If we wanted to say something about art that we could be quite certain was true, we might settle for the assertion that art is intentional. And by this we would mean that art is something we do, that works of art are things that human beings make. And the truth of this assertion is in no way challenged by such discoveries, some long known, others freshly brought to light, as that we cannot produce a work of art to order, that improvisation has its place in the making of a work of art, that the artist is not necessarily the best interpreter of his work, that the spectator too has a legitimate role to play in the organization of what he perceives. [. . .]
Though much is unclear about the notion of activity, one thing seems clear. From the fact that art is something that we do, it follows that in the making of art a concept enters into, and plays a crucial role in, the determination of what is made: or, to put it another way, that when we make a work of art, we make it under a certain description – though, of course, unless our attention is drawn to the question, we may not be in a position to give the description. Indeed [. . .] the truth seems to me to be that, on any given occasion, or in the case of any given work, there will be more than one concept involved, and the concepts that are involved will form some kind of hierarchy, with some concepts falling under others, either more or less organized. In the existence of such a conceptual hierarchy, regulative in the production of works of art, we find, I maintain, the justification for talking of a theory of art. In this lecture I want to look at the matter rather more generally, and I shall ask you to consider with me a theory of modern art. The theory I have in mind could not be called the theory of modern art, for there is evidently no such thing, but I would claim that it is the dominant theory. [. . .]
My suggestion is this: that for the mainstream of modern art, we can postulate a theory that emphasizes the material character of art, a theory according to which a work of art is importantly or significantly, and not just peripherally, a physical object. Such a theory, I am suggesting, underlies or regulates much of the art activity of our age, and it is it that accounts for many of the triumphs and perhaps not a few of the disasters of modern art. Within the concept of art under which most of the finest, certainly most of the boldest, works of our age have been made, the connotation of physicality moves to the fore.
The evidence for such a theory at work is manifold, the inspiration of the theory can be seen in a wide variety of phenomena which it thereby unifies: the increasing emphasis upon texture and surface qualities; the abandonment of linear perspective, at any rate as providing an overall grid within which the picture can be organized; the predilection for large areas of undifferentiated or barely fluctuating colour, the indifference to figuration; the exploitation of the edge, of the shaped or moulded support, of the unprimed canvas; and the physical juxtaposition of disparate or borrowed elements, sometimes stuck on, sometimes freestanding, to the central body of the work, as in collage or assemblages. These devices have, beyond a shadow of a doubt, contributed decisively to the repertoire of European art since, say, 1905: and any dispute about the presence of some such theory as I have produced would, I imagine, confine itself to the issue of how central these devices, and the modifications in art that they have brought about, are thought to be.
I want therefore to turn away from any central discussion of the theory to the qualifications that need to be entered if the theory, or the formulation of it, is to be adequate. I shall bring what I have to say under three general considerations. But in doing so, I shall give the theory itself a small twist towards greater specificity. I shall consider it exclusively in relation to painting, and I shall understand it as insisting upon the surface of a painting. In the context of a painting, for ‘physicality’ read ‘possession of a surface’.
The first consideration, is this: The theory that I have been suggesting emphasizes or insists upon the physicality of the work of art, or the surface of the painting – emphasizes or insists upon them, but (this is the point) the theory did not discover or invent them. I am not, of course, making the self-evident point that even before 1905 paintings had surfaces. I am making the somewhat less evident point that before 1905 the fact that a painting had a surface, or the more general fact that works of art were physical, were not regarded as accidental or contingent facts about art.
For to read certain critics, certain philosophers of art, even certain contemporary artists, one might well think that before the beginning of the twentieth century the concept of art was totally without any connotation of materiality. Of course – it is conceded – in making their pictures earlier artists recognized that they were making physical objects. But for them the picture and the physical object were not equated, and the manipulation of the medium was seen more as a preliminary to the process of making art rather than as that process itself. For the picture was conceived of as something immaterial that burgeoned or billowed out from the canvas, panel or frescoed wall that provided its substrate. These things – canvas, panel, wall – were necessary for its existence, but it went beyond them, and the concept of it bore no reference to them.
Such a view of the past, which is artificially sustained by the very careless and utterly misleading use of the term ‘illusionism’ to characterize all forms of figurative painting, indeed all forms of representational painting, seems supported neither by empirical nor by theoretical considerations. Furthermore, there is a body of evidence against it. There are various moments in the history of European painting since the high Renaissance, when artists have shown a clear predilection for the values of surface, and they have employed selected means to bring out the physical quality of what they were working on or with. Take for instance, the emergence of the brush-stroke as an identifiable pictorial element in sixteenth-century Venetian painting; the free sketching in of landscapes in the background of seventeenth- and eighteenth-century painting; or the distinctive use of cropped figures set up against the tdg^ of the support in late nineteenth-century Parisian art. Now, of course, there is this difference: that to the earlier painters these devices were no more than a possible employment of painting, and for them the constraints of art lay elsewhere. It was, for instance, optional for Velazquez or for Gainsborough whether they expressed their predilection for the medium. What was necessary within their theory of art was that, if they did, it found expression within the depiction of natural phenomena. For, say, Matisse or Rothko, the priorities are reversed. But none of this suggests that the earlier painters thought that what they were doing was ancillary to painting. Between them and us what has happened is that some connotations of art that were previously recessive have moved to the fore, and vice versa.
The second consideration that touches upon the theory of modern art I have proposed seems to strike somewhat deeper. It is this: The theory emphasizes the physicality of art; it insists upon the fact that a painting has a surface. Indeed, by a ready trick of exaggeration the insisted-upon fact that a painting has a surface – a fact which, as we have seen, earlier generations did not overlook – can convert itself into the thesis that a painting is, or is no more than, a surface. Which gives us an extreme version of the theory, though not one unknown. However, a necessary modification is effected – once it is recognized that, in talking of a surface, the theory is irreducibly or ineliminably referring to the surface of a painting. In formulating the theory we can safely drop the phrase ‘of a painting’ only when the dropped phrase is understood. If the phrase is not merely dropped, but drops out of mind, the theory becomes incoherent. For to talk of a surface, without specifying what kind of surface it is, which means in effect what it is a surface of, picks out no kind of object of attention.
Perhaps one way of bringing out this consideration is to go back to the nature of the theory. For the theory, as we have seen, provides us, if it is adequate, with those concepts under which a certain form of art – the art of our day -has been produced. However, it is clear that no one could set himself to produce a surface, unless he had some answer to the question what it was the surface of: nor could he endeavour to accentuate or work up a surface unless, once again, he thought of himself as accentuating or working up the surface of this or that kind of thing. To put it another way round: the instruction ‘Make us aware or conscious of the surface’ given in the studio, would take on quite different significances, if said, to someone throwing a stoneware pot, to someone painting in oil on primed canvas, to someone carving in marble, or to someone Working in fresco. (Think, for instance, whether, in conformity to the instruction, the surface should be made smooth or rough.) Each of the recipients of the instruction would, in effect, fill it out from his knowledge of what he was doing, before he obeyed it. And, if he didn’t know what he was doing, if, for instance, he was a complete beginner who hadn’t as yet grasped the nature of the activity on which he had launched himself, he could not obey the instruction at all. It wouldn’t be, simply, that he wouldn’t know how to do what he had been asked to do: he wouldn’t know what he had been asked to do. For him the instruction would mean about as much as ‘Make it average-sized’.
The examples I have given might be misleading in one respect: for they might suggest that the further specification that is required before it becomes clear how the surface is to be worked refers exclusively to the material. The artist, in other words, needs to know what the surface is of just in the sense of what it is made of. This, however, would be erroneous. Not merely is this no more than part of what is required, but it is misleading even as to that. The distinction we need here is that between a material and a medium. A medium may embrace a material, though it may not, but, if it does, a medium is a material worked in a characteristic way, and the characteristics of the way can be understood only in the context of the art within which the medium arises. 1 ‘Fidelity to material’ is not so much an inadequate aesthetic, as some have thought, it is rather an inadequate formulation of an aesthetic: Bernini and Rodin were not faithful to marble, though they may have been faithful to marble as a material of sculpture. Given this distinction, the barest specification of the surface must be by reference to the medium which makes up or is laid on the surface. [. . . ]
The third consideration that I want to raise in connection with the theory of modern art is this: The theory insists upon the physicality of the work of art
– upon for instance, the surface of the painting. And this I have equally put by saying that the theory insists upon the fact that the painting has a surface. But from this it does not follow that a painting produced in conformity with this theory will insist upon the fact that it has a surface. Yet this is sometimes thought to follow both by critics and by artists: and consequential distortions are produced both in criticism – so that for instance, it is thought good enough to say of a painting that it insists upon the fact of its surface – and in art itself
– so that we are confronted by objects which seek to acquire value from this insistence. [. . .]
Perhaps the best way of bringing out this consideration is to show how the theory, as it stands, can lead to anything but boring art. For the theory, in asserting that a painting has a surface, draws the painter’s attention to the surface in a way or to a degree not contemplated by his predecessors. But if the theory were that the painting should assert that it has a surface, then not merely would no premium be placed on the use of the surface but the effect of the theory might well be to work against the use of the surface. For it might be felt that any such use would only interfere with the clarity or the definitive-ness of the painting’s assertion. The fact of the surface might become eclipsed, wholly or partially, by the use of the surface. And in an aesthetic situation where the fact of the surface is reckoned the important thing, the use of the surface begins to look diversionary at best and probably hazardous.
To talk of the use of the surface and to contrast this with the fact of the surface, and to identify the former rather than the latter as the characteristic preoccupation of modern art, attributes to modern art a complexity of concern that it cannot renounce. For it is only if we assume such a complexity that there is any sense in which we can think of the surface as being used. Used, we must ask, for what? And the answer to this has to lie in that complexity of concern. – The point, I must emphasize, would not be worth making were it not for the widespread confusion which equates the autonomy of modern art with its single-mindedness, even with its simple-mindedness. To talk of the autonomy of art is to say something about where its concerns derive from, it is to say nothing about their number or their variety.
To talk of the surface being used, rather than of its existence being asserted, as a characteristic of modern painting, is not a point to make in the abstract. The point cannot be grasped without some kind of incursion into the substantive issue. What therefore I should like to do for the rest of this lecture is to consider three paintings, and try to make the point in relation to them. It is no accident that these paintings are amongst the masterpieces of twentieth-century art.
The first painting is La Fenetre Ouverte painted by Matisse in Tangiers in 1913. Some of the things that I shall say about it will apply to the other great open-window paintings of Matisse, for instance the sombre La Porte-fenetre of 1914 or the painting entitled La Fenetre, or Le Rideau Jaune. In this painting we discern, amongst other things, Matisse’s recurrent concern with the nature of the ground. Now, if we consider what is not so much the earliest painting we have, though it is often called that, as the precursor of painting – I refer to the cave art of the early Stone Age – there is no ground, there is simply the image. 2 With the introduction of the ground, the problem arises, How are we to conceive of the ground in a way that does not simply equate it with the gap between the figures or the absence of depiction? In the history of European painting we can see various answers to this question. One answer is for the painter to equate the ground with the background or, if this term is taken broadly enough, with the landscape, and then to organize the detail that this equation is likely to impose upon him in a hierarchical fashion, detail subsumed within detail, in a Chinese box-like fashion. This answer we can see as given in some of the finest achievements of European art – for instance, in such different kinds of work as the masterpieces of Van Eyck and Poussin. Another answer is to regard the ground as providing, still through representation, not so much content additional to the central figures, but a space in which the central figures are framed. Now for a variety of reasons neither of these two classic answers is open to Matisse. For Matisse – and here he exhibits two of the main thrusts of twentieth-century art – dispenses both with the notion of detail in the traditional sense and also with the commitment to a unitary and ordered spatial framework. And so the question returns. How is the ground to be conceived of except in purely negative terms? How – which is an extension of this question – is the frontier of the ground, of the line which encloses it, not to seem quite arbitrary? And it is at this point, to find an answer to this question, that Matisse resorts to the surface. It is here that he uses the surface. For what he does is to associate the ground so closely with the surface – by which I mean that he charges the surface in such a way that it barely involves a shift of attention for us to move from seeing a certain expanse as ground to seeing it as surface – that we fully accept the size of the surface as determining the extent of the ground. To understand Matisse’s use of the surface, we might say that through it he reconciles us to the ground without our hankering after any of the classic ways of treating the ground that Matisse has forsworn.
There is perhaps another line of thought in La Fenitre Ouverte which is worth pursuing. What we see through the open window is a view. Now, at any rate for a painter there is, perhaps, a certain absurdity in thinking of a view as a view on to – well, a view on to nothing, which is rather what treating what is depicted through the frame of the window as mere ground implies. The view is – of course in a rather special, one might say in a rather professional, sense – an object. And now we can see a secondary use that Matisse makes of the surface. For by emphasizing the surface where it coincides with the ground, he leads us towards this painter’s way of looking at or considering the view. Of course, Matisse isn’t a clumsy painter, and he avoids that use of the surface which would make it look as though the open window were filled with a solid object. It is made clear, at one and the same time, that the view isn’t an object but that it is as though it were.
The second painting I want to consider is one of Morris Louis’s later canvases [Alpha Phi, 1961]. Louis’s work at this stage was largely dominated by one preoccupation – apart, that is, from his interest in the physical look of the picture or how the surface looks. And this preoccupation can be described from two different points of view. From one point of view, it is a concern with colour: from another point of view it is a concern with patches – where patches are contrasted both with volumes, which are three-dimensional, and with shapes, which, though two-dimensional, are seen as suspended in, or visibly inhabit, three-dimensional space. Louis, in other words, wanted to introduce colour into the content of his paintings but to as great a degree as is humanly possible -or, better, visibly possible – he wanted consideration of the spatial relations between the coloured elements, or the bearers of colour, to recede. Now, I do not think that it is quite correct to say – as Michael Fried does in his otherwise perceptive account of these paintings 3 – that Louis’s patches are non-representational: that is to say, I do not think that Louis wants us exclusively to see stained parts of the canvas. He seeks a form of representation where the representation of space or of anything spatial is at a minimum. And to achieve this effect, he uses the surface in such a way that so long as we look centrally at one of the patches we see it representationally. But, as our eyes move towards the edge of the patch, the representational element diminishes, and we become dominantly, then exclusively, aware of the canvas. In other words, representation gets negated at the very point where questions of spatiality – how does this patch stand to the next? – would begin to arise. The overall effect is that, in looking at Louis’s patches, we seem aware of them as though they were embedded in, or pressed down upon, the surface – an effect, which, incidentally, we find, in a highly figurative context, in some of Goya’s paintings. The surface, then, is used to control or to limit the operation of representation, so that colour can be encountered in what we might call a ‘pure’ mode: as predicated of extended but non-spatial elements.
The third painting that I want to consider is one of Rothko’s canvases from the Four Seasons series, now hanging in the Tate [Red on Maroon, 1969]: to my mind, one of the sublimest creations of our time. In comparison with Louis, even with Matisse, Rothko uses the surface in a highly complex way. And I shall only give one hint of how we might think of this. The greatness of Rothko’s painting lies ultimately, I am quite sure, in its expressive quality, and if we wanted to characterize this quality – it would be a crude characterization – we would talk of a form of suffering and of sorrow, and somehow barely or fragilely contained. We would talk perhaps of some sentiment akin to that expressed in Shakespeare’s Tempest ~ I don’t mean expressed in any one character, but in the play itself. However, the immediacy of Rothko’s canvas derives from the way in which this expressive quality is provided with a formal counterpart: and that lies in the uncertainty that the painting is calculated to produce, whether we are to see the painting as containing an image within it or whether we are to see the painting as itself an image. Whether we are to see it as containing a ring of flame or shadow – I owe this description of the fugitive image to the brilliant description of the Four Seasons paintings by Michel Butor in his essay ‘Rothko: The Mosques of New York’ 4 – or whether we are to look upon it as somewhat the equivalent of a stained glass window. 5
Now, it is to bring about this uncertainty, as well as to preserve it from, or to prevent it from degenerating into, a mere oscillation of perception, which could, if I am right, be highly inimical to Rothko’s expressive purpose, that he uses the surface as he does. For the use of the surface, or the way it manifests itself to us, simultaneously suggests forms within the painting and imposes unity across the painting. It suggests light falling upon objects and light shining through a translucent plane. Wherever a definitive reading begins to form itself, the assertion of surface calls this in doubt.
It is only now, when we have taken note of other, or more working, aspects of the theory of modern art as I have suggested it, that it seems to me appropriate to observe an aspect that might have seemed to some worthy of earlier attention: I mean the way it is likely to give rise to objects that manifest the only kind of beauty we find acceptable today.
* Stanley Cavell, Must We Mean What We Say? 1969.
On this, see Meyer Schapiro, ‘On Some Problems in the Semiotics of Visual Art: Field and
3 Vehicle in Image-Signs’, Semwtica, I, no. 3, 1969, pp. 223-42.
4 Michael Fried, Three American Painters, Boston, 1965, pp. 19-20.
5 Michael Butor, Inventory, trans. Richard Howard, London, 1970.
The original version of this paper was published in Studio International, vol. 180, no. 928, London, December 1970.