PRINT May 1972


GRIDS, WITH MODULES AND SERIES, have become important modes of organization for recent art. One is tempted to talk here of grid structures; but if the word “structures” is to have any precise definition we must distinguish it from things that are more properly frameworks. Grids can constitute structures or can, more often, be simply frameworks. To assume that a linear surface organization when visible in a work of art is its structure is to follow in part the Florentine precept that aspects other than drawing are somehow accessory to the work’s substance. Of course, much 20th-century art demands attention in more or less Florentine terms, and does so because of the enormous significance of Cubism. When Cubism exaggerated contrasts of light and dark it maximized the skeletal aspects of traditional painting and made drawing a central issue for those who followed its lead. Impressionism, however, narrowing or, at times, suppressing values, turned painting into an affair of surfaces to which linear figuration as such was subsumed. And yet, the grid, as currently used, owes as much to Impressionist sources as to Cubist ones. For example, Sol LeWitt’s gridded wall drawings are really a kind of updated Pointillism—they create a regularized vibrancy of sensation due to their small “crosshatched” scale. His later work confirms this reading. Moreover, many current grid paintings are quasi-Impressionist. The grid is used to decentralize the picture surface: to create an integral surface of closely similar elements so repeating themselves as to produce a uniform textural block. This allover expansiveness comes most evidently by way of Pollock, to whom those that do use the grid as a primary (structural) device are more often indebted than to such late Cubists as Ad Reinhardt. For it is the reduplicative surface that matters here far more than the mere fact of rectilinearity. A mediator in this respect is Philip Guston—at the time of his “plus-and-minus” type paintings. These, however, also return us to Cubism by way of Mondrian and, in so doing, locate the late ’40s fusion of Cubism and Impressionism that still directs the course of American art.

But when can we talk of grid structures? Toward answering this, Lawrence Alloway’s definition is useful to consider:

The field and the module (with its serial potential as an extendable grid) have in common a level of organization that precludes breaking the system. This organization does not function as the invisible servicing of the work of art, but it is the visible skin. It is not, that is to say, an underlying composition, but a factual display.1

The grid as a “factual display” is very much to the point, as is the notion of it being a potential attribute of the module. What exactly is meant by the “visible” and “invisible” organization? If we take, for example, a painting by Agnes Martin and mentally remove the grid (its “visible skin”), any “underlying organization” that remains would indeed be invisible, for the painting is all grid. There are more general problems in the idea of things “underlying” art, but for pre-sent purposes what is significant is that grids such as Martin’s are all that is “displayed.” They do not “service” other pictorial components (for none exist) and must, therefore by a simple act of deduction, be counted as structural.2 In contrast, the early paintings of Larry Poons merely use grids as a framework or scaffolding to be departed from. To apply the word “structure” to Poons’ grids is (as he himself said) to think of structure as something that can be counted: “good painting has a structure but it’s a structure that’s integral to the painting and not to any rules.”3 This points to one peculiarity in the use of grids in painting: what results looks often too much like a game.

I said that a painting by Agnes Martin is all grid and that the grid, therefore, can perform no “servicing” function. But strictly speaking this is not true. What the grid does service is the literal object on which it is drawn. For the Minimalist sensibility, the grid system supports “the flatness and delimitation of flatness” of the painting in question, its “minimal conditions” for being a painting.4 The most striking example of this at work (and perhaps it is meant as a pun on the issue itself) is Robert Ryman’s Stretched Drawing, where the regular grid is distorted slightly from the canvas being pulled around the stretcher bars. Here, the shape of the support questions the regularity of the grid. Interesting, certainly, but a very restricted interpretation of what painting offers. The grid (as so often) merely signals art’s surfaceness, not working for it in a fully pictorial way.5 The grid as a convention (used to get an unconventional look), enfeebled a lot of art in the great graph paper period of the ’60s, and (to judge from recent painting) isn’t over yet. The problem with grids as structure is that though they can produce works of considerable charm, this is only achieved by so reducing the scale of their divisions that the effect is more properly that of “cross-hatching.” This is the quasi-Impressionism I referred to in LeWitt, that also characterizes Agnes Martin’s somehow light-filled paintings; but most often as the scale increases, so—in direct proportion almost—does one’s boredom. They become just feats of ideation, toy versions of art-making.

A notable exception to this is Mondrian’s work, and specifically his regular grid and “checkerboard” paintings of 1918–1919. In these, however, the grid is modulated by varying thicknesses of line or by enclosed squares of color. The grid, here, is a framework. But these paintings have a further significance in that they comprise the very small minority of his works that can be considered allover or nonrelational in effect; in this they are unlike most Cubist-derived art. We can, therefore usefully separate grids as frameworks into those that work to cohere a surface and those that fracture it. Hence, even within works using visually intact grids, we may observe the familiar distinction between relational and non-relational distribution. For example, with Ellsworth Kelly or Richard Anuszkiewicz the grid might be complete but still subverted to form identifiable subgroupings within it. For these artists, however, part-to-part relationships can be understood as part of their indebtedness to European abstraction. But what is interesting about much recent grid practice is that it is relational without being directly “European” or Cubist in approach. Prime examples here would be the use of the fragmented grid in the work of such different artists as Kenneth Noland and Alan Shields. While John Walker’s recent work has shown grid fragments used in paintings: as distinct pictorial elements, among others, to reinforce the uprightness of the surface. The regular grid, one might say, is but an aspect of repetitiveness, and developed as such in the ’60s to “objectify” the repetitive organizations of Abstract Expressionism. Recent usage, however, shows a turn to the relational on the one hand. And, on the other, an exploitation of the grid to merely inaugurate paintings that finally devalue its significance as a formal component or that use it as literally a framework, that is as a place on which to hang the crucial forms of the painting.

Poons’ early work devalued the grid as an intact entity, painting it out while reserving its surface-descriptive function in the ellipses of color that remained. Some recent grid-based painting follows this approach — Donald Lewallen’s for example. His loose brushing and the pearlescent shimmer of his paint visually dissolve the rectilinear framework. More literally frameworks are the grids of Joan Snyder—whose charged brushmarks are almost “written” between the drawn lines—and of Patricia Lasch—where the grid is truly a graph paper chart for presenting numerical data of an autobiographical kind. Their work, and some others’, belongs to a special aspect of grid usage, best thought of as grids as “containers,” that keeps cropping up in 20th-century art, and necessitates at this point a short historical exegesis.

Traditionally, the grid has two main functions: as a measuring and as a transfer device. Although a grid may be used primarily as a tool for measurement or proportioning, if it contains any added imagery its transfer function is inevitably present. A useful model here is the drawing frame. The borders of the frame, like a viewfinder, select a sample of information; the gridded divisions allow it to be transferred (at any scale) to gridded drawing paper. And graph paper itself has the same implication. It seems no accident, therefore, that it was invented (ca. 1800) in a period that saw a great boost in the inventions for mechanical reproduction (between lithography and photography).6 Placing data in separate but juxtaposed squares (in “time modules” as it were)7 when used within art, is to assert its “informational” attributes. It implies that the art consists of two things: the regular framework, and the information put into this framework.

I have written on several occasions of the development from Analytical to Synthetic Cubism as representing a changed way of viewing the world:8 how, in Synthetic Cubism, paintings become in a new way “containers” for signifying elements. Although these elements do, of course, comprise any given painting, their significative (no longer depictive) function allows them to be read (far more than previously) as independent units of information. So far as grids are concerned, Synthetic Cubism extended “allover” the centralized Analytic form—most explicitly in Mondrian, whose grids are themselves signifying elements, as symbols of cosmic states. By the ’30s, however, the allover grid was used merely as a framework—within which the significant information was placed. Important here are the “pictographs” of Kandinsky and Torres-Garcia, where the grid is a passive receptacle for containing “meaningful“ forms. Undoubtedly, Surrealist “boxed objects” had no small influence on this development—as is shown by a seminal American work in this vein, Gottlieb’s Sea Chest of 1942. Here a box placed frontal to the picture plane contains a selection of rocks and shells, and this inaugurated Gottlieb’s own pictograph paintings. Other early works by Abstract Expressionists can usefully be considered in this light (including even Pollock’s Guardians of the Secret). But only when Abstract Expressionist painters managed without a differentiated formal and “meanings” system did their art achieve maturity. This isn’t to say .that such work can’t be successful: no systems in art are intrinsically better than any others—and some good paintings emerged from this approach. And yet, the principle of “containment” has seemed increasingly antipathetic to major art.

Taking this development in retrospect (and in a drastically simplified manner) what seems to have occurred is that first the small-scale expansive network of Analytical Cubism gave way to the manipulation of large, more visually discrete units in the Synthetic period. This in turn allowed these elements to be thought of as individually meaningful; and the exposed full grid of the ’30s objectified their informational containment within the art work. In Pollock the “grid” reemerged as a natural attribute of the forms of painting: to create an allover small-scale texture that repudiated all implications of containment, though the art here (as with all grids as structures) was in large measure a matter of defining containment in such a way that the painting would not seem an excerpted unit. The emergence of real grids and grid structures in the ’60s objectified (or literalized) Pollock’s approach. The recent trend to use grids as, literally, frameworks marks a return to an informational or signifying sensibility: of grids, once more, as containers. Not, of course, for all artists—or even for most—but the ideational grid work I mentioned does seem of this kind, and allies itself therefore to recent Conceptualist concerns; most especially, I suppose, to those who use the grids of maps and charts to present art rather than create it. Once again, though, the system itself is not what tests this art. Patricia Lasch and Joan Snyder both use the grid as a “note pad” kind of framework. But for the trite “diaries” of the former, we have in Snyder truly expansive painting—where the just visible grid (sometimes denoted by a filled-in square), though functioning to contain the separate units, is itself spread by these units to create a space the opposite of contained. As with Poons’ early paintings, the grid sets up the art—to which it finally surrenders. That a painting like Snyder’s Whole Segments can recall Hofmann (and even has just a flavor of Matisse’s The Moroccans) shows that grids needn’t always make for so austere an art as we might assume.

A typology of grid uses would be incomplete without consideration of how the grid is comprised. The grid is an attribute of the module, consisting of units of fixed dimension arranged in a horizontal-vertical pattern (whether presented upright or diagonally), and the orthogonal grid is created when the modular units are themselves rectilinear. But this “additive” definition presumes the network of vertical-horizontal lines to be negative elements—merely the junctions of modular units—and for most grid painting the linear is clearly a positive component. That is, the lines are a method of dividing a preexistent surface: a subtractive, not cumulative, system. And its use, as we have seen, has been to “display“ the surfaceness of a painting or to so organize a surface as to allow the dispersion of pictorial elements within its “framework.” The nonrelational aspects of subtractive grid systems lent themselves very readily to Minimalist art (from which they passed into some recent allover painting); but equally relevant to Minimalism is the cumulative grid—the most compressed and homogeneous series into which modules can be arranged.

This is most evidently the realm of sculpture, for the approach is directly analogous to building or construction. Here there are plenty of true Constructivist precedents, though most are architectural. While sculpture as construction is now the dominant mode, only rarely does it use a grid arrangement—for such a unified system would divert attention from the juxtaposition of elements (the true subject of Constructivist work) to the elements themselves (for the juxtapositions would be nonhierarchical and therefore neutral) and hence to the object they comprise. In consequence, when Minimalist sculpture did use grid (and gridlike) systems it approached far more closely the condition of “non-art” than did Minimalist painting. No matter how clearly stated a grid is in painting, the “object” is clearly a painting. As we have seen, full grids in painting are but an objectification of the functional rectilinearity of form natural to painting on a flat rectangular surface. To use a (three-dimensional) grid in sculpture is to assume it conforms to the cube in the same way painting does to the plane. And the very fact that Minimalist sculpture did often conform to a rectilinear mode shows just how much it owed to painting conventions. Since sculpture itself offered few models, the Minimalists looked sometimes to architecture, but more crucially to painting—or, at least, to painting conventions: to the “minimal conditions” of painting. “Wraiths of the picture rectangle and the Cubist grid haunt their work,”9 whether they liked it or not.

Most evident here is Carl Andre’s ground plate pieces which do in fact look like picture rectangles created by a regular grid. Except, of course, it is a cumulative (quiltlike or mosaiclike) method, rather than the subtractive one that most often belongs to grid painting. LeWitt’s work, however (here I am thinking of the modular space-frame structures), though also cumulative in approach doesn’t really look so. Since the three-dimensional grid is a positive element in the “cages,” and since each individual cubic space shares boundary lines with its neighbor, the grid somehow describes a unitary space rather than comprising it—although, of course, the total space only exists by virtue of its cumulative planning. Thus, while these sculptures of LeWitt seem further from painting than Andre’s (less flat), they are actually far closer to the subtractive methods of grid painting—dividing and defining spaces by “displaying” their composition.

Nevertheless, painting can make use of a cumulative grid method. Warhol’s banked rows of “redundant” images are a good example of this. Individual, self-contained units are put together in a theoretically infinite continuum. These can end up to form a rectangular image, as most do; but equally, since it is the addition that makes the pictures, they can quite naturally have an irregular perimeter. When Warhol’s paintings are like this, they are “incomplete” towards the bottom, thus expressing the boredom in such a repetitious activity. Occasionally he has capitalized on the durational implication of this, making the work purposefully narrative: charting a brief time span with changed poses in each frame (Ethel Scull 36 Times) or even something approaching a life span, as in the biographical Rauschenberg.10 All cumulative grid art has this temporal suggestiveness, though it rarely specifies it in so literal a way. Grid painting is generally only cumulative when its units are informational containers like Warhol’s; and it is not quite a tautology to suggest that when such painting does use a cumulative method it somehow demands that it be “read” in an informational way. The grid as a “containing” framework is of obvious relevance here. Torres-Garcia’s “pictographs” are really not so very different in principle from Warhol’s work. They share an obsession with personalized icons built up to form a “meaningful” display—and the suggestiveness of massed icons is given, on balance, more significance than what—as paintings—they are.

Even “cumulative” painting that is abstract doesn’t escape either the temporal or the informational implications I have noted. To think of Donald Cole’s work, which abuts rectangles of patterned canvas, or of Gary Stephan and Charles Fahlen who make open gridlike plastic wall hangings, is to be aware that each element is meant individually, as well as for the configuration it helps to create. Moreover, works like this so evidently grow (in a way we can intuit) as they are made. The same might be said of Mary Heilmann’s work, though it derives more obviously from Jasper Johns whose tessellated map and number paintings are a primary source for most informal grid work.

If grids had their heyday in Minimalism and in the ’60s, clearly they are not worked out yet. They remain to “control” the randomness in quite a lot of post-Minimalist activity. Their neutrality permits a rudimentary ordering that doesn’t look composed and their stability is a quickly grasped support for the improvisatory and the gestural. This, however, works two ways. It may be the means to nudge into existence an art that justifies itself irrespective of systems; or it may be a prop around which to decorate a surface, with the art no more than in-filling. And, unfortunately, the latter is all too common. The majority of grid work ends up not moving, but merely pleasant. To take some graph paper, strengthen some of its lines, and start filling in some of the squares can’t really help but produce an attractive effect—but more often than not the result will just be good design. Ending on this pessimistic note is unfair to those who have used grids to considerable and sometimes to major effect, but it was not the grid itself that made their art—it merely initiated it.

John Elderfield



1. Lawrence Alloway, Systemic Painting, The Guggenheim Museum, New York, 1966.

2. Lucy Lippard has written of artists including Martin that “the grid per se is of absolutely no importance . . . providing, as it does, merely an armature. . . .” This, however, is to assume it is always a framework—and Martin’s work belies this. Lippard’s statement comes from her catalogue introduction to the exhibition “Grids” (Institute of Contemporary Art, University of Pennsylvania, Philadelphia, January 27 - March 1) which prompted this essay.

3. Larry Poons to Phyllis Tuchman, “An interview with Larry Poons,” Artforum, December, 1970, p. 45.

4. For “minimal conditions”: Clement Greenberg, “After Abstract Expressionism,” Art International, October, 1962, and Michael Fried’s comments in his “Art and Objecthood,” Artforum, June, 1967.

5. I discuss the grid as a contextual symbol in “Mondrian, Newman, Noland: Two notes on changes of style,” Artforum, December, 1971.

6. And as with the photograph, with the grid as a transfer device “the implied presence of the rest of the world, and its explicit rejection are as essential in the experience . . . as what it explicitIy presents.” (Stanley Cavell, The World Viewed, New York, 1971.)

7. The phrase is James W. Davis’ whose “Unified drawing through the use of hybrid pictorial elements and grids,” Leonardo, Winter, 1972, contains a useful section on grid functions in Western art.

8. Especially “The Language of Pre-Abstract Art,” Artforum, February, 1971, and most recently in a discussion of Leger’s place in this development (Artforum, April, 1972).

9. Clement Greenberg, “Recentness of Sculpture,” in the catalogue, American Sculpture of the Sixties, Los Angeles County Museum of Art, 1967.

10. John Coplans discusses this aspect of Warhol’s work in “Andy Warhol: The Art,” in his Andy Warhol, New York, 1971.