Pictures and Their Surfaces: Wollheim on 'Twofoldness'

Edward Winters

University of Westminster

The aim of this paper is to provide a philosophical account of our appreciation of representational painting. In so doing I shall hope to characterise the nature of our experience of paintings in such a way that there are consequences for the contemporary practice of Fine Art. My aim, then, is to provide a philosophical argument which could be of value to the critic and the parctitioner alike. In attempting to apply the philosophy of art to its objects, I hope to follow Richard Wollheim, whose work in the aesthetics of painting continues to be both revelatory and instructive.

There is a tendancy, nowadays, to think of pictures (and to analyse them) in terms of their content alone. This is particularly true of the vast amount of work being published in the emerging field of Visual Culture. A great deal can be said for this and for the political and moral insights that have been unearthed by such analysis. However, concentrating solely upon our apprehension of pictorial content leaves out of consideration an important aspect of pictures. Pictures are objects. Moreover, they are objects that have been brought into existence to serve our purposes. There are many projects we humans have to which pictures can be recruited. In what follows I shall be arguing that our appreciation of representational painting requires a special kind of attention to both the depicted scene and to the surface of the picture. It is my task to illucidate the nature of that attention. Let us begin by looking at Richard Wollheim's development of the twofold thesis, since that thesis in both its versions insists upon our attention being distributed between surface and content.

I

Wollheim's position has changed regarding the twofold thesis. Whereas he once held that our appreciation of a representational painting involved two distinct experiences, one as of the surface, the other as of the representational content, he has come to hold that it involves one experience with two aspects.[1] His former position regarded the two experiences as occurring simultaneously. This 'two-experience' view was itself a reaction to Gombrich's illusionistic account.[2]

Gombrich has argued that seeing a painting requires that I direct my attention either at the content of the picture or at the paint marks which constitute the picture.[3] He claims that we cannot direct our attention toward both simultaneously. His argument for this is that the picture is ambiguous between the marks and the content in the same way that the Jastrow Duck/Rabbit figure is ambiguous between being seen as a duck-picture and being seen as a rabbit-picture. Wollheim points out that this argument is flawed. Gombrich is only properly entitled to say that the analogy is held between seeing the black line which forms the ambiguous figure and whichever of the duck-picture or the rabbit-picture I see the black line as. It is true to say that I cannot see the black figure as both a duck-picture and a rabbit-picture at the same time, but it does not follow that I cannot see the black figure and the duck-picture (or the rabbit- picture) at the same time. If, when looking at a portrait of Marilyn Monroe, my attention switches between paint marks and Miss Monroe, as Gombrich contends, then when I am attending to the content, it must be experientially indistinguishable from my seeing that woman, Marilyn Monroe. This is rarely, if ever, the case. Indeed where it is the case (say where I take a painting of Marilyn Monroe to be Marilyn Monroe) I am at best described as having made a mistake; and certainly no analysis of such a phenomenon could be offered as an analysis of what it is to appreciate a picture.

It is a feature of our looking at pictures, in general, that we recognise that we are in the presence of a picture and not in the presence of the depicted content. Moreover, this recognition is part of the description of the experience we have when looking at pictures. So Gombrich's view is deficient in this respect. Wollheim, recognizing this deficiency develops the twofold thesis. The twofold thesis acknowledges the fact that when we look at pictures, as pictures, we do not suffer the effects of a visual illusion. The twofold thesis of pictorial understanding, in general, is developed in ways which pull into focus our appreciation of representational painting, in particular, as aesthetic objects. In this paper I hope to sustain the insights that Wollheim has revealed in his work on representation, but only after making certain criticisms of his thesis as a whole.

The twofold thesis, as has been mentioned, comes in two versions, the second developed to cope with criticisms of the first. In both versions Wollheim holds that when we say that we see the brush marks across the canvas, we are making a genuine perceptual claim. We are required by the twofold thesis to attend to the pictorial content at the same time. So when we say, for instance, that we see Marilyn Monroe in the picture, we are also making a genuine perceptual claim.

In the first version the descriptions of these two perceptual claims are characterised differently. As such, they are thought to be two experiences bound together in the peculiar perception of the painting. We have 'two projects' in mind when we look at paintings. We look at the paint surface and attend to its properties. We also see in the surface objects represented therein. The practice of painting has developed out of our ability to see surfaces representationally. So that now we see the surfaces of paintings as having been prepared for us in such a way that certain objects show up in the surface.

The artist... exploits twofoldness to build up analogies and correspondences between the medium and the object of representation [but he] cannot be thought content to leave the two visual experiences in such a way that one merely floats above the other. He must be concerned to return one experience to the other.[4]

Now, as has been mentioned, it is a mark of Wollheim's thesis that he puts so much weight on the activity of painting. Painting, that is, 'acquires a rationale' and as it does so the practice of painting brings with it constraints upon the spectator which in turn feed back and place constraints upon the painter.

Malcolm Budd finds Wollheim's 'seeing-in' account wanting because at least one of the two experiences - one of the surface of the painting, the other of the depicted content - is left uncharacterised. In effect, the two experiences are to be identified with their respective objects, (i) the patterned surface, and (ii) the depicted scene. The second of these experiences, that of the depicted scene, is endured on the basis of the first experience, that of the coloured surface. This second experience is as of an absent content. How then are we to characterise it? It cannot be in terms of the face-to-face seeing of the depicted content or else it will fall into the illusionism Wollheim has been concerned to reject. So Wollheim's account is unable to provide an explanation of one of the two component experiences that twofold seeing, in its first characterisation, requires.

Wollheim's second view, as amended in Painting as an Art, now holds that there is but one experience which divides between two aspects. Michael Podro's argument supporting this amended view again makes much of the institution of painting.[5] So the theory of how we appreciate paintings diverges from, although it may rest upon, a view of what it is to understand pictures generally. It is only after the point of divergence that our study belongs to aesthetics and not to, or not only to, epistemology or the philosophy of perception and the philosophy of experience that follows from it; and it is only after this point of divergence that peculiarly aesthetic considerations mark out this form of representation from other representational practices within Visual Culture taken more broadly.

Podro claims that there is a symmetry between the representational content and the seen surface.[6] This symmetry holds between the look of the objects depicted and the look of the surface; not the merely brute fact of the marked surface, but the appearance of the surface. So when we see the head of a woman in a painting we see the look of the woman and at the same time, as part of the same experience, we see the look of the surface. What is the relevant look of the surface? According to Podro, we see the surface as an object brought about as the result of a painting procedure. So it is not just that there is a surface, an array of colour patches. Rather, it is that such a surface can be seen as the culmination of an activity. That is, we are required, in order to see the surface as the surface of a painting, to consider the achievement of the artist in contriving such a surface. The activity of painting is integral to our experience of the painting as a painting.

Before we press on with the nature of looking at representational paintings, however, it is important to return to the basic conception of pictorial understanding. The artist exploits twofoldness in his fabrication of a painting, but, according to Wollheim, twofoldness is in evidence in all pictorial understanding and so prefigures the development of pictures into aesthetic representations.

In characterising the nature of this twofoldness, Wollheim wants to analyse pictures in terms of seeing-in rather than in terms of seeing-as.^[7] Wollheim's three arguments for the seeing-in analysis are as follows:

(A) There is a difference of the range of the things that can be seen-in as opposed to seen-as. Given that what I see is a particular, then what I see it as must also be a particular; but I can see states of affairs in particulars.

(B) If I see a as x, then I must be able to say which bits of a (up to the whole) that I see as x. Wollheim calls this the localisation thesis. This

constraint cannot be put upon seeing-in. If I see a cardboard shape as a man, then I must be able to say which bit of the shape (up to the whole) that I see-as his head; but I could not be asked to localise the degradation of the young man that I see-in the paintings by William Hogarth called The Rake's Progress.

(C) If I see a as x, then my attention switches between a and x. Seeing x in a, I am allowed to attend to both features of a and features of x at one and the same time.

Wollheim thus secures 'seeing-in' over 'seeing-as'. However, even if we accept (A) we can still reject (B). For, provided we can see particular parts of the surface as depicted particulars, then we can account for our experiences as of states of affairs in pictures in that such states are constituted by depicted particulars. A spectator imagines that the storm is gathering in that she sees some combinations of marks on the picture surface as clouds and others as bent trees; and on this basis she has an experience as of the storm approaching and the wind blowing. This seems to be enough to reject (B) - at least in its spirit - for whilst we have admitted in our acceptance of (A) that states of affairs are not localised, the depicted particulars in virtue of which depicted states of affairs obtain, or out of which they are composed, are localised. Since there is no particular location of the gathering of the storm, it follows that the spectator could not be expected to localise that part of the descriptive content when looking at the surface of the picture.

However, in grasping localised depicted particulars, the spectator thereby grasps their non-localised contributions to the state of affairs under view. Hence depicted states of affairs become tied down to their constituent depicted particulars. Is this not the case with states of affairs in general? Or else a spectator would not be able to see the gathering of the storm in the natural landscape. (Seeing states of affairs in the world seem equally superveinient upon particulars. A spectator has to learn that the growing gusts of wind and the darkening cloud are indicative of an approaching storm; she learns to connect the particulars in this further interpretative act of seeing the approaching storm.)

A difficulty arises here, to which we must return. It has been thought that paintings can represent objects of a certain kind, without representing any particular object of that kind. A painting can, it is supposed, represent a man, but no particular man. This is, of course, threatening to the view being put forward here. We shall, in die course, connect this difficulty to a more general difficulty in the theory of perception.

Accepting (A) but rejecting (B) we lose our grip on the force of (C). Why must her attention switch between a and x when she sees a as x? Why can she not attend to both features of a and x at one and the same time? Rejecting (B) permits us to reject (C). We shall return to this below.

The later version of the twofold thesis has one experience with two aspects, the one recognitional, the other configurational. This single complex experience is now to account for the nature of seeing pictures. We have but a single experience in front of the picture, but this experience is an amalgam of the two aspects, neither of which can be identified independently of the other. That is, neither aspect - since it is a constitutive aspect of a single complex experience - can be regarded as commensurate with, respectively, an experience of the depicted scene or the experience provided by the face-to-face seeing of the patterned surface. (This is what the introduction of the double aspect view was meant to secure over the double experience view). Budd now argues this version inherits problems of its predecessor. For if, in the complex double aspect experience, we are unable to describe either of its aspects in isolation from the other, how then are we to identify the recognitional aspect? The price to be paid for gluing the two aspects together in a single indivisible experience is that we are no longer in a position to account for either aspect in terms of the experience of its face-to-face counterpart. We cannot describe the picture in terms of its depicted content since this is to ignore the surface as that aspect of the experience intrudes upon the complex experience under view.

Further, the problem is exacerbated, according to Budd, by Wollheim's tolerance of the relative weights which attach to either of the two aspects of the experience. According to Wollheim either aspect can assume prominence and the recessive aspect can be thinned to the point of evaporation. When this occurs seeing-in is replaced by a different kind of experience; presumably a simple experience with non-aspectual content. It is at the point at which either aspect vanishes that Budd believes the view is shown to be unsustainable. For when the configurational aspect is attenuated to this point we are left with an experience as of face-to-face seeing the depicted scene. If the receding aspect has not transformed the remaining recognitional aspect into something of a different kind, then the remaining experience is as of seeing face-to-face the depicted scene. If this is an experience, then it looks as if the aspect from which it emerged was also an experience, and so we have reverted to the former account of twofold seeing, with all its attendant problems.

Now consider the recession of the recognitional aspect to the point at which it disappears. What then are we left with? The incommensurability of the configurational aspect with the seeing of the surface face-to-face, leaves us with no conception of what we are looking at when the recognitional aspect has disappeared. As Budd puts it,

The face-to-face experience is one of seeing a flat differentiated surface without seeing anything in it... [I]t must consist in the visual awareness of a two-dimensional coloured expanse...If the phenomenology of the configurational aspect of seeing-in is incommensurate with the phenomenology of such a face-to-face experience, the configurational aspect cannot be the visual awareness of a two-dimensional coloured expanse. But it cannot involve an awareness of depth, which is confined to the recognitional aspect of the seeing-in experience. Hence it consists of nothing at all.^[8]

III

These problems loom large for the twofold thesis. In part, however, I believe it is because Wollheim has wrongly characterised the recognitional aspect, and in so doing has distorted the nature of the configurational aspect.

Wollheim has maintained that our appreciation of pictures in each fold is perceptual. I think that this is the source of a major problem for his account. 'Perceiving' is the present participle of an achievement verb, just as is 'remembering'. If a perceives x, or remembers y, then x or y are related to a in some appropriate causal way, which in turn means that x exists or that y occurred. However, seeing a picture as depicting unicorns cuts off that part of the experience from perception. A spectator cannot perceive unicorns because it is not possible for unicorns to stand in the required relationship with the perceiver. It is at this point, pace Wollheim, that we need to introduce the notion of imaginative seeing. In consideration of 'seeing-as' and 'seeing-in', 'seeing-as' seems initially preferable as the direct object of perception is the picture surface. What it is ‘seen-as’ has yet to be accounted for but in preferring 'seeing-in', Wollheim commits himself to the object of perception being something that on occasion just could not be seen.

Let us, then, turn to our conception of seeing and consider how that notion is to be used when reflecting upon the nature of looking at pictures. Seeing is a perceptual state. When asked to describe the processes involved in seeing, we would have to draw upon both human biology and physics. We would expect to ground our account in some causal story that knits together the behaviour of

light and our sensitivity to it. Our seeing objects in the world, that is, is made available to us (at least in part) because external objects absorb and reflect light of

different wavelengths. We are sensitive to these different wavelengths and our perceptual ability to see those external objects depends upon the activity of reflected light upon our optical apparatus. Seeing objects in the world is founded upon our capacity to discern their shape and colour. An account of the perceptual state of seeing, therefore, will set out to explain what mechanisms are ordinarily put to work by our nervous system and what standard external conditions must prevail for our optical apparatus to apprehend objects in our environment successfully.

In order to provide an account of our access to the world beyond we need

some such externalist account of perception. At its weakest (and that is all that is

required for present purposes) we shall say that features of our environment play

some appropriate causal role in our perception of that environment. Successful

episodes of perception, that is, require features of the environment to show up in those episodes and for those features to be causally responsible (in some appropriately specified way) for the perceptual episodes under consideration. Perception, that is, provides access to features of a subject's environment.

However, we use the verb 'to see' in a variety of ways and in ways which are at odds with this account. On occasion someone might see a face in the clouds, a unicorn in a painting, some monstrous green slime in her dream, and she might start when she sees a strange old man in the shadows (which turns out to be a pile of clothes bundled together for collection by the rag-man). She may picture, in a daydream, the anxious look on her lover's face as he waited for her response to his proposal.

In our primary attempt to account for seeing, however, we were concerned with providing an explanation for cases where the spectator actually sees what is in her environment. In these deviant cases, where she 'sees' a face in the clouds, a unicorn, an animated green slime, a stranger in the shadows and her lover's face, what she 'sees' is not really there to be seen. So here we have a class of cases that we might label 'non-veridical seeing'. (There are distinctions to be made between these various examples. For present purposes, however, we shall look at those relevant to our task in focusing upon our appreciation of pictures.)

Seeing the man in the clothes pile is just a mistake. The subject's perceptual state, given by the kind of account put forward in normal circumstances, is such as to have led her to the false judgement that what she is seeing is an old man. Her experience was as of an old man but what she saw was a bundle of clothes. Closer inspection reveals that she had mistaken the bundle for a stranger. 'I thought I saw an old man. What in fact I saw was that heap of clothes', replaces the original perceptual report. Correcting the original report will rely upon there being other evidence, either perceptual (looking closer) or non-perceptual (relying upon the reports of others or weighing up the possibilities - 'there surely couldn't be an old man there, it must be a play of light'), which provides an alternative account for the experience undergone. Such evidence, in the case of perceptual mistakes, is likely to rely upon there being more detailed observation grounding the perceptual report which thereby secures the attendant veridical experience. Further observation thereby rules out the mistaken attribution. It is not that we need further evidence upon which to base an inference. Rather, in the case of possible mistakes, various experiential content is commensurate with the visual phenomena present in the perceptual state (i.e. shape, colour, distance etc.). In the case of the mistake what the subject sees supports a variety of experiential content, none of which is inferred from the perceptual state, but each of which may be represented to the subject immediately. Further observation allows the subject to amend her experience in light of new evidence. ('The bundle of clothes in the corner, in that light, just looked as would an old man', for instance. Here the subject does not renounce what was first 'seen' at the level below content attribution.)

It may be useful here to distinguish between perceptual presentation and perceptual representation:

In having perceptual experience the world seems to us to be a certain way; it presents itself to our experience as containing various objects and properties. Experience, we may say, represents the states of affairs so presented (or apparently presented): perceptual representation is the converse of perceptual presentation. The way in which experience represents the world constitutes its content, the way it makes things seem. The content of an experience determines what it is as of - how the world would actually be presented if the experience were veridical. ^[9]

In the case of seeing the rags and mistakenly identifying them with an old man the spectator is presented with perceptual content which might best be described in terms of volumes, sizes, colours, distances and lighting conditions. To the extent that the spectator will agree to these descriptions her perceptual content is veridical. Only when, on the basis of what she sees, she judges her perception to be of an old man does she move from veridical perception to non-veridical experience. Her experience is as of an old man; an old man is represented by her in her experience. This is not to claim that in judging that she sees an old man she makes an inference. She simply mistakes what she sees. Her seeing is a complex matter. In one sense what she sees is presented to her (her seeing is veridical) but in judging the object of her perception to be an old man she makes a mistake. There is something over there, and that something is what she sees; but she does not see an old man even though an old man is represented in her experience.

We return here to the threat encountered when considering depicted states of affairs. It was admitted that depicted states of affairs depended for their apprehension depicted particulars. Paintings, we are told, are able to represent kinds of object, but no object in particular of the kind depicted. What of the old man in the corner. Experience of the old man in the corner is not experience of an old man. (It's just a bundle of rags in the corner.) However, the experience the spectator undergoes is as of a particular old man: that old man in the corner. So too with painting. A painting cannot represent a particular unicorn, since that kind of object has no particulars. However, the experience the spectator has is as of a particular unicorn: that unicorn under the pale moon gazing at the virgin.

What might we learn from our mistakes? The phenomenology of mistaking the rags for the man is unlike that of 'seeing' the pictorial content in the picture in front of us. We do not, as we have seen, mistake a picture of Marilyn Monroe with Miss Monroe, herself. However, it would be hasty to leave the example there. For the possibility of mistaking a picture for its content, whilst at odds with the experience of looking at a picture, ought not to be ignored. It is this possibility that needs explanation if we are to do justice to the kind of experience engendered when looking at some, if not all, pictures.

We might now return to the bundle of rags. Suppose that the spectator, having corrected her mistaken experience, now recognises that these are the rags she had previously piled ready for collection. Can she not now re-create the visual part of her experience by attending to features of the rags and seeing them as an old man in the corner? It seems plausible that she can. But here she would have to mask out certain features that have now shown up in the corrected perception. Perhaps in attempting to mask out certain features, she has to give undue attention to other features in her perceptual state. So, here she fastens features of the willed experience as of the old man to parts of her perceptual field, whilst attempting to pay no heed to other features - particularly those features which provided evidence for the corrected veridical experience. The difference between her initial mistaken experience and this later contrived experience is that this latter is detached from any functional engagement with her environment. (It does not form the basis for a perceptual judgement.) It is known by the agent to be non-veridical and, further, it is not an illusion that the agent undergoes. Unlike the original mistaken experience, this newly contrived experience is imaginative. It is created by the spectator to fit the features of her perceived environment. In her mistaken perception, she had judged that a particular man was in the corner. Now she imagines a man in the corner. Which man? That one there in the corner with these features (the same features as 'the man' in her mistaken perception').

No-one has ever seen a unicorn. However, when a spectator 'sees' a unicorn in a picture, she does so because of certain lighting conditions. It is because the picture surface is illuminated that she sees it. The standard behaviour of light as it is variously absorbed and reflected by the picture surface enters into an account of what it is for her to have an experience as of the unicorn. That account will include her normal sensitive reception of the picture surface. Others too can 'see' a unicorn in the particular picture at which they look. Moreover, when each reports, 'It's a unicorn' no one is making a mistake. Here, with this particular example of a non-veridical perception we return to our problem.

From the discussion thus far, however, we might now begin to consider some basic truths about the spectator's appreciation of the picture in front of her. She sees the picture surface in virtue of its being illuminated and in virtue of its variety of marks. On the basis of seeing the variously coloured surface she has an experience as of the unicorn in the picture. We are now in a position to note that her seeing the surface of the picture relies upon a straightforward perceptual account of light reflecting from the surface and exciting her nervous system by way of her optical apparatus. The unicorn, however, is not so illuminated. If the unicorn is lit dramatically from the right, say by a strong moon in an otherwise darkened sky, this will be contained in the report of what is experienced in the picture even if the subject is looking at the paint surface lit by a fluorescent light source from the left. That is, the illumination within the painting is independent of the illumination of the painted surface. The contents of pictures, that is, are not illuminated at all. (The illumination, being part of the content, cannot be specified independently of that content and so cannot be called upon to account for what is seen - any more than the appearance of green slime in a dream can be explained by its ability to absorb and reflect light of certain wavelengths. For the dream slime neither reflects nor absorbs light of any wavelength.)

Consider a subject sitting in her study looking at a series of etchings, say Rembrandt's Crucifixion series. The intense light that bathes the stricken Christ is seen in the picture. But the white of the paper that contrasts with the aquatint to give that effect is nothing other than creamy paper reflecting the forty watt light-bulb of her angle-poise lamp. Here we might remember the adage of Hans Hoffmann, quoted by Wollheim, that 'in nature light creates colour but in painting colour creates light'. We want to say that the strange light within some paintings, no matter how dull the condition of light by which their surfaces are illuminated, is created in the mind of the viewing subject. To adapt an aphorism of Wittgenstein we might say, 'It is as if an image came into contact, and for the time that we attend to the pictorial content remains in contact, with our visual impression of the picture surface.' Depicted content, as we have seen, like dreams, hallucinations and visualisings, requires no source of light for its comprehension. Unlike dreams, hallucinations and visualisings, it requires for its substratum a coloured surface which does require illumination. So our understanding of depiction lies somewhere along the line of intersection between perception and imagination.

If what has been said thus far is plausible, then there are consequences for the aesthetics of the visual arts. In this conclusion I gesture toward them.

With what are we to replace 'seeing-in'? So far, it might seem, I have been advocating a return to 'seeing-as'; but I prefer the locution 'seeing-under', where what is seen, the coloured surface, falls under an imaginative description, itself in part constitutive of the experience endured. My preference here, for this terminology, is explained by the range of description available to us when speaking of pictures and their content. Moreover, if such a neutral view of pictorial comprehension can be given, then a supplementary account of pictorial purposes promises to provide the necessary distinctions to be made between our responses to pictures of different kinds. The proposal is that we regard pictorial seeing in terms of the perception of a physical object (the picture) together with the wilful attribution of imaginative descriptions to that object. Hence the favoured terminology is 'seeing-under'. However, at times it may seem more natural to use 'seeing x in a', at others 'seeing a as x'. Provided that we remain aware that the primary object of perception is the picture and that the imaginative seeing of the pictorial content is given under a description, these two locutions should remain harmless enough.

So far, then, I can see a picture and my experience of that picture can be captured by such divergent descriptions as, 'It's a man', 'It's the misery of war', 'It's the moment at Emmaus at which we realise, along with the apostles, that we are in the company of Christ'. To the extent that we are able to assent to these descriptions, we are able to see the picture in front of us. We see the picture under these descriptions. These three examples variously describe a particular, a state of affairs and an event.

It is with our rejection of (C), however, that we can come to see the significance of our purposes with regard to our aesthetic appreciation of paintings. For once we abandon the claim that each content description needs to be localised in the painted surface, then the claim that 'seeing a as x' necessitates a switching of attention between a and x becomes a mere stipulation. Here we might feel the need to focus upon a feature of our attention to pictures as it might vary upon different occasions. Sometimes, when I look at a picture, my attentive thought (and hence experience) is wholly occupied with content. At others, say when I am cleaning a patch of a painting, my thought is occupied solely with the patterned surface.

Let us remind ourselves that in looking at pictures we enter into an imaginative relation with a physical object - the picture surface. Being imaginative, the experience undergone is subject to the will. Seeing the picture as thus and so is part of our project. Hence our experience can 'filter in' and 'filter out' features of the surface as they fit with our project. On occasion, when looking at the sports pages, I can look at a footballer in the execution of some contorted move without paying heed to the surface of the picture - without noticing that it is made up of tiny black dots on a creamy paper. My project, in looking at the photograph is to recapture a moment of last night's game. Features of the picture as an object do not enter into the descriptions that complete my project. The question of the transparency of pictures, it seems to me, arises not out of the logic of pictures, but rather out of the nature of our purposes in looking at them. (Since the nature of pictorial seeing, on my view, is wholly imaginative, it follows that we need to look at the nature of our imaginative practices to answer all sorts of questions that have usually found themselves asked in terms of the structure of pictures.)

Sometimes, the imaginative description of the perceived surface will mask out features of the surface and at other times it will not. The consumer of pictorial pornography has as his purpose sexual arousal and its eventual gratification. The imaginative experience he seeks contains no description of the pictorial surface but only of its content. To the extent that the perceived surface shows up in his experience, it is a hindrance to the fulfilment of his project. In this instance the nature of that project provides the basis upon which he directs his willed experience. He sees the picture and thinks of it thus. This thinking is a wilful form of seeing. It is experiential. By contrast the builder looking at the architect's plan of a kitchen layout is exercised by the requirement to place taps, drains and water pipes in places which are rationally determined. On looking at the plan he can see solid wall and cavity and can interpret according to features of the drawing that show up in his experience. That is, the experience the builder has is determined by him under a different mode from that which determines the consumer of pictorial pornography. The builder, that is, makes cross references between drawing and kitchen and notes correspondences between the two. His drawing enables him to interpret features of the space in which he stands. These two examples show that there are different purposes with which we look at representations. In each case the purpose is pursued by imaginatively construing the picture surface under some description appropriate to the spectator's project.

Looking at paintings engages our understanding of the making of the picture as an art. Our pleasure in our appreciation of painting emerges from our seeing the surface of the painting as an object manufactured by an artist in such a way that we have experiences as of the content. We have the experience as of the content at one and the same time as we perceive the picture surface. That experience is fastened to the perception of the surface. Unlike the spectator's consumption of pictorial pornography, where the picture surface is masked out of experience, the picture surface of representational painting is the object of attention of the spectator. The experience is akin to a mental image which suffuses the perception of the picture surface. So I have but one perception; and upon the substratum of this perception an imaginative experience supervenes.

If what has been written here has plausibility, it follows that the enterprise of making pictures within the medium of paint exploits an aesthetic dimension in terms which will require more than the content analysis offered by Visual Cultural Studies. Painting, too, will be imunised from some of the criticisms that ignore the fact of the painted surface as a constituent of the aesthetic significance of painting.

I would like to add, at this point, as an afterword, that as a practising painter, I have taken nothing but encouragement from the many writings on painting as an art by Richard Wollheim.

[1] See Richard Wollheim, Painting as an Art, (London: Thames and Hudson, 1987), p.360n.

[2] E. H. Gombrich, Art and Illusion, (2nd edition) (London: Phaidon, 1962)

[4] Richard Wollheim, Art and its Objects, second edition, (Cambridge University Press, 1980), p.224.

[5] See Michael Podro, 'Depiction and the Golden Calf', in Norman Bryson, Michael Ann Holly and Keith Moxey, (eds) Visual Theory, (Oxford: Polity Press, 1991).

^[7] Richard Wollheim, Art and Its Objects, op. cit. p.224.

^[8] Malcolm Budd, 'On Looking at a Picture' in Jim Hopkins and Anthony Saville (eds.), Psychoanalysis, Mind and Art: perspectives on Richard Wollheim, Aristotelian Society Series, volume 11, (Oxford: Basil Blackwell, 1992), p.272.

^[9] Colin McGinn, Mental Content, (Oxford: Basil Blackwell, 1989) p.58.