Art in an Expanded Field Wittgenstein and Aesthetics

a bstr act This article reviews the various ways in which the later writings of Ludwig Wittgenstein have been employed to address the question “What is Art?” These include the family resemblance model, the cluster concept model and the form of life model. The article defends a version of the form of life approach. Also, addressed the charge that it would have been more profitable had aestheticians explored what Wittgenstein actually said about art instead of trying to extrapolate from his writings an approach to what Nigel Warburton calls the art question. k ey wor ds Wittgenstein, the art question, family resemblance, cluster concept, form-of-life


I. Introduction
I am not a Wittgenstein scholar. I am a student of the philosophy of art. So what this article will discuss primarily is the influence of Wittgensteinian ideas on the philosophy of art. As has been noted by others, 2 this influence is most evident regarding questions about the way or ways in which we go about identifying art. This, of course, is somewhat arresting, since this is not a topic that appears to concern Wittgenstein himself when he broaches the topic of art. Indeed, some observers have commented that it might be more profitable for philosophers of art to attend to what Wittgenstein said about art rather than attempting to invent a Wittgensteinian approach to the issue of the identification of art. That, it has been suggested, would have brought something new, exciting, and even pathbreaking to the study of the philosophy of art. 3 Instead, philosophers of art have deployed Wittgenstein's thought to their perennial obsession with the question of "What is art?" when they might, it has been recommended, have struck out in more promising heretofore unexplored directions.
I agree that the Wittgenstein's influence on philosophers of art has been to suggest ways in which to go about identifying art. In what follows, I will examine three approaches to this question that claim roots in Wittgenstein's work. They are what might be called the family resemblance approach (which is most frequently associated with Morris Weitz), the cluster concept approach (which has been defended recently by Berys Gaut), and the forms of life approach (which, I believe, was first suggested by Richard Wollheim in his book Art and its Objects). Of the three approaches, I favor the forms of life approach, not only as the most promising application of Wittgenstein's thought to this problem, but also as at least part of the most effective solution to the question of how to identify art. To this end, I will attempt to expand and defend a modified version of the position introduced by Wollheim. But I will also address the charge that philosophers of art would have been better advised to explore what Wittgenstein had to say about art 4that that would have opened up new and more fruitful avenues of philosophical research. In contrast, I will maintain that (1) what Wittgenstein says about how we respond to art and conduct aesthetic debate is not as original as alleged, (2) that there is a standing branch of the philosophy of art concerned with the phenomena that fascinated Wittsgenstein, namely, metacriticism, and (3) that Wittgenstein's conception of what goes on in the appreciation and criticism of art, although insightful as far as it goes, does not go far enough and, for that reason, is at least incomplete.
So what follows comes in four major parts, specifically: sections on the family resemblance approach, the cluster concept approach and the forms of life approach, followed by a brief and somewhat critical comment upon the purported originality and fecundity of Wittgenstein's remarks on aesthetics.

II. The Family Resemblance Approach
Although advanced by several thinkers, the family resemblance approach to identifying art is probably most frequently associated with Morris Weitz's widely anthologized article "The Role of Theory in Aesthetics." 5 In that article, Weitz attempted to extrapolate from his understanding of Wittgenstein's discussion of games to a position about the way in which to go about identifying artworks. It may seem odd that rather than focus upon what Wittgenstein said about art, Weitz contrived to use other parts of Wittgenstein's philosophy to deal with an issue he never broached. But such behavior is common in aesthetics, as when in his best known articles, Clement Greenberg turned to Kant's First Critique for constructing his theory of fine art rather than to his Third Critique.
Weitz's article was quite ambitious. As is well known, it proposed an argument against what Weitz conceived of the traditional way of doing the philosophy of art, namely the attempt to define art. He also introduced an alternative way of identifying art to the definitional approach, namely the family resemblance approach. And lastly, he offered a reconception of what traditional definitions were up to, thereby salvaging what was valuable in them. In particular, he maintained that once the traditional theories of art were reconceived as art criticism -pointing to often neglected features of emerging art -supposed theories of art, like formalism, could be seen to make a salutary contribution to the appreciation of challenging, innovative art, like Neo-Impressionism.
Although Weitz notes that all attempts to define art in the past have failed, he does not rest his case against the possibility of defining art on inductive grounds. Rather, he believes that he can demonstrate logically the impossibility of the attempt by philosophers of art to construct theories of art in terms of identifying the necessary and sufficient conditions for qualifying for the status of art. Weitz's argument against the possibility of defining art was called "the open-concept" argument. Basically the argument maintained that "the very expansive, adventurous character of art, its ever-present changes and novel creations make it logically impossible to ensure any set of defining properties." 6 The thrust of this argument can be expanded formally as a reduction ad absurdam in the following manner: 1.
Art is creative, constantly open to innovation.

2.
If something is creative, constantly open to innovation, then it cannot be defined in terms of necessary and sufficient conditions. 3.
Suppose that art can be defined.

4.
Then art is not creative, constantly open to innovation.

5.
Therefore art is not art.
Supposing that the open-concept argument is successful, it is incumbent upon Weitz to produce an alternative, nondefinitional way for identifying art, since we obviously succeed in doing so. Imitating Wittgenstein's discussion of games, Weitz maintains that the way in which we identify artifacts as artworks is by means of family resemblances. Taking some body of established artworks as paradigms, we assess new candidates in terms of their resemblance to the paradigms. Abstractions by Mondrian are art because, like paradigmatic representational works by Poussin, they have a discernible compositional structure. The more similarities between a candidate and paradigmatic artworks, the stronger the grounds for deciding that the former is an artwork. Moreover, the fact that there is a putatively successful alternative to the definitional approach provides additional grounds for abandoning it, especially in light of all of the successive attempts to discover the definition of art. Finally, Weitz has some kind words for the various essentialist attempts to define art. They failed to do what they tried to do; for, it couldn't be done, given the open-concept argument. But the theories concocted by these philosophers were not utterly uninformative. They were just not informative in the way in which their proponents imagined. Where they were valuable, they were valuable as art criticism, rather than as art theory. The expression theory of art, for example, may educate viewers, steeped in expectations of realism, in the emotive power of distortion, thereby enabling them to appreciate the value of a painting by El Greco.
Perhaps ironically, this reconstruction of the value of traditional philosophies of art may be the element of Weitz's philosophy which is of lasting merit. For, the open-concept argument and the family resemblance approach are both embattled.
The open-concept argument is in trouble virtually every step of the way. The first premise -that art is creative, constantly open to innovation -is dubious. One problem, among others, is that it is parochial. At times, especially in the West, we have prized art for being creative and innovative. That is certainly especially the case since the onset of the avant-garde in the nineteenth century. We fetishize, what Irving Howe once called, the tradition of the new. But there are periods in history and places where innovation was not considered the sine qua non of art; consider long stretches of Chinese art, not to mention our own middle ages. Indeed, in some situations, artistic innovation was considered the antithesis of the going art, as in the case of the realist interlude during the reign of Akhenaten in ancient Egypt.
Likewise, the second premise of the argument is troubling. For it is not the case that if something can be defined in terms of necessary and sufficient conditions, that that precludes creativity and innovation. Both George Dickie's earlier Institutional Theory of Art and its later incarnation as the Art Circle as well as the theory of art found in Danto's Transfiguration of the Commonplace provide necessary and conjointly sufficient conditions for art, but there is no artistic innovation that they preclude. Indeed, Danto explicitly, and, I think, successfully argues that, on his view, art can look like anything. None of these theories block artistic creativity in any respect. So, there are theories that define art in terms of necessary and sufficient conditions that are not impediments to creativity. Therefore, the second premise is false.
But perhaps a deeper problem with the open concept argument occurs as we move from the first two premises to the third premise. For, it appears that in the first two premises when we speak of art, we are speaking of the practice of art. Why? Because it makes sense to say that the practice of art is open to constant innovation. Individual artworks, save various forms of environmental-process art may be open to innovation, but most individual artworks are not. Richardson's Clarissa is not now open to innovation, nor is Leonardo's Virgin and Child with Saint Anne.
However, when we are asked to suppose that art can be defined in the third premise of the argument, it would appear to be the case that we are supposed to be thinking in terms of individual artworks, since this is typically what traditional art theorists like Clive Bell were attempting to define. Yet, if premise three is about defining artworks, then art in that premise is not referring to that which art refers in the first premise. Moreover, the upshot of this equivocation is that the argument fails to secure a contradiction, since there is no absurdity in asserting that art qua practice is not the same as art qua individual artwork.
Of course, the problems with the family resemblance approach are even better appreciated than the problems with the open-concept argument. The approach recommends choosing a set of paradigms against which to measure candidate works in terms of the affinities they bear to the paradigms. However, we are not given instructions about how to assemble the class of paradigms. So, at the very least, the family resemblance approach is radically incomplete. Moreover, the paradigms possess a large number of properties, presumably many of which are artistically irrelevant. Kliest's Marquise of O is written in German, but that provides no grounds for thinking that a recent issue of the Berliner Zeitung is an artwork because it resembles Kliest's masterpiece in respect of its language. Here the ready solution might appear to be to specify exactly in what ways candidates have to resemble the paradigms, yet that seems like a return to laying down definitional conditions. But without any constraints on which resemblances count, the family resemblance approach will lack the means to segregate artworks from everything else, since everything resembles everything else in some respect. Alien sex toys will resemble Michelangelo's David with respect to being material objects, as well as in an indefinitely large number of other artirrelevant properties, but that is scant reason to think them art. And, the same for virtually everything else in the universe.
Given the inadequacies of the family resemblance approach, let me turn to the cluster concept approach.

III. The Cluster Concept Approach
The cluster concept approach has been developed by Berys Gaut. 7 Gaut claims inspiration from Wittgenstein's discussion of the proper name "Moses", to wit: "By 'Moses' I understand the man who did what the Bible relates of Moses, or, at any rate, a good deal of it. But how much? Have I decided how much must be proved false for me to give up my proposition as false?" 8 Following Searle's refinement of this notion with respect to proper names, Gaut characterizes a cluster concept as one according to which there are multiple criteria for the application of a concept, none of which, however, is necessary. 9 That is, more precisely: "A cluster account is true of a concept just in case there are properties whose instantiation of an object counts as a matter of conceptual necessity toward falling under the concept." 10 On Gaut's conception of a cluster concept, there are several criteria for the application of a concept. If all the relevant criteria are instantiated, the object falls under the concept; that is, the sum total of the criteria are jointly sufficient for subsumption under the concept. As well, the disjunctive set of all the relevant criteria are disjunctively necessary for the application of the concept, although no individual criterion is necessary. Moreover, even if fewer than all of the pertinent criteria obtain, that may still be sufficient for the concept applying to the object. 11 Like Weitz, Gaut regards the cluster concept approach as non-definitional, since he denies that there is any set of individually necessary conditions which are also conjointly sufficient for art status. A cluster concept is not, in other words, a real or essential definition.
Employing this structure with respect to the concept of art, Gaut maintains that the pertinent cluster of criteria with regard to the application of the concept of art are: (1) that x possess positive aesthetic properties; (2) that x be expressive of emotion; (3) that x be intellectually challenging; (4) that x be formally complex and coherent; (5) that x have the capacity to convey complex meanings; (6) that x have a point of view; (7) that x be the result of an exercise of a high degree of skill; (8) that x be the result of an exercise of imagination; (9) that x be a member of a recognizable art form; (10) that x be the product of an intention to create a work of art. 12 In contrast to traditional approaches to identifying art, the cluster concept enumerates a large number of potential criteria for art status, whereas past theories of art usually emphasized one or two conditions for art status. Thus, the cluster concept approach has more heuristic power than its traditional predecessors just because it countenances more dimensions of artistic creation and appreciation. Thus, it situates art in an expanded field.
Nevertheless, as conceptually complex as the cluster concept approach is, I think that the version that Gaut presents is still unconvincing. According to Gaut, if an object satisfies all of the criteria he enumerates then that is sufficient for a candidate to count as an artwork. But is that so? Consider this counterexample.
A baker makes a cake to commemorate his anniversary of his marriage to his wife. They initially celebrated that event by taking a car trip during which they visited many cities. So the baker makes the cake in the shape of their car. Since his wife is an amateur cryptographer, he notes each of the cities they visited in alternating colors, along with remarks about what they saw there in code on the side of the cake. Arguably, since our cook regards baking as an art, this cake meets all of Gaut's criteria for art status, although this is not a result that most art lovers and philosophers of art would abide by.
The cake possesses positive aesthetic properties; it is pleasing to the eye. It expresses an emotion, namely love. It is intellectually challenging; it is in a non-obvious code. It is formally complex and coherent -the name places and accompanying observations are in alternating colors. It has the capacity to convey complex meanings, an affectionate history of the beginnings of their marriage. It has a point of view: that of the baker in love. It evinces great skill and imagination -there never was an anniversary cake like this one before. It belongs to a recognizable art form, specifically sculpture. And the baker intended it to be a work of art. But I doubt that few cognoscenti would accept it as such. Thus, Gaut's version of the cluster concept of art fails on its own terms.
Gaut also maintains that his ten criteria are disjunctively necessary for art status. That too may be disputed. There are interpretations of Duchamp's readymades that deny that Duchamp intended them as anything more than artworld pranks. That is, he did not intend them to be artworks nor did he intend them to make any challenging or complex statements. At the same time, something like his industrially produced, canine grooming comb lacks positive aesthetic qualities, evinces no great skill in its production, expresses no emotion, lacks a formal structure of the relevant sort, belongs to no recognizable art form and possesses no point of view. Suppose that all of this, as some have maintained, is true. Nevertheless, such works have still been accorded art status. Thus, it is at least possible that the entire set of criteria advanced by Gaut as constituting the cluster concept of art is not disjunctively necessary. So again, Gaut's version of the cluster concept of art fails.
Gaut's response to this, I predict, will be to say, as he has in the past, that even if his version of the cluster concept of art fails, that does not imply that the cluster concept approach to identifying art has failed. His only purpose was to show that the cluster concept approach may be a promising line of inquiry to explore in the search for a viable avenue for identifying candidates as artworks.
However, saying that demonstrating the failure of his version of the cluster concept approach does not preclude the possibility that some cluster concept approaches may succeed is not very compelling dialectically, since one can say, with equal logical propriety, that showing that this or that definitional approach fails, does not preclude the possibility that some version will succeed some day. Both claims are equally lackluster. In both cases, the burden of proof shifts to the proponent of the approach in question. That is, the definitionalist must come up with a convincing definition, while someone like Gaut must come up with a persuasive cluster concept. Failure to do so in either case, compromises the approach in question, suggesting that we look for another way in which to go about identifying art.
So, with that in mind, let me now turn to the forms of life approach.

IV. The Forms of Life Approach
The best known statement of the forms-of-life approach occurs, I believe, in section 45 of Richard Wollheim's Art and its Objects. 13 There, Wollheim asserts, that "Art, in a Wittgensteinian sense, is a form of life." I concede that this is a controversial interpretation of what Wittgenstein may have meant by the notion of "a form of life." Newton Garver would challenge Wollheim's understanding of this concept inasmuch as he believes that the phrase refers to human life in general. Others might argue that art is at best part of a form of life. Since I am not a Wittgenstein scholar, I will not enter this dispute. Rather, I begin with the assumptions that Wollheim's view is a reasonable, though perhaps not textually definitive, understanding of Wittgenstein's conception, and that Wollheim's articulation of this notion is widely shared among philosophers of art, albeit, perhaps, because of Wollheim's championing of this interpretation. For Wollheim, the conviction that art is a form of life has a number of interesting consequences. One is that "art is essentially historical." That is, the way in which to understand an evolving form of life --the way in which to grasp its unity --is to comprehend its history. For, among other things, it is by means of understanding the history of a form of life that we are able to specify a method for identifying artworks in a manner that is not definitional. Instead ascertaining membership in the relevant form of life is a historical matter.
Wollheim's suggestion about the way in which to go about doing this is as follows: "the method might take this form: that we should, first, pick out certain objects as original or primary works of art; and that we should then set up some rules which, successively applied to the original works of art, will give us (within certain rough limits) all subsequent or derivative works of art." 14 I believe that with certain modifications, including augmentations, Wollheim's proposal can be shown to be one -I stress the singular here -method for identifying candidates as art works. Moreover, this method should be especially attractive to Wittgensteinians, since it reflects the way in which the denizens of this form of life -which was christened as the artworld by Wollheim's colleague at Columbia University, Arthur Danto -actually go about establishing that candidates are artworks.
Mention of Danto here warrants a short detour for the purpose of distinguishing Danto, and for that matter Dickie's, artworlds from Wollheim's forms of life. Danto claimed that establishing candidates to be art works involved connecting them to an atmosphere of art theory and art history. One problem with Danto's early formulation was an apparent overemphasis on the role of theory. Too much art appeared to be created without the benefit of theory and, in any event, there was also some question about what would constitute an art theory. Indeed, can this be done without courting circularity? Like Wollheim, Danto would have been better off highlighting the importance of history and, again like Wollheim, invoking that which Danto regarded as enfrancishing art theories as operating as elements of the internal history of the art world. However, even with that emendation, Danto's approach would have differed from Wollheim's, since Danto, at least until the publication of After the End of Art, wanted appeal to the atmosphere of the artworld to figure as a necessary condition in his definition of art, as propounded in his Transfiguration of the Commonplace.
Simlarly, although there are some correspondences between both Dickie's earlier Institutional Theory of Art and his later account of the Art Circle, including the importance that he, like Wollheim, lays on the reciprocal relationship between the artist and his audience, nevertheless, Dickie's approach differs from Wollheim inasmuch as Dickie remains committed to definition as the means for identifying art whereas Wollheim, in what he believes to be the spirit of Wittgenstein's anti-definitionalism, eschews it. Thus, although Wollheim, Danto, and Dickie, each, in different ways, aspired to secure the identification of artworks in an expanded field -including that of the history and practices of the artworld -Danto and Dickie aspired to achieve this in terms of definitions of the art work, whereas Wollheim proposed doing it by means of a nondefinitional method.
As we have seen, that method involves presuming that we know certain historical works are artworks and then tracing prospective candidates for the status of artworks back to these primary artworks in virtue of certain rules. One modification to Wollheim's approach here that I would recommend is to abandon talk of rules in this regard and to speak instead of certain recurring scenarios. That is, the way in which we connect what Wollheim calls the derivative art works to the primary ones is by means of certain narratives that show how the later works emerged from the earlier ones.
A form of life, as I think Wollheim understood it, is a structure of possible moves. The game of chess stipulates the general ways in which certain chess pieces can move; then the disposition of pieces at a given point in any chess game further constrain the range of viable movement. Within the artworld form of life, the course of artistic development is both facilitated and constrained by the antecedent conditions of history.
One example that Wollheim mentions is what he calls supersession. He remarks "… in the art of our day one work of art generates another by supersession of its most generative or its overall properties, e.g., Pont-Aven as the successor of Impressionism, hard-edge painting as the successor of abstract expressionism." 15 Here, of course, it is strained to think of supersession as a generative rule. Rather, as I have already suggested, it is more profitable to construe it as a scenario, or narrative type or genre, or story form.
Expanding upon Wollheim's comments on supersession, the story, I think, goes like this: abstract expressionism exploited strategies of busyness and messiness for expressive purposes. Pollack's use of the drip is an example here. This, in turn, generated an almost predictable reaction formation in the next wave of artists who emphasized hard-edged patterns, such as the neatly linear, geometrical forms of certain minimalist painters who bracketed subjectivity in favor of an aura of objectivity.
The narrative of the reaction formation is a recurring genre with respect to modern art. It was mobilized once again to explain the emergence of post modern art which was said to be a reaction to the austerity of minimalism which was accused of banishing concern with non-painterly content, like politics.
If I interpret Wollheim correctly, he believes that we identify candidates as artworks -such as early hard-edge paintings -by means of art historical narratives, such as the story-form of supersession to which he alludes. That is, when a candidate seems likely to threaten prevailing expectations about art, a narrative, like that of supersession, is invoked in order to show the ways in which the candidate follows from or is generated by some antecedent artworld conditions -which, in the modern artworld, would appear to include the abiding pressure "to make it new." In this matter, I think that Wollheim can be defended on solid Wittgensteinian grounds. Namely, this is one way in which players in the artworld language game, conceived of in terms of a form of life, do actually go about establishing that a candidate is an artwork. Stories, like that of supersession, may be told after a candidate arrives on the scene and some skeptic challenges its status as art. But quite often the story is told, so to speak, proleptically in manifestos by the prospective artists or in critical pieces by advocates of the new work. One need only look to the leading art magazines, gallery catalogues and the like in order to confirm that this is a recurring move in the language game of the artworld form of life.
Of course, supersession is not the only available move in this language game. New artworks emerge from past work through a number of different processes and are, in consequence, identified by various different narrative strategies. At the most fundamental level of artistic transformation from earlier art to successive art, the process is one of repetition, albeit typically with studied variations. Art at time T2, for example, may repeat the basic ingredients of its predecessors (at time T1), but, in addition, may simplify them, exaggerate them, or elaborate upon them. For instance, the Corinthian column is basically a column, but one that elaborates on its earlier Doric and Ionic models. Likewise, one may embrace the structures of a certain form of art while adding or subtracting from them. George Blanchine maintained the figures of the classical ballet, but added unprecedented speed to their execution.
Another way in which later art plays off of earlier art involves the subsequent work from one art form appropriating données and developments from another artform as the ballet Giselle helped itself to motifs from German Romantic poetry and when Nijinsky imitated ancient Egyptian imagery to create the angularity of the postures in his Afternoon of the Faun. Repetition of this sort, which I call the interanimation of the arts, often occurs when the avant-garde of one artform apes parallel developments in the avant-garde of another artform as when minimalist choreographers of the early seventies, like Lucinda Childs and Laura Dean, with their repetitive, regular movement patterns, with slight accentual variations, echoed the minimalist music of composers like Steve Reich and Philip Glass.
Repetition in various forms is one of the ways that later art is generated by earlier art and a narrative of one of these types is a way in which one might go about establishing that some newer candidate belongs to the order of art. On the face of it, this may seem like a reversion to the family resemblance approach, but it is not, because the repetition scenario is not simply a matter of taking note of a resemblance between earlier and later works; it requires a causal connection -a line of influence or lineal descent -as well. Although when later art repeats, albeit with variations, ingredients from antecedent art, we rarely have cause to explicitly establish its art status -since the relation between the two is usually evident. Nevertheless, if called upon to justify the attribution of art status to certain works, one strategy for doing so is the narrative of repetition, as when one points to the use of Renaissance deep space in many of Dali's otherwise enigmatic paintings.
In addition to narratives of repetition, there are also those of amplification. Amplification involves a transition from one stage of art historical development to another where the later stage solves some problem that beset the earlier stage. Picasso's use of collage -his attaching objects to the surface of a painted panel -for example, can be represented as a solution of how to call attention to and acknowledge the perceived, dual existence of paintings as simultaneously two and three dimensional. Thus, challenged to establish the art status of Picasso's 1926 Guitar -a collage of canvas, wood, rope, nails, strings, newspaper, tacks and knitting needle on a painted surface -one would point out that this combination of real objects and a painted surface was a way of negotiating a problem that loomed large in the earlier avowed masterpieces of Cubist art, not only those of Picasso by also those of Braque, and, as well, in pre-Cubist painting, including classic works by Manet and Cézanne.
Hybridization is another artworld narrative that enables us to identify emerging candidates as artworks, as well as contested works of the past, in light of their predecessors. Early performance art, notably Happenings, involve the transposition of the concerns of painting to a three dimensional arena whereas installation art may mash up an array of art forms ranging from sculpture and architecture, to video, photography and the essay. In the case of hybridization, pre-existing art forms and media -often along with their associated preoccupations -are combined to make something new, although recognizably descended from previously acknowledged, going artistic concerns.
I have replaced Wollheim's notion of supersession with a more variegated list of art historical scenarios because I, as I suppose Wollheim would agree, believe that there are many more than one narrative type available for identifying art. One scenario that I have not discussed yet, but which seems to me the closest one to Wollheim's example of supersession is what I call repudiation. This is a recurrent gesture especially in the art of the twentieth century and now our own.
Repudiation involves emerging artists making work that rejects reigning art practices and their assumptions and biases in favor of a new artistic regime as when expressionists abjured realism and its commitment to objective verisimilitude. This is revolutionary art. However, it is important to remember that even the most revolutionary art does not sever its relation to the past entirely. At the same time that the German expressionists embraced distortion and repudiated realism, they simultaneously appealed to precedents among recognized masterpieces in the tradition such as Mathias Grunewald's Isenhim Altarpiece in which the figure of the crucified Christ is twisted unnaturally for expressive effect.
Practitioners of repudiation not only assault dominant artforms of the present, but, at the same time, strike alliances with selected art of the past in order to recuperate dimensions of that art, such as expressiveness, that have been allegedly neglected or even repressed in the art that the emerging artists aspire to replace. Thus the art historical narrative of repudiation not only illuminates what the new candidates reject, but also its art historical predecessors and inspirations as a way of establishing the unity and the continuity of the artworld form of life.
So far I have enumerated several art historical narratives -or genres -which can be deployed for the purpose of identifying/establishing candidates as artworks, including repetition (in several forms including the interanimation of the arts), amplification, hybridization, and repudiation. I do not claim that this exhausts the available identifying narratives. Future study will undoubtedly benefit from close attention to the language games that are actually in operation in the artworld form of life. Nor would I want to deny that these identifying narratives can be conjoined and maybe even combined on particular occasions in various different ways for the purpose of demonstrating the art status of emerging work or historically contested work.
As Gregory Currie notes "Narratives are artefactual representations which emphasize the causal and temporal connectedness of particular things, especially agents; they are exquisitely suited to the representation of motive and action." 16 The identifying narratives that I have introduced as a way of filling out Wollheim's Wittgensteinian conception of the ways in which artworks are identified in the form of life of the artworld are agent-centered. These narratives strive to represent the relation of the actions and intentions of emerging artists to their heritage in a way that shows them to be involved in an intelligible continuation of that tradition.
The means available for demonstrating that continuity include at least the narratives of repetition, including interanimation, amplification, hybridization, and repudiation.
The artist is a person who arrives at some juncture in the history of the artworld resolved to continue the practice by creating something which either re-enacts, with variations, the basic données of the practice or which re-directs it. The artist chooses her means for implementing her commitment on the basis of a plausible assessment of the state of the artworld and of the alternative, existing choices for realizing her aims. These choices, of course, run the gamut from elaborate repetitions to repudiations. Identifying narratives track these choices, and, where supported by the facts, establish these choices as artworks by representing them as the outcomes of art-generating actions.
Let me conclude this section of my talk by attempting to dispel three objections to this framework for identifying art which might have occurred to you. First, you may be worried that this account makes it too easy to establish any candidate as an artwork, since these narrative forms can be imposed willy-nilly on any string of events. However, I maintain that the identifying narratives that I have in mind are genuine historical narratives and, to that extent, they must fit the facts as they stand. Moreover, this guarantees that these narratives cannot be spun at will, since they will have to square with all of the salient facts.
Second, you may suspect that this approach leaves too much to the artist's conception of the situation. What if he has an off-the-wall take on the art historical situation and the alternatives therein? Here, I want to re-emphasize that we are talking about plausible artistic assessments and that that will protect us from absurd results.
Last, you might notice that this cannot be the only way in which we identify candidates as artworks, since there may be things, like prehistoric cave paintings, that are recognized as artworks but which are not susceptible to the narrative approach. That is true. That is why I began by emphasizing that this is one way we have for identifying artworks. There are others. However, it seems to me that the narrative approach is perhaps the most prominent way we have for establishing candidates to be artworks in our ongoing artworld form of life. Furthermore, I contend that this was what Wollheim was getting at with his appeal to Wittgenstein at the end of the first edition of his Art and Its Objects.

V. Wittgenstein and Appreciation
So far I have briefly addressed a few objections to the narrative framework for identifying art. But another objection -one I referred to in my opening remarks -charges that mining Wittgenstein for a solution to the problem of identifying art fails to make contact with what Wittgenstein had to say about art and, moreover, were attention paid to Wittgenstein's express concerns about art, the philosophy of art would be enriched beyond the aesthetician's obsessive and perhaps sterile preoccupation with the question "What is Art?" The parts of Wittgenstein at issue here are his comments on aesthetics in his Lectures on Aesthetics, Psychology and Religion and the essay "Wittgenstein's Lectures in 1930-1933" by G. E. Moore. 17 The relevant passages concern what might be called aesthetic appreciation or aesthetic judgment. Wittgenstein stresses that the notion of beauty is rarely pertinent when it comes to rendering aesthetic judgments. More often we are concerned with what is right (or wrong), appropriate (or inappropriate) with regard to a composition. Furthermore, when we disagree about the appropriateness of a composition, Wittgenstein proposes that we manage those disagreements by what I shall call demonstrative and comparative criticism.
According to Moore That is, Wittgenstein emphasizes the dialogical dimension of aesthetic disagreement. Aesthetics is or should be concerned with particular art works and the appreciation thereof and when we disagree about said works, the aesthetician should elucidate the ways in which we can conduct those disagreements in a persuasive manner. Two strategies that Wittgenstein recommends are gesturing to certain details in a work, or perhaps to the relations between various details in the work saying, perhaps, look here and then there --until your interlocutor grasps your point. This gesturing can be literal -a matter of ostension -or by means of words and descriptions, including metaphorical ones. This is demonstrative criticism. Its aim is to get your interlocutor to see things.
Comparative criticism involves getting others to see your point by comparing the work of art in questions with further works, either by the same artists or by other artists. Think of the old two-projector lectures in the art history classes of yesteryear. Works are juxtaposed in such a way as to make convergences and differences leap out. That too is a way to persuade another to see the work as you see it. It is a matter of showing.
Wittgenstein, it has been said, opened up this unexplored area of aesthetics upon which, unfortunately, philosophers of art did not follow up, preferring to remain mired in the question of "What is art?" Commentators like Kjell Johannessen, 19 suggest that it would have been more fruitful had philosophers of art built upon these unprecedented remarks by Wittgenstein rather than in using Wittgenstein in the way they did in pursuit of their enduring questions.
But I disagree with this assessment for several reasons. First, Wittgenstein's remarks about criticism were not unprecedented. Consider these observations by Clive Bell: A good critic may be able to make me see in a picture that had left me cold things that I had overlooked, till at last, receiving the aesthetic emotion, I recognize it as a work of art. To be continually pointing out those parts, the sum or rather the combination , of which unite to form significant form, is the function of criticism. But it is useless for a critic to tell me that something is a work of art; he must make me feel it for myself. This he can only do by by making me see; he must get at my emotions through my eyes. 20 Here Bell, like Wittgenstein a fellow acquaintance of Moore, is contending that critical disputes be settled by demonstrative criticism, such as pointings. If we disagree that a candidate lacks significant form -which would seem to be something like Wittgenstein's notion of appropriateness -we must show it to our interlocutor in a way in which he or she grasps it -sees it in a compelling manner -gripped viscerally by a feeling of aesthetic emotion. Moreover, it seems likely that Bell would be happy to include among the strategies of pointing description, including metaphorical description , and comparison, so long as they enable the critic to make us apprehend what is appropriate (and inappropriate) in a particular artwork. Thus, Wittgenstein's observations about aesthetic judgment are not unprecedented.
Furthermore, it is not the case that subsequent philosophers of art abandoned these concerns of Wittgenstein and Bell. For, these concerns fall under the label of metacriticism. Moreover, there are metacritics who would appear to be kindred spirits to Bell and Wittgenstein, such as Sibley and Isenberg. I do agree that the recognition of the importance of demonstrative criticism and comparative criticism is a major contribution to the philosophical understanding of our engagement with artworks. At the same time, however, I would argue that this is not all there is to criticism or to appreciation. For there is an extremely significant dimension of appreciation and criticism that these two practices do not encompass. What I have in mind is what we can call in general purpose-driven criticism of which thematic criticism is one of the most common sorts.
Purpose-driven criticism involves hypothesizing a purpose or set of purposes to a work so that the unity of the artwork becomes apparent as we realize that the parts hang together or work together in order to realize an overarching purpose. Often the purpose in question is the communication of a theme. Postulating a certain theme as the aim that a certain artwork is committed to promote is one of the most popular and effective way in which critics and appreciators alike come to understand why the characters, incidents, plot conflicts, descriptions, and so on in a work have been chosen and fit together. Themes colligate the parts of a work under a concept or a set of concepts. Unlike demonstrative criticism or comparative criticism, purpose-driven criticism, including theme-driven criticism, is much more a matter of theorizing (with a small t) than simply an exercise in seeing.
Thus, to the extent that Wittgenstein neglected the relevance of these sorts of criticism, his account of aesthetic appreciation and judgment, although laudable as far as it went, is and was incomplete. Moreover, this incompleteness, added to the fact that Wittgenstein's remarks were hardly unprecedented, suggest that subsequent philosophers of art may not have been completely unjustified in exploring other ways in which to apply Wittgenstein's thought to the philosophy of art. Notes 1. This paper was given as a talk at the conference on Wittgenstein and Aesthetics at the University of Southampton in the summer of 2010. The author has benefitted greatly from the comments from other conference participants, although they are not responsible for the defects in this paper.