Dickie's Institutional Theory of Art — Epoche B2
Whose Artworld? Dickie's Institutional Theory and the Problem of the First Artwork What this review argues George Dickie's institutional theory is popularly summarised as the claim that art is whatever the artworld says it is [1] . Defenders reply that this misreads a classificatory definition as an evaluative one, and they are right; that part of the defence should be conceded at once. My claim is that the defence is usually allowed to stop there, and that stopping there conceals a harder problem. The theory cannot say where an artworld comes from, and therefore cannot say what made the first artwork art, or what makes art in societies whose practices do not include an institution of the kind Dickie describes [2] . His answer, given in The Art Circle (1984 [3] ) and not in the early papers, is that the definitions of art, artist, public and artworld are frankly circular and are to be accepted as such. That converts an objection into a stipulation. Since later relational accounts keep Dickie's central insight without the circle [4] , the circle is a defect rather than, as he maintains, an informative feature. The cause: why a relational definition was wanted Two pressures produced the institutional theory. The first was the readymade. Duchamp's Fountain (1917) is a manufactured urinal. Nothing in how it looks distinguishes it from the fixtures still in the shop. If what makes something art is a property you can see in it, Fountain is not art; and by the 1960s the art world had settled the question in practice. The second was Morris Weitz's argument in "The Role of Theory in Aesthetics" (1956 [5] ) that art is an open concept. New forms are constantly created that no prior definition anticipated, so any set of necessary and sufficient conditions must either exclude them or be so loose as to say nothing. Weitz concluded that definition was the wrong project. Arthur Danto's "The Artworld" (1964) suggested a way to keep the project while accepting Weitz's evidence: look for the art-making property somewhere other than in the object. For Danto that somewhere was theory and history — what one has to know in order to see a Brillo box as a work. Dickie took the relational move and gave it a different relatum. In "Defining Art" (1969 [6] ) and Art and the Aesthetic: An Institutional Analysis (1974), the artworld is a social institution, and art status is conferred by persons acting on its behalf. The effect, and the misreading it produced The immediate effect was the charge of arbitrariness: if status is conferred, anything can be art, so the definition is empty and the theory licenses whatever the powerful decide. Two things dispose of this. Dickie's definition is classificatory. It says what makes something a work of art, not what makes one good, and he insists that most art is bad and is art nonetheless. So "anything can be art" is not the shocking consequence it is presented as; it is the consequence any account must accept once it accepts Fountain . And the artworld is not a body with officers. It is a loose established practice with roles — maker, presenter, public — that anyone may occupy. Dickie's frequent example is that a Sunday painter who hangs a canvas in a living room for family to look at has done everything the definition requires. No gatekeeper is involved. The reductionist reading therefore fails. It attacks a picture of the artworld as an authority, which was never the theory. The problem that remains: where the practice comes from Now the harder question. Conferral requires an artworld. An artworld is a practice with the roles already in place. So what was the status of the first object made and presented in this way, before there was any established practice for the maker to act on behalf of? The same question arises without any appeal to prehistory. Consider the shadow-puppet theatre of Java and Bali. Wayang kulit has been performed for something over a thousand years, with trained specialists, a repertoire, standards of execution transmitted through apprenticeship, and audiences competent to tell a fine performance from a poor one. It had no galleries, no critics in the modern sense, no institution devoted to art as a category, and for most of its history no participant would have described the practice using anything corresponding to our concept of art. In 2003 UNESCO recognised it as intangible heritage, which is a conferral of exactly the sort Dickie describes, made by a body wholly outside the practice. Dickie's theory must give one of two answers, and both are damaging. It can say the dhalang and the village audience constitute an artworld public: people prepared in some degree to understand what is presented to them. That is a natural thing to say, and it is what Dickie's 1984 definitions permit. But it drains "artworld" of content. If any practice with skilled makers and competent audiences counts, the institutional condition amounts to "someone made this to be understood by people able to understand it", and the theory has stopped distinguishing art from cookery, oratory, or well-argued litigation. Or it can hold the institutional requirement to something more specific — a practice organised around art as a category — in which case wayang was not art before that framework reached it, and became art through recognition by outsiders. That is not a conclusion anyone should want, and it is the arbitrariness charge returning at a place where it does apply. Dickie's answer, and why it does not help The Art Circle meets this directly. Dickie replaces conferral with a set of five interlocking definitions — of artist, work, public, artworld and artworld system — and concedes that they are circular. He argues that the circularity is not vicious because the definitions are informative taken together: they display the structure of a practice whose elements are constituted by their relations to one another, in the way that "buyer", "seller" and "sale" are. The analogy is instruc