Is a pile of bricks in a gallery art? A painting made by a chimpanzee? A cave drawing whose makers had no word for art? This level asks the philosopher's version of the question: is there something that all and only artworks have in common?

Art as imitation

The oldest answer is that art imitates. The Greek word was mimesis. The imitation theory holds that art is essentially the representation of reality, and for two thousand years it seemed obvious: paintings show scenes, statues show bodies, plays show actions. Plato used the idea against the arts. In Book X of the Republic he argued that a bed made by a carpenter is already a copy of the true Form of Bed, so a painting of that bed is a copy of a copy. The painter, in Benjamin Jowett's translation, is "thrice removed from the king and from the truth." Plato distrusted art because it copies appearances, which are themselves copies of the Forms.

Aristotle defended imitation instead. In the Poetics he observed that people learn by imitating from childhood, and that "objects which in themselves we view with pain, we delight to contemplate when reproduced with minute fidelity," such as the forms of despised animals and of dead bodies (S. H. Butcher's 1895 translation). For Aristotle, we enjoy imitations because we learn something from looking at them.

The trouble came with music and, later, abstract painting. A string quartet or a canvas of colored squares imitates nothing obvious, yet both are plainly art.

Art as expression

The Romantics proposed that art expresses feeling. William Wordsworth called good poetry "the spontaneous overflow of powerful feelings" in 1800. Leo Tolstoy built a full theory on the idea in What Is Art? (1897): "Art is a human activity, consisting in this, that one man consciously, by means of certain external signs, hands on to others feelings he has lived through, and that other people are infected by these feelings, and also experience them." Tolstoy defined art as the passing on of felt emotion from artist to audience. In the 1930s the English philosopher R. G. Collingwood refined the view: for him, expression is not unloading a ready-made feeling but working out, in the medium itself, what one feels.

The standard objection is simple. Critics of expression theory note that artists can make moving work without feeling the emotion it expresses. A composer on a deadline may write a heartbreaking song while feeling only tired.

Art as form

The English critic Clive Bell, writing in 1914, claimed that what all real art shares is "significant form." For Bell, significant form is the arrangement of lines and colors that stirs aesthetic emotion. Subject matter, he thought, did not matter; a battle scene and a bowl of apples are equal as art if their forms move us equally. Formalism helped audiences accept abstract painting. Critics reply that ignoring subject matter makes nonsense of much art, from religious icons to Goya's war prints, and that the definition is circular: significant form is whatever causes aesthetic emotion, and aesthetic emotion is whatever significant form causes.

Giving up on a definition?

In 1956 the philosopher Morris Weitz argued that the search was hopeless. Borrowing Ludwig Wittgenstein's remarks on games, he called art an open concept. Weitz held that art cannot be defined because artworks share family resemblances rather than one common essence, and because artists keep stretching the concept in ways no definition could foresee.

Others answered with definitions based on context rather than appearance. Arthur Danto and George Dickie argued, in different ways, that what makes something art is its place in an art world of history, theory and institutions. A middle path, the cluster account associated with Berys Gaut, lists typical criteria such as expressing emotion, being intellectually challenging, showing skill, originality, and belonging to an established art form. On the cluster account, something counts as art if it meets enough of a list of typical criteria, though no single one is required.

Worked example: the chimpanzee's painting

In the 1950s a chimpanzee named Congo, studied at London Zoo by the zoologist Desmond Morris, made hundreds of paintings. In 2005 three of them sold at auction in London for about fourteen thousand pounds. Are they art?

Imitation theory says no: they depict nothing. Expression theory hesitates. Morris reported that Congo seemed absorbed and would refuse to add anything once he judged a picture finished, but did he mean to pass on a feeling to anyone? Formalism may say yes, if the bold, fan-shaped strokes have significant form; some viewers say they do. A context theory points to the auction house and the gallery shows. The cluster account counts criteria: some form and a kind of skill, but little intended meaning, no tradition and no audience in Congo's own mind. Each theory captures something real about the case, and none captures everything, which is exactly what Weitz would have predicted.