Depiction

7. Resemblance revisited

The intuition that pictures look like what they depict was clarified above as the claim that depicting and depicted resemble each other in respect of some visible property. The discussion of seeing-in provides an opportunity to present the intuition another way. Depiction should be understood by characterizing our experience of it. Thus the better way to understand the intuition is as advocating a certain account of seeing-in: that it is the experience of likeness. (See, however, Hyman 2006) for arguments that this move is neither necessary nor wise.)

So understood, the view fares somewhat better with the first problem presented in §2. Resemblance may require there to be two things, one resembling, one resembled; the experience of resemblance does not. For instance, I may experience a sound I hear as resembling another, even if that other sound is merely one I have on occasion conjured in my imagination, and even if I am fully aware of this. This opens the way for a single account of depiction, whether or not there exists something depicted. In either case, the picture is experienced in a special way. It is seen in a way somehow involving the thought of the thing depicted (existing or not), and the proposal is to understand that way of seeing the picture as an experience of resemblance.

However, this still leaves the second problem entirely untouched. In what respect is resemblance experienced? For it is no more plausible to say that a picture need be experienced as resembling what it depicts in shape, colour, material composition and the like than it is to say that the two must resemble each other in those respects.

The answer lies in a property of things that we regularly perceive, but rarely articulate. Consider the fact that in one way a disk seen at an oblique angle may look elliptical, even if it is clear to us that its 3-D shape is a flat circle. What lies behind this elliptical look is our perceiving a shape property of an unusual kind. This property is unusual in that it is relative to a point - in this case, the point from which the disk is seen. In fact, it is the property of subtending a certain solid angle at that point. For convenience, call this property the ‘outline shape’ the disk has at that point. (Though we might just as easily follow Thomas Reid (1764) in calling the property ‘visible figure’, and might equally well understand it as he did, as a matter of the directions from the relevant point in which parts of the object lie.) Outline shape is not simply the shape of an object’s silhouette. For if the disk is marked with concentric circles, each of those will also subtend a certain solid angle at the point, and the outline shape of the whole may be taken to include these nested sets of angles too. So understood, outline shape provides a feature we see objects as having, and a respect in which it is plausible that pictures and their objects resemble each other. Why not, then, suppose that to see things in pictures just is to see the pictures as resembling those objects in that respect?

Seeing-in, then, is essentially the experience of likeness in respect of outline shape. Depiction may then be understood as that representation which works through the deliberate generation of this experience. This leaves many details to work through, and many objections to counter. Nonetheless it is a position with great promise. It succeeds in accommodating the fact that seeing-in forms an integrated whole. For in general to experience one thing as resembling another in some respect is to undergo an experience that is integrated in the right way. (It is not as if one, say, sees the resembling thing and quite separately experiences, or visualizes, or thinks about that which it resembles.) The difficulties facing partial illusion and Walton are thus avoided. But the position also succeeds in explaining why depiction has the features noted above (§3). For instance, the fact that all depiction is from a point of view stems from the fact that outline shape is itself relative to a point. To talk of an object’s outline shape is already to invoke a particular point in its surroundings, the point at which it has that outline shape. So any experience of something else (some marks, for instance) as resembling that object in respect of that outline shape already involves that point. Thus any experience of seeing something in a surface will be perspectival: it will present that thing from a point in its surroundings. And if seeing-in is necessarily perspectival, so will depiction be. While the details of this and other explanations would need working out, it is clear that there are at least the materials here for something satisfying. Prima facie, then, the view thus offers a way to perform all the most pressing tasks encountered above - accommodating our intuitions about depiction, explaining various of its features, and characterizing the experience to which it gives rise.