CLEMENT GREENBERG
COLLAGE
The quintessential essay on Cubism and probably the most important single essay about 20th
Century painting.
-- TF
Pablo Picasso, Still Life with Piano, 1911
COLLAGE WAS A major turning point in the evolution of Cubism, and therefore a major turning
point in the whole evolution of modernist art in this century. Who invented collage--Braque or
Picasso--and when is still not settled. Both artists left most of the work they did between I907
and 1914 undated as well as unsigned; and each claims, or implies the claim, that his was the
first collage of all. That Picasso dates his, in retrospect, almost a year earlier than Braque's compounds the difficulty. Nor does the internal or stylistic evidence help enough, given that the
interpretation of Cubism is still on a rudimentary level.
The question of priority is much less important, however, than that of the motives which first
induced either artist to paste or glue a piece of extraneous material to the surface of a picture.
About this, neither Braque nor Picasso has made him-self at all clear. The writers who have tried
to explain their intentions for them speak, with a unanimity that is suspect in itself, of the need
for renewed contact with "reality" in face of the growing abstractness of Analytical Cubism. But
the term "reality," always ambiguous when used in connection with art, has never been used
more ambiguously than here. A piece of imitation-woodgrain wallpaper is not more "real" under
any definition, or closer to nature, than a painted simulation of it; nor is wallpaper, oilcloth,
newspaper or wood more "real," or closer to nature, than paint on canvas. And even if these
materials were more "real," the question would still be begged, for "reality" would still explain
next to nothing about the actual appearance of the Cubist collage.
There is no question but that Braque and Picasso were concerned, in their Cubism, with holding
on to painting as an art of representation and illusion. But at first they were more crucially
concerned, in and through their Cubism, with obtaining sculptural results by strictly
nonsculptural means; that is, with finding for every aspect of three-dimensional vision an
explicitly two-dimensional equivalent, regardless of how much verisimilitude might suffer in the
process. Painting had to spell out, rather than pretend to deny, the physical fact that it was flat,
even though at the same time it had to overcome this proclaimed flatness as an aesthetic fact and
continue to report nature.
Neither Braque nor Picasso set himself this program in advance. It emerged, rather, as something
implicit and inevitable in the course of their joint effort to fill out that vision of a "purer"
pictorial art which they had glimpsed in Cézanne, from whom they also took their means. These
means, as well as the vision, imposed their logic; and the direction of that logic became
completely clear in 191 I, the fourth year of Picasso's and Braque's Cubism, along with certain
contradictions latent in the Cézannian vision itself.
By that time, flatness had not only invaded but was threatening to swamp the Cubist picture. The
little facet-planes into which Braque and Picasso were dissecting everything visible now all lay
parallel to the picture plane. They were no longer controlled, either in drawing or in placing, by
linear or even scalar perspective. Each facet tended to be shaded, moreover, as an independent
unit, with no legato passages, no unbroken tracts of value gradation on its open side, to join it to
adjacent facet-planes. At the same time, shading had itself been atomized into flecks of light and
dark that could no longer be concentrated upon the edges of shapes with enough modeling force
to turn these convincingly into depth. Light and dark in general had begun to act more
immediately as cadences of design than as plastic description or definition. The main problem at
this juncture became to keep the "inside" of the picture--its content--from fusing with the
"outside"--its literal surface. Depicted flatness--that is, the facet-planes--had to be kept separate
enough from literal flatness to permit a minimal illusion of three-dimensional space to survive
between the two
Braque had already been made uncomfortable by the contraction of illusioned space in his
pictures of 1910. The expedient he had then hit upon was to insert a conventional, trompe-l'oeil
suggestion of deep space on top of Cubist flatness, between the depicted planes and the
spectator's eye. The very un-Cubist graphic tack-with-a-cast-shadow, shown transfixing the top
of a 1910 painting, Still Life with Violin and Pitcher, suggests deep space in a token way, and
destroys the surface in a token way. The Cubist forms are converted into the illusion of a picture
within a picture. In the Man with a Guitar of early 1911 (in the Museum of Modern Art), the
line-drawn tassel-and-stud in the upper left margin is a similar token. The effect, as distinct from
the signification, is in both cases very discreet and inconspicuous. Plastically, spatially, neither
the tack nor the tassel-and-stud acts upon the picture; each suggests illusion without making it
really present.
Georges Braque, Man with a Guitar, 1911
Early in 1911, Braque was already casting around for ways of reinforcing, or rather
supplementing, this suggestion, but still without introducing anything that would become more
than a token. It was then, apparently, that he discovered that trompe-l'oeil could be used to
undeceive as well as to deceive the eye. It could be used, that is, to declare as well as to deny the
actual surface. If the actuality of the surface--its real, physical flatness--could be indicated
explicitly enough in certain places, it would be distinguished and separated from everything else
the surface contained. Once the literal nature of the support was advertised, whatever upon it was
not intended literally would be set off and enhanced in its nonliteralness. Or to put it in still
another way: depicted flatness would inhabit at least the semblance of a semblance of threedimensional-space as long as the brute, undepicted flatness of the literal surface was pointed to
as being still flatter.
The first and, until the advent of pasted paper, the most important device that Braque discovered
for indicating and separating the surface was imitation printing, which automatically evokes a
literal flatness. Block letters are seen in one of his 1910 paintings, The Match Holder; but being
done rather sketchily, and slanting into depth along with the depicted surface that bears them,
they merely allude to, rather than state, the literal surface. Only in the next year are block
capitals, along with lower-case letters and numerals, introduced in exact simulation of printing
and stenciling, in absolute frontality and outside the rcpresentational context of the picture.
Wherever this printing appears, it stops the eye at the literal plane, just as the artist's signature
would. By force of contrast alone--for wherever the literal surface is not explicitly stated, it
seems implicitly denied--everything else is thrust back into at least a memory of deep or plastic
space. It is the old device of the repoussoir, but taken a step further: instead of being used to push
an illusioned middleground farther away from an illusioned foreground, the imitation printing
spells out the real paint surface and thereby pries it away from the illusion of depth.
The eye-undeceiving trompe-l'oeil of simulated typography supplements, rather than replaces,
the conventional eye-deceiving kind. Another literally and graphically rendered tassel-and-stud
imbeds flattened forms in token depth in Braque's Portuguese (1911), but this time the brute
reality of the surface, as asserted by stenciled letters and numerals, closes over both the token
illusion of depth and the Cubist configurations like the lid on a box. Sealed between two parallel
flatnesses--the depicted Cubist flatness and the literal flatness of the paint surface--the illusion is
made a little more present but, at the same time, even more ambiguous. As one looks, the
stenciled letters and numerals change places in depth with the tassel-and-stud, and the physical
surface itself becomes part of the illusion for an instant: it seems pulled back into depth along
with the stenciling, so that once again the picture plane seems to be annihilated--but only for the
fraction of another instant. The abiding effect is of a constant shuttling between surface and
depth, in which the depicted flatness is "infected" by the undepicted. Rather than being deceived,
the eye is puzzled; instead of seeing objects in space, it sees nothing more than--a picture.
Through 1911 and 1912, as the Cubist facet-plane's tendency to adhere to the literal surface
became harder and harder to deny, the task of keeping the surface at arm's length fell all the more
to eye-undeceiving contrivances. To reinforce, and sometimes to replace, the simulated
typography, Braque and Picasso began to mix sand and other foreign substances with their paint;
the granular texture thus created likewise called attention to the reality of the surface and was
effective over much larger areas. In certain other pictures, however, Braque began to paint areas
in exact simulation of wood graining or marbleizing. These areas by virtue of their abrupt density
of pattern, stated the literal surface with such new and superior force that the resulting contrast
drove the simulated printing into a depth from which it could be rescued--and set to shuttling
again--only by conventional perspective; that is, by being placed in such relation to the forms
depicted within the illusion that these forms left no room for the typography except near the
surface.
The accumulation of such devices, however, soon had the effect of telescoping, even while
separating, surface and depth. The process of flattening seemed inexorable, and it became
necessary to emphasize the surface still further in order to prevent it from fusing with the
illusion. It was for this reason, and no other that I can see, that in September 1912, Braque took
the radical and revolutionary step of pasting actual pieces of imitation-woodgrain wallpaper to a
drawing on paper, instead of trying to simulate its texture in paint. Picasso says that he himself
had already made his first collage toward the end of 1911, when he glued a piece of imitationcaning oilcloth to a painting on canvas. It is true that his first collage looks more Analytical than
Braque's, which would confirm the date he assigns it. But it is also true that Braque was the
consistent pioneer in the use of simulated textures as well as of typography; and moreover, he
had already begun to broaden and simplify the facet-planes of Analytical Cubism as far back as
the end of 1910.
Pablo Picasso, Man in Hat, Charcoal and collage, 1912
When we examine what each master says was his first collage we see that much the same thing
happens in each. (It makes no real difference that Braque's collage is on paper and eked out in
charcoal, while Picasso's is on canvas and eked out in oil.) By its greater corporeal presence and
its greater extraneousness, the affixed paper or cloth serves for a seeming moment to push
everything else into a more vivid idea of depth than the simulated printing or simulated textures
had ever done. But here again, the surface-declaring device both overshoots and falls short of its
aim. For the illusion of depth created by the contrast between the affixed material and everything
else gives way immediately to an illusion of forms in bas-relief, which gives way in turn, and
with equal immediacy, to an illusion that seems to contain both--or neither.
Because of the size of the areas it covers, the pasted paper establishes undepicted flatness bodily,
as more than an indication or sign. Literal flatness now tends to assert itself as the main event of
the picture, and the device boomerangs: the illusion of depth is rendered even more precarious
than before. Instead of isolating the literal flatness by specifying and circumscribing it, the pasted
paper or cloth releases and spreads it, and the artist seems to have nothing left but this undepicted
flatness with which to finish as well as start his picture. The actual surface becomes both ground
and background, and it turns out--suddenly and paradoxically--that the only place left for a threedimensional illusion is in front of, upon, the surface. In their very first collages, Braque and
Picasso draw or paint over and on the affixed paper or cloth, so that certain of the principal
features of their subjects as depicted seem to thrust out into real, bas-relief space--or to be about
to do so--while the rest of the subject remains imbedded in, or flat upon, the surface. And the
surface is driven back, in its very surfaceness, only by this contrast.
In the upper center of Braque's first collage, Fruit Dish (in Douglas Cooper's collection), a bunch
of grapes is rendered with such conventionally vivid sculptural effect as to lift it practically off
the picture plane. The trompe-l'oeil illusion here is no longer enclosed between parallel
flatnesses, but seems to thrust through the surface of the drawing paper and establish depth on
top of it. Yet the violent immediacy of the wallpaper strips pasted to the paper, and the only
lesser immediacy of block capitals that simulate window lettering, manage somehow to push the
grape cluster back into place on the picture plane so that it does not "jump." At the same time,
the wallpaper strips themselves seem to be pushed into depth by the lines and patches of shading
charcoaled upon them, and by their placing in relation to the block capitals; and these capitals
seem in turn to be pushed back by their placing, and by contrast with the corporeality of the
woodgraining. Thus every part and plane of the picture keeps changing place in relative depth
with every other part and plane; and it is as if the only stable relation left among the different
parts of the picture is the ambivalent and ambiguous one that each has with the surface. And the
same thing, more or less, can be said of the contents of Picasso's first collage.
In later collages of both masters, a variety of extraneous materials are used, sometimes in the
same work, and almost always in conjunction with every other eye-deceiving and eyeundeceiving device they can think of. The area adjacent to one edge of a piece of affixed
material--or simply of a painted in form--will be shaded to pry that edge away from the surface,
while something will be drawn, painted or even pasted over another part of the same shape to
drive it back into depth. Planes defined as parallel to the surface also cut through it into real
space, and a depth is suggested optically which is greater than that established pictorially. All
this expands the oscillation between surface and depth so as to encompass fictive space in front
of the surface as well as behind it. Flatness may now monopolize everything, but it is a flatness
become so ambiguous and expanded as to turn into illusion itself--at least an optical if not,
properly speaking, a pictorial illusion. Depicted, Cubist flatness is now almost completely
assimilated to the literal, undepicted kind, but at the same time it reacts upon and largely
transforms the undepicted kind--and it does so, moreover, without depriving the latter of its
literalness; rather, it underpins and reinforces that literalness, recreates it.
Out of this re-created literalness, the Cubist subject re-emerged. For it had turned out, by a
further paradox of Cubism, that the means to an illusion of depth and plasticity had now become
widely divergent from the means of representation or imaging. In the Analytical phase of their
Cubism, Braque and Picasso had not only had to minimize three-dimensionality simply in order
to preserve it; they had also had to generalize it--to the point, finally, where the illusion of depth
and relief became abstracted from specific three-dimensional entities and was rendered largely as
the illusion of depth and relief as such: as a disembodied attribute and expropriated property
detached from everything not itself. In order to be saved, plasticity had had to be isolated; and as
the aspect of the subject was transposed into those clusters of more or less interchangeable and
contour--obliterating facet-planes by which plasticity was isolated under the Cubist method, the
subject itself became largely unrecognizable. Cubism, in its 1911-1912 phase (which the French,
with justice, call "hermetic") was on the verge of abstract art.
It was then that Picasso and Braque were confronted with a unique dilemma: they had to choose
between illusion and representation. If they opted for illusion, it could only be illusion per se--an
illusion of depth, and of relief, so general and abstracted as to exclude the representation of
individual objects. If, on the other hand, they opted for representation, it had to be representation
per se--representation as image pure and simple, without connotations (at least, without more
than schematic ones) of the three-dimensional space in which the objects represented originally
existed. It was the collage that made the terms of this dilemma clear: the representational could
be restored and preserved only on the flat and literal surface now that illusion and representation
had become, for the first time, mutually exclusive alternatives.
In the end, Picasso and Braque plumped for the representational, and it would seem they did so
deliberately. (This provides whatever real justification there is for the talk about "reality.") But
the inner, formal logic of Cubism, as it worked itself out through the collage, had just as much to
do with shaping their decision. When the smaller facet-planes of Analytical Cubism were placed
upon or juxtaposed with the large, dense shapes formed by the affixed materials of the collage,
they had to coalesce--become "synthesized"--into larger planar shapes themselves simply in
order to maintain the integrity of the picture plane. Left in their previous atom-like smallness,
they would have cut away too abruptly into depth; and the broad, opaque shapes of pasted paper
would have been isolated in such a way as to make them jump out of plane. Large planes
juxtaposed with other large planes tend to assert themselves as independent shapes, and to the
extent that they are flat, they also assert themselves as silhouettes; and independent silhouettes
are apt to coincide with the recognizable contours of the subject from which a picture starts (if it
does start from a subject). It was because of this chain-reaction as much as for any other reason-that is, because of the growing independence of the planar unit in collage as a shape--that the
identity of depicted objects, or at least parts of them, re-emerged in Braque's and Picasso's
papiers collés and continued to remain more conspicuous there--but only as flattened silhouettes-than in any of their paintings done wholly in oil before the end of 1913.
Analytical Cubism came to an end in the collage, but not conclusively; nor did Synthetic Cubism
fully begin there. Only when the collage had been exhaustively translated into oil, and
transformed by this translation, did Cubism become an affair of positive color and flat,
interlocking silhouettes whose legibility and placement created allusions to, if not the illusion of,
unmistakable three-dimensional identities.
Synthetic Cubism began with Picasso alone, late in 1913 or early in 1914; this was the point at
which he finally took the lead in Cubist innovation away from Braque, never again to relinquish
it. But even before that, Picasso had glimpsed and entered, for a moment, a certain revolutionary
path in which no one had preceded him. It was as though, in that instant, he had felt the flatness
of collage as too constricting and had suddenly tried to escape all the way back--or forward--to
literal three-dimensionality. This he did by using utterly literal means to carry the forward push
of the collage (and of Cubism in general) literally into the literal space in front of the picture
plane.
Some time in 1912, Picasso cut out and folded a piece of paper in the shape of a guitar; to this he
glued and fitted other pieces of paper and four taut strings, thus creating a sequence of flat
surfaces in real and sculptural space to which there clung only the vestige of a picture plane. The
affixed elements of collage were extruded, as it were, and cut off from the literal pictorial surface
to form a bas-relief. By this act he founded a new tradition and genre of sculpture, the one that
came to be called "construction." Though construction, sculpture was freed long ago from strict
bas-relief frontality, it has continued to be marked by its pictorial origins, so that the sculptorconstructor Gonzalez, Picasso's friend, could refer to it as the new art of "drawing in space"--that
is, of manipulating two-dimensional forms in three-dimensional space. (Not only did Picasso
found this "new" art with his paper guitar of 1912, but he went on, some years afterwards, to
make some of the strongest as well as most germinative contributions to it.)
Neither Picasso nor Braque ever really returned to collage after 1914. The others who have taken
it up have exploited it largely for its shock value, which collage had only incidentally--or even
only accidentally--in the hands of its originators. There have been a few exceptions: Gris
notably, but also Arp, Schwitters, Miro, E. L. T. Mesens, Dubuffet and, in this country, Robert
Motherwell and Anne Ryan. In this context, Gris's example remains the most interesting and
most instructive.
Braque and Picasso had obtained a new, self-transcending kind of decoration by reconstructing
the picture surface with what had once been the means of its denial. Starting from illusion, they
had arrived at a transfigured, almost abstract kind of literalness. With Gris it was the opposite.
As he himself explained, he started out with abstract flat shapes, to which he then fitted
recognizable images and emblems of three-dimensionality. And whereas Braque's and Picasso's
subjects were dissected in three dimensions in the course of being transposed into two, Gris's
first Cubist subjects tended--even before they were fitted into the picture, and as if preformed by
its surface--to be analyzed in two-dimensional and purely decorative rhythms. It was only later
that he became more aware of the fact that Cubism was not a question of decorative overlay and
that the resonance of its surfaces derived from an abiding concern with plasticity and illusion
which informed the very renunciation of plasticity and illusion.
In his collages almost more than anywhere else, we see Gris trying to solve the problems
proposed by this fuller awareness. But his collages also show the extent to which his awareness
remained incomplete. Because he continued to take the picture plane as given and therefore not
needing to be re-created, Gris became perhaps too solicitous about the illusion. He used pasted
paper and trompe-l'oeil textures and lettering to assert flatness, but he almost always completely
sealed the flatness inside an illusion of conventional depth by allowing images rendered with
relatively sculptural vividness to occupy, unambiguously, too much of both the nearest and
farthest planes.
Because he shaded and modeled more abundantly and tended to use more explicit color under his
shading, Gris's collages seldom declare their surfaces as forthrightly as Picasso's and Braque's
collages do. Their total presence is thus less immediate and has something about it of the
removedness, the closed-off presence, of the traditional picture. And yet, because their
decorative elements function to a greater extent solely as decoration, Gris's collages also seem
more conventionally decorative. Instead of the seamless fusion of the decorative with the plastic
that we get in Picasso and Braque, there is an alternation, a collocation, a mere juxtaposition of
the two; and whenever this relation goes beyond juxtaposition, it leads more often to confusion
than to fusion. Gris's collages have their merits, but only a few of them deserve the unqualified
praise they have received.
Juan Gris, Le Journal, 1916
But many of Gris's oils of 1915-1918 do deserve their praise. In all Justice, it should be pointed
out that his paintings of those years demonstrate, perhaps more clearly than anything by Braque
or Picasso, something which is of the highest importance to Cubism and to the collage's effect
upon it: namely, the liquidation of sculptural shading.
In Braque's and Picasso's very first papiers collés, shading stops being pointillist and suddenly
becomes broad and incisive again, like the shapes it modifies. This change in shading also
accounts for the bas-relief effects, or the velleities to bas-relief, of the first collages. But large
patches of shading on a densely or emphatically patterned ground, such as woodgrain or
newsprint, tend to take off on their own when their relation to the model in nature is not selfevident, just the way large planes do under the same circumstances. They abandon their
sculptural function and become independent shapes constituted by blackness or grayness alone.
Not only did this fact contribute further to the ambiguity of the collage's surface; it also served
further to reduce shading to a mere component of surface design and color scheme. When
shading becomes that, all other colors become more purely color. It was in this way that positive
color re-emerged in the collage--recapitulating, curiously enough, the way "pure" color had
emerged in the first place for Manet and the Impressionists.
In Analytical Cubism, shading as shading had been divorced from specific shapes while retaining
in principle the capacity to inflect generalized surfaces into depth. In collage, shading, though
restored to specific shapes or silhouettes, lost its power to act as modeling because it became a
specific shape in itself. This is how and why shading, as a means to illusion, disappeared from
the collages of Braque and Picasso, and from their Cubism, never really to reappear. But it was
left to Gris, in his pictures of 1915-1918, to elucidate this process and its consequences for all to
see--and, in doing so, to produce, finally, triumphant art. Gris's Cubism in this period--which is
almost as much Analytical as it is Synthetic--separated, fixed and immobilized, in oil on wood or
canvas, some of the over-lapping stages of the transformation that Cubism had already
undergone in Braque's and Picasso's glued and pasted pictures. The cleanly and simply contoured
solid black shapes on which Gris relied so much in these paintings represent fossilized shadows
and fossilized patches of shading. All the value gradations are summed up in a single, ultimate
value of flat, opaque black--a black that becomes a color as sonorous and pure as any spectrum
color and that confers upon the silhouettes it fills an even greater weight than is possessed by the
lighter-hued forms which these silhouettes are supposed to shade.
In this phase alone does Gris's art, in my opinion, sustain the main tenor of Cubism. Here, at last,
his practice is so completely informed by a definite and steady vision that the details of execution
take care of themselves. And here, at last, the decorative is transcended and transfigured, as it
had already been in Picasso's, Braque's and Léger's art, in a monumental unity. This
monumentality has little to do with size. (Early and late, and whether in Picasso's hands or
Braque's, Cubism has never lent itself with entire success to an outsize format. Even Léger's
rather splendid big pictures of the late 1910S and early I920S do not quite match the perfection
of his smaller scale Cubism of 1910-1914.) The monumentality of Cubism in the hands of its
masters is more a question of a vision and attitude--an attitude toward the immediate physical
means of pictorial art--thanks to which easel paintings and even "sketches" acquire the selfevident self-sufficiency of architecture. This is as true of the Cubist collage as of anything else in
Cubism, and perhaps it is even truer of the collage than of anything else in Cubism.
I959
-- Art and Culture, substantially revised from an article in Art News, September, 1958