Representational seeing and expressive perception are, within pictorial art, required to conform to a standard of correctness, and this standard of correctness goes back to the intentions of the artist in so far as they are fulfilled. (p. 89)
Perception is the focal point in Wollheim's account of pictorial meaning. Perception allows us to see what a painting is like and through it we can grasp what the picture means because of what it is like. In the perception of paintings the distinction between seeing and knowing has been problematic. In painting accounts of perception have fluctuated between the painted surface (seeing) and cultural background (knowing). Wollheim grounds pictorial meaning in an experience to be had through perception, only what can be seen in the painting carries meaning. Semioticians argue that reference to individual perception is meaningless because it is outside the social formation. Pictorial perception is grounded in our social formation. Wollheim's notion of what one does in order to grasp the meaning of a picture is this:
I came to recognize that it often took the first hour or so in front of a painting for stray associations or motivated misperceptions to settle down, and it was only then, with the same amount of time or more to spend looking at it, that the picture could be relied on to disclose itself as it was (p .8)
We learn about a painting by looking at it and this looking is informed by external evidence, what we actually see in the picture must overrule any evidence external to that painting. The evidence of the eyes is always primary. The spectator, Wollheim insists, must have the required sensitivity and he must have the required information. What he does not need is what is required of the hearer of a language: knowledge of rules or conventions which he applies to the painting to extract its meaning.
Criticizing Lessing's notion of perception, Wollheim stresses that our visual apparatus is not just optical eye but an embodied eye:
"..it is only a piece of theory, an epistemological presupposition, that leads us to think that there is available a neutral description drained of emotion that fits the original perception we have of such objects." (Wollheim 1973 p. 95)
In looking at paintings there is a sense in which the spectator has a relationship to the object that carries meaning that is distinctive of perception. Vision is intrinsically perspectival. We perceive the painting from the location of our bodies in space and time.
Wollheim argues that in looking at a painting we are not trying to see the artist's intention. This seems to answer criticisms of other intentionalist accounts of meaning that place knowledge of the artist's intention as the determinant of meaning:
"..if the spectator has, in front of the picture and caused by it, the experience that the artist intended him to have, this is enough. There is no reason why recognition of the artist's intention should have to play a part in this causal story." (p. 96)
Wollheim is concerned to distinguish between our natural ability to see things in the world and cognitive seeing that we use in decoding a sign:
"Do we , when we look at such things [as maps signs etc], see whatever they are of in their surface, or do we just see the things as marks, which the, in virtue of the knowledge of the system to which they belong, recognize to be signs of what they are of?" (p. 60)
Only natural seeing allows us to perceive what Wollheim defines as a representation. Cognitive seeing requires us to have a learned skill that allows us to decode, to recognize what the image is a sign of. According to Wollheim the power of painting rests upon this subsidiary capacity of perception he calls 'seeing in'. Pictorial perception involves seeing one thing in another, we move from seeing one thing in the figure to seeing a different, and incompatible, thing in it. Seeing in is subject to a criterion of correctness that comes from the fulfilled intentions of the artist. But if seeing in is necessary to our grasp of pictorial meaning, how do we explain the paintings that do not conform to Wollheim's requirements? How do� trompe l'oeil paintings and Barnet Newmans, whose correct perceptions, according to Wollheim, is not characterized by twofoldness carry meaning? That they are meaningful under Wollheim's scheme in virtue of their falling under the concepts painting and art is without question. Wollheim does not explain how their meaning is conveyed through perception.
Another form of perception that accounts for expressive meaning in painting Wollheim calls expressive perception. Expressive perception is the capacity which allows us to see a painting as expressing feelings and emotions. In expressive perception the emotion flows from us to what we perceive to but it is also responsive to how the world looks. Once the emotion is invoked it penetrates perception and affects how we perceive what we perceive:
" We see emotion" (Wittgenstein 1967 p. 41e)
Bryson takes the Wittgensteinian position that all seeing (and meaning) is seeing as, painting comes already interpreted: "there is no spectator who looks at a painting who is not already engaged in interpreting it." Bryson (1990 p. 10) Seeing, on this view, is cognitive and, he argues, allows the admission of historical, social or gender differences in the perception of art:
"Something is recognized through interindividual links and the social consensus that exists between persons, and makes no reference to an inner state that is not embodied in signs." (p. 43).
The double negative is rather puzzling. It seems to say that inner states can be embodied into a sign and they are thus made manifest. This seems to concur with Wollheim's view that the mental states of the artist are available to the spectator through looking at the painting.
Bryson does not dispute that a spectator looking at a picture and identifying the same content which someone else finds in it, could have the same experience. But we can arrive at the belief that the spectator has successfully recognized the painting only by looking at his recognition-behaviour and not by referring to the spectator's mind. The congruence between picture and mental event cannot be thought the criterion for recognition.
By drawing an analogy with Wittgenstein's pain argument, Bryson argues that in looking at paintings the spectator has neither knowledge of, nor cannot infer a mental event in the artist's mind which then can be compared with that painting. Even if there is an innate state of perception modified by culture, as Wollheim believes, that state is irrelevant and undemonstrable because Bryson believes that:
" ...coherent propositions could not be constructed involving that state" (p. 62).
He adds that, just because someone has an experience, to generalize from one case to an entire species cannot be regarded as a valid procedure. He does not deny the possibility of either congruent private experiences or universal visual experience. What he denies is the validity of propositions concerning the existence and nature of that universal visual experience.
The criterion of recognition always involves more than one spectator: it is essential that an agreement exists to attach a term of recognition to the image in regularized and consensual fashion. Mental events may accompany recognition,
"...but in so far as recognition in any way submits to criteria of correctness it comes to operate within the world as practical activity, and no longer concerns the noumenal field: the latter may assume any guise we may attribute to it." (Bryson 1986 p. 50)
Wollheim refers to the cross-cultural studies in perception by W. Hudson. There is strong evidence to suggest that the identification of lines with contours is an innate function of normal human vision see also Kennedy J.M. (1974). Evidence against the thesis that pictorial recognition depends upon language comes from animal studies. Fodor's sheep, monkeys and pigeons.  on perception has shown patterns of visual cognition that point to a primitive nature that is separate from social conventions. J. Halverson's (1992) findings support Wollheim's argument in showing that even in highly conventional representations a certain figure is recognized at a perceptual level. The symbolic value or reference that figure may have is a matter of social convention, involving social cognition rather than visual perception. If some figures become stylised beyond recognition by a cultural outsider it is because they lack sufficient isomorphism with the represented object. He concludes that there is no reason to suggest that the pictorial perception of paintings, of drawings and of Palaeolithic depictions, is a culturally specific skill, conventions seem to be secondary:
"Culture would be relevant to pictorial perception only in the sense that recognition of the object depicted would depend on familiarity with the object, which would of course be determined by the life experiences afforded by a particular culture."(p. 401)
Wollheim allows that both kinds of perception are partly shaped by social influences, what he proposes is backed by empirical evidence and theoretical argument. Bryson believes that the social formation is the criterion for correct perception. One view must be mistaken. If we believe current empirical research Bryson seems to be factually wrong.
Wollheim agrees with Gombrich (1960) that the perception of paintings depends upon the information we have about it, there is no 'innocent eye', perception depends upon the spectator getting hold of 'cognitive stock' [i] . This is evidence of how the painting came to be made. In interpreting paintings by looking directly at them, a variety of background information is brought to that experience. But not just any piece of information will do:
"...any information of which the spectator has need must be information that affects what he sees when he looks at the picture: because it is only through what can be seen when the picture is looked at that the picture carries meaning." (p. 95)
Wollheim argues that if the information does not allow us to see something in the painting then it is background information that may be of historical or biographical but not aesthetic value. Hearsay or independent knowledge of what the artist intended can serve as background information in shaping or forming how the spectator sees the painting. But it oversteps its legitimate role when it leads the spectator to say or think things about the painting that he does not see when he looks at it. Nevertheless the distinction between what is there and what is background is always a relative matter until we come to a cut-off point. This is analogous with what one would think, in philosophy, as the proper way of distinguishing between what is given and what is inferred, or what is directly or indirectly perceived.
Evidence about the artist's, conscious or unconscious, state of mind at the time of producing the work guides us in our expressive perception of the work. Wollheim looks at some pictures executed by Manet in the first few months after Manet's wife had died in great pain. Once we have this information, then: "...we are likely to start perceiving these pictures as of a piece with the emotions that filled Monet's life at this period. We shall see these pictures as expressing sorrow, and regret, and the slow recovery from sorrow sustained by the abandonment of old regrets. We shall see them as works of mourning, which is what I take them to be." (p. 95)
Cognitive stock allows us to see something in the painting cognitively, as something else. That something else must always be what the artist intended us to see it as.
Perception and Knowledge
As we have seen, Wollheim's claim is that we directly see what the artist intended by studying the artist's painting and its context intensively over a period of time. Danto (1981), in contrast, seems to suggest that to understand painting we can dispense with the eyes. Danto proposes a number supposedly indistinguishable paintings that consist of identical squares of red paint painted by various artists with titles like 'Israelites Crossing the Red Sea ' , ' Kierkegaard's Mood ' , ' Red Square ' and so on. From this Danto makes two assumptions. First that a work of art can be identical in all its physical and perceptual particulars to an object that is not a work of art. Second, and this is the crucial point in relation to Wollheim's thesis, two different works of art can be physically and perceptually identical. Perception cannot distinguish one painting from another. It is beyond the study of appearances, and by analogy the study of paintings, that knowledge can be had, not through visual evidence alone.
David Carrier (1993) notices that Danto's argument that two visually identical works can really be different directly contradicts Wollheim's account of meaning grounded in perception. He suggests these are unreconcilable views. Wollheim believes that, when properly identified, the visual appearance of the painting tells us what its aesthetic qualities are. What cannot be seen or imagined in the picture is not part of that picture's meaning. Danto argues that perception does not reveal the true content of the painting, perceptions do not provide direct access to knowledge. he denies this by showing that a painting is not identifiable by reference to its appearance.
If Danto's argument is valid then, Carrier argues, one central assumption of Wollheim's thesis in Painting as an Art is false. I disagree, Wollheim, as we have seen, does not believe that we can have access to a neutral description of the world. Tilghman (1984) convincingly argues that the titles provide interpretations of the paintings that enable us to see them differently, under different interpretations our visual experience is quite different:
"The argument that they are visually indiscernible can only appear to be the result of a philosophical prejudice that fails to take account of the complexity of the concept of seeing." (Tilghman 1984, p. 135)
[i] Both Wollheim and Danto use the notion of seeing in 'terms of' proposed by Hanson, N.R. (1961) Patterns of Discovery (Cambridge: Cambridge University Press) Hanson's theory that perception is 'theory loaded' is a particular version of the notion that what is visible is relative to background information.