Like anything at all, art is hierarchical, and, like any hierarchy, it can unfold itself in different directions presenting an infinity of its complementary aspects. That is why there is no closed definition of art; in a discourse, the only realistic approach is to compare art with anything else indicating the distinctions specific for that particular angle of view.
In the modern society, art is commonly regarded as a professional occupation of a number of specially trained people producing articles (works) of art (such as paintings, fiction books, or musical compositions) or organizing spectacular performances (exhibitions, installations, concerts etc.). That is, art is thus defined as an activity, and we need to guess, why other (and often very similar) activities are not exactly art.
On the other hand, we can admire the perfection of somebody's regular work and call it artistic. In this sense, art is taken as an aspect or quality of any activity at all, rather than a separate activity. This raises the question of the sound criteria to distinguish the level of art from mere smartness.
Also, we can find some natural phenomena or artifacts exceptionally impressive and attribute an artistic character to these outer things, which are considered to be beautiful or ugly on their own account. Once again, one has to explain this inner aestheticism in contrast to pure functionality.
All the three connotations are present in the idea of art, as well as many others, not mentioned here. Instead of approaching each aspect in a random manner, collecting the appropriate phenomenology (which, too, is a legitimate choice, in an appropriate context), unism is to consider them as an expression of the universal integrity of the world and the universality of human (conscious) activity.
First, we observe that each time we do anything, we, first, produce some material changes in the world (otherwise, there would be no activity at all), and second, we observe, consider and mean ourselves as the producer (otherwise, there would be no consciousness). The second, reflexive aspect of activity could be taken as an object for a special activity producing the very ways of production; however, this new activity does not coincide with the original and needs to get objectified in its turn in order to adjust the ways of reflection. A whole hierarchy of reflective activities grows in this way, and art is to be located somewhere within this hierarchy. That is, the primary definition of art is that it belongs to the domain of cultural self-reflection, along with its other modes.
However general, this preliminary idea suggests a number of important corollaries. Thus, it is important that art cannot exist outside a particular culture, and hence it does not belong to things as they are, regardless of the way we use them. The world on itself does not possess any aesthetic quality; it is people who treat it that way. As the complementary aspect of the same, any natural thing at all can be made into an article of art by involving it in some reflexive activity. There is no absolute beauty, or absolute sublimity; everything depends on the historical and cultural circumstances.
Further, the cultural dependence of art results in a difference of aesthetic feelings and norms in different societies or epochs. An artefact can be regarded as a work of art as long as it represents a certain mode of action typical of the current community; however, it may have no aesthetic value (and even no value at all) for those who do not share the same cultural heritage. This also means that the perception of art and artistic creativity is never innate, it must be socially cultivated. In particular, one can grow accustomed to the ways of an alien culture and gradually develop the appreciation of its art. Sometimes, however, the load of traditional education would not permit one to accept too "exotic" kinds of art, though one might rationally admit their presence. Tastes differ.
This brings us to comparing art to science. Though science is obviously yet another reflexive activity, its product is often acceptable by very different social layers, regardless of nationality or individual preferences, and this "portability" of science opens up the industry of education, including art schools. Science is not a matter of taste, it refers to objective necessity. On the other hand, there is yet another area of reflection, which is akin to science in its pretense to objective universality, while requiring a subjective choice, an active social position; this is what we call philosophy. We find that these three kinds of reflexive activity belong to the same level of hierarchy, which we could denote as analytical reflection. The common feature of all such activities is that they are intended to produce something that represents the ways of human activity rather than satisfies their immediate needs; however, this representation is not intended to directly influence any further activity, and its practical value is to yet be revealed.
Within the analytical level, there is a hierarchy of cultural phenomena representing the different forms of reflection, and there are no rigid boundaries. Though the distinction of art, science and philosophy is of a very fundamental nature, other hierarchical structures might be preferable for certain applications. Specifically, we distinguish the levels of analytical reflection by the type of the product. Thus, in art, any material thing at all can be used to represent anything else, to become its image. On the contrary, the product of science is restricted to a very limited range of standard forms specially designed for maximum portability; such an abstract image of reality is called knowledge. Finally, philosophy combines the generality of scientific notions with the versatility of the artistic image producing universal categories and categorical schemes. A philosophical category cannot be expressed in a formal way, like scientific knowledge; it can only be conveyed by the specific modes of action. On the other hand, such an activity is no longer a mere image of something, since it refers to a universal principle, representing the very activity of reflection.
In this way, we obtain a criterion of art, to distinguish it from the many art-like phenomena. When some human activity is aimed to creating something valued for its ability to represent the typical modes of other activities without developing the very way of representation in a systematic manner, we are certain to observe there some artistic quality; in the appropriate social conditions, the product of such an activity will be treated as a work of art. However incomplete, this criterion is definite enough, without exaggerating the differences between the levels of reflection. Art differs from science or philosophy in a relative way, and any rigid rubrication would be ultimately inadequate. Within this approach, one can speak, say, about the art of physical experiment, as well as the study of social phenomena by means of art.