What Ever Happened to Anti-Essentialism ?

How should one explain the relative disappearance of a major preoccupation of English-speaking Analytical Philosophy in the late 1950s/early 1960s: an anti-essentialist response to the question, ‘What is art?’, typified in papers by Kennick and Weitz? Minimally, anti-essentialism denies the widely-held assumption that something must be in common between all the instances where (in our case) the term “[fine] art” or the concept art is rightly ascribed, in virtue of which all are called ‘art’; a stronger version urges that, in fact, there is no essence to (our

BJA of what had been a major preoccupation in English-speaking Analytical Philosophy in the late 1950s and early 1960s: namely, an anti-essentialist response to the question, 'What is art?', with papers by Kennick 1 and Weitz 2 perhaps the most famous. 3ow did the situation Carroll accurately described come about?What does it mean for contemporary philosophical aesthetics?The disappearance from the mainstream of such apparently important sets of arguments rarely permits clear and uncontentious explanation.Certainly that disappearance seemed not to reflect the prevalence of major counter-arguments.But, with some unpicking, Carroll's conjectures might offer some consensus; I follow him in taking as my examples these papers by Kennick and Weitz.
Carroll calls the views under discussion 'neo-Wittgensteinian', explicitly claiming that both Weitz and Kennick attempted '. . . to demonstrate that a definition of art (in terms of necessary and sufficient conditions) was impossible', implicating Wittgenstein's discussion of 'games', 4 its resolution deploying (in some way) the notion of 'family resemblance'. 5This 'neo-Wittgensteinian' title has some justice: Weitz explicitly quotes Philosophical Investigations and uses the expression 'family resemblance'; 6 Kennick merely mentions Wittgenstein, recognising his own deployment of that expression as '. . . to torture a phrase of Wittgenstein's'. 7However, both Kennick 8 explicitly and Weitz by allusion employ the idea of open texture, derived from Waismann, and treated as somehow equivalent to Wittgenstein's thoughts on family resemblanceeven though open texture was rightly acknowledged as Kneale's translation of Waismann's coining; 9 Waismann's debt was to some prior writing by Wittgenstein. 10arroll remarks that the 'neo-Wittgensteinian belief . . .that art cannot be defined . . .[was] soon contested', initially citing two such forces of contestation. 11The first comes from Arthur Danto's 'The Artworld' 12 as Carroll acknowledges, 13 in that work Danto appeared to argue '. . .that artworks have at least one necessary condition -that they be enfranchised by art theories'; since then, Danto has defended two or, arguably, three. 14As I remember the situation only a few years later, Danto's explicit essentialism was not stressed: his primary insight was taken to be his emphasis on 'confusable counter-part' objects (read, at the time, as anti-essentialist), undermining the commitment to many characteristics typically thought shared by artworks, by showing how non-art 'real things' might also share those characteristics. 15Such ideas point away from characteristics or properties of art of the kinds discussed by Weitz and Kennick, characteristics central to anti-essentialist critiques.
Second, more specifically, Mandelbaum urged that the charges laid by anti-essentialist writings, and especially their (supposed) deployment of 'family resemblance', could be met. 16In Mandelbaum's view, '. . .family resemblances, properly so-called, are governed by conditions': 17 indeed, that '[i]t is a necessary condition of authentic family resemblances that they require a genetic foundation'. 18Yet then, taking '. . . the family resemblance analogy seriously . . .would incur a reversion to the project of defining art'. 19And Mandelbaum's arguments were central to the justification offered for the attempted definition(s) of 'art' by George Dickie, 20 in taking Mandelbaum's idea of 'unexhibited properties' to offer '. . . the possibility that the relevant features for defining art might not be manifest at all, but rather underlying, non-manifest properties'. 21Even those actively contesting Dickie's putative definition(s) appear to have adopted this aspect of his project, seemingly showing the anti-essentialist considerations to lack power.Typical responses assumed this outcome, rather than rehearsing it -actual arguments were not considered.
While neither conjecture alone explains the relative disappearance of the anti-essentialist arguments, taken together they suggest how these argumentative strategies might be set aside, largely without due consideration.As Carroll comments, the net effect has been that '[s]ince the early nineties (and perhaps earlier), interest in the question 'What is art?' has slackened somewhat'. 22But the anti-essentialist insight was fundamental in philosophy (especially in aesthetics), so this 'slackening' was mistaken; and recognising this sheds light on others implicated in Carroll's account.
Here, I begin by sketching the version of essentialism under consideration, and outlining the views of first Kennick ( §III) and then Weitz ( §IV), before commenting both on their arguments and on their (differential) dependence on Waismann and, especially, Wittgenstein ( §V).Then I suggest why the setting-aside of their concerns is neither explained by the arguments standardly used nor warranted in its own right ( §VI).

II
Since these writers from the 1950s are rightly described as 'anti-essentialist', exactly what conclusions count as 'anti-essentialist'? Minimally, an antiessentialist denies the widely-held assumption that there must be something in common between all the instances where one rightly ascribes (in our case) the term '[fine] art' or the concept art, something in virtue of which one calls them all 'art'.Anti-essentialists oppose this assumption of essentialism, denying that '. . .there must be something in common to them all, or we should not call them by the same name'. 23Here one might with profit distinguish two versions of the anti-essentialist claim: 24 one denies that it is necessary for artworks to have something in common (besides being called 'artworks'); the other denies that it is necessary for artworks to have some essence in common, of the sort that might be captured by a definition: that is, by non-circular necessary and sufficient conditions.But, for Kennick at least, these clearly amount to the same thing.Indeed, Kennick quotes, as a stalking horse, just such an assumption from De Witt Parker: . . .there is some mark or set of marks which, if it applies to any work of art, applies to all works of art, and to nothing else -a common denominator . . .which constitutes the definition of art, and serves to separate . . . the field of art from other fields of human culture. 25e 'common denominator' here is just what is meant by the term 'essence', and such essences are identified with a set of conditions individually necessary and jointly sufficient.And rightly, for, after all, they were supposed (by, for example, De Witt Parker) to provide 'the definition of art'.Then, in a sub-heading, Kennick urges that one should 'question' the tempting assumption of such an essence or common denominator (the kind a genuine definition might identify). 26Further, granting that no such essence is needed may explain why this minimal anti-essentialism might rightly be confused with the view, above, that artworks need have in common nothing but being called 'artworks'.For our anti-essentialists are contesting the account of concept-use whereby a justification for the use of a particular term (here, 'art') is required -in effect, they are responding to the well-known view of Bell (quoted below) that to use that term with no such common denominator is to 'gibber'! 27But, they urge, no such 'common denominator' or essence is needed.
Then, a stronger version of anti-essentialism argues that, as a matter of fact, there is no essence to (in our case) art: both Kennick and Weitz endorse this version of anti-essentialism explicitly, perhaps in addition to the other.Thus Kennick (1965: 21)   31 in explaining an open concept as one whose '. . .conditions of application are always emendable and corrigible'; 32 and identifying '. . .a logically vain attempt . . . to conceive the concept of art as closed when its very use reveals and demands its openness'. 33xplained that way, concepts with open texture (or open concepts) lack an essence in the relevant sense; as such, genuine definitions of them will be impossible.Hence acknowledging the lack of a definition is not a step additional to identifying such open-ness.Then, further, Weitz stresses that, since art's '. . .ever-present changes and novel creations . . .makes it logically impossible to ensure any set of defining properties', the concept 'art' is especially open. 34or Kennick, the anti-essentialist project is characterised as demonstrating: . . .that the search for essences in aesthetics is a mistake, arising from a failure to appreciate the complex but not mysterious logic of such words and phrases as 'art', 'beauty', 'the aesthetic experience', and so on. 35t Kennick lays out fully the remainder of a project shared with Weitz, by recognising: If the search for the common denominator of all works of art is abandoned, abandoned with it must be the attempt to derive the criteria of critical appreciation and appraisal from the nature of art. . . .Traditional aesthetics mistakenly supposes that responsible criticism is impossible without a set of rules, canons, or standards applicable to all works of art. 36 is a short step to see what, say, Kennick finds valuable here for, if art has no essence or nature, appeal to such an essence cannot ground artcriticism.He identifies the assumption under criticism as the (mistaken) idea that 'responsible criticism is impossible without standards or criteria universally applicable to all works of art': 37 instead, '[c]riticism has in no way been hampered by the absence of generally applicable canons and norms . . .'. 38 Then Bell's claim that 'art is Significant Form' (for instance) is really an attempt to guide appreciation, functioning positively to change taste at the time.For Bell's discovery was '. . .a new way of looking at pictures', directing attention to features of Cézanne as against features of the descriptive painting of the time, typified by Frith's Paddington Station. 39Like Kennick, Weitz aims '. . . to reassess . . .[theorizing's] role and its contribution in order to show that it is of the greatest importance to our understanding of the arts'. 40For both, then, are concerned to defend the possibility of art-criticism, despite the anti-essentialist conclusion.Of course, '[a]ny critical judgement, to be justified, must be supported by reasons . . .'. 41 But, since this does not require ' . . .general rules, standards, canons, or laws, applicable to all works of art by which alone such [art-critical] criteria can be supported', future aesthetic theories for the arts need no longer regarded as definitions; and, hence, not as failed definitions. 42However, our concerns in anti-essentialism mean that such issues can be set aside here.

III
Why might one think that art lacks an essence?What considerations are set aside here?Both Kennick and Weitz begin from the failure of aesthetics to date to find any such essence for art: '[i]f nothing else does, the history of aesthetics itself should give one enormous pause here . . .[since] . . .we seem no nearer our goal today than we were in Plato's time'; 43 indeed 'no such definition has been found . . .' despite '. . .all the fruitless scrutinising that has already been done' in the search for an essence or definition. 44n example (at least for visual art) is provided by Bell's essentialist assertion that '. . .either all works of visual art have some common quality, or when we speak of 'works of art' we gibber'; and his conclusion that "'Significant Form" is the one quality common to all works of visual art'. 45With no independent grip on that notion, what possible help is Bell's conclusion?
To show that nothing of the form assumed (or on offer) could be helpful, Kennick offers what is, in fact, the centre of his argument here -his famous 'warehouse example': Imagine a very large warehouse filled with all sorts of thingspictures of very description, musical scores for symphonies and dances and hymns, boats, houses, churches and temples, vases, books of poetry and prose, furniture and clothing, newspapers, postage stamps, flowers, trees, stones and musical instruments.Now we instruct someone to enter the warehouse and bring out all the works of art it contains.He will be able to do this with reasonable success, despite the fact that . . . he possesses no satisfactory definition of Art in terms of some common denominator, because no such definition has yet been found. 46at could this example show?First, an account of art like (say) Bell's cannot really resolve the practical plight of the person required to enter the warehouse.Of course, such a person would not be 100% successful in bringing out all and only artworks: probably some artworks would be left behind, and some decorative objects that are not art brought out, while any 'confusable counterpart' cases doubtless remain puzzling.Still, the instruction seems intelligible: '. . .we are able to separate those objects which are works of art from those which are not because we know English . . .[even though] . . .no such definition [of "art"] has yet been found'. 47ut, second, the best conclusion from that argument is that any putative definition (exemplified by Bell on Significant Form) is no better than simply relying on one's pre-theoretical intuitions about art.Thus Kennick stresses the comparative uselessness of any putative definition, expressly borrowing, and redrafting, a sentence from Waismann: If anyone is able to use the word 'art' or the phrase 'work of art' correctly, in all sorts of contexts and on the right sorts of occasions, he knows 'what art is', and no formula in the world can make him wiser. 48person with a typical (passing?)experience of art '. . .knows a work of art when he sees one, but has little or no idea what to look for when he is told to bring an object that possesses Significant Form' (Kennick, 1965: 6).So our person, typically puzzled, would fare no better asked to bring out, say, all the objects with Significant Form.Kennick makes a slightly stronger claim, perhaps overstating the case: . . . the notion of Significant Form is clearly more obscure than is that of Art or Beauty, as the example of the warehouse . . .amply illustrates; the same holds for Expression, Intuition, Representation, and all the other favoured candidates of the aestheticians. 49, if the instruction to bring objects with Significant Form from the warehouse is no better than that to bring out the artworks (especially if many works of recent art are included: urinals, snowshovels), must it be worse?At best, the example indicates only that these other concepts (with Bell's 'Significant Form' the primary example) do not provide more grasp on the identification of artworks than our understanding of art.But perhaps a definition of art is not needed, since we can manage as well without these putative definitions, even though we have no such definition (a fact supported by the continuing search for such a definition).That conclusion accords with our earlier weaker anti-essentialism: the thesis is not (or not necessarily) that there is no such essence, but rather that one has no justification for assuming there must be one.Moreover, from the perspective of identifying artworks, a definition in terms of hidden characteristics, characteristics we did not know, fares no better than no definition at all.
Yet can one infer, with Kennick, that the person '. . .possesses no satisfactory definition of Art' from the fact that 'no such definition has been found'? 50Focusing on the practical role played by such a definition in, in this case, identifying artworks, the inference seems warranted -he can hardly have a definition aesthetics lacks!But we must be careful here with what people do or do not know.As we will see ( §VI), the sort of 'enfranchisement by art theories' that, for Carroll, appeared as '. . .at least one necessary condition' for artworks in Danto (1964/1989) might be accommodated in some less explicit way. 51Yet, thus far, much in Kennick is at least consistent with a defence of the weaker anti-essentialism identified above, where the issue is a wholly general one -against the assumption that there must be such essences.
However, when Kennick urges that '[t]he trouble lies not in the works of art themselves but in the concept of Art', does he really think the concept of art is special in this way? 52Or is he merely reflecting the context of his discussion?For Waismann, open-ness is warranted by '. . . the essential incompleteness of an empirical description', such that '. . . it is always possible to extend the description by adding some detail or other'. 53For '[e]very description stretches . . .into a horizon of open possibilities'. 54Then no genuine definition will be possible for any concept with open texture, since '. . .no definition of an empirical term will cover all possibilities'. 55And Waismann specifically mentions, as 'closed' in this sense, concepts from mathematics where one '. . .could construct a complete definition, i.e. a thought model which anticipates and settles once for all every possible question of usage'. 56f some concepts are open, others not, Kennick could align art with some other The reference to Wittgenstein is considered below (see §V).What should be made of the others?First, for Weitz: . . .aesthetic theory is a logically vain attempt to define that which cannot be defined, to state the necessary and sufficient properties of that which has no necessary and sufficient properties, to conceive the concept of art as closed when its very use reveals and demands its openness. 59e essentialist assumption requires that what is common to whatever (say, art) uniquely identify it: it should provide '. . .a common denominator, so to say, which constitutes the definition of art, and serves to separate . . . the field of art from other fields of human culture'. 60So, to count as a definition, conditions both necessary and sufficient are required.Thus necessary conditions for art not jointly yielding a sufficient condition, for example those urged by Danto (below), would fail to identify an essence for, say, art: they could not constitute a (genuine) definition.
Here, for Weitz, the open-ness of the concepts explains the lack of definability, such that: [a] concept is open if its conditions of application are emendable and corrigible. . .I can list some cases and some conditions under which I can correctly apply the concept of art but I cannot list all of them, for the important reason that unforeseeable or novel conditions are always forthcoming or envisageable. 61is parallels Waismann's claims, quoted above, that '[v] . . . the very expansive, adventurous character of art, its ever-present changes and novel creations . . .makes it logically impossible to ensure any set of defining properties. 64t (at best) this seems to be about our ensuring, rather than there being, such properties: a definition unknown to the practice-community (for art, say) would not alter this point.
The second reason involves Weitz's use of Wittgenstein's discussion of games (at Philosophical Investigations § §66-67), which becomes integrated into Weitz's general position because, when asked how the concept art resembles the 'concept' game, Weitz replies that '. . . the basic resemblance between these concepts is their open texture'. 65Then: . . .'[i]f one asks what a game is, we pick out sample games and describe these, and add 'this and similar things are called 'games'.'This is all we need to say and indeed all any of us knows about games. 66

Weitz concludes:
Knowing what a game is (is) not knowing some real definition or theory but being able to recognize and explain games and to decide which among imaginary and new examples would or would not be ['rightly'?]called games. 67Does this follow from anything Weitz said, or quoted from Wittgenstein?It would follow that no definition was required in order that one's claims be true/false of their intended object, a claim that Kennick extracted from his warehouse example -that definitions would not be useful to us in these situations.But, unlike the appeal to (Waismann-ian) open-ness, this position is consistent with there being a definition unknown to us -the relevant concept could still be deployed with confidence and accuracy in a wide variety of circumstances; although, again like the warehouse example, with no suggestion of always being right.But, insofar as this is a general comment about concept-mastery, it just reflects the more general difficulty (minimal anti-essentialism, as explained above).By contrast, the discussion of open texture is more specific in claiming open-ness for certain concepts (including art) while denying it to others.
Both Kennick and Weitz refer to the past failures of aesthetic theorists to find a factor common to all cases of art, mentioning the classic objections to definitions -counter-cases that show putative definitions being either too tight or too loose and/or the circularities indicative of emptiness.But of what exactly is that evidence?Here, both Kennick and Weitz stress our capacity to use concepts (especially art) without having a definition: such definitions form no part of our practice with such concepts.But that alone cannot show that there are no such definitions, or common factors -and that is one way to put Mandelbaum's 'insight": namely, that family resemblances are governed by conditions.
Notice, first, that although Lopes calls the remarks of both Mandelbaum and Danto '. . .responses to anti-essentialism' (meaning explicitly Kennick and  Weitz), neither is responding to our anti-essentialists: properly understood, neither set of arguments will undermine the anti-essentialist contention. 68anto seems simply to attempt to change the topic, while importing the essentialist assumption, while Mandelbaum's worries about 'family-resemblanceconcepts' reflects directly nothing in either paper -although part of the blame here must fall on those anti-essentialists.Thus, 'Mandelbaum's criticisms of the family resemblance model . . .disclosed a disturbing oversight in the rhetoric of Weitz and Kennick' 69 only if they were deploying these conceptsdefinitely not true of Kennick (remember '. . . to torture a phrase of Wittgenstein's'), 70 and arguably not of Weitz.But, second, the real force of Mandelbaum's remarks must await brief consideration of Wittgenstein's actual arguments, to which we now turn.

V
These anti-essentialist views are regularly identified as 'neo-Wittgensteinian'; 71 but, at the time, Wittgenstein's ideas were comparatively newly published.Since then, the comparative disappearance of Wittgenstein's ideas from analytic philosophy in general, and philosophical aesthetics in particular, has meant that arguments (or, at least, examples) from his work are not given due attention.For example, the notion of 'family resemblance' is widely misunderstood, in aesthetics as outside it.Confusing presentations by our typical anti-essentialists, together with misunderstandings (in some cases, or some places) of what those argumentative strategies could demonstrate, have not helped here.For, to put it bluntly, Waismann was not Wittgenstein.
The relevant sections of Wittgenstein's Philosophical Investigations are regularly treated as if independent of his general discussion; and widely misrepresented, not least as elaborating a newly-discovered logical tool -familyresemblance-concepts. 72But in these sections, and those following (as well as elsewhere), Wittgenstein aims to expose as unjustified just that assumption that there must be something in common when the same term (say, 'game') is used!For, recognising the '. . .craving for a definition' 73 as a 'craving for generality', 74 Wittgenstein is keen to limit the pretensions of philosophy here -philosophy's standards cannot be higher than those used normally: 'In philosophy . . .we cannot achieve any greater generality than we express in life and in science'. 75Hence Kennick and Weitz are entitled to the conclusion that, for concepts such as 'art', the absence of a definition from our explanatory practices really does permit the setting aside of the need for a definition (although, naturally, not to conclude that one is impossible).
In PI §66 Wittgenstein famously affirms that games have no common property; generally he is more circumspect, since his target is only that there need be none.Once this is granted, the ability to give a definition (in terms of conditions individually necessary and jointly sufficient) cannot be necessary for understanding -at least if we do understand games.As Wittgenstein urged elsewhere, when confronting a counter-case, and noting the absence in this case of what we naïvely took for a property common to all games, a commitment to essentialist assumptions may lead us to postulate one -but one that is difficult to grasp, and awaiting discovery. 76That must be a false step, if 'Look and see' is really the right advice -for one can only 'look and see' what is in plain sight.Here, Wittgenstein requires a property on account of which the term 'game' is rightly applied to all his diffuse examples.In reality, his dogmatic insistence that there is nothing in common among, say, games is best seen as the conclusion that, even were some common feature discovered, it need not define 'game' (it might, for instance, play no role in our explanation of the word 'game').Certainly, we are not awaiting such a discovery.
Further, an example from Wittgenstein makes one element of our practice clear -that differences among games are recognised contextually: 77 hence, '[s]omeone says to me: "Shew the children a game".I teach them gaming with dice, and the other says "I didn't mean that sort of game".'An exhaustive account, defining the term 'game', is not needed to make good sense of this.On the other hand, it asks what is involved in, say, 'not meaning that sort of game': when did one mean or not mean whatever?Wittgenstein ponders: What if one were to ask: 'When are you able to play chess?Always?While you are playing it?And during each move?' -How strange that being able [knowing how] to play chess should take such a short time, and a game of chess so much longer! 78 this way, Wittgenstein foregrounds our practices of explaining: what has no role in such practices cannot be drawn upon to offer an account of whatever it is.
Moreover, as noted above, Wittgenstein's target in these sections is the 'must' in the assumption that there must be something in common when the same term (say, 'game') is used: why must there be something in common?The person asking the question seems unable to think of any other way to explain using the same word.So Wittgenstein offers an alternative model; and, since the essentialist had talked in terms of similarities, so does Wittgenstein's alternative proposal.He imagines a rope made of overlapping threads: Thread A overlaps with thread B, thread B overlaps with thread C, thread C with D, and so on: threads A and D are part of the same rope but they do not overlap with one another.So here is an alternative version of 'unity-indifference', one without essences.If some 'similarity-based' explanation must be offered here, why not this one?And, of course, faced with this alternative, one cannot insist that it must be this way or that: it could (at least) be either of these ways.
Since such similarities might be thought to form a family, Wittgenstein offers the term 'family resemblance' to characterise this situation.Some writers have imagined that this constitutes a thesis of Wittgenstein's: the discovery of family-resemblance concepts.Such a suggestion does not fit with his use of the idea.
Those believing in the importance of 'family-resemblance concepts' would expect to find that, having discovered or invented family resemblance, this notion would loom large in the rest of Wittgenstein's text.Instead, the term 'family' occurs a couple of other times, but only in the ordinary sense: this idea of 'family resemblance' does not recur in Philosophical Investigations.Why?Because it was only an argumentative strategy against the essentialist who said, 'there must be something in common'; and Wittgenstein in effect replied, 'If you think of it this way, you will see that the 'must' is unjustified'.
First, Wittgenstein has only offered an alternative.Except in his more dogmatic moments, he has not (yet) even suggested that the essentialist is wrong in his conclusion about there being something in common in any actual case, but only shown that the essentialist argument cannot generate the conclusion that there must be such a common factor.Hence, as Wittgenstein says, one needs to 'look and see' in any particular case.And so, in effect, here there is a kind of 'battle of the metaphors' for being securely grounded: • Frege (writing about the special case of mathematical concepts) urged that a concept was like a field; if it was not bounded on all sides, it was not a field at all. 79• Wittgenstein's picture of the overlapping threads above: like the essentialist, this offers shared elements, but without guaranteeing that they will be shared by all cases -the model shows how this assumption might be avoided. 80 John Wisdom offered the metaphor of a horse: someone might urge 'there must be something that a horse stands on, that supports it, otherwise it would fall over', but Wisdom recognised that a blacksmith lifts each leg of the horse when shoeing it -without that proving it is not supported by those legs! 81 And there are many more! 82econd, this emphasis on 'look and see', once adopted, also impacts generally: one can only 'look and see' if what is required is in plain sight.But Wittgenstein recognised that as a requirement on the usefulness of the supposed common element.And this provides his justification for setting aside (any) definitions unknown to the speaker.Thus, suppose that the putative definition of 'sport' from Bernard Suits indeed offered conditions individually necessary and jointly sufficient, thereby having an exact fit on the concept 'sport': what progress is made? 83Wittgenstein's thought is that someone who learned such a definition on Thursday would still have understood sport on Wednesday: that, in typical cases, our understanding concepts did not (and could not) await the uncovering of such hidden connections -an implied criticism both of the project of definition and of its results.
Then, third, notice how Wittgenstein's account of family resemblances turns on such 'looking and seeing': 84 Lord Peter Wimsey is proud of having inherited the slender, elegant 'Wimsey hands', visible in family portraits back through the ages. 85Yet his brother and sister have not.Of course, there will be a genetic explanation of all these facts -but being in the (biological) family does not guarantee one the 'Wimsey hands'; and they are what might be noticed here, as Harriett Vane recognises them in Lord St. George, Peter Wimsey's nephew.However Wittgenstein's view does not lend itself to standard forms of exposition: what does one write to make plain that certain assumptions are unjustified, while having no intention of offering alternative assumptionsor, more exactly, offering alternatives only to undermine the thought that the initial assumptions must be adopted, that no others are possible?Further, despite rejecting the '. . .misguided craving for exactness' that Wittgenstein came to find in his own earlier views, his later philosophy does involve some generality. 86Very often, questions asked by philosophers '. . .can be answered only by describing quite particular uses of signs, hence only from case to case'. 87For, at best, one can understand '. . .how concepts would apply in, or to, particular circumstances for which they were designed', 88 where a chain of examples as 'objects of comparison' 89 is offered in a person-specific and problem-specific manner, with a therapeutic intent, recognising what Charles Travis has called 'occasion sensitivity'. 90VI So, properly understood, Wittgenstein's own arguments do undermine the assumption that there must be something in common to all instances of using the same word (say, 'art') simply by showing that, in any case, other models might be deployed.More specifically, the considerations raised by Mandelbaum can be set aside: Wittgenstein never aimed to introduce 'family-resemblance-concepts' into the philosophical lexicon.Hence it is no criticism of him that such a notion, if introduced, would be problematic.Further, the injunction to 'Look and see!' is relevant here simply because the essentialist requires, not just that there be an essence, but that one calls it art because of the essence -hence that one's use of the term 'art' draw, if implicitly, on this essence.An unknown (or not yet known) essence will not meet that need.
Our classic anti-essentialism in Kennick and Weitz sometimes requires only the kind of weaker anti-essentialism mentioned above, the kind really deserving the name.Yet sometimes both seem convinced by Waismann's claims for the 'open-ness' of concepts that 'art' cannot be defined: by definition, such concepts are not amenable to definition.Further, Weitz suggests a special open-ness for the concept 'art', '. . .since new cases can always be envisaged or created by artists . . .'. 91 Of course, when they were writing, Wittgenstein was not well-understood.Yet the central anti-essentialist thrust remains in Kennick and Weitz; and most factors mentioned by Carroll need not have been daunting.Further, the anti-essentialist argument requires that the absence of an essence should not preclude critical reasoning -both Kennick and Weitz meet this challenge in ways not undermined by the criticisms of their papers noted above.
The case of Danto remains to be considered.Strictly speaking, the suggestion that, to meet our requirements, any essence be known, or explicit enough to explain our practices, counts against both Danto's initial account ('the artistic enfranchisement of these objects') and his more recent version ('embodied meaning . . .[and] wakeful dreams'. 92These conditions, while suggestive of the nature of art, are not candidate elements for definitions.The pair of necessary conditions that become 'embodied meaning' is spelled out so that '[t]o be a work of art is (i) to be about something and (ii) to embody its meaning'. 93At best, they were only necessary, with the suspicion of circularity attaching to the notions both of meaning and of embodiment: can they really be elaborated without becoming, respectively, art-type meaning and art-type embodiment, especially when their introduction deploys cases from artworks?Such a suspicion of circularity will attach even more strongly to the idea of 'wakeful dreams', especially once Danto insists that '[t]he perception [of art] is shared in a way dreams never are shared . . .'; so in what sense exactly are they dreams? 94We may feel unsure whether a non-circular explanation could be given, since none is provided.Clearly, the expression 'wakeful dreams' is evocative; Danto did not live to put much flesh on its bones, except that aiming at a '. . .condition that captures the skill of the artist' -not, of course, the skill in manufacture that might be shared with the manufacturer of a 'confusable counterpart', but rather as elaborating '. . .all the different ways artists have found to dream-ify' while granting that this '. . . is not possible to catalogue'. 95I doubt such a condition could be made specific enough to hunt for candidate counter-cases; and, if not, it cannot really be part of any account sufficient for art-status.
What of the three-seater?Although '[o]bjects similarly to it were to become accepted [as art] in the next generation of artists', de Kooning did not develop in this direction; nor would such a development make obvious sense, given how de Kooning's development is presently regarded. 103Here Danto rightly quotes Wölfflin's slogan: 104 'Not everything is possible at every time', concluding (a) that '. . .this particular object can be a work of art only if it is a de Kooning'; 105 but (b) '. . .there is no way it can be that'. 106Hence it is not an artwork, despite the 'atmosphere of theory' that seems to surround it.So agreeing with Danto about this example highlights a limitation in his claims about conditions necessary for art-status.
Danto rightly recognises that, with the history of art as currently written, '. . .there is no space in his [de Kooning's] corpus for an object of 1954 like this'. 107Hence substantial re-writing of de Kooning's corpus, and the trajectory of his art, would be required in order that de Kooning were, after all, rightly seen as a precursor of (say) Jasper Johns.While not impossible, for us (as for Danto) it seems improbable.Yet the history of art of a particular kind (or, picking up Danto's point, which brings such a kind 'within the fold' -as Duchamp did for Ready-mades) is not written once-and-for-all.After all, the sorts of huge changes in the future of art required to bring it about that, in 1954, de Kooning might create something precursive of Jasper Johns (or whomsoever) may not be unimaginable -although we think them unlikely.But granting this fails to regard these conditions as uncontentiously definitive of art.
We learn from genuine anti-essentialists (such as Wittgenstein, and perhaps Kennick) that there is no reason to claim that there must be an essence for art (or set of conditions individually necessary and jointly sufficient for deploying the term 'art'), and that definitions of (say) art are not necessary to our art-theoretic practice: the real outcome of Danto's institutional tendencies can be similarly regarded without his essentialist gloss.For the connection of artworks to art-theory (or art-history more generally conceived) is an insight.Can it offer genuinely necessary conditions -let alone a sufficient condition -for arthood (as opposed to, say, ones merely defeasibly necessary)? 108Our discussion of anti-essentialism shows why that question deserves a negative answer.gmcfee@graham-mcfee.co.uk

NOTES
concepts, but not all, in drawing attention to the open-ness of the concept art.Yet why should such 'open-ness' apply to art?And what exactly justifies his confidence in '. . . the essential incompleteness of an empirical description'? 57IV If Kennick's view is as sketched above, how should we characterise that of Weitz?More emphatically than Kennick, he claims that art '. . .has no set of necessary and sufficient properties'; and, although speaking chiefly of open concepts, his allegiance to Waismann and to open texture is explicit. 58He offers no specific reasons of his own for doubting the possibility of an essence: rather, having presented a version of Wittgenstein's account, Weitz raises the open status of the concept 'art', as though to explain what is really shown by Wittgenstein.However, lacking even citations to Waismann, he cannot assume without argument Waismann's confidence in '. . . the essential incompleteness of an empirical description'.Finally, Weitz seems to offer an especially art-specific reason in addition to those noted above.

which all the definitions of the aestheticians can do nothing at all to re- move', thereby granting to of such putative definitions an '. . . essential incom- pleteness'; 30 while Weitz echoes the claim by Waismann that
'. . .definitions of open concepts are always corrigible and emendable' agueness can be remedied by giving more accurate rules, open texture cannot . . .definitions of open terms are always corrigible and emendable'. 62So open concepts (or concepts with open texture) are precisely those lacking conditions individually necessary and jointly sufficient.For Weitz at least, granting that 'art' is an open concept, or a concept with open texture, contrasts it with closed concepts in mathematics. 63But why should one accept that art is an open concept?In effect, Weitz offers two sorts of reasons: one, specific to art, amounts to a positive reason towards another view: