"What is art?" Turning to philosophy for answers

Thomas Cole, Ruins in the Campagna di Roma, Morning, 1842.
This post is introduces definition of what art is. I'll introduce different theories art and consider their respective merits and pitfalls. To start we will need to have a clear idea on what we hope to achieve with a definition of art and what sort of thing that definition would need to be.
An important distinction to make is the one between nominal and real (or essential) definitions. A nominal definition defines the idea that a word stands for, while a real definition defines what it is to be what that word refers to. A real definition for X would identify a property (or set of properties) that each and every X has and that only Xs have. For example, a real definition for blue would be the light waves with wavelengths in the 450–495 nanometer range. While a nominal definition may state that blue is the color associated with the sky and the sea. To be the color blue is to be (to reflect) light in the 450–495 nm range, not to be the color of the sky or sea (which not even always blue). Now when we move our considerations from color to art, the real definition of art seems to be our true goal. Other definitions of art, like in its use as praise (“Wow, your painting of those flowers is a work of art!”) or derision (“Wow, your painting of those flowers is a work of art!”), fail to provide both sufficient and necessary properties for an artwork. That being said, there is only so much a definition can do. We should not , for example, expect it necessary for a definition to explain why a art matters or why we create it.
The have been many attempts to provide a theory of art, going as far back as Plato and continuing into the current era. Some early definitions of art include:
  • Art as imitation or representation.
  • Art as a medium for emotional expression.
  • Art as 'significant form'.
These definitions all have an immediate draw, but upon closer look one can see that these views lacking. By these accounts many non-art objects would count as art (like a nicely made advertisement or a sports car) and possibly some artworks like Duchamp's Fountain, Warhol's Brillo Boxes , and other conceptual pieces would not count as art.

Can art be defined?

With the difficulties faced in defining art with an appropriate scope, one has to question the possibility of defining art at all. In his The Role of Theory in Aesthetics, Morris Weitz argued that any real definition of art would fail because works of art are related to one another like a family rather than by some rigid set of properties. This family resemblance relationship (taken from Wittgenstein) proposes that groups given a common name and thought to connected by a common, essential feature are, rather, connected by a network of overlapping and criss-crossing features. Much like a family whose shared characteristics like: build, height, eye color, facial features, overlap and criss-cross throughout their family tree.
Let's consider Wittgenstein's example, games, which exhibit this familial resemblance to one another. There are many kinds of games: ball games, card games, board games, and so on, that fail to be united by a ubiquitous trait. When asked to define what makes something a game one might say “It is a competition with winners and losers.” While this may work for games like chess, it doesn't seem to work for games like catch or games with a single participant. Another may offer skill as a definition, but we can turn to them with games like rock-paper-scissors or Russian roulette. For any uniting feature offered, there will be a game that lacks this feature (or a non-game that has this feature). To know what a game is not to have a real definition of it, but be able to take new examples and being able to determine whether they are games or not. Art shares this quality of Weitz proposes an open definition of art where, upon experiencing a new art-candidate, one has to make a decision whether or not it counts on art based on its similarities to past artworks. In doing so the number of properties that one associates with art is, which accounts for the expansion of art from the fine arts to the multitude of art forms accepted today. He concludes that while theories of art fail to provide a real definition of art, they retain value as suggestions to reconsider what we consider in deciding whether something is art, and can be seen as reactionary pieces to the times.
Weitz provides quite a compelling challenge to any theory of art, which combined with the challenge of placing artworks like Duchamp's Fountain and Warhol's Brillo Boxes, led aestheticians to definitions of art that of two major kinds: functional (being defined by what it does or is intended to do) and relational (being defined by its standing to other things). These approaches hope to avoid the issues of past definitions by focusing on non-perceptual properties of art rather than something percerptual like form.

New Theories of Art

Most functional definitions of art deal aesthetic properties as being central to art's function. A popular functionalist theory of art is Monroe Beardsley's intentional account; he defines art as an arrangement intended to be capable of giving an aesthetic experience made valuable by its aesthetic qualities (or an arrangement that belongs to a class of arrangements generally intended to have said capability).
This view seems to fall into the same traps as earlier definitions of art in that it can be said to be too wide and too narrow. The functionalist has some responses to this. To being to narrow the functinoalist can respond with a wider definition of aesthetic properties that includes non-perceptual qualities that would give conceptual piece like those of Warhol's and Duchamp's proper due as art. The functionalist could also double down and claim that pieces these do not constitute art, but are comments on art. In response to the functionalist account being too broad, the functionalist can dismiss things like nice cars or elegant mathematical equations using a distinction between first and secondary functions and their effects on art status.
Of the relational theories of art there are two major strains, procedural (how art is given art-status) and historical (how art is related to past art).
The popular proceduralist theory of art is George Dickie's institutional theory of art, which states that an artwork is an artifact that is presented by the artist to an Artworld audience. He later revised this theory into a set of interlocking definitions: An artist is someone who knowing creates art. Art is an artifact of a kind to be presented to an Artworld public. A public is a set of persons who are prepared to (partially) understand an object that is presented to them. The Artworld is the totality of all Artworld systems. An Artworld system is a framework for the presentation of a of art by an artist to an Artworld public.
It is obvious that the second argument is circular, but Dickie's argues that it still positively represents art and the manner in which it exists. The circularity of the argument is a reflection of art's nature. Another common objection to Dickie's definition is that this Artworld social structure is hard to distinguish from other similar social structures, making it fail to properly distinguish art from non-art.
The historical theory of art defines something as art if it stands in a certain relation to past artworks. Proposed relations have been intentional (present art has been made with the intent as being regarded in the same way as past art), functionalist (present art succeeds in performing one of the functions of an established art form), or stylistic (present art has been made in a similar style as past art).
Both kinds of relational definitions have fallen to similar criticisms. For one they lack an account of the original artworks or Artworld that future art stands in relation to. Relational definitions also have a universality problem, they seem to suggest that there is one narrative of art and fail to account for art from other cultures and histories outside the traditional canon. They seem to exclude the possibility of a lone artist outside of the art-history narrative. These criticism have been meet by providing a functionalist account for the 'original' artworks and for artworks from independent Artworlds. Generally, there have recently been moves to hybrid theories of art, as these relational and aesthetic definitions do not seem to necessarily conflict or in some cases resemble/invoke each other.
I tend to agree with Weitz's approach to the issue, that a real, distinct definition of art cannot be made, in that art is not a concept with distinct boundaries. However, theories of art can be useful in defining what we tend to think of as art and in providing us with fresh perspectives on what art can be.
Do you think either the relationalist or functionalist definitions succeed? Or has it correctly shown been that real definitions of art are impossible? Are theories of art even valuable or a waste of time?
Further Reading:
The Role of Theory in Aesthetics – Morris Weitz
Definitions of Art – Stephen Davies
The Artworld – Arthur Danto