Carl Andre and Film
Carl Andre has never made a film. But film can tell us something about Andre’s work.
The work of Andre that I consider here is the work with which he rose to prominence in the late 1960s. I would call this his classic period. And I consider only the work that is characterized by being predominantly in the horizontal. During his classic period Andre did work that was not in this mode, but when we think of Andre, it is the horizontal work from his classic period that comes to mind. The horizontal work is the essential work. This is what Andre himself recognized when he said that his work was “more like roads than like buildings.” He was saying that his work doesn’t concern itself with mass or volume, but with extension in the horizontal. And he was also pointing to his work’s being linear.
The vast majority of these works consists of a quantity of identical units: bricks, concrete blocks, metal plates. And in almost all cases these units are connected to each other in the same way, but more about the manner of connecting later on.
Another name for such units is a module. A module is a standardized form or shape. The term implies that the shape is simple. Shapes that are identical can be readily combined with each other in different ways, and simple shapes are the ones that can combine with each other in the greatest number of ways.
Almost all the modules that Andre used in his classic period are rectangular (and by rectangular I also mean square). A rectangular module is as flexible in its possible arrangements as any other, if not more so.
I once heard Andre speak with real feeling about bricklaying. I heard for the first time the terms header, stretcher, and Flemish bond. Stretchers are bricks laid with the long side showing; headers are bricks laid with the end showing. Flemish bond is a pattern for laying bricks. With the help of Google, I found more names for patterns for bricklaying: common bond, running bond, herringbone, English bond, stack bond. These are several ways that a brick, a rectangular module, can be combined with other bricks. (For paving, where the bricks don’t have to support the load that a building does, there are more patterns.) Many of these patterns show both the ends of the bricks and the sides of the bricks. In all of these patterns the bricks from one layer, or course, to the next are staggered in relation to each other. This is true for all the patterns but the last, the stack bond.
As the name suggests, in the stack bond the bricks are stacked one directly on top of another. In the stack bond the bricks are all stretchers: you see only their long sides, so you see only one shape repeated again and again. We know that bricks are modules, but the stack bond makes us see this fact as other bricklaying patterns do not. The bricks line up with each other in the horizontal, and they line up with each other in the vertical. The joints between the bricks form a set of parallel lines that run without interruption from side to side and another set of parallel lines that run without interruption from top to bottom. The lines are an index that attests to the uniformity of the units that produced them.
A module usually implies orderly relations among the repeated shapes, but this does not have to be the case, as Andre himself demonstrated in Spill (Scatter Piece) from 1966. But this work is an exception; as a general rule Andre’s works are systematic arrangements of a module.
There are many more possibilities for arranging rectangles than a few patterns for building or paving with bricks. Andre’s work that is not in the horizontal, above all the work that uses modules of timber, explores in a limited way the range of relations and the ways of connecting that the module makes possible. But in the horizontal work that is composed of rectangular modules in regular relations, which is to say almost all of the horizontal work, the ways of connecting the modules are drastically limited. Instead of a range of relations, there is (with a handful of exceptions) only one. I don’t know all of Andre’s work, but to the extent that I know it, this is the case.
Of all the possible patterns that were available to him to arrange the rectangle, Andre in the horizontal work chose almost exclusively the stack bond. The edge of one brick lines up with the edge of another. The edge of one steel plate lines up with the edge of another. It doesn’t matter that the work is horizontal, not vertical like a brick wall. If a number of identical rectangular units are fitted together with their edges aligned, it’s a stack bond.
As I have noted, a brick wall built with a stack bond creates two systems of parallel lines that are at right angles to each other and that intersect. Another name for this arrangement of lines created by the stack bond is the grid.
We tend to think of a grid as something that comes into existence all at once. The word brings to mind an array of cells already in their relation of vertical and horizontal alignment. I have come to the grid through bricklaying and the stack bond to emphasize that a grid can also be the result of building up an array unit by unit.
The stack bond, or grid, is so pervasive in Andre that it raises the question: Why is this so? What is the origin of this pattern that so gripped Andre?
A stack bond can in principle be as simple as a single stack of bricks. It’s no longer a grid, but it is still a stack bond, if in its simplest form. Let’s call this a stack. I know this term is used to describe some of Judd’s work. But I want to use it in talking about Andre because in his work we find the direct adjacency of the modules that the word implies. Judd’s stacks should have been called exploded stacks.
Just as the stack bond as a grid occurs repeatedly in Andre, so does the stack bond in its simplest form, the stack. The joint catalogue for Andre’s shows at the Gemeentemuseum in The Hague and the Van Abbemuseum in Eindhoven in 1987 reproduces nearly 30 of them, the earliest being 35 Timber Row (1968). The catalogue entries describe these pieces as rows or lines, but for the sake of continuity in my discussion, here I call them stacks. The modules are made of a variety of materials: timber, cement blocks, metal plates, bales of hay, pieces of chalk. The examples that the catalogue reproduces are all horizontal except for one, but in principle they are all the same as a stack of bricks. (The earliest stack I found in the catalogue that on the basis of the description was unmistakably a stack was 34 Bay State Hard Line (1965). But there was no reproduction.)
Andre had used a form of the stack in his early work, and it is very likely that he used the term. 12 Dialogues, 1962-1963 reproduces twelve written exchanges between Carl Andre and Hollis Frampton. It includes photographs by Hollis Frampton of Andre’s early work, much of it destroyed. The captions for the photographs of three pieces made in 1959, Pyramid (Triangular Base), Pyramid (Square Plan)[sic], and Pyramid, describe the pieces as stacks. They are stack-like, but they are not stacks like the stack bond. The ends of the elements are aligned, but the edges of the elements are not aligned with each other, and the elements are not identical. But in any case, the term is there. The editor of this book was Benjamin H. D. Buchloh. It is difficult to imagine that the descriptions were prepared without Buchloh’s consulting with Andre.
Stacks, repeated modules in a row or a line, isolate the distinctive property of the stack bond. A stack is a succession of identical rectangular units joined edge to edge with their free edges bounded by two parallel lines. This relationship of parts also describes the relation of frames in motion picture film. It doesn’t matter if the proportions of Andre’s rectangles range from the square to the extremely elongated. In the edge-to-edge alignment of a repeated module, Andre’s row or line pieces and a strip of motion picture film are identical. Both are stacks.
Film consists of a module, the frame. You can cut the frames apart and reassemble them in new combinations. But there are limits to how you can connect one frame to another frame. You can connect frames only at two edges, their tops and bottoms. This limitation in how you can connect the modules in film is forced on you by what film is, a long strip with a width that is fixed so it can pass through a series of machines.
In film, the frames must be in a relation that is the stack; there is no other choice. Andre was free to organize the relations among his modules in any way he pleased, but of all the possible relations available to him, the relation he chose almost exclusively in preference to all others was the stack. And, as we will see in a moment, the stack is also the grid in another form. The conclusion I come to is that the essential model of the relations in Andre’s essential work refers to film.
A strip of motion picture film is inscribed with time. The passing of time is what the succession of adjacent frames in a strip of film registers. The succession of adjacent modules in Andre’s stacks likewise suggests the passing of time.
When Andre said that his pieces were “more like roads than like buildings,” he was pointing to their extendedness, their linearity. The linearity of a road is not built all at once. You start at one end and add to its length. Andre’s stacks could only have been constructed in a similar way. The length was created by adding one module after another. You put down the first module, then you put down the next module adjoining the first one, you put down the next module adjoining the module you just put down, and so on. The process of making one of these pieces is so evident from looking at the work that it is possible to visualize someone making it, putting down the modules one after the other. The modules are added one after the other successively in time, just as frames in motion picture film are added one after the other when they are exposed in the camera.
The longer the stack, the more time it takes to build it, and the greater the evidence of the presence of time. And the longer the stack, the greater the suggestion that Andre intended that the work should be understood as embodying the passing of time, the same passing of time that is inscribed in a succession of motion picture frames.
This has to have been Andre’s intention because a very long stack, like a road, cannot be experienced all at once but only in time. Consider two pieces reproduced in the catalogue I mention above. Sixty-seventh Copper Cardinal (1974) is a stack of sixty-seven modules that are each 50 centimeters square. The overall length of the piece is 3,350 centimeters, or nearly 110 feet. Secant (1977) is a stack of one hundred modules of Douglas fir that are each 1 foot square and 3 feet long. The modules are laid end to end. The overall length of the piece is 300 feet.
When a piece is that long, you can’t see it all at once. You may be able to take it all in, but you can’t see all of it equally well. You may have a good view of the end you are standing at, but the piece stretches off into the distance, and the other end is far away. To see all of the piece equally well, you have to walk along it. As you walk from one end to the other, you experience one module after another, just as you do frames in a film. It doesn’t matter that when you watch a film you are motionless and the modules move and so succeed one another on the screen, and that in Andre’s work the modules are motionless and the movement is supplied by you. In both cases you can experience the succession of the modules only through movement, and movement requires time. The film moves in time, or you move in time. You experience film in time, you experience Andre in time.
Secant in relation to its length is very narrow. This alone is enough to remind us of film, and it comes in addition to the work’s being made of modules and being of such a length that we must read it in time.
White Chalk Run (1972), another piece reproduced in the same catalogue, is not as long as either of these examples, but it embodies the same property as Secant, extreme narrowness in relation to its length, that is characteristic of film. The work is a line of modules of white chalk. Each is 1/2 inch square and 3 1/2 inches long. There are thirty-six modules, and they are laid end to end. The overall length of the work is 126 inches, or 10 1/2 feet. The piece is narrower than 16mm motion picture film. A piece of 16mm film the same length as White Chalk Run would last seventeen and one-half seconds. If you were to disregard the length and take each module as having the duration of a frame of film, the film would last one and one-half seconds; less, but it is duration nonetheless.
It’s possible to say that pieces of such length are exceptional in Andre’s oeuvre and so can’t be made so much of. But I would say that even if such long pieces are exceptional, it does not matter. These pieces only act out what the stack pieces in principle imply, that identical elements one after the other inscribe time, just as they do in film. Time is equally in the stack pieces that are shorter, even much shorter, than the examples I offer here.
Just as time is in Andre’s stacks, it is also in his grids. In the criticism of modernist painting, the grid is without time. It comes into existence all at once, it is perceived all at once. But if the grid is timeless, it can also imply time, just as Andre’s stacks do.
A bricklayer builds a brick wall on the pattern of the stack bond brick by brick, course by course, and in the end he has made a grid. Laying bricks takes time. (So does writing. Andre’s poems are written on a typewriter because they depend on the grid that is already inscribed within the mechanism of a typewriter, and that a typewriter automatically produces. Of course letters in the rows are aligned, but the letters are also aligned in columns. And what takes time to write takes time to read.)
A grid can also be thought of as a long line of modules broken into segments of equal length and then arranged side by side. The reverse is equally true. A grid can be turned into a line by arranging one after another the stacks that compose it.
On this principle, film and the grid are each a form of the other. A grid can be turned into a film, and a film can be turned into a grid. You can compose an array of frames in a grid, then scan the grid line by line to turn the frames into a film. This transforms an array into duration. You can divide a film into stacks of equal length and arrange them side by side to form a grid. This transforms the duration of the film into an array.
Both of these operations have been performed by filmmakers. Peter Kubelka composed his film Arnulf Rainer as a grid, then transformed the grid into the film, scanning from left to right and going down the rows. The grid in this case was a framework for a kind of score. Paul Sharits has transformed some of his films into grids by cutting the prints up into segments of equal length, then mounting the segments one next to the other on transparent plastic sheets.
These examples illustrate the ambiguity of the grid with respect to time. The grid can be without time, or it can be inscribed with time, depending on the specific instance, and in these two examples it is inscribed with time. In the Sharits pieces you see the frames all at once, as you would cells in any grid, but you also understand that the frames are in a sequence. You can scan the frames one after the other and so have a sense of the experience of watching the film. For the images to be in their correct orientation, the strips must be vertical. So instead of reading row by row, you read column by column. In a grid, time doesn’t go only from left to right, row by row, it can go from top to bottom, column by column. Sharits’s title for each of these pieces is Frozen Film Frames. The frames are frozen, so instead it is the viewer’s eye that moves. In composing a film on the grid, the grid helps to organize patterns, and a film turned into a grid brings out patterns that you wouldn’t see if you looked at the film as a continuous line.
Andre’s stacks are inscribed with time from their resemblance to strips of film, and so are his grids. This is equally true of his pieces made of rectangular modules in which the overall shape is a triangle with one side that is stepped. They could be either compositions for films or the frames of a film arranged to bring out a certain feature, for example, successive shots are one frame shorter than the one before, or one frame longer.
It is not news that time is present in work that we call Minimalist. Michael Fried says that duration underlies all of Minimalist art. In his 1967 critique of Minimalism, “Art and Objecthood.” Fried says, “[T]he experience in question [the beholder’s encounter with Minimalist work] persists in time [emphasis in the original][.]” In Fried’s view, this persistence in time is a sign of the theatricality that is the defect of Minimalist art, for which his term is literalist art.
I want to suggest that Andre’s device of repeated modules placed edge to edge so unmistakably enacts the form of film, which we know is both made in time and experienced in time, that we see time directly inscribed in Andre’s work as we do not in the work of other Minimalist artists.
Duration is inscribed in Andre in a form that in its specificity is exceptional. And the exceptional form in Andre of this property that Fried attributes to all of Minimalist art is accompanied by a second exception, the complete absence of another property that Fried sees in all Minimalist work: the anthropomorphic. According to Fried, this anthropomorphism comes from the fact that Minimalist works are perceived as hollow. I would say it’s difficult to suppose that relatively thin plates of metal or bricks or timbers are hollow. I would say that the contrary is true, that we know (rather than merely sense) that they are solid. Fried also says that we feel confronted by Minimalist works, as if by a person, but I wonder if we feel confronted by something that is underfoot and that we can walk on, something so marked by flatness, the two-dimensional, that its main sense is that of extent.
We don’t walk on film, but film is otherwise the same as Andre’s horizontal work. Film isn’t hollow, and it’s flat. Like the metal plates, film has thickness, but it is so negligible that we don’t think about it. Film is in effect two-dimensional. It’s a matter of width and, above all, length, just as Andre’s stacks are. I would say that on these two key points Andre’s work is an exception to what Fried says are the general properties of Minimalism.
The film that Peter Kubelka made from his grid was of finite duration. The extension in time of the film corresponds to the extension in space of the grid. The grids that Paul Sharits turned his films into were films of finite duration. The extension in space of the grids corresponds to the extension of the films in time. Andre’s stacks are of finite length, and so are the stacks that his grids can be turned into, so I want to say that the duration implied by any of Andre’s works is as finite as it is in any film.
But Fried has already disagreed with me. He says, “[T]he repetition of identical units…carries the implication that the units in question could be multiplied ad infinitum.” He continues in a footnote, “That is, the actual number of such units in a given piece is felt to be arbitrary, and the piece itself—despite the literalist preoccupation with holistic forms—is seen as a fragment of, or cut into, something infinitely larger.”
To give Fried a chance to be utterly emphatic on this point, here is an earlier passage, only the beginning of which I gave above. Fried says, “[T]he experience in question [the beholder’s encounter with Minimalist work] persists in time, and the presentment of endlessness that, I have been claiming, is central to literalist art and theory is essentially a presentment of endless or indefinite duration [emphases in the original].”
I suggest that Andre is as much an exception on this point as he is on the two I’ve mentioned. Fried could well be right in saying that the number of units in Andre’s stacks is arbitrary. But I would not follow him in his assertion that the arbitrary number of units implies that the units in Andre could be multiplied ad infinitum, such that the work is of endless duration. I think the form of the stack suggests otherwise. The stack is not a field of objects that we wander among; it’s not an arrangement of volumes that surround us; it’s a line. It’s a line of identical flat rectangles edge to edge on the floor. The line, or stack, starts here, it goes there, and it stops. How Andre’s work was made is how we experience it. We start here, we go there, we stop.
Just as we traverse the stack, we traverse the grid, module by module, reading down the rows or across the columns. You start in one corner, you move through the successive modules until you come to the last one, then you stop. To me, something that starts and stops suggests not endlessness but finite duration. I think this would be my sense of Andre’s work in any case, and the sense of its being finite is doubled by its form being identical to a strip of motion picture film.
Any film, any composition in film, is of finite duration. A film has a finite number of modules, just as Andre’s pieces have a finite number of modules. The duration of most films is governed by what happens within the frames that are its module. When the requisite relations have been enacted among the modules, the film ends. By convention, most commercial films accomplish this in about two hours. But to the extent that Andre’s modules are identical, a skeptic could say that nothing happens in them to enact relations among them and so govern the length of the stacks that they constitute, as things happen among the modules of a film and so determine its length. (This is not an issue in the pieces where the modules are of different materials, for example, the grids made of as many as six different metals. The different materials are in themselves enough to suggest an articulation in time.)
Commercial films are not composed of arbitrary quantities of frames. The lengths of the shots are not determined in advance, they are long or short as they need to be to tell the story. But there are non-commercial films that are founded on precise relations among quantities of frames. These quantities may be arbitrary, but they are the basis of the films they compose. The duration of the films is not endless or indefinite; they are specific lengths. The films are not fragments of something infinitely larger than they are; they are films, whole and complete. Peter Kubelka has made films like this, and Paul Sharits made films like this. Hollis Frampton, Andre’s friend in the 1950s and 1960s, made films like this. Andre’s essential device, which reproduces the appearance of specific lengths of film, recognizes this kind of filmmaking.
It is well known that for several years in the early 1960s Carl Andre worked for a railroad. In his own words, he was a “freight brakeman and conductor.” It has been noted that Andre’s employment during this time brought him into sustained contact with the module. Hollis Frampton describes Andre’s work for the railroad as “assembling freight trains.” Frampton goes on to say, “The baldest guesswork would suggest that [Andre’s] earlier intimations of modular and isometric structures found abundant examples among the boxcars and crossties of New Jersey.”
Hollis Frampton and other commentators who have said comparable things are right. Railroad ties and boxcars are modular, but there is more to be said about them.
As important to Andre’s work as these examples of modules, if not more so, is the set of relations of which they are part. Ties are laid at a regular interval and span the relatively narrow width between two long parallel lines that the rails form. The rails are almost, but not exactly, at the ends of the ties that connect them. Aside from this detail, railroad tracks and a strip of motion picture film are schematically the same. Both consist of a module repeated at a regular interval, and the modules are bounded by two parallel lines. And in both railroads and film, the width is called the gauge.
Boxcars when they are coupled together are more adjacent to each other in relation to their size than railroad ties. The couplings connect them to each other, but even so their edges (in this case their ends) are not aligned with each other; they are separated by the gap that the couplings occupy.
There can be little doubt that Andre’s daily familiarity in the early 1960s with the modules of railroad ties and boxcars helped to lay the foundation for the work that was to come. But it might be more accurate to say that what Andre found on the railroad was a combination of two relations among parts that he had already made the basis of some of his work.
As I noted, Andre had earlier made work that was stack-like and, as I suggested, he probably used that term to describe them. In the Pyramid series (1959), the ends of the elements are aligned, and their edges touch. But the elements are stepped in relation to each other, so they are not aligned edge to edge. And because the elements are not identical, they cannot be called modules. Untitled (1960) is a spiral of rectangular steel elements that are identical and so are modules, but neither the ends of the modules nor their edges are aligned.
These pieces between them contain two of the three relations of parts that together are the foundation of Andre’s classic work: the module, and the ends of elements being in alignment. But one relation is in one group of works, and the other relation is in a separate work. The two relations are not combined with each other in a single work. And in any event there is a third relation that is missing: modules aligned edge to edge.
Railroad ties combine the two relations that Andre had used separately. Ties are modules and their ends are aligned. But ties are not aligned edge to edge. Between the ties there is a gap. Boxcars are modules, but their edges (their ends) are not aligned with each other. Still missing is the third relation, modules aligned edge to edge. To remind ourselves, and to look ahead a little, when the edges of modules are aligned, their free edges are also aligned. When bricks are in a stack, their ends line up. So this third relation of edges in alignment contains within it the relation of ends in alignment that Andre had already used in his work and had seen on the railroad.
In the early 1960s Hollis Frampton started to work with 16mm film. I suggest that Andre’s own work and, even more so, his work on the railroad prepared him to recognize as already familiar the relations of the parts that he found in film. As I noted, film resembles railroad tracks in consisting of a series of modules repeated at a regular interval bounded by two parallel lines. The width of railroad tracks is called the gauge, and so is the width of film. I suggest that Andre recognized in the relations between the frames of 16mm film the two devices he had used separately in his work, modules and elements with their ends aligned, and that he had found on the railroad in a combined form, modules with their ends aligned.
But Andre also found something in 16mm film that he had not used in his earlier work and that he had not found on the railroad. In his work and in railroad ties the elements or modules are not aligned edge to edge. But in 16mm film they are. The frames in 16mm film are exactly aligned edge to edge in the plane of the film. And so the free edges of the frames are aligned as well. The line between the frames is so narrow that it is no more than a boundary between two adjacent modules. The equivalent in Andre’s work of the lines between the frames is the joints between the modules as they are exactly aligned edge to edge.
The set of relations between the modules in Andre’s stacks repeats the set of relations between the modules in 16mm film. And the stack implies the grid. With the discovery of modules aligned edge to edge in 16mm film, Andre found the third and last term in the set of relations that are the foundation of his classic work.