Art

In plain terms

For most of history nobody needed a definition. Art was skilled making, and the skill was visible in the object.

That stopped working in the twentieth century, when things appeared in galleries that were indistinguishable from ordinary objects. If a box of scouring pads in a gallery is art and an identical box in a warehouse is not, then whatever makes the difference is not in the object.

Three kinds of answer followed. That art is defined by what it does - expresses feeling, presents form. That it is defined by a relation to an institution - the artworld confers the status. Or that it cannot be defined at all, and is a family of overlapping practices held together by history rather than by a common feature.

The first has counterexamples, the second explains the mechanism without justifying it, and the third gives up the question.

As a philosopher would put it

Traditional definitions were functional: art imitates, expresses, or presents significant form. Weitz's 1956 argument was that art is an open concept whose creativity guarantees new cases, so no set of necessary and sufficient conditions can hold - a Wittgensteinian position that dominated for a decade.

Danto's response, prompted by Warhol's Brillo Box in 1964, is the pivot. Two objects can be perceptually indistinguishable while one is a work and the other is not, so what makes the difference is not perceptible: it is an atmosphere of theory and a knowledge of art history - the artworld.

Dickie turned this into the institutional theory: an artwork is an artefact on which some person acting on behalf of the artworld has conferred candidacy for appreciation. The objection is circularity and permissiveness. Levinson's historical definition replaces institutions with intention - something is art if intended for regard in a way earlier artworks were correctly regarded - which handles outsider art better and inherits the problem of the first case.

What is actually disputed

Whether art can be defined.

Anti-essentialists hold that the concept is open by nature: art expands by producing cases that no existing definition anticipated, so any definition is a snapshot. Definitions are then descriptive of a period rather than of a kind.

Institutional theorists accept a definition and locate it outside the object, which explains the mechanism accurately and provides no standard - anything the artworld accepts is art, including whatever it accepts next, which makes the theory unable to criticise its own subject.

Historical definitions do better on that count and face the origin problem: if something is art by being intended for regard as earlier art was regarded, the first artworks cannot be accounted for.

And all three are Western in construction. Applying the category to objects made in traditions with no separate notion of fine art is a classification decision, and museums have been making it for two centuries.