literary or architectural achievement (see Lopes 2014, 97-100). This, one might think, saves the achievementist having to specify what is achieved in the right kind of achievement. Whatever is achieved in artistic achievement in virtue of which the achievement is artistic, it is strongly correlated with artistic merit, unlike what is achieved in, say, jurisprudential achievement. Perhaps artistic achievement magnitude also has the right determinants. Boswell's biography, one might think, would have been an equally great literary and artistic achievement if he had not faced the difficulties mentioned above.
This might seem like the obvious form of achievementism to adopt. But such an approach faces a challenge of its own.

The order of explanation problem
The challenge is to show that the artistic-achievement theory gets the order of explanation right. A natural thought is that a work is a great artistic achievement partly because it has great artistic merit. But it does not have great artistic merit even partly because it is a great artistic achievement. Nor is having great artistic merit identical to being a great artistic achievement. Rather, artistic merit is prior to artistic achievement. A similar objection applies to the view that a work has merit at least partly because creating it was a great artistic achievement. It seems, rather, that writing Ulysses was a great artistic achievement at least partly because Ulysses has great artistic merit.
This objection is plausible in its own right, but we can also argue for it. In other cases, kinds of goodness are prior to the corresponding kinds of achievement. Take scientific theories that are scientific achievements. The factors that affect how good a theory is-the degree of understanding it provides, the diversity of phenomena it unifies, whether its theorems were arrived at from axioms that were known or merely postulated, etc.-also affect how great a scientific achievement it is. This can hardly be a coincidence. But these features do not make it a better theory because they make it a greater scientific achievement. A theory is not better for providing great understanding because this makes it a greater scientific achievement. Rather, providing great understanding makes the theory a greater scientific achievement because this makes it a better theory. Likewise, the features that make a theory better can also make the process of arriving at it a greater scientific achievement. Arriving at a theory that unifies more diverse phenomena is, ceteris paribus, a greater scientific achievement than arriving at one that unifies less diverse phenomena. But it would be bizarre to suppose that unifying more diverse phenomena makes a theory better because this makes the process of arriving at it a greater scientific achievement. It makes the process a greater scientific achievement because it makes the theory better. 8 Similar points could be made about being a great diplomatic achievement and being good diplomacy; being a great achievement in parenting and parenting well; being a great achievement in engineering and being a good work of engineering; etc. So goodness of a kind is sometimes prior to achievement of that kind. This is not conclusive evidence that this is always so. But it is some evidence that it is in the nature of kinds of achievement to be responsive to the corresponding kinds of goodness, not the other way around.
An achievementist who does not invoke artistic achievement could avoid this problem. For example, she might specify the right kind of achievement by direct reference to what is achieved, such as aesthetic value. She might say that merit is determined by how great an aesthetic achievement the thing is, not how great an artistic achievement it is. Aesthetic value is prior to aesthetic achievement. But artistic merit, she might claim, is not. 8 Achievementists influenced by the view that knowledge is an achievement might reply as follows. According to that view, an achievement is a success (e.g. believing truly) due to a non-Gettiered exercise of ability. Now, suppose a scientist arrives at a true theory from a non-Gettiered exercise of cognitive ability. This makes it a better theory than it would be if her exercise of ability had been Gettiered. Therefore, being an achievement makes it a better theory. This argument, if sound, would not show that being a greater achievement makes a theory better, but only that being an achievement does. However, it fails to show even this. It is plausible enough that being a success from a non-Gettiered exercise of ability makes a theory better. But this does not mean that being an achievement makes it better, because being an achievement is a different property from being a success from a non-Gettiered exercise of ability. To return to Keller's example from Sect. 1, adding milk to coffee can be a successful non-Gettiered exercise of ability (no gust of wind blew the stream of milk off course and back again) without being an achievement, and without the success (the milky coffee) being one. Therefore, being a success from a non-Gettiered exercise of ability is not the same as being an achievement. The former property may make a theory better, but this does not establish that the latter does.
An achievementist who takes this approach must, of course, avoid the wrong kind of achievement problem. Avoiding this without appealing to artistic achievement is a formidable challenge. Take the aesthetic-achievement theory just outlined. Some qualities can make a work a greater aesthetic achievement without giving it more artistic merit. The elegance of a geometrical proof inserted into a novel might not give the novel more artistic merit. But it might make the novel a greater aesthetic achievement. Perhaps the proof is exceptionally elegant, and it required great skill and effort to make it so. And perhaps, without the proof, the novel would have had little aesthetic value, and giving it that aesthetic value would have taken little skill or effort. It seems to me that, in such a case, the novel would be a greater aesthetic achievement with the proof than without it, even if the proof had a detrimental effect on the novel's unity. But the novel would not necessarily have more artistic merit.
If this is right, the achievementist has three options. First, she can try again to identify a suitable kind of achievement, other than artistic achievement, that avoids the order of explanation problem. Second, she can try to argue that artistic achievement is not enhanced by artistic merit, after all. Third, she can argue for a more nuanced position suggested by weak achievementism.
A weak achievementist can preserve the idea that it is artistic achievement whose magnitude affects merit. And she can agree that a work is a great artistic achievement partly because of its artistic merit. She can accept this because she believes in achievement-independent merit. That is, she believes works can have a certain degree of merit not in virtue of how great an achievement they are. So the weak achievementist can say that a work is a great artistic achievement partly because it has great achievement-independent merit. She cannot agree that it is a great artistic achievement because or partly because it has great overall merit. Its overall merit is partly determined by how great an artistic achievement it is. But she can say that, when a work is an artistic achievement, it has a certain degree of merit not in virtue of how great an artistic achievement it is. This degree of achievementindependent merit affects how great an artistic achievement it is. So some merit is indeed prior to artistic achievement. That is the truth in the order of explanation objection. Strong achievementists and identity achievementists, by contrast, cannot allow that any merit is prior to the right kind of achievement. But weak achievementists can say both that artistic achievement is the right kind of achievement, and that some of a work's merit affects how great an artistic achievement it is.
These responses to the order of explanation and wrong kind of achievement problems would require further development. I do not know whether any of them can ultimately be made to work. Even if one of them works, however, achievementism faces further problems.

The lucky merit problem
Strong achievementism and identity achievementism imply that anything with artistic merit is an achievement. But there are counterexamples to this claim.
Consider comic performances. It is a familiar fact that timing is crucial in comedy. Delivering a line at just the right moment can make a significant difference to how funny the delivery is.
A performer could, however, deliver a perfectly timed line by luck. Imagine an actor with no sense of comic timing. She just counts to two in her head before delivering her line. On this occasion, that is exactly the right time to deliver it. As a result, her delivery of the line is hilarious.
This actor's delivery could have artistic merit. She delivered the line well, though it is no credit to her. Her delivery was better than it would have been if she had waited a second longer. These facts suggest that her delivery had some artistic merit.
However, a performance that is successful by luck (of this kind) is not an achievement. This is therefore a counterexample to identity achievementism and strong achievementism. If there are non-achievements with merit, then having merit is not the same as being an achievement. Nor does everything with merit have all of its merit in virtue of being an achievement of a certain magnitude. Nor does achievement magnitude enable other properties to confer merit whenever they confer merit. 9 This objection leaves weak achievementism unscathed. Weak achievementists can allow that non-achievements have merit. The next objection, however, applies to all forms of achievementism. And it is most difficult for weak achievementists to answer.

The one explanation too many problem
To articulate this objection, we must first note some features of achievement in general. The magnitude of an achievement, of whatever kind, is determined by at least two sorts of factor. On the one hand, there is something about the agent's task or her performance of it. The difficulty of the task, the difficulty for the agent of performing it, the effort required, the effort actually exerted, or the ability, skill, or excellence displayed are all plausible candidates. It is a matter of debate which of these it is. 10 But let us call whatever it is about the task or the performance of it that enhances achievement magnitude ''the performance factor''.
The performance factor is not all that affects achievement magnitude. It is a greater achievement to save ten thousand lives than to save twenty lives, even if the two achievements were equally difficult and skilful, required equal effort, and so 9 The same objection applies, mutatis mutandis, if we imagine something that is an achievement, but which has some merit-conferring properties through luck. Perhaps the actor's overall performance that night was an achievement, though her comic timing was perfect only through luck. Her perfect timing would still give her overall performance some merit. But her timing would not make her performance a greater achievement (nor, obviously, would it be perfectly timed in virtue of how great an achievement it was). 10 See James (2005, 438-440), Hurka (2006, 221), Tasioulas (2006, 251-254), Portmore (2007, 4, 7, 10), Reginster (2007, 42-49), Greco (2010), Pritchard (2010, 68), Bradford (2015a, Chap. 2), von Kriegstein (2017. Some of these philosophers consider whether such factors are conditions of being an achievement at all, not whether they enhance achievement magnitude.
forth. Two physicists, each with different evidence, might make equally rational, skilful, and excellent use of their evidence, and develop theories that were equally difficult to develop. But due to differences in their evidence, one might develop a better theory: it is more accurate and provides greater understanding. It is the greater scientific achievement.
These cases suggest that the greatness of an achievement is affected by more than the performance factor. What also matters is how much is achieved. The person who saves ten thousand lives achieves more, as does the physicist who provides greater understanding. That is why their achievements are greater than those of their counterparts, even if performance factors are the same. The same effort, ability, and skill achieves greater success.
Achievementism in aesthetics has only been advanced to explain differences in artistic merit when there are differences in performance factors: when, for instance, greater creativity went into an original than a copy. However, if achievementism is true, it should also explain differences in merit when there is no difference in performance factors, but one achievement is greater because more is achieved. It should then be plausible that this work is better because it is a greater achievement. It will not only be true that it is better because it is superior in respect of what makes it the greater achievement (e.g. it is more beautiful). It will also be true that it is better because it is the greater achievement.
We find such cases when technology enables artists to achieve more with the same level of ability, effort, and difficulty. For instance, the Stradivarius is said to have a finer sound quality than other violins. A performance on a Stradivarius might be a more beautiful rendition of a piece than an equally skilful, difficult, and effortful performance on another violin. The sound might be sweeter and fuller. The performance might also better express the spirit or character of the piece. This performance achieves greater beauty and better expression, not because more skill went into it, but because a better instrument was used.
Similarly, recent films have better special effects than older films, due to new technologies. A new film's superior special effects could manifest the same ability, require the same effort, and be as difficult to achieve as those of an older film. But greater realism might be achieved in the new film. As a result, other artistic merits might also be achieved to a higher degree. Greater dramatic tension might be achieved in the film in which the monster is more realistically rendered, and other characters' reactions to it might be more believable. 11 In each of these pairs, the work in which more is achieved is a greater achievement (if it is an achievement) because more is achieved. The performance on the Stradivarius is a greater aesthetic achievement, as well as a greater achievement in expressing the spirit of the piece. It could therefore be a greater musical achievement overall. The film with the better special effects is a greater achievement in cinematic realism and a greater achievement in horror. It could therefore be a greater cinematic achievement overall. 12 Now, other things being equal, the greater achievement in these pairs also has more artistic merit. The new film has more merit because its special effects are more realistic, it has greater dramatic tension, and its characters' reactions are more believable. The performance on the Stradivarius has more merit because the sound is sweeter, fuller, and more beautiful, and because it better expresses the spirit of the piece.
The achievementist can agree with these explanations. But she must say that there is more to it than this. First, in her view it is also true that the more realistic film and the more beautiful performance have greater merit because they are greater achievements. Second, she must give a certain explanation of why the properties I have mentioned (greater realism, greater beauty, and so forth) give these works greater merit. She must say that, although the two films are made with equal skill, the new film's greater realism makes it better because it makes the film a greater achievement. Likewise, being more beautiful gives the equally skilful performance more merit because this makes it a greater achievement.
Are these claims plausible? My own impression is that they are not. I do not think they would recommend themselves, as explanations of why these works have greater merit, or why greater realism and greater beauty give a work greater merit, to someone not already persuaded of achievementism. Each seems to be one explanation too many. At least, they are less plausible than achievementist explanations of cases in which performance factors differ. That in itself is hard to explain if achievementism is true.
It is especially hard for a weak achievementist to defend these claims. She believes in achievement-independent merit. Suppose she says that the new film's greater realism gives it more achievement-independent merit than the older film's realism gives the older film. Since other things are equal, it follows that the new film has more total merit than the older film. So by the weak achievementist's own theory, the claim that the film is a greater achievement is not needed to explain why it has more total merit.
The weak achievementist must argue that this claim is needed to explain how much more total merit the new film has. She must argue that its greater realism gives it more achievement-independent merit and more achievement-dependent merit than the older film. So she must somehow establish that the new film surpasses the older film in total merit by more than it surpasses it in achievement-independent merit.
It may be possible to defend this rather baroque picture. But it is difficult to see its appeal. Again, we appear to have one explanation too many. Indeed, we now appear to have one explanandum too many. The weak achievementist must insist that the work that is the greater achievement is not only better: it is so much better that its betterness cannot be fully constituted by its greater achievement-independent merit. This additional fact, unlike the fact that the work is better, is not obvious. So weak achievementists not only offer an implausible explanation of why the work is better: they must also establish that there is a further fact requiring explanation. Weak achievementism looked like a potential solution to the order of explanation and wrong kind of achievement problems, and avoided the problem of lucky merit altogether. But it seems to be in the worst shape of any form of achievementism when it comes to the one explanation too many problem.
One might think that accepting achievementism's explanation of these cases is a price worth paying if its explanation of other cases is sufficiently good. If achievementism gives an unrivalled account of differences in merit when performance factors differ, perhaps we should accept its account here, even if it is not independently plausible. In the next section, however, I shall argue that achievementism is not the best explanation even of cases where performance factors differ.

A simpler hypothesis
The fact that achievement is only used to explain differences in merit when there are differences in the performance factor suggests a simpler hypothesis. Perhaps it is just the performance factor, or some feature that accompanies it, that explains differences in merit when achievementism seems plausible. In this section, I shall defend this claim.
We can explain these cases, I suggest, by appeal to a difference in the excellence of the artists' perceptions and acts, including their mental acts. Excellences include qualities like dextrousness, courageousness, and imaginativeness. They are forms of skilfulness, virtuousness, or aptitude. Excellence may or may not be the performance factor. I wish to remain neutral about this. But differences in excellence (e.g. degree of skilfulness) are often accompanied by differences in features that are plausible candidates for the performance factor (e.g. the difficulty of the artist's task). So if differences in excellence account for these differences in merit, that would explain why differences in achievement magnitude seem to do so.
Many achievementists would accept that the excellence of the artist's acts enhances the work's merit. In fact, they must accept this if they believe that excellence enhances achievement magnitude. But in their view, excellence enhances merit because it enhances achievement magnitude. That claim is false, if my arguments succeed. An excellence theory of merit can be, but should not be, combined with achievementism. 13 Two clarifications are in order. First, my claim will be that differences in the excellence of the artists' acts and perceptions in creating their works explain differences in merit. I am not appealing to differences in the excellence of artists themselves. 14 For instance, in comparing a van Gogh and a forgery, I am not contrasting how imaginative van Gogh was and how imaginative the forger was. I am contrasting the imaginativeness of van Gogh's acts, in creating his painting, with the imaginativeness of the forger's acts, in creating hers. An imaginative person does not always act imaginatively, and an unimaginative person may sometimes act imaginatively. It is the acts that matter for my purposes.
Second, I will appeal to acts and perceptions whose excellence is manifested in the work, as intelligent films manifest the intelligence of the thinking that went into them. I am not appealing to acts that merely formed part of the work's causal history, but whose excellences are not manifested in the work (e.g. the film director's skilful driving on the way to the studio). I do not have an answer to the difficult question of what it is for a work to manifest an excellence, but I can point to central cases of what I have in mind: intelligent films manifest the intelligence of the director's thinking, skilfully carved sculptures manifest the skilfulness of acts of carving, and perceptive satire manifests the perceptiveness of the writer's observations of the world. Commonly, when an act's excellence is manifested in a work, we can apply to the work itself an adjective that can also be used to attribute the excellence to the act ('intelligent', 'imaginative', 'perceptive', etc.). For bearers of merit that are themselves acts (e.g. a dancer's performance), I shall also appeal to excellences they possess (e.g. skilfulness), not merely excellences they manifest. 15 Examples used to support achievementism include the following: Liszt. A performance of Liszt's Transcendental É tudes at normal speed is not as good as a performance at dazzling speed. (Dutton 1979, 304) Improvisation. Some rehearsed musical performances are not as good as they would have been if improvised. (Dutton 1979, 312) Demoiselles. Picasso's Demoiselles d'Avignon has great merit, despite its hasty composition and stylistic inconsistencies. But it would have had much less merit if it had been a late Cubist painting instead of an early one. (Currie 1989, 34-35) These cases are easily explained by differences in excellence. In Liszt, the performance at dazzling speed is more virtuoso than the slower one. In Improvisation and Demoiselles, the superior work manifests greater excellence of conception (e.g. imaginativeness and independence of mind) than the inferior work.
What was imaginative in early Cubism, for instance, would not be imaginative in late Cubism. The superior work manifests these greater excellences of conception in addition to whatever aptitudes of execution the inferior work manifests. So it manifests greater excellence overall. The following cases are similar: Vermeer. Han van Meegeren's twentieth-century Vermeer forgeries are not masterpieces. But they would have been if they had been genuine Vermeers. (Dutton 1979, 302-303) Giotto. A modern fake Giotto might be nearly worthless artistically, even if it is as good or better at representing volume and movement than any work of Giotto's time. But it would, rightly, be highly praised if it were a genuine Giotto. (Currie 1989, 35-36) These cases are underdescribed. Not enough information is given to make it clear that the better work is the greater achievement. But the natural ways of filling out the cases to establish this involve differences in excellence. Perhaps the paintings would have manifested greater imaginativeness (e.g. about how to represent Christ), perceptiveness (e.g. about how to represent volume and movement), or insight (e.g. into human psychology) if they had been genuine. What is imaginative, perceptive, or insightful earlier in history might not be later, when artists have more examples to learn from. If the cases are filled out like this, the judgement that they would have been greater achievements if original would be vindicated. But they would also have manifested greater excellence: in addition to manifesting the technical excellences they actually manifest, they would have manifested these other excellences. Currie (1989, 36-39) presents a more unusual case: Guernica. Picasso's Guernica is a great painting. But suppose a painting that looks the same and belongs to the same artistic category is produced independently by a Martian child. The Martians have vastly greater artistic abilities than we do, such that the child's work is unremarkable. The Martian painting would be inferior.
Currie thinks this shows that the difference between the level of ability that went into the work and the level prevailing in its community of origin affects its merit. This difference is greater in Picasso's case than in the Martian's case, so the Picasso is superior. Currie concludes from this that differences in achievement explain differences in merit. One might think Currie himself has shown how an excellence theory should explain this example: namely, by appeal to the difference between the level of excellence the work manifests and the prevailing level. However, it is not clear that we should accept that explanation, because the case is seriously underdescribed. For instance, the Martian child, though as technically skilled as Picasso, may be less imaginative. Perhaps the ideas that it was imaginative for Picasso to have are not imaginative ones for Martian children to have. If so, the explanation of the difference in merit is simply that the Picasso manifests greater excellence. Alternatively, suppose the same levels of the same excellences-the same degree of imaginativeness, technical skilfulness, and so on-went into both paintings. Is the child's work inferior? It is not clear to me that it is. Rather, this seems to be a scenario in which Martian children produce works as good as Picasso's, and Martian adults produce works far better even than Picasso's. Indeed, it is otherwise hard to see how it could be true that Martian adults have vastly greater artistic abilities than humans.
Consider a final example, presented by Huddleston (2012, 712-713): Tempest. A machine randomly shoots paint onto a canvas. It was not designed for this: it just happens to do it. One day it produces an array of colour that looks just like Giorgione's Tempest. This array has many of the same aesthetic values as the Tempest. But it has no artistic value, even though the Tempest has great artistic value in virtue of these aesthetic values.
Why doesn't the array's aesthetic value give it artistic value? ''The most intuitive explanation'', Huddleston writes, ''is that Giorgione's painting is a great human achievement and the array is not. When aesthetic value is a value of art as art, it is only because this aesthetic value has been achieved by an artist'' (2012, 713). However, a difference in excellence explains this case as well as a difference in achievement does. The aesthetic values of Giorgione's painting manifest the excellence of his actions, but the array's aesthetic values do not manifest the excellence of any actions. This suggests that the notion of achievement is not needed to explain the cases it has been used to explain. They can be explained as cases in which the better work is better because it manifests or instantiates greater excellence. Moreover, this explanation appeals to factors that everyone, achievementists included, should agree make a difference to merit anyway. The perceptiveness, creativity, intelligence, and skilfulness of an artwork or performance evidently make a difference to its artistic merit. In appealing to excellence to explain these cases, we are not appealing to anything that we aren't already committed to thinking does explanatory work.
These considerations count against achievementism in their own right. They also strengthen the one explanation too many objection. We have no reason to bite the bullet by endorsing achievementist explanations of the cases described in the last section if achievementism does not give the best explanation of other cases.

Failing better
An appeal to excellence not only explains the cases achievementism explains. It explains more besides.
One set of facts that achievementism does not explain are certain facts about artistic failure. Experimentation in the arts does not always succeed, but the attempt can still be creative, bold, and intelligent. Some failures manifest excellence. And at least sometimes, excellent or competent failure is better than inept failure. At the end of the detective story, it turns out that the dog did it. This may not work as an ending for that story, but at least it shows a certain inventiveness. The ending might be better than if the butler did it, which would show a total lack of imagination. An excellence theory has no trouble explaining this.
Achievementism, however, will struggle. Achievement requires success at what the agent's effort, ability, or excellence is directed to, or in the difficult task she undertakes. The presence of excellence is not enough. An archer's skilful shot is not an achievement if, due to a gust of wind she couldn't have foreseen, she misses the target altogether. A legislator's herculean and ingenious efforts to pass a bill are not, and do not result in, an achievement if the bill does not pass, and if her efforts achieve no other successes. Similarly, an improviser's unfunny remark, a plot that fails to cohere, and an architectural design that fails at harmonizing the requirements of classical architecture with the functional requirements of the building are not achievements, even if they were intelligent and imaginative attempts to be funny, to achieve coherence, or to harmonize competing demands.
This makes it hard for achievementism to explain why artistic failures that are excellent or manifest excellence in their failures can have greater merit in virtue of this than inept failures. The work that fails excellently is not an achievement in the respect in which it fails (e.g. the ending of the story), and it might not be an achievement overall. So it is not a greater achievement (overall, or in this respect) than the work that fails ineptly. And even if it were an achievement overall, only successful excellence, or manifesting it, makes something a greater achievement, if (manifesting) excellence of some kind is the performance factor. But being excellent and manifesting excellence, even in what a work fails at, can give a work artistic merit.
Two clarifications are worth making about this objection. First, it does not imply that being excellent or manifesting excellence are not themselves kinds of success. Indeed, when instantiating or manifesting excellence gives a work merit, it plainly is a kind of success. The claim is, rather, that success at what the excellence aims at (a coherent plot, a plausible but hard to predict solution to the mystery, etc.) is necessary for achievement.
Second, the objection does not imply that achievement requires success in everything the agent is trying to achieve. The success required for achievement might only be the attainment (through sufficient effort or ability) of a sufficiently high standard for the task, even if the agent was aiming higher. Finishing fourth in an Olympic sprint can still be an achievement if the sprinter ran sufficiently fast, even though she was trying to run fast enough to win. She attained a high enough speed, through her athletic ability, to be credited with an achievement.
An achievementist might respond to this objection by adopting a conception of success such that failures like the above are always at least minor successes of a relevant kind. But this will be a difficult case to make if success is conceived in the ways suggested by discussions of achievementism. One of Dutton's most influential claims is that we need to know the artist's task to evaluate her work, because this tells us what kind of achievement her work represents. Her task might be to set a poem to music or to harmonize warm and cool colours (e.g. the red of the Virgin's robe and the blue of her cloak) (Dutton 1979, 306-307). If success at tasks like these is required for achievement, it looks like there can indeed be failures that manifest aptitude but are not successes of a relevant kind. A painting that is a clever but unsuccessful attempt to harmonize warm and cool colours can be better than a failure that is not even clever. But the cleverness does not make the painting a greater achievement.
This objection provides counterexamples to strong achievementism and identity achievementism. Those forms of achievementism are false if there are nonachievements with artistic merit. They are also false if any property unrelated to the magnitude of the achievement gives something artistic merit.
These are not counterexamples to weak achievementism, which does not claim to account for all the merit of everything with merit. But this objection still provides a reason to reject weak achievementism. It shows that an excellence theory can explain facts that weak achievementism cannot. An excellence theory also explains the cases weak achievementism explains, as I argued in Sect. 3.1. An excellence theory is therefore a better theory than weak achievementism. 16

The prospects for an excellence theory of merit
If the above is right, achievement is a red herring. What matters is excellence. But accepting this does not require us to abandon the attractive features of achievementism.
We can agree with achievementists that we need to know what the artist's task was to evaluate her work. One reason we need to know this is that it tells us what kind and degree of aptitude and ineptitude her work manifests. Alex Neill gives the example of a poem that seems dreadfully sentimental, but which we discover is an academic exercise in sentimentality: ''What before seemed terribly heavy-handed … now seems rather deft. What before looked hideously inappropriate to the occasion now strikes us as nicely judged'' (Neill 1999, 206). On learning the artist's task, we do learn, as achievementists say, what sort of achievement the poem represents. But we also learn what kind and degree of aptitude or incompetence it manifests.
Relatedly, achievementism can seem plausible because it sounds right to say that ''what was achieved'' makes a difference to merit. But as noted above, being an achievement and being achieved are different. The fact that certain aims were 16 In the arguments above, the examples are performances, artworks, or their parts (e.g. a performer's delivery of a line). An achievementist might hold that acts of creating certain artworks, or acts of conceiving of them, are not performances, artworks, or parts of them, but can have artistic merit. (I thank an anonymous referee for suggesting this.) After all, some properties of acts of creation and conceptionsuch as skilfulness, sensitivity, and imaginativeness -are shared by performances, and can give performances artistic merit. One might conclude that they do the same for creatings and conceivings. One might hold that these properties give creatings and conceivings merit because they make these acts greater achievements. Now, I do not share the sense that acts of creation and conception, when not themselves performances, artworks, or parts of them, can have artistic merit. But suppose I am wrong. The excellence theory would still be the better theory of the merit of creatings and conceivings. As in other cases, it provides a simpler explanation of the impact upon merit of the excellences of these acts. And if excellent creatings and conceivings that are failures (such as having the inventive idea to make the dog the murderer in a detective story) can have artistic merit, an excellence theory can explain this, but achievementism cannot. achieved-realized through the artist's own efforts or abilities, rather than (e.g.) by accident-does matter. But an excellence theory allows that. Indeed, if it can matter that something was achieved even when it does not constitute or enhance an achievement, an excellence theory is better placed to accommodate this.
Finally, one might be drawn to achievementism because it grounds artistic merit in a finally valuable property-being a (substantial) achievement. But if excellences of acts, such as courageousness, brilliance, and creativity, have final value when directed toward artistic ends, an excellence theory retains this feature of achievementism. If the excellence of such acts has final value, their excellence makes them worth performing for their own sakes. This would be an extension of the virtue ethicist's claim that an act's virtuousness makes it worth performing for its own sake. 17 This is not the place to argue that excellences do have final value. But they are serious candidates for finally valuable features.
However, one might wonder whether an excellence theory also inherits some of achievementism's problems. Take the order of explanation problem. An analogous objection to the excellence theory would say that a work is (e.g.) imaginative partly because it has merit. But it does not have merit even partly because it is imaginative.
This objection is unconvincing for two reasons. First, some artworks do seem to have merit partly because they are imaginative. Blake's Marriage of Heaven and Hell and Joyce's Finnegan's Wake, for instance, are good partly because they are imaginative. So the objection's second claim is dubious. Second, it is false that a work is imaginative partly because it has merit. I have argued elsewhere that an artwork is imaginative in virtue of features which it was plausible to believe would give the work merit, where this was not an obvious way of giving the work merit to think of (Grant 2018, 335-336;cf Grant 2013, Chap. 3.5-3.6). The Bilbao Guggenheim is an imaginative building in virtue of its shape, because giving it that shape was a plausible but unobvious way of making it expressive, and thereby giving it merit. One needn't accept the details of this account to accept the core point. A work is imaginative in virtue of features that (in cases of success) also give it merit. And the explanation of why these features make it imaginative must mention that the features were conferred to give it merit. But the work is not imaginative even partly because it has merit. This provides a model for at least some other excellences that one might think are vulnerable to an order of explanation objection. A painting can be skilful in virtue of features that also give it merit; these features make it skilful because giving it those features was a way of giving it merit that satisfied certain conditions (e.g. giving it those features would be very difficult for most people). An account with a similar shape could be given for other properties, such as intelligence. This does not show that all excellences escape an order of explanation objection. But many important ones do.
The problem of lucky merit does not arise for a weak excellence theory. A weak determination excellence theory could accept that not everything with artistic merit has all of its merit in virtue of instantiating or manifesting excellences. Therefore, properties unconnected to excellence (such as a performance's luckily perfect timing) can confer merit, on this view.
An excellence theory can also avoid an analogue of the one explanation too many problem. The performance on the Stradivarius that is just as skilful as the one on the lesser instrument is better because it sounds sweeter and fuller. An excellence theory does not require us to say more. The greater sweetness does not make it a more skilful performance. So an excellence theory does not imply that the greater sweetness gives it greater merit by making it more skilful.
A remaining challenge is to solve an analogue of the wrong kind of achievement problem. The geometrical proof inserted into the novel manifests the intelligence of the writer's thinking. But this needn't give it merit. Why not? Similarly, Boswell demonstrated extraordinary perseverance over personal difficulties to write his biography of Johnson. Perseverance is plausibly an excellence. But his biography would have had no less merit if he had not had a drinking problem or lost his wife, and so had not needed as much perseverance to write it. Again, why not? What kind of excellence makes a difference to artistic merit? This problem does not make it doubtful that the skilfulness of many musical performances or the imaginativeness of many stories gives them merit. It still seems true that manifesting or instantiating various excellences confers merit. But if it sometimes makes no difference to merit, a developed excellence theory should explain why.
A full-dress excellence theory would need to answer other questions. The discussion above suggests that the only viable excellence theory will be a weak theory, because a strong or identity excellence theory will be vulnerable to the lucky merit problem. It will presumably be a weak determination theory, since it seems clear that works can have merit in virtue of being, for instance, imaginative. Whether manifesting excellence also enables other properties to confer merit is a question that remains to be considered. It also remains to be considered whether the correct weak excellence theory will be a direct or an indirect theory. Does the imaginativeness of a novel give it merit directly? Or does its imaginativeness confer merit only by giving the novel other properties that confer merit (such as affording a more valuable experience)?
A difference in excellence can be a difference in degree or a difference in kind. One act might be, on balance, more excellent than another. Or they might have different excellences (e.g. one is deft and the other is bold), even if neither is more excellent. Accordingly, an excellence theory might say only that manifesting more excellence can give a work more merit. Or it might say, in addition or instead, that manifesting certain excellences (such as creativity) gives a work more merit than manifesting others (such as technical skilfulness), even if there is no difference in degree of excellence manifested. There might be a hierarchy of excellences. So the theory can be quantitative, qualitative, or mixed. In arguing that an excellence theory is better than achievementism, I have appealed only to quantity. But an excellence theory might need a qualitative component.
A developed excellence theory would also need to say whether the same excellences make the same difference to merit regardless of the kind of art. Creativity is sometimes said to be less valued in some cultures (e.g. ancient Egypt), and to make less of a difference to the merit of their art. An excellence theory should say whether this is true.

Conclusion
Achievementism faces several problems. First, it is not clear what sort of achievement it ought to appeal to. Artistic achievement seems to be enhanced by artistic merit; other kinds of achievement seem to be enhanced by factors that make no difference to artistic merit. Second, strong and identity achievementism falsely imply that non-achievements cannot have merit. Third, achievementism is implausible when the better work is a greater achievement and what I have called ''performance factors'' are the same. The features that make it better (e.g. sounding sweeter) make it a greater achievement in such cases, but do not make it better because they make it a greater achievement. Fourth, differences in aptitude and excellence provide better explanations of all the differences in merit that achievementism explains. Fifth, they also explain why excellent artistic failures can be better than inept ones, whereas achievementism does not. Differences in excellence often accompany differences in achievement magnitude, which explains why achievementism can seem right.
We should consider how artistic merit relates to excellence, not achievement. The resulting theory could retain achievementism's attractive features. It faces challenges of its own, notably the wrong kind of excellence problem. But it would be an improvement on achievementism. That in itself would be at least a minor achievement.