DOES AESTHETICS REST UPON A MISTAKE?

John Dilworth

The mistake in question is that of supposing that there are works of art, which could serve as the main subject-matter of aesthetics. But alas, I come bearing sad tidings. Contrary to what has generally been assumed, there are no works of art. Not a single one. But wait, the news gets worse.  Not only aren't there any artworks, but it turns out that there could not be any works of art, either.  It is impossible for anything to be a work of art, and hence necessarily there are no such things.

I suppose that at first my thesis is bound to sound incredible. But the traditional view that there are works of art (or that artworks do exist) turns out to be a house of cards built upon assumptions which are highly questionable.

Let me start with what would seem to be the strongest class of cases for the traditional view, those in which certain physical objects are commonly identified as works of art - paintings, sculptures and the like. For any given painting, a short argument seems to guarantee its existence as a work of art or artwork. The painting exists, since it is a physical object. The painting is identical with a work of art. Therefore the work of art exists.

I do not deny for a moment that the painting and other such artifacts exist. I shall continue to use the term 'artifact' for such uncontroversial objects associated with works of art (the term could usefully be thought of as a condensed form of 'art artifact'.) My disagreement with the tradition instead focusses initially on premiss two of the short argument, namely the claim that the artifact in question is identical with a work of art. Once this premiss has been shown to be false, it will be possible to examine the question of the existence of artworks unimpeded by the undisputed fact of the existence of artifacts.

In denying the identity of artifacts and works of art, I shall remain non-committal on ontological issues such as that of whether there could be objects which have some kind of being or ontological status in spite of their not existing. I shall also leave it open that apparent references to or talk about artworks might more properly be explained or analysed away in terms of some other, non-object-centered mode of discourse.

I.   Artworks Are Not Identical with Artifacts

Consider that favorite example, the painting Mona Lisa by Leonardo da Vinci, which hangs in the Louvre Museum in Paris. The painting itself is an artifact which undoubtedly exists, and whose continued existence is assured by a system of transparent screens and barriers to prevent any harm coming to it from potentially hostile persons. This artifact of course has a multitude of physical properties, including its dimensions, weight, the physical and chemical composition of its canvas and pigments, and so on.

On the other hand, consider also the artwork Mona Lisa. This work shows us a woman with an enigmatic smile, situated in an Italian landscape which recedes in three-dimensional space behind her. Discerning viewers will probably note too that these elements are arranged in such a way, and are so modulated by Leonardo's characteristic style, that each heightens the aesthetic effect of the others in producing what is generally agreed to be an artistically very satisfying work of art.

My claim at this stage quite simply is that the artifact Mona Lisa and the artwork Mona Lisa cannot be identical, because different things are true of each of them.Consider for example the woman in the artwork. The artwork certainly has the property of containing a woman with an enigmatic smile, and indeed this is arguably an essential property of the artwork, since without it we would no longer acknowledge the artwork Mona Lisa to be the same artwork. On the other hand, the artifact Mona Lisa contains no such woman.

Admittedly the painting-artifact does include a woman-shaped area of paint which visually resembles a woman in some ways, but mere visual resemblance does not a woman make. The artifact at best contains a woman only in some very loose, metaphorical sense which is in no way equivalent to the way in which the artwork necessarily contains that woman who is the Mona Lisa.

Actually the modal point concerning the necessity of the woman to the artwork can be used to clinch the difference here. Suppose that some vandal managed to breach the security precautions at the Louve, and proceeded to scrape off that area of the paint which represents the woman. In such a case, the painting-artifact Mona Lisa would still retain its identity as the painting Mona Lisa executed by Leonardo (etc.) It would have been grossly disfigured, but it is still the same painting nevertheless.  On the other hand, the artwork Mona Lisa would no longer be identifiable at all under such circumstances: a necessary feature of its identity would have been removed.

Thus it follows that the artifact and artwork differ in at least one property. Even if it were granted that the artifact in some loose and popular sense does contain a woman, that woman lacks the property of being essential to the identity of the artifact. Whereas on the other hand for the artwork, the woman does possess the property of being essential to the identity of the artwork. Hence the artifact and artwork cannot be identical.

There are plenty of easier examples which could have been chosen to demonstrate the non-identity of artifacts and artworks. For example, in the artwork the Italian landscape clearly extends behind the woman in three-dimensional space, whereas of course on the painted surface of the artifact the landscape and the woman are all located in the same two-dimensional plane of the canvas. The artifact (or at least its surface) involves only two spatial dimensions, whereas the artwork is extended in three-dimensional space. Hence, again, the artifact and artwork cannot be identical.

I conclude that the non-identity of artifacts and artworks in such typical cases has been established, at least provisionally. Thus it is now possible to raise the question of whether artworks exist independently of the undoubted fact of the existence of artifacts.

II.  Evidence Against the Existence of Artworks

Evidence counting against the existence of artworks is surprisingly easy to find, once one approaches the issue without presuppositions. Indeed, the evidence is so strong that the issue of how people could have failed to give due weight to it becomes a pressing one. Thus, to lend credence to my account, I shall try to explain how this massive evidential neglect or discounting could have happened, as part of my presentation of the non-existence evidence.

Let me start with some commonplaces which everyone knows to be true. Everyone knows that the character Hamlet in Shakespeare's play is not a real person, but merely a fictional character, and hence that Hamlet himself does not actually exist. And similarly for the other characters in the play, the actions they perform and so on - everyone knows that none of these actually exist.

But why then are people (whether theoreticians or just people in general) so convinced that the play Hamlet itself does exist? After all, a commonsense or culturally unprejudiced view of the play would be that it is simply made up of or comprised of all of the characters and their actions and interactions. Since none of those elements or parts exist, should it not follow automatically that the play itself does not exist?

However, as a counterweight to that natural conclusion there is the not insignificant issue of something else which everyone knows, namely that some real person (generally assumed to be Shakespeare} did actually write the play Hamlet. Also the work in question has an unbroken performance tradition in hundreds of theatre companies from the day of its first performance until now. Thus its status as a real, unquestionably existing cultural entity seems as secure as anything could be.

Given these two kinds of facts about Hamlet - the undeniable fictionality of its characters etc, versus the equally undeniable reality of the play as a cultural institution - virtually everyone seems to have decided to have their cake and eat it too. In effect the assumption is made that the play or artwork Hamlet must be identical with the unquestionably existing cultural entity with which it is associated.

However, I have already discredited this assumption, with the aid of my initial distinction between artifacts and artworks and my demonstration that they are not identical since different things are true of each. From this perspective, the non-existence of the characters and actions in Hamlet can be taken at face value as near-conclusive initial evidence that the artwork Hamlet itself does not exist.  At the same time, my view explains the undeniable reality of Hamlet as a cultural institution in terms of  the undeniable reality of the artifact Hamlet, which is demonstrably not identical with the artwork Hamlet.

Similar points could  be made about other kinds of examples, including paintings such as the Mona Lisa. Since the woman seen in the picture does not exist any more than the landscape in the background and other details of the painting, there is no reason to deny the obvious conclusion that the artwork Mona Lisa does not exist either, once we clearly distinguish existent artifacts from the artworks with which they are associated. However, note that even if the painting depicts a real person or landscape, the current issue is about the status of the images of these things as they occur in the painting, not about the existence of any corresponding real-life entities they might depict.

III.   Artworks Necessarily Do Not Exist

The issue can now be raised about what kind of non-existence is being dealt with in the case of artworks. I shall initially discuss this question in terms of artworks which involve fictional entities, and assume for the present that results which hold for them can be generalized to apply to all artworks.

In the general case of artwork-derived fictional entities such as the character Hamlet, it is generally assumed that their non-existence is a contingent matter. Hamlet, it is assumed, does not exist in the real world, but he could have existed, and if the world were different in various possible ways, Hamlet would have existed. Using possible-worlds jargon, it is assumed that there are possible worlds in which Hamlet exists, even though our actual world is not one of those possible worlds.

However, it seems not to have been realized that there is an impenetrable roadblock to establishing identity links between the character Hamlet and an actual person Hamlet in other possible worlds.  To see this, note first that mere similarity (even exact similarity) is not sufficient to link an entity existing in one world to similar entities in other worlds - there must be a justifiable claim of identity, i.e., that it is actually the same identical entity which is being claimed to occur in each possible world. Secondly, such identity-claims require reasonably clear identity-criteria for an entity across worlds, and more specifically the absence of any identity-criteria which would prevent an entity from existing in a given world.

In the case of a fictional character such as Hamlet, there is at least one indisputable essential characteristic which he possesses, namely the obvious one of being a particular identifiable fictional character, who is identical with the fictional character Hamlet in the play Hamlet created by Shakespeare. Any world in which this characteristic is true of no entity in that world could not be a possible world in which Hamlet exists.

To put the matter more briefly, being identical with the fictional character Hamlet is an essential characteristic which must be possessed by any entity being claimed to be Hamlet in any world.

So far I have said nothing controversial or even substantive about Hamlet and possible worlds, since I have merely been spelling out the standard assumptions about identification of a particular entity across possible worlds as applied to Hamlet.

At this point it becomes unavoidable to consider metaphysical questions such as that of what kinds of changes between the actual world and other possible worlds would be so extreme as to prevent an entity from existing in those possible worlds. More traditionally, such questions have been discussed in terms of the essence or essential characteristics which must be possessed by a particular entity in every world in which it occurs.

Does the fictional character Hamlet have such an essence? I would claim that he does. Whatever else might be true of Hamlet, it is part of his essence that he is a particular fictional character occurring in the play Hamlet.

Any object in any world which failed to have that characteristic could not justifiably be identified as Hamlet. It should be noted too that this is not just a metaphysical point about Hamlet's essence, but also an epistemological point about the conditions under which an attribution of 'Hamlet-hood' to an entity could adequately be justified.

To see that both of these points are true, contrast the case of Hamlet with that of some real person.  Given the facts about human biology, it is normally the case that any human being is a particular person created by her two parents. But we do not take this as an essential characteristic of that person, because we assume that we could identify that person as herself independently of that fact. (She might have had other parents, i.e., there are other possible worlds in which she exists in which she has other parents.)

On the other hand, in the case of Hamlet there would be no way in which to identify Hamlet as Hamlet in other possible worlds independently of his characteristic of being a fictional character as found in the play Hamlet. Without that anchor there would be no way in which to distinguish close Hamlet look-alikes (persons with very similar characteristics to Hamlet's, but who nevertheless are different persons not identical with Hamlet) from the real thing.

Note that this essence of Hamlet is derived in a way interestingly different from traditional claims about the essences of individuals. Traditional claims are usually based on strong ontological intuitions about the essential nature of a kind of being. But with fictional entities, it is rather their epistemological precariousness which guarantees that they have an essence of a certain kind. Our knowledge about the character Hamlet is highly dependent upon the specific facts about his creator, the play in which he occurs, and so on. It is our inability to identify him independently of the precariously few particular facts about him which we do know which underwrites and guarantees the attribution to him of the above-mentioned essence.

I can now proceed to a conclusion about Hamlet's status in other possible worlds. It has been shown that it is of the essence of Hamlet to be a particular fictional character created by Shakespeare in any world in which Hamlet could be identified as Hamlet. Moreover, I have also independently shown that fictional characters do not exist in the actual world. But exactly parallel arguments about the non-existence of fictional entities in the actual world would also show that such entities do not exist in any possible world in which they occur as fictional entities. Hence in particular Hamlet will not exist in any world in which he occurs, since in any such world it must be true of him (as already shown) that he is a fictional character.

The outcome of this chain of reasoning is that Hamlet does not exist in any possible world in which he occurs.  But by definition, contingent non-existence requires existence in some possible world.  An entity such as Hamlet which does not exist in any possible world (or equivalently, which is non-existent in every possible world) must instead be necessarily non-existent.  Hence Hamlet and other artwork-derived fictional entities are necessarily non-existent. And as sugggested at the beginning of this section, this result could be extended to apply to all artworks: all of them are necessarily non-existent.

Nevertheless, in keeping with my stated policy of trying to explain the wide prevalence of alternate views, the question must be asked of why is it widely assumed that fictional characters as found in artworks are only contingently non-existent.

I think the answer lies in a conflation of imaginary or fictional entities in general with the specific cases of particular fictional entities as occurring in artworks which have close ties to particular artifacts created by artists.  In the first place, many fictional entities are not thought of as being particular individuals, but rather as being natural kinds, for example unicorns.   Unicorns only contingently fail to exist because any object which satisfied the general characteristics of unicornhood (being a horse-like animal with one horn, etc.) would actually be a unicorn.

In other cases such as the Tooth Fairy or Santa Claus the fictional entity is being assumed to be an individual, but there is nothing to prevent (in some possible world or worlds) a particular individual from satisfying the required characteristics of that individual.  In their case too, any object which (uniquely in that world) satisfies the general descriptive characteristics of the object in question would count as an actual instance of the (in our world fictional) entity in question.

Contrast all of these cases, however, with artwork-derived fictional entities.  These entities are not free-floating imaginary entities, such that anything satisfying their descriptive characteristics would count as an instance of them.  Instead they are inextricably linked to particular circumstances and events in the creation and continuing existence of their corresponding artifacts in our actual world, which do not allow the artworks to identifiably occur other than as fictional entities in any other world.

In conclusion, then, not only are there no artworks, but there could not have been such things either.  Thus aesthetic theory must somehow adjust itself to the necessary non-availability of its traditional cornerstone.  But how it might do so is a story for another occasion.