PRINT November 1965

Jules Olitski’s New Paintings

The human body is the best picture of the human soul.

JULES OLITSKI’S NEW PAINTINGS are executed in a soak-and-spray technique (mostly spray) of his own invention. I want to begin this essay by comparing their structure with that of paintings by Noland and Stella, both of whom in different ways have developed an unmistakably modernist structural mode I have elsewhere described as “deductive.”1 There are several reasons for starting here rather than, say, with an attempt to describe either their color or the spray technique itself. First, much of the animus behind the development of this technique has been a considered reaction against deductive structure. Second, this revulsion has stemmed largely, but not wholly, from what Olitski regards as the grave limitations deductive structure imposes upon the painter as colorist, so that to understand the relation of Olitski's paintings to deductive structure would be to understand something about both their color and their means of execution. Third, I will argue that the structural conception behind his new paintings—despite their deep and, I think it can be said, successful opposition to a deductive mode of pictorial organization—is grounded in essentially the same awareness of, and attitude toward, the framing-edge that works of Noland and Stella (the latter extrapolating from Barnett Newman) were the first to insist upon. Moreover, I will suggest that the structure of Olitski's finest recent pictures is vital to their quality and power; that however moving, daring or sensuous their color, it alone would not be enough to account for their convincingness as paintings. The implication here is that our experience of pictorial structure is crucial in determining whether or not we find a given painting convincing: not because structure is in any general sense more important than color, but because manipulations of structure have played a central role in the dialectical revisions of its own conventions by which modernist painting has periodically renewed and redefined itself, and hence renewed and redefined our conception of what ambitious painting is. Finally, I believe that an examination of the structure of Olitski's spray paintings will provide a context in which the significance of the new technique for his own aspirations as a painter becomes clarified.

The initial development of what I have called deductive structure in the work of Noland and Stella appears to have been compelled by the deep need each seems to have experienced for a new kind of pictorial organization, grounded in and lucidly evincing a more acute awareness of the shape and size of the picture-support than had been the basis of any previous painting. Cubism, of course, had manifested (though to a lesser degree) an awareness of the shape of the support, by aligning elements within a given picture to a rough congruence with the framing-edge. And in general deductive structure's insistence on acknowledging the literal character of the picture-support amounts to nothing so much as a return to and reaffirmation of Cubism's implicit but decisive interpretation of the half-century of painting between Manet's first great seminal pictures of the early 1860s and the late works of Cézanne in terms of a growing consciousness of the literal character of the picture-support and a draining of conviction in traditional illusionism.2 Cubism's interpretation of this painting consisted for the most part in its increasingly perspicuous declaration of the flatness of the support. But flatness is an essentially tactile characteristic; and the dissolution of tactility in much of the finest painting just prior to that of Noland and Stella—most notably in the work of Newman and Louis—meant that flatness was no longer something an advanced painter had to, or perhaps even could, establish positively. Even less could it be violated, however ambiguously, by means of illusionistic incursions into a fictive tactile space. Rather, the phenomenological import of Newman's and Louis' pictures—as well as, from our present vantage anyway, Pollock's 1946–50 all-over drip paintings—can be characterized as a denial of tactility in all its manifestations, including traditional illusionism: the integrity of the surface in their paintings is taken for granted at the same time as its actual texture is both subsumed and dissolved in the creation of a space addressed to eyesight alone, which itself may be thought of as establishing an unprecedented illusion. More than any other single factor, the dissolution of tactility in the work of Newman and Louis was instrumental in engendering the new and more acute awareness of the shape and size of the picture-support that has been the basis of deductive structure; because shape and size are also literal characteristics of the support, but they differ from the characteristic of flatness in this critical respect, that whatever tactile connotations they may have do not come into conflict with the exclusively visual or optical illusionism I have tried to describe.

The fact is significant that, following Newman's and Louis' dissolution of tactility, other literal characteristics of the support—its weight, for example—did not become the focus of a new and more acute awareness on the part of advanced painters. Even more significant is our own lack of surprise at the fact, our inability to conceive of modernist painting transforming itself to the degree that considerations such as weight could ever be pictorially relevant—an inability that I believe is not a failure of imagination. That is, I want to suggest that Noland and Stella could not have based their paintings on other literal characteristics of the support and still have made good paintings, or even objects that would compel us to take them seriously as paintings. And this, if true, would suggest that deductive structure represents something other than the outcome of a decision by Noland and Stella to base the structure of their paintings on the shape and size of the support: not because they did not make such a decision, but because a decision to base their paintings on any other literal characteristics of the support would have resulted in bad paintings or none at all. The sense of something having been decided for him vibrates in Noland's remark that, after making accomplished but finally derivative paintings for several years, he broke through to what he had been after all along when he “discovered the center” of the canvas, which in practice resulted in his coming to locate the central point of concentric or radiating elements at the exact center of the support. He does not speak of having decided to do this, though there may be sense in saying that he must have done so: perhaps the sense there is in saying that the center was there before it was discovered. The crucial distinction remains between one's reasons for deciding to do something and the source of one's conviction that something one has just done or learned amounts to a discovery about something or oneself or both. One is here speaking of modernist painting as a cognitive activity, an activity whose aim is knowledge; or perhaps one can say, an activity whose aim is the kind of conviction associated now only with knowledge and not to be found in religion or morality or politics. It may be this aim, the demand for this satisfaction, which distinguishes this art from the art contemporary with it. (Noland's discovery could only have been made out of his deep knowledge, verging on vicarious re-creation, of both traditional and modernist painting, together with what had been his utter refusal to rest content in his own work with less than truly significant achievement. The composer Leverkuhn in Doctor Faustus: “Art would like to stop being pretence and play, it would like to become knowledge.”) What Noland discovered when he found the center of the canvas was nothing less than how to make paintings in which he believed and by which he was moved: specifically, by rigorously and lucidly relating the elements within the painting to the shape of the support.

The size of the support, its actual dimensions, is the other characteristic already noted whose discovery and emphasis yielded a structure on which these ambitions could rely. The dimensions became important when, around 1962, Noland changed the format of his paintings from concentric rings which nowhere made contact with the limits of the support to vertical stacks of chevrons which ran off the framing-edge. The first of these paintings, in which the bottom point of one chevron was anchored to the midpoint of the bottom framing-edge while the boundaries between chevrons were allowed to intersect the limits of the canvas anywhere at all, have always seemed to me structurally inconclusive—except for a very few in which the boundary between two of the chevrons runs into the top two corners of the canvas. Whether or not Noland himself felt that these particular paintings succeeded more than the others, he appears to have moved quickly to the conclusion that within the chevron format the crucial points, equivalent to the center of the canvas for the concentric rings, were the two upper corners of the support. (Unlike the mid-point of the bottom framing-edge, which only related the chevrons to the picture-support by dint of symmetry—so to speak, at one remove—the corners amount to firsthand, immediate, physical features of the edge itself.) As long as Noland ran the outer boundary of his bottom chevron into these corners his paintings succeeded as structures: to the extent that he was able to dispense with lateral symmetry in the interests of giving his paintings a powerful sideways impulsion that, almost incredibly, belies the static implications of their deductive formats. Some of Noland's subsequent diamond-shaped paintings achieve an even more unexpected lateral thrust than the asymmetric chevrons; and in general one can, I think, describe Noland's work over the past two years as an attempt to find within the rigor and lucidity of a deductive mode of pictorial organization sufficient room for the free, spontaneous and, above all, lyric exercise of his superb and wholly distinctive coloristic gifts.

Stella on the other hand has always been more interested in structure for its own sake, virtually without regard for its coloristic implications. He was the first painter really to insist upon the framing-edge as a primary object of modernist pictorial concern, in his black stripe paintings of 1959 (toward which he had been working since the year before) and even more explicitly in his aluminum and copper paintings on shaped supports dating from 1960 and '61 respectively. In the latter two series, as in his magenta paintings with open centers of 1963, the outer edge of the support is made to generate a succession of parallel or concentric stripes until the entire surface is filled: as in Noland's paintings starting with the first cornered chevrons, the framing-edge assumes an active structural significance without precedent in modernist painting. Deductive structure activates the edge of the support much as Cubism activated its surface; but in Cubism the various elements relate two-dimensionally chiefly to one another, whereas Stella's stripes or Noland's chevrons or bands of color are related as a single integral entity to the framing-edge conceived and experienced as a whole. Further, in the absence of tactility there can be no relative placement in depth, and this means that their paintings are no longer conceived as boxes or flat surfaces or combinations of the two within or upon which, with varying degrees of consideration for the shape of the support, different elements are placed in such a way as to balance off against each other, as has been the case in both traditional and modernist composition up to the advent of deductive structure (or at any rate up to the optical painting that preceded it). Noland's and Stella's paintings are not composed: a given picture does not, as regards what I have been calling structure, represent the achievement of a state of equilibrium for the beholder's experience of the work.3 Their paintings refuse to be seen in these terms; and if we insist on approaching them in this way—or, more precisely, if we give in to what amounts to a strong and deep-running inclination to impose compositional considerations upon them—they will seem, or be, trivial and vacuous. (This inclination cannot be explained simply as the result of our having been conditioned by the painting of the past, in which composition has always played a role of enormous importance. It has to do with whatever facts about our natural history urged to the institution of that particular artistic tradition in the first place.) Above all they refuse the distancing, the estheticizing of the work of art through the appraisal and appreciation of its merely “formal” or “decorative” qualities, its subtle contrivance and clandestine unity, that is implied by seeing in compositional terms. Instead they demand to be experienced in their terrific immediacy and directness, as of a gesture or cry—though of course this demand can be refused, as can the demand of a gesture or cry, and distanced, estheticized. This immediacy and directness are the fruit of their unparalleled explicitness (they conceal nothing, they consist entirely in the simple and completely lucid presentation of all they are), which is itself in turn, finally and astonishingly, the achievement of deductive structure.

Olitski's new paintings are made in opposition to deductive structure; in fact, they are almost certainly the first paintings to acknowledge the existence of deductive structure as an achievement the ambitious modernist painter—whatever the particular cast of his own sensibility—cannot simply ignore. The immediate grounds for Olitski's antagonism are not hard to define: roughly, he refuses to accept what seem to him the intolerably constraining coloristic implications of the division of the canvas into clearly delimited elements without which what I have been calling deductive structure would be inconceivable. In a painting organized according to deductive principles one must know beyond a doubt where, say, one chevron ends and the next (or the bare canvas) begins; and this is made known by a distinct and even change in color (or by an abrupt transition to the raw canvas). Within a single chevron or bar there may be some fluctuation of color, but not much; certainly not enough to raise the slightest query as to where the limits of that element lie. Moreover, the colorist who works in a deductive mode always works within a format of some kind. This is not at all the same as saying that in his pictures color is secondary. In Noland's paintings, for example, there are two ways in which color actually dictates to structure: by compelling or at least strongly influencing gross changes of format (e.g., from concentric rings which virtually require raw canvas to be left between them, to chevrons, which do not) and by controlling variations within a given format (e.g., in the size of the raw canvas intervals between concentric rings, chevrons or bands). But the fact remains that some sort of distinction can always be made between format and color. Whatever may be their wider and deeper resonance, Olitski's new paintings represent a deliberate attempt to abolish this distinction, to obliterate the division of the canvas into clearly delimited areas of uniform color. It is an effort to allow color to take upon itself the full burden of pictorial structure.

This, so I am arguing, is a large part of the animus behind Olitski's brilliantly improvisatory technique in his most recent paintings. He draws a length of unprimed and unsized canvas through a trough containing acrylic paint and places it on the floor. Almost always working rapidly, in an impassioned attempt both to elicit and to record his most intuitive, evanescent coloristic impulses, he sprays into the still wet canvas acrylic paint of new colors from one, two or three spray guns powered by an electric air-compressor, continuing, usually, to a point beyond which one can no longer determine the original color of the ground.4 The result is that the entire surface of the canvas is filled with tiny flecks of different colors which, depending on the wetness of the surface at the moment they were sprayed on, are distinct or slightly blurred or almost wholly dissolved into adjacent colors, and depending on the size of the droplets in a given burst of spray, fluctuate in size from extremely fine points to larger (though still quite small) splashes of pigment. Facture too is capable of enormous variation, depending not only on the considerations just mentioned but also on the angle of the bursts of spray to the picture surface (the astonishing facture of some paintings is the consequence of spraying the paint parallel to and slightly above the surface, onto which it falls of its own weight) and the quantity of paint driven into it. When, later, having been cut and stretched, the canvas is seen at a considerable distance—say, thirty feet—a given portion of it may appear deep blue; but as one walks nearer one becomes aware that it also contains minute flecks of green and red and perhaps other colors as well, and that these either increase or diminish in quantity as one explores the immediate environs of the “blue” area, whose blueness consists simply in the predominance of blue flecks which themselves extend in reduced numbers into areas that, at a distance again, appear green or red or some other color. Depending partly on the colors used and partly on facture, individual paintings achieve, to different degrees, an illusion of depth, but an illusion addressed to eyesight alone. At its most powerful—in those paintings in which blue, green or deep red, as opposed to bright pink, yellow, etc., play an important role—this illusion far exceeds in depth and force of conviction any that can be found in the painting of Newman, Louis or Noland. More than these men Olitski is intent on driving color back into the ground, by the sheer force of this visual illusionism. At a distance, the surface of the canvas virtually disintegrates in one's perception of simultaneously dissolving and condensing clouds of color swarming and flowing into and over and behind one another at indefinite depths—an illusion that resembles the Northern Lights (not the hanging-curtain but the more indefinite, flickering kind) more closely than any other natural phenomenon of which I can think.

Because in these paintings all lines of demarcation between colors are dissolved, not only is a deductive format avoided, but anything that could count as a format is ruled out from the start. Whereas deductive structure—while not actually asserting the flatness of the support—consists in relating clearly delimited elements two-dimensionally to the framing-edge, Olitski's disintegration of the surface, together with his universal interpenetration of different colors, gives to a color's illusionistic existence in depth at least the importance and reality given to its apparent extension in two dimensions. It is not true of these paintings, as it demonstrably is true of Newman's and Still's, that “(Color) has . . . to be uniform in hue, with only the subtlest variations of value if any at all, and spread over an absolutely, not merely relatively, large area. Size guarantees the purity as well as the intensity needed to suggest indeterminate space: more blue simply being bluer than less blue.”5 Instead, the spray technique enables Olitski simultaneously to obtain constantly fluctuating variations in value, often of a quite dramatic sort, and to subsume them in an illusion of sheerly optical space, as connoting not tactile or sculptural qualities (this is one important difference among many between Olitski's paintings and pointillist practice) but what can perhaps best be described as variations in the visual density of that space. Moreover, the intensity of a particular color in his recent paintings is not proportional to its two-dimensional extension, as the remarks by Clement Greenberg just quoted suggest is the case for Newman and Still; in fact, it is hard to know what ought to count as the “extension” of a color in Olitski's paintings. For example, standing back, the color blue might appear to be restricted to a small portion of a given canvas; but on closer looking it might turn out that in fact most of the surface of the canvas contains tiny flecks of blue paint, which one registers not merely as blue flecks but as what the color blue now means. In this instance one could equally well, it seems to me, consider the “extension” of the color blue to consist either in its illusory restriction at a distance or in its actual dispersal over most of the canvas at close range, and that far from having to choose between them we must try to keep both of them in mind. But both construals of the term point up the fact that the intensity of a particular color in Olitski's spray paintings is not a function of its “extension” at all, but rather of its concentration or density (what might be called its “intension”). More blue is bluer than less blue only if we understand “more” and “less” to pertain to the concentration of blue particles in a given portion of the canvas, and “is” to mean something like “appears at a distance.” Even at a distance, however, we are conscious to varying degrees of the mutual interpenetration of color and color, amounting to a kind of heightening and intensifying of our visual experience beyond the apparent intensity of single colors. Up close, all the particles, whatever their color, are equally intense, and one's attention focuses instead on the infinitely varied and—in his most successful paintings—infinitely compelling minute particulars of their simultaneous juxtaposition. Both at a distance and at close range what Olitski has called the “in-ness of color in color” displaces the two-dimensional extension of color basic to the work of Newman and Still, and to the paintings of Noland and Stella as well.6

But despite the fact that, in all these respects, Olitski's spray paintings stand in fundamental opposition to deductive structure, I want to suggest that they would have been inconceivable without Noland's and Stella's prior achievement of it. This is not to imply that they are nothing more than an attempt to negate as many of the preconditions and characteristics of deductive structure as could be opposed at one go. On the contrary, they represent—as anyone familiar with Olitski’s development over the past three years or more is bound to recognize at once—an inspired and wholly unpredictable fruition of the pictorial aspirations expressed with passion, resourcefulness and even humor in his previous work (including, as we shall see, the copiously tactile “matter” paintings of the mid-1950s). In fact, it is precisely the inner logic of his own development that has brought Olitski to a position that can be described as one of fundamental opposition to deductive structure. At the same time, however, this development was influenced and almost certainly accelerated by his growing recognition that deductive structure was something an ambitious painter, whatever his personal inclinations, could not safely ignore: and this, I want to say, amounted to the recognition that, in the light of deductive structure, various aspects of his own previous painting could be seen as, if not actually ambiguous, equivocal or otherwise unsatisfactory, then at least as not realizing his deepest pictorial aspirations as fully and perspicuously as he now saw to be possible.

During the two years before his new work, in a number of paintings made by rubbing and staining mostly acrylic paint into unsized and unprimed canvas, Olitski modulated from one color to another without leaving anything like a boundary between them. At first, as in the exquisite Fatal Plunge Lady (1963), the modulation occurs between two colors close to one another in hue, in this instance between rose-brown and orange-brown; but in Hot Ticket (1964) the broad vertical curtain of color that occupies most of the canvas suddenly and dramatically inflects from intense green down through deep blue to bright red—three colors, incidentally, that are used often in the spray paintings. The structure of these paintings, clearly not deductive, requires for its apprehension a more acute attendance to the interaction of colors than any painting since Matisse. One's conviction before a painting like Fatal Plunge Lady is that the colored areas have only this moment assumed their final configurations, in response to forces at work within the picture-rectangle that are the direct expression of the painter's coloristic intuitions. Nothing about these paintings feels pre-set or pre-established (though in fact Olitski has tended at any time to work within a fairly restricted repertoire of shapes and even arrangements among shapes), and the final “rightness” of their structure is experienced in terms of the momentary configuration of a unique sensibility. Moreover, the emphasis is on the singularity and relative autonomy of the colors and shapes in these paintings, as if Olitski were concerned with the mutually repulsive rather than attractive properties of colors. So that while interacting with one another to create a precarious and fugitive condition of equilibrium, individual colors are, in effect, isolated more tellingly than if each were seen in actual isolation: our perception of each becomes one of recognizing the color in question, as if for the first time.

Fatal Plunge Lady and Hot Ticket thus anticipate the spray paintings in three very general and closely related respects: in the strongly coloristic basis of their structure, in their reliance on an intuitive rather than a pre-established organization of the surface, and in their refusal to be seen simply in terms of the relations among the different colors involved. But there is also a vital difference between these periods: the earlier paintings are composed while the spray paintings are not. Or rather, when seen in the context of the achievement of deductive structure in the work of Noland and Stella, it becomes clear that an important part of what one means when one describes the structure of Fatal Plunge Lady, Hot Ticket, etc., as not deductive is that it remains compositional: its success depends in large measure on the creation of a state of equilibrium through the balancing off of different elements against one another. In this sense, the structure of these paintings does not oppose deductive structure so much as it is opposed by it. Not that this hurts the paintings in question; it doesn't largely because the sense in which they remain compositional is more coloristic than in the case of any previous painting, Matisse included. But I want to suggest that Olitski's experience of deductive structure over a period of years may have enabled or perhaps compelled him to see certain aspects of his own work as compositional which he had not previously considered so and which, without the perspective provided by deductive structure, might not be see-able as such even today; and that at the same time as Olitski felt himself grow increasingly opposed to what seemed to him many of the implications of deductive structure, he nevertheless became increasingly convinced by its antagonism, in the name of rigor, explicitness and (most convincing of all) immediacy, to composition. Or to put it another way: to the extent that, under pressure from deductive structure, the organization of his own paintings came to strike him as at bottom compositional, his aspirations toward a wholly coloristic, intuitive and non-relational structural mode came to seem not more than limitedly or equivocally realized in his previous work. But the more this sense of limitation or equivocation grew, and the more it came to be associated with an effort to compose, the more nakedly, radically and inescapably these aspirations began to oppose themselves to deductive structure. I am suggesting then that increasingly during the past several years, and decisively within the past year, realizing his pictorial aspirations and opposing deductive structure became for Olitski one and the same enterprise; and, moreover, that this identification brought with it a considerable clarification of the formal task at hand. In particular, it urged on him the ambition discussed earlier: to obliterate all lines of demarcation between different colors or between a color and the blank canvas; that is, to do away with all edges or limits of any sort within the painting, and instead to fill the entire surface of the canvas with modulations of color into color, such as occurred for the first time in his work (to my knowledge) within single elements in paintings like Fatal Plunge Lady and Hot Ticket

Around the beginning of 1965 Olitski executed by the same means he had used in Fatal Plunge Lady and Hot Ticket at least one rather small painting—it was in the back room at Poindexter's during his last show there—whose entire surface was impregnated with paint and which modulated from one color to another roughly along a diagonal between two corners. But the familiar methods—rubbing, brushing, sponging or rolling paint into canvas—now came to prove unsatisfactory, chiefly for two practical reasons. First, as in the paintings just referred to, these techniques were capable of modulating from one color to another only along more or less definite axes, thereby opening the door to other than coloristic structural considerations. And second, they required much more premeditation and permitted far less reliance on instantaneous color intuitions than Olitski wanted. The second point relates back not only to Louis and Pollock but, perhaps even more strikingly, to the desire for rapidity of execution which Delacroix continually expressed. Baudelaire, in his great memorial essay published within a few months of the painter's death on August 13, 1863, adverted to this desire more than once:

He once said to a young man of my acquaintance: ‘If you have not sufficient skill to make a sketch of a man throwing himself out of a window, in the time that it takes him to fall from the fourth floor to the ground, you will never be capable of producing great machines.’ This enormous hyperbole seems to me to contain the major concern of his whole life, which was, as is well known, to achieve an execution quick and sure enough to prevent the smallest particle of the intensity of action or idea from evaporating.7

For both Delacroix and Olitski, at a certain point in their respective developments, a rapid or better still an instantaneous means of execution became a major desideratum; and as the example of Olitski makes particularly clear, the need in both cases was not simply, or primarily, for a means of recording fugitive impressions so much as it was for a means of eliciting them in the first place. Toward this end Delacroix kept his brushes and paints meticulously clean, prepared his palette with care (Baudelaire describes it as like “an expertly matched bouquet of flowers”), investigated the chemical properties of pigments and worked at great speed—and Olitski began seriously to explore the possibilities of a spray technique.

The paintings that within the past nine or ten months have resulted from that exploration are, I believe, among the most beautiful, authoritative and moving creations of our time in any art. They differ from his own previous work in their complete freedom from anything that smacks of eccentricity, caprice or artiness; and they are distinguished from paintings in deductive formats by their ease, sensuousness and accessibility. Olitski's paintings are more like natural objects, to be contemplated or enjoyed or embraced or possessed; most of all they are like the human body when it is neither possessed by nor making meaning as in a gesture or cry. Where Olitski's paintings differ from natural objects is in the risk of failure which they run. That is, they only aspire to the condition of natural objects; their success or failure in realizing this aspiration depends on considerations of art—in particular, on considerations of pictorial structure. More precisely, I want to argue that, despite their hostility to deductive structure, Olitski's spray paintings depend for their success upon the new and more acute awareness of the shape and size of the support embodied for the first time in deductive structure: and not simply in their avoidance of composition but, more positively, in the role played in these paintings by the framing-edge. There are of course certain obvious and important differences between the respective functions of the framing-edge in Noland's or Stella's paintings on the one hand and Olitski's on the other. For example, it would make no sense to say of Olitski's paintings that pictorial structure is either generated by or deduced from the framing-edge. But it seems to me true of his most successful spray paintings that the entire “contents” of a given picture relate as an integral entity to the framing-edge conceived and experienced as a whole—which, I suggested earlier, is one of the achievements of deductive structure. In fact no paintings, Noland's and Stella's included, have ever been put under greater pressure by considerations of shape and size; or, more accurately, have ever put these considerations under greater pressure. When the structure of a given spray painting by Olitski strikes us as problematic, our difficulty never concerns the location of something “within” the painting—as for example one might want to see the boundary of a particular chevron in one of Noland's first chevron paintings run into the upper corners of the canvas instead of just missing—but rather the location of the edge itself. Similarly, our conviction in front of his most successful paintings is that the framing-edge has been arrived at by the colors themselves: as if the paintings in question only happened to end up rectangular in shape and of a certain size. (There is just a trace here of the “pretence and play” which deductive structure annihilates, and which the Devil in Doctor Faustus scornfully refers to as the introduction of “prescribed and formalized elements . . . as though they were the inviolable necessity of the single case.'' But the kind of necessity at work in Olitski's paintings is not that of composing.) Moreover, when two paintings very similar as regards both facture and color but different in shape and size strike one as unequal in quality, it is never because the less successful one appears composed, but because in it the painter has failed to possess the edge as completely and convincingly as he has possessed its surface. (In such instances the sprayed canvas has felt to me like background in traditional painting.) Structural considerations of this kind must lie behind Olitski's extraordinary success in a number of paintings (e.g., Hidden Combination, on the cover of this issue) whose dimensions—roughly, between six and eight feet high by no more than two feet wide—have never until now been made to yield work of comparable quality. In other paintings, perhaps less structurally convincing to begin with, Olitski has drawn in pastel after the spray has dried. At first the pastel was applied only along the edges of the canvas in the neighborhood of one or two of its corners; this inevitably strengthened the picture structurally at the same time as it enabled Olitski to specify particular colors with greater definiteness than the spray technique alone tended to admit. But in one painting—as well as, I understand, in others that I have not seen—Olitski has worked in pastel well inside the edges of the canvas: again, partly in a linear mode that evinces a sharp concern for the shape and size of the support.

Shortly before he began working on the spray paintings, in a conversation with Noland and the English sculptor Anthony Caro in Noland's studio in Shaftsbury, Vermont, Olitski remarked that he would like to make paintings that would consist of nothing but some colors sprayed into the air and remaining there. Whatever feats of engineering this would entail, objects of this kind would avoid the issue of pictorial structure altogether; at any rate, there would be nothing that it makes sense to think of as a picture-support or framing-edge, and in particular nothing literal in them but paint itself. This underscores Olitski's belief in the primacy of paint (as opposed, say, to drawing) in painting—a belief expressed by his work as early as the midfifties, in the “matter” paintings referred to earlier. But a spray of colored paint suspended in the air would simply be a kind of natural object, like water from a fountain; and although we would doubtless take pleasure in it, we would not begin to know how to be convinced by it as a painting (For example, can we imagine two such sprays, one of which strikes us as a better or more successful painting than the other?) And this suggests that the pictorial structure of Olitski's spray paintings is not merely crucial to their quality as they stand, but vital to our experiencing them as paintings in the first place. In his fourth “Discourse” Reynolds writes:

Though it be allowed that elaborate harmony of colouring, a brilliancy of tints, a soft and gradual transition from one to another, present to the eye, what an harmonious concert of music does to the ear, it must be remembered, that painting is not merely a gratification of the sight. Such excellence, though properly cultivated, where nothing higher than elegance is intended, is weak and unworthy of regard, when the work aspires to grandeur and sublimity.8

For modernism as well as for Neo-Classicism, painting is not merely a gratification for the sight It is a dialectical, cognitive and (in the deep sense of the word) conventional enterprise which, in Jules Olitski's new paintings, has given us works to match the art of the museums in quality, conviction and—perhaps most astonishing of all—naturalness

Michael Fried



1. See the catalog to the exhibition “Three American Painters,” Fogg Museum, Cambridge, Mass., 1965.

2. The definitive essay here is Clement Greenberg's “Collage,” in Art and Culture, Beacon Press, Boston, 1961.

3. The same holds for the work of the English sculptor Anthony Caro, except that there is no sculptural equivalent to deductive structure.

4. Olitski often uses three spray-guns, two at a time, with as many as nine screw-on cans of paint, each a different color, at his disposal.

5. Greenberg, “After Abstract Expressionism,” Art International, Vl/8, October 25, 1962, p. 29.

6. This is as good a place as any to acknowledge my immense debt to Jules Olitski. Many of the ideas put forward in this essay are his, just as others belong to Noland and Stella. At many points, however, I have speculated about Olitski's development without checking my formulations with him. I also want to thank Stanley Cavell for innumerable criticisms and suggestions both of substance and of phrasing which I have had no scruples about accepting.

7. The Mirror of Art, Anchor Books, New York, 1956. p. 331.

8. Discourses on Art, Collier Books, New York, 1961, p. 64.