We will now look at a second conception of art -- one that is distinctly modern -- the theory that art is essentially an "expression of emotion." Although it has forerunners in antiquity and may be traced to isolated statements throughout the Renaissance and medieval period, and even back to Roman times, the theory held no place of importance among philosophers until the latter part of the nineteenth century. Just as the modern imitation theory of art is historically tied to the predominantly representational art of the Renaissance and Academic period of the seventeenth and eighteenth centuries, so the expression theory is linked to the rise of Romantic art in the nineteenth century. While imitationist theories try to account for art in terms of the external, objective world imitated in works of art, expressionist theories direct attention to the inner, subjective world of human emotions, feelings, and sentiments.

As we saw in the last chapter, there is an ambiguity in describing what an artwork represents. When, in describing Raphael's Madonna of the Meadow, we say that John the Baptist is holding a shepherd's crook, it is not clear whether we are talking about John himself (the subject matter) or a figure that appears in the central left-hand portion of Raphael's painting (the content). Similarly, when we talk about music in terms such as stately, exuberant, melancholy, or gay, it is often unclear whether we are describing the emotion aroused in us by the music or the character of the music itself. Ordinary language is very economical, and so we use the same words to describe both the object represented and the artistic representation of that object, and similarly, both the emotion expressed and the artistic expression of that emotion. As we saw in the previous chapter, problems naturally arise from this sort of ambiguity.

The central problem of expression, therefore, parallels that of representation. In fact these are distinguishable but not separable aspects of a unified whole which we must ultimately see properly coordinated in our account. Just as the object represented can be seen as (1) the object itself and (2) the object as represented in a given artwork, so we can distinguish expression as concerned with (1) emotion in its pure mental state ("in one's head") and (2) emotion as expressed in a given work of art. The autonomy/heteronomy problem springs in either case from (1) and is solved by (2). It is only in the first sense that we suppose the artwork to be separated from its representational or expressive content.

In the theory of expression the autonomy problem is known as "the problem of expression." How can a private, internal, mental sensation appear in aesthetic experience as part of a public, external, physical object such as a painting or piece of sculpture? That is, how can the emotion expressed in art be understood in both senses (1) and (2) above? A work of art is a physical object, or, in the case of theater and music, a public performance. The feeling or mood that these works generate, on the other hand, is a private mental process that goes on inside people's heads. How, then, can the one be said to "express" the other? By separating the two so sharply, we are puzzled how to put them back together again. As the philosopher Bernard Bosanquet put it, "How can the feeling be got into the object?"

The problem is that, stated in this way, it doesn't seem possible, and yet we know quite clearly from our own experience that it surely does occur. In ordinary aesthetic experience we find a Bonnard warm and cheerful, a van Gogh intense and exciting, a symphony by Tchaikovsky somber and melancholy. And yet this common occurrence has been made to appear utterly impossible -- a typical philosophical problem. Like the imitationist theory, expressionism is based on an obvious and seemingly harmless truism, that art has something to do with our feelings, that works of art typically engage our emotions, seem to be about our feelings, and often bring about such emotions and feelings. This is less a theory of art than a statement of elementary fact. As the expressionist R. G. Collingwood put it,

Since the artist proper has something to do with emotion . . . what is it that he does?...Nothing could be more entirely commonplace than to say he expresses them. The idea is familiar to every artist, and to every one else who has any acquaintance with the arts. To state it is not to state a philosophical theory or definition of art; it is to state a fact or supposed fact about which, when we have sufficiently identified it, we shall have later to theorize philosophically. For the present . . . we have to decide what it is that people are saying when they use the phrase. Later on, we shall have to see whether it will fit into a coherent theory.2

At first, then, the expressionist uses the term expression simply to record the obvious fact that art has something to do with human emotions and feelings. But by articulating this in terms of "expression" a special and by no means obvious slant is placed on this intuitively obvious assumption which suggests that art is the venting or exhibition of emotion. And this slant is largely determined by the ordinary meaning of the term expression...