« Features

On the Trail of the Unicorn: Trying to Define Art

By John Valentine

There have been many attempts to capture and analyze this elusive animal in the history of philosophical thought. Plato and Aristotle, for example, took the route of necessary and sufficient conditions. That is, they assumed that for any item correctly to be designated as art it had to be 1) an artifact, and 2) an imitation. Hence, their use of the term mimesis: art is essentially any humanly-made object that copies natural forms or those of the human world. Thus, these two characteristics were thought to be jointly necessary and sufficient to tell us what art is. This approach is over two thousand years old and still has it devotees in the contemporary art world.

The problem, of course, with this ancient definition is that societies have evolved and we now accept abstract and non-representational works as art. This ancient, two-place predicate is susceptible to many counter-examples. And so are various other two-place predicate attempts. Consider: art is 1) an artifact, 2) that is physically beautiful. But one sees immediately here that beauty is highly subjective and there can exist works of art that are not regarded as physically beautiful. Another attempt: art is 1) an artifact, 2) that is expressive of emotions. We find this approach in philosophers such as Leo Tolstoy and R.G. Collingwood. But, aside from the issue of exactly how an inanimate work of art can express emotions, we have the significant objection that not all works of art are meant to be expressive of emotions nor need they be taken that way by audiences. And finally, it has been suggested that art is 1) an artifact, that 2) is creative. The issue now, I think, is obvious. If “creative” means “unique/never experienced before,” then this unduly restricts the domain of art. And also, even with a weaker definition of “creative,” there are clearly many works in the art world that are commonly accepted as art but show no creativity whatsoever (knock-offs, etc.).

John Valentine. Beginning Aesthetics: An Introduction to the Philosophy of Art. Dubuque, IA: Kendall Hunt Pub Co. First published in 2001.

My brief survey of two-place predicate attempts to define art is not meant to be exhaustive. I have outlined some of the main contenders. However, a sea change took place in the 1960s. The philosopher Arthur Danto, following a line of thinking in Marcel Duchamp and Maurice Mandelbaum, wrote a famous article entitled “The Art World” in which he claimed that art is any object (artifact) presented to or found within the ‘atmosphere of theory’ in the contemporary art world. Noël Carroll formulated Danto’s approach to defining art in more rigorous terms: “X is a work of art if and only if (a) X has a subject (b) about which X projects an attitude or point of view (c) by means of rhetorical (usually metaphorical) ellipsis, (d) which ellipsis requires audience participation to fill in what is missing (interpretation), (e) where both the work and the interpretation require an art-historical context.”1 This complicated-sounding formula is actually fairly straightforward. It says that a work of art must have a subject (it must be about something) toward which the artist is expressing an attitude; the technique of expression involves ellipsis or some kind of shortening or coding of the subject and point of view; the shortening or coding requires that audiences produce an interpretation or theory of what is ‘being said’ in the work; and this latter task can only be accomplished in a given art-historical setting.

An example would be useful at this point. Consider Duchamp’s Fountain. Some critics have claimed that the urinal as presented in 1917 had a subject, namely the basic idea of art itself; that Duchamp was radically critiquing-in Dada-esque terms-prevalent ideas about art by means of the ellipsis or enigmatic quality of the urinal itself, which ellipsis virtually forced audiences to interpret the work in critical terms, and this interpretation could only be accomplished within the relevant art historical time period (the early days of Dada and the turn-of-the-century atmosphere of theories about art in the art world). If Danto is right about this, Fountain could not have been presented as a work of art in 1817 because the art world simply had no nest of incipient ideas at that time regarding readymade objects as art. And to try to define art by means of its formal qualities (line, shape, color, texture, etc.) allegedly is defeated by Danto’s so-called ‘visually-indistinguishable-pairs’ argument. Danto claims that, of two visually identical objects (for example, snow shovels), one would be art if it were presented within the art world, while the other would not be art if presented, say, at a hardware store. The only explanation for this, he says, is that the former shovel (In Advance of the Broken Arm?) is found and interpreted within the matrix of theories making up the art world, whereas the latter shovel lacks this crucial context and is not therefore art.

Marcel Duchamp (1887-1968), Fountain, 1917, urinal signed R. Mutt. Published in The Blind Man, No. 2, page 4, New York, May 1917, Editors: Henri-Pierre Roche, Beatrice Wood, and Marcel Duchamp. Caption read: "Fountain by R. Mutt, Photograph by Alfred Stieglitz, THE EXHIBIT REFUSED BY THE INDEPENDENTS." Source: Wikimedia Commons.

Danto’s theory was influenced by a statement Duchamp made concerning readymades: “No beauty, no ugliness, nothing particularly aesthetic about them.” Danto was also influenced by an argument put forward by the philosopher Maurice Mandelbaum to the effect that art could be defined, not by reference to physical characteristics of putative art objects, but by the ideational context in which they are found/presented.2 These earlier ideas evolved into Danto’s full-blown definition of art, and also influenced George Dickie’s later approach to defining art.3 The ‘art world’ or ‘institutional’ definition is now widespread in many cultural and academic settings.

But there are problems here. Danto and Dickie do not tell us about how the art world originated historically. There’s an odd circularity in saying that an item can only be art in terms of its presence in a pre-existing art world when we don’t know with exactness the origin of that world. Which came first: putative art objects or the art world itself? Also, both philosophers are vague about defining the term ‘art world’ with precision, and whether there is only one art world or many (Danto was often accused of equating the ‘art world’ with ‘the world of art in New York City’). However, I think that to track down the unicorn of an adequate definition of art we will have to return to the issue of formalism and the formal properties of putative art objects.

First, however, a brief point about nomenclature. Most philosophers today believe that, whatever else art is, it must at least be an artifact (something made by humans). In the past decade or so, scientists have discovered the incredible abilities of many other animal species (see Virginia Morell’s excellent book Animal Wise). Among these abilities is the widespread appearance of an aesthetic sensibility and the likelihood that many animals can make art-objects, at least in a basic way. While these claims are still being debated, I shall choose to take the high road and substitute the neologism ‘sentifact’ for the old word ‘artifact.’ A sentifact is any object created by a sentient being. Thus, Koko’s gorilla paintings and the nests made by male Bower Birds (among other animal examples) would be sentifacts. But could they be art?

Let’s try to answer this by re-considering the issue of defining art by way of formal qualities such as line, shape, color, texture, smell, and so on. Duchamp tells us there is “nothing particularly aesthetic” about readymades, but everything here depends on what he meant by the word ‘aesthetic.’ Indeed, he clarified this in an interview given in 1915: “If only America would realize that the art of Europe is finished-dead-and that America is the country of the art of the future…Look at the skyscrapers! Has Europe anything to show more beautiful than these?”4 In other words, Duchamp was bored with the old 19th Century Salon standards of beauty and aesthetics. But he apparently overlooked or ignored the root meaning of the Greek word aisthētikos, namely ‘that which is perceived by the senses.’ We may be able therefore to make a distinction between low-end sensory (or formal) qualities of an intentionally made object (the primordial aesthetics of the object, if you will) and higher-end aesthetic qualities, such as beauty, elegance, harmony, mystery, and so on. Higher-end aesthetic qualities are generally more complex than low-end ones and thus require some kind of critical interpretation, undoubtedly via a web of theories in the existing art world. With this distinction in mind we might offer the following definition of art: art is any intentionally made sentifact that has been produced as a candidate for low-end aesthetic notice. But something is still missing.

The missing ingredient is what I call a ‘perceptual shift.’ Let’s first consider Jastrow’s famous duck/rabbit image. Looked at from a certain perspective and with the concept of a duck in mind (either explicitly or implicitly), one sees a duck. Likewise, from a different perspective and with the concept of a rabbit in mind, one sees a rabbit. One can shift back and forth between the two images rather easily, but they cannot be seen simultaneously. We can apply this generic sort of perceptual shift to defining art. There’s no better example than a readymade. Viewed from a utilitarian perspective, Fountain is clearly a urinal. But viewed from the perspective of low-end aesthetic properties, Fountain can be shifted, as it were, to a candidate for aesthetic notice, wherein its functionality is kept in the background. When so shifted, we can say that we are perceiving Fountain as art simpliciter; that is, we are not asking about the critical meaning of the piece or whether it is good or bad art. Via the perceptual shift from function to form, we are saying that the piece is art in merely a low-end classification sense (art vs. non-art). Additionally, it should be noted that Duchamp also mentioned the possibility of ‘reverse readymades”: take a painting by Rembrandt and use it as an ironing board! Aside from the obvious absurdity of this, one can see how the perceptual shift can move fluidly from function to form, and back again to function.

Now we can put all these aspects together: art is any intentionally made sentifact that has been produced as a candidate for low-end aesthetic notice by means of a perceptual shift. Notice that this definition can be applied outside of any existing art world. It has the virtue of being phenomenologically foundational, and need not address the issue of whether or not the putative art object ‘has a meaning’ that must be interpreted within a nest of existing art world theories. Interpreting art is not the same as defining it in the low-end classification sense.

Although the ‘sentifact theory’ of art has other complexities, I shall briefly consider four major objections to it. The first is that the theory cannot account for performance or conceptual art where the idea behind the piece is crucial (consider the work of Joseph Beuys). But, as I’ve suggested, the low-end formal qualities of the art object are phenomenologically foundational; they are place-holders establishing the artness as such of the object. What the performance or conceptual work means, or what the artist is ‘trying to say,’ is a high-end critical function. It is not crucial in defining art, although it may be in terms of understanding it.

Secondly, one might argue that the sentifact theory is too broad; it would allow anything and everything to be art. But notice that the theory rules out naturally occurring objects for the status of art-they are not sentifacts. Thus, the theory preserves the semantic sortal distinction between art and non-art: if everything is art, nothing is. Also, art in the 21st century has become so diverse and multi-faceted that we need a very broad definition of it. The sentifact theory provides a non-circular and very broad definition. (An example of a circular and useless definition of art would be: “art is anything that artists say it is.”)

Thirdly, one might wonder if ideas as such can be art. Are ideas intentionally made sentifacts that have been produced as candidates for low-end aesthetic notice by means of a perceptual shift? George Dickie has led the way here by arguing that artifacts do not have to be physical, they can be mental. (He gives the example of a poem as a non-physical artifact.) If this is true, then sentifacts can be mental. If a sufficiently intelligent sentient being (such as a human or other higher primate or male Bower Bird, etc.) is capable of making something like a concept or image emerge in its consciousness, then it might also be capable of doing a sort of intuitive perceptual shift where it is mentally considering only the imagined low-end aesthetics of the concept or image. In the case of humans, for example, take the example of someone imagining the formal qualities of an Arabian stallion. What he or she is imagining is indeed a mental sentifact that satisfies the criteria of the sentifact definition of art. He or she creates the idea of the stallion, notices the imaginary formal qualities of the stallion in their ‘mind’s eye,’ and is shifting away from the horse as a functional or rideable object to the horse as aesthetic form. And we might note that Koko (the lowland gorilla) not only makes paintings, she also names them. One only has to read Virginia Morell’s book to see how we humans have vastly underestimated the capabilities of non-human animals due to our persistent speciesism.

Greg Eltringham, Mission Accomplished, 2010, acrylic on canvas, 65 x 91. Photo: John McKinnon

The last objection I shall consider to the sentifact theory has to do with intentionally made sentifacts that are produced with only functionality in mind. Could they still be correctly designated as art? If the original maker of the sentifact does not do the perceptual shift from function to form, someone else could do it in their stead. I call this process ‘Duchampification.’ For example, Brillo Boxes at the grocery store could be re-purposed as art objects simpliciter if we titled and named them-either individually or collectively-and pointed or referred to their low-end formal aesthetic properties via the perceptual shift. This is somewhat similar to what Duchamp did with the transition from the urinal to Fountain; there could be an intentional re-making of the Boxes as candidates for aesthetic notice. We can do this even if it was not done originally by the makers of the Boxes. This might also pertain to non-human animal cases where we are not sure if the animal in question has an aesthetic sensibility, but of course one must be careful about this assumption. Thus, an ant hill might be Duchampified just in case we believe that ants cannot do the requisite job required by the sentifact theory.

In conclusion, I believe the sentifact theory has some potential to be the lost unicorn in the search for art’s definition. It has the virtue of avoiding many of the pitfalls of ‘art world’ theories and in no way threatens future artistic creativity. It also provides some important necessary and sufficient conditions for defining art rather than saying art is an open concept (i.e., one with no definition) or art is “anything you want it to be.” Time will tell if we are coming closer to our rare and hard-to-find unicorn.


1.    Carroll, N. Cited in Stecker, R. Aesthetics and the Philosophy of Art. Lanham: Rowan and Littlefield, 2005, p. 94.

2.    See Mandelbaum. M. “Family Resemblances and Generalizations Concerning the Arts,” in Neill, A. and Ridley, A. The Philosophy of Art: Readings Ancient and Modern. New York: McGraw-Hill, 1995.

3.    See Dickie, G. Introduction to Aesthetics: An Analytical Approach. New York: Oxford University Press, 1997.

4.    Duchamp, M. Cited in Danto, A. “Marcel Duchamp and the End of Taste: A Defense of Contemporary Art.” News, TOUT-FAIT: The Marcel Duchamp Studies Online Journal. 2000. Issue 3: 1-15.

John Valentine is a professor of philosophy at the Savannah College of Art and Design, where he has taught since 1990. His publication credits include the textbook Beginning Aesthetics (McGraw-Hill, 2007); articles in the Florida Philosophical Review, the Southwest Philosophy Review, the Journal of Philosophical Research, the Journal of Speculative Philosophy and The Philosopher; poetry published in various journals, including The Sewanee Review, The Midwest Quarterly and the Southern Poetry Review, among others; and six chapbooks of poetry.

Tags: , ,

Related Articles

Leave a Reply

You must be logged in to post a comment.