In my last post I argued that Morris Weitz's famous essay on the role of theory in aesthetics points the way to a more positive approach to the issue of defining art. He offers us a way of reinterpreting the great theories of art so that they are seen as successes, i.e. as what he called honorific definitions of art. This was the main point of this truly seminal essay, and yet oddly, in the course of history, it has been ignored.
Joseph Margolis in "The Importance of Being Ernest about the Definition and Metaphysics of Art," Journal of Aesthetics and Art Criticism 68:3 (2010) returns to Weitz's essay but seems sometimes more interested in the question of whether Weitz misinterpreted Wittgenstein than in whether or not he was right about the role of theory in aesthetics. I do not even think we should judge essays of this sort on the basis of whether or not they interpret a great philosopher correctly: such works should not be judged as though they intended to be mere examples of "secondary literature." Important philosophers tend to misinterpret other important philosophers: creative mis-interpretation is part of what goes into thinking great thoughts. Margolis, ironically, is guilty of this too: a truly great thinker (certainly one of the finest mind in aesthetics) who often reads others insensitively.
Margolis begins his essay with the traditional negativism about definition: "The philosophy of art may be doomed, again and again but always once and for all, to define what it is to be 'a work of art'..." (215) In my last post I argued that the history of efforts to define art has been a history of successes if one interprets these efforts as presenting us with honorific definitions of art that participate in various multidisciplinary efforts at self-understanding understood in terms of their original place at particular times in history. My take on this is something like what Hegel referred to by saying that in a culture Spirit comes to self-understanding through dialectic, but without the transcendent metaphysics or any notion of progress towards an end-point (there is progress, but it is relative.) Oddly, in Margolis's effort to show how Weitz misinterprets Wittgenstein, he reconstructs, by accident, Weitz's own discovery, i.e. that definition of art in terms of necessary and sufficient conditions is not going to happen, but that supposed definitions of this sort are best seen as honorific definitions that focus our attention on one central property or type of property. I argued that this is done through definition as metaphor. Margolis observes that Wittgenstein is "plainly open to admitting a great many different kinds of definitions," is willing to impose limitations on open concepts such as "game" for some limited purposes, and that Aristotle's definition of tragedy, which is an attempt to capture the essence of tragedy, would not therefore be problematic for Wittgenstein. Again, I am not interested here in Wittgenstein interpretation: if Margolis is right, however, about Wittgenstein, so much the better, since the position is, I am convinced, right. More interesting is that Margolis misinterprets Weitz since he says "what Weitz might regard as failed essential definitions of the realist kind may, in Wittgenstein's tolerant sense, actually be successful in their own way..." (319) The point attributed to Wittgenstein was exactly Weitz's point when he brought up that these failed essential definitions of the realist kind should be seen as honorific definitions and therefore successful in their own way! Margolis thinks that Aristotle's definition is successful in this way, and so do I, and so, as I read Weitz, does Weitz. Well, as it turns out, once we get past all of this silliness about correct readings of Wittgenstein and Weitz, Margolis is deeply right about definition, and right when he goes on to say that Stephen Davies (and many other contemporary philosophers of art) makes a serious mistake when he insists that a definition of art must be in terms of necessary and sufficient conditions. Margolis says that Davies "agrees with Weitz about what a proper definition requires" (219) which is not quite right since, although Weitz believes that a proper "real" definition of art would be of this sort, a proper honorific definition of art does not have to meet this standard. As Margolis observes, Wittgenstein is tolerant of many different kinds of definitions. (219). Further, "a 'defective' definition of the essentialist sort...may, nevertheless, be a complete success in terms of a reasonable reading of its philosophical contribution." (219) Weitz (and I) would agree with this point, although I would say that there is another, more valuable, kind of essentialism, in which essences are quasi-fictional and changing objects that can actually be captured quite well by honorific definitions. Aristotle's definition of tragedy is one of these! So, where Margolis speaks of the skewed influence of Weitz's misinterpretation of Wittgenstein, I would speak of the skewed influence of misreadings of Weitz himself. Margolis' take on this is that "real" definitions need not be "essentialist" i.e. in terms of necessary and sufficient conditions. I would make almost the same point using different language, namely that essentialist definitions of art need not be "real" definitions, i.e. definitions in terms of necessary and sufficient conditions.
In the end, I am not opposed to Margolis' idea that a definition is "realist" if it addresses the "nature" of a thing, although I find it odd when he says that a "real" definition (in his sense of "real," not in Weitz's) "needs only to be 'usable' in the way ordinary usage tolerates." (220) Well, that would make it "real," I suppose, but not particularly good or true, even in a pragmatist sense of "true." A good definition must not only be usable and tolerable to ordinary usage (actually, that might be a serious disadvantage!): it should be true in the sense of powerful and fruitful in the realm of reasoned discourse. Aristotle's definition of tragedy, again, fits the bill (sometimes more successfully, sometimes less so at different times in history), but so too does Clive Bell's formalist definition of art and Tolstoy's expressionist definition of art, and so also Collingwood's Crocean "intuitionist" version of expressionism.
Margolis is right on the mark in my view when he says that Berys Gaut's cluster account view of defining art does not show how it is "genuinely serviceable as a replacement for a definition." (220), although one could say, to be fair to Gaut, that cluster accounts could serve as one kind of definition among the many kinds that Margolis, through Wittgenstein, is otherwise allowing. Margolis also thinks that the cluster account really only captures the kind of usage of terms we find in the teaching of those terms to children.
Despite my criticisms I think this essay to be one of the most insightful (could we expect less from Margolis?) in recent aesthetics. Margolis is dead right when he says "I see no reason why one must choose, disjunctively, between Aristotle and Nietzsche...": the rest of the paragraph is wonderful, but I have no time to type it out in full.