Film Form

by Sergei Eisenstein

  Montage thinking—the height of differentiatedly sensing and resolving the “organic” world—is realized anew in a mathematic faultlessly performing instrument-machine.

  Recalling the words of Kleist, so close to the Kabuki theater, which was born from marionettes:

  . . . [grace] appears best in that human bodily structure which has no consciousness at all, or has an infinite consciousness—that is, in the mechanical puppet, or in the god.7

  Extremes meet. . . .

  Nothing is gained by whining about the soullessness of Kabuki or, still worse, by finding in Sadanji’s acting a “confirmation of the Stanislavsky theory”! Or in looking for what “Meyerhold hasn’t yet stolen”!

  Let us rather—hail the junction of Kabuki and the soundfilm!

  The Cinematographic Principle and the Ideogram

  IT IS a weird and wonderful feat to have written a pamphlet on something that in reality does not exist. There is, for example, no such thing as a cinema without cinematography. And yet the author of the pamphlet preceding this essay* has contrived to write a book about the cinema of a country that has no cinematography. About the cinema of a country that has, in its culture, an infinite number of cinematographic traits, strewn everywhere with the sole exception of—its cinema.

  This essay is on the cinematographic traits of Japanese culture that lie outside the Japanese cinema, and is itself as apart from the preceding pamphlet as these traits are apart from the Japanese cinema.

  Cinema is: so many corporations, such and such turnovers of capital, so and so many stars, such and such dramas.

  Cinematography is, first and foremost, montage.

  The Japanese cinema is excellently equipped with corporations, actors, and stories. But the Japanese cinema is completely unaware of montage. Nevertheless the principle of montage can be identified as the basic element of Japanese representational culture.

  Writing—for their writing is primarily representational.

  The hieroglyph.

  The naturalistic image of an object, as portrayed by the skilful Chinese hand of Ts’ang Chieh 2650 years before our era, becomes slightly formalized and, with its 539 fellows, forms the first “contingent” of hieroglyphs. Scratched out with a stylus on a slip of bamboo, the portrait of an object maintained a resemblance to its original in every respect.

  But then, by the end of the third century, the brush is invented. In the first century after the “joyous event” (A.D.)—paper. And, lastly, in the year 220—India ink.

  A complete upheaval. A revolution in draughtsmanship. And, after having undergone in the course of history no fewer than fourteen different styles of handwriting, the hieroglyph crystallized in its present form. The means of production (brush and India ink) determined the form.

  The fourteen reforms had their way. As a result:

  In the fierily cavorting hieroglyph ma (a horse) it is already impossible to recognize the features of the dear little horse sagging pathetically in its hindquarters, in the writing style of Ts’ang Chieh, so well-known from ancient Chinese bronzes.

  But let it rest in the Lord, this dear little horse, together with the other 607 remaining hsiarig cheng symbols—the earliest extant category of hieroglyphs.

  The real interest begins with the second category of hieroglyphs—the huei-i, i.e., “copulative.”

  The point is that the copulation (perhaps we had better say, the combination) of two hieroglyphs of the simplest series is to be regarded not as their sum, but as their product, i.e., as a value of another dimension, another degree; each, separately, corresponds to an object, to a fact, but their combination corresponds to a concept. From separate hieroglyphs has been fused—the ideogram. By the combination of two “depictables” is achieved the representation of something that is graphically undepictable.

  For example: the picture for water and the picture of an eye signifies “to weep”; the picture of an ear near the drawing of a door = “to listen”; a dog + a mouth = “to bark”; a mouth + a child = “to scream”; a mouth + a bird = “to sing”; a knife + a heart = “sorrow,” and so on.1

  But this is—montage!

  Yes. It is exactly what we do in the cinema, combining shots that are depictive, single in meaning, neutral in content—into intellectual contexts and series.

  This is a means and method inevitable in any cinematographic exposition. And, in a condensed and purified form, the starting point for the “intellectual cinema.”

  For a cinema seeking a maximum laconism for the visual representation of abstract concepts.

  And we hail the method of the long-lamented Ts’ang Chieh as a first step along these paths.

  We have mentioned laconism. Laconism furnishes us a transition to another point. Japan possesses the most laconic form of poetry: the haikai (appearing at the beginning of the thirteenth century and known today as “haiku” or “hokku”) and the even earlier tanka (mythologically assumed to have been created along with heaven and earth).

  Both are little more than hieroglyphs transposed into phrases. So much so that half their quality is appraised by their calligraphy. The method of their resolution is completely analogous to the structure of the ideogram.

  As the ideogram provides a means for the laconic imprinting of an abstract concept, the same method, when transposed into literary exposition, gives rise to an identical laconism of pointed imagery.

  Applied to the collision of an austere combination of symbols this method results in a dry definition of abstract concepts. The same method, expanded into the luxury of a group of already formed verbal combinations, swells into a splendor of imagist effect.

  The concept is a bare formula; its adornment (an expansion by additional material) transforms the formula into an image—a finished form.

  Exactly, though in reverse, as a primitive thought process—imagist thinking, displaced to a definite degree, becomes transformed to conceptual thinking.

  But let us turn to examples.

  The haiku is a concentrated impressionist sketch:

  A lonely crow

  On leafless bough,

  One autumn eve.

  BASHŌ

  What a resplendent moon!

  It casts the shadow of pine boughs

  Upon the mats.

  KIKAKU

  An evening breeze blows.

  The water ripples

  Against the blue heron’s legs.

  BUSON

  It is early dawn.

  The castle is surrounded

  By the cries of wild ducks.

  KYOROKU2

  The earlier tanka is slightly longer (by two lines):

  O mountain pheasant

  long are the feathers trail’st thou

  on the wooded hill-side—

  as long the nights seem to me

  on lonely couch sleep seeking.

  HITOMARO[?]3

  From our point of view, these are montage phrases. Shot lists. The simple combination of two or three details of a material kind yields a perfectly finished representation of another kind—psychological.

  And if the finely ground edges of the intellectually defined concepts formed by the combined ideograms are blurred in these poems, yet, in emotional quality, the concepts have blossomed forth immeasurably. We should observe that the emotion is directed towards the reader, for, as Yone Noguchi has said, “it is the readers who make the haiku’s imperfection a perfection of art.” 4

  It is uncertain in Japanese writing whether its predominating aspect is as a system of characters (denotative), or as an independent creation of graphics (depictive). In any case, born of the dual mating of the depictive by method, and the denotative by purpose, the ideogram continued both these lines (not consecutive historically but consecutive in principle in the minds of those developing the method).

  Not only did the denotative line continue into literature, in the tanka, as we have shown, but exactly the same method (in its depictive aspect) operates also in the most perfect e
xamples of Japanese pictorial art.

  Sharaku—creator of the finest prints of the eighteenth century, and especially of an immortal gallery of actors’ portraits. The Japanese Daumier. Despite this, almost unknown to us. The characteristic traits of his work have been analyzed only in our century. One of these critics, Julius Kurth, in discussing the question of the influence on Sharaku of sculpture, draws a parallel between his wood-cut portrait of the actor Nakayama Tomisaburō and an antique mask of the semi-religious Nō theater, the mask of a Rozo.

  The faces of both the print and the mask wear an identical expression. . . . Features and masses are similarly arranged although the mask represents an old priest, and the print a young woman. This relationship is striking, yet these two works are otherwise totally dissimilar; this in itself is a demonstration of Sharaku’s originality. While the carved mask was constructed according to fairly accurate anatomical proportions, the proportions of the portrait print are simply impossible. The space between the eyes comprises a width that makes mock of all good sense. The nose is almost twice as long in relation to the eyes as any normal nose would dare to be, and the chin stands in no sort of relation to the mouth; the brows, the mouth, and every feature—is hopelessly misrelated. This observation may be made in all the large heads by Sharaku. That the artist was unaware that all these proportions are false is, of course, out of the question. It was with a full awareness that he repudiated normalcy, and, while the drawing of the separate features depends on severely concentrated naturalism, their proportions have been subordinated to purely intellectual considerations. He set up the essence of the psychic expression as the norm for the proportions of the single features.5

  Is not this process that of the ideogram, combining the independent “mouth” and the dissociated symbol of “child” to form the significance of “scream”?

  Is this not exactly what we of the cinema do temporally, just as Sharaku in simultaneity, when we cause a monstrous disproportion of the parts of a normally flowing event, and suddenly dismember the event into “close-up of clutching hands,” “medium shots of the struggle,” and “extreme close-up of bulging eyes,” in making a montage disintegration of the event in various planes? In making an eye twice as large as a man’s full figure?! By combining these monstrous incongruities we newly collect the disintegrated event into one whole, but in our aspect. According to the treatment of our relation to the event.

  The disproportionate depiction of an event is organically natural to us from the beginning. Professor Luriya, of the Psychological Institute in Moscow, has shown me a drawing by a child of “lighting a stove.” Everything is represented in passably accurate relationship and with great care. Firewood. Stove. Chimney. But what are those zigzags in that huge central rectangle? They turn out to be—matches. Taking into account the crucial importance of these matches for the depicted process, the child provides a proper scale for them.*67

  The representation of objects in the actual (absolute) proportions proper to them is, of course, merely a tribute to orthodox formal logic. A subordination to an inviolable order of things.

  Both in painting and sculpture there is a periodic and invariable return to periods of the establishment of absolutism. Displacing the expressiveness of archaic disproportion for regulated “stone tables” of officially decreed harmony.

  Absolute realism is by no means the correct form of perception. It is simply the function of a certain form of social structure. Following a state monarchy, a state uniformity of thought is implanted. Ideological uniformity of a sort that can be developed pictorially in the ranks of colors and designs of the Guards regiments . . .

  Thus we have seen how the principle of the hieroglyph—“denotation by depiction”—split in two: along the line of its purpose (the principle of “denotation”), into the principles of creating literary imagery; along the line of its method of realizing this purpose (the principle of “depiction”), into the striking methods of expressiveness used by Sharaku.*

  And, just as the two outspreading wings of a hyperbola meet, as we say, at infinity (though no one has visited so distant a region!), so the principle of hieroglyphics, infinitely splitting into two parts (in accordance with the function of symbols), unexpectedly unites again from this dual estrangement, in yet a fourth sphere—in the theater.

  Estranged for so long, they are once again—in the cradle period of the drama—present in a parallel form, in a curious dualism.

  The significance (denotation) of the action is effected by the reciting of the Jōruri by a voice behind the stage—the representation (depiction) of the action is effected by silent marionettes on the stage. Along with a specific manner of movement this archaism migrated into the early Kabuki the ater, as well. To this day it is preserved, as a partial method, in the classical repertory (where certain parts of the action are narrated from behind the stage while the actor mimes).

  But this is not the point. The most important fact is that into the technique of acting itself the ideographic (montage) method has been wedged in the most interesting ways.

  However, before discussing this, let us be allowed the luxury of a digression—on the matter of the shot, to settle the debated question of its nature, once and for all.

  A shot. A single piece of celluloid. A tiny rectangular frame in which there is, organized in some way, a piece of an event.

  “Cemented together, these shots form montage. When this is done in an appropriate rhythm, of course!”

  This, roughly, is what is taught by the old, old school of film-making, that sang:

  “Screw by screw,

  Brick by brick . . .”

  Kuleshov, for example, even writes with a brick:

  If you have an idea-phrase, a particle of the story, a link in the whole dramatic chain, then that idea is to be expressed and accumulated from shot-ciphers, just like bricks.8

  “The shot is an element of montage. Montage is an assembly of these elements.” This is a most pernicious make-shift analysis.

  Here the understanding of the process as a whole (connection, shot-montage) derives only from the external indications of its flow (a piece cemented to another piece). Thus it would be possible, for instance, to arrive at the well-known conclusion that street-cars exist in order to be laid across streets. An entirely logical deduction, if one limits oneself to the external indications of the functions they performed during the street-fighting of February, 1917, here in Russia. But the materialist conception of history interprets it otherwise.

  The worst of it is that an approach of this kind does actually lie, like an insurmountable street-car, across the potentialities of formal development. Such an approach overrules dialectical development, and dooms one to mere evolutionary “perfecting,” in so far as it gives no bite into the dialectical substance of events.

  In the long run, such evolutionizing leads either through refinement to decadence or, on the other hand, to a simple withering away due to stagnation of the blood.

  Strange as it may seem, a melodious witness to both these distressing eventualities, simultaneously, is Kuleshov’s latest film, The Gay Canary [1929].

  The shot is by no means an element of montage.

  The shot is a montage cell.

  Just as cells in their division form a phenomenon of another order, the organism or embryo, so, on the other side of the dialectical leap from the shot, there is montage.

  By what, then, is montage characterized and, consequently, its cell—the shot?

  By collision. By the conflict of two pieces in opposition to each other. By conflict. By collision.

  In front of me lies a crumpled yellowed sheet of paper. On it is a mysterious note:

  “Linkage—P” and “Collision—E.”

  This is a substantial trace of a heated bout on the subject of montage between P (Pudovkin) and E (myself).

  This has become a habit. At regular intervals he visits me late at night and behind closed doors we wrangle over matters of principle. A
graduate of the Kuleshov school, he loudly defends an understanding of montage as a linkage of pieces. Into a chain. Again, “bricks.” Bricks, arranged in series to expound an idea.

  I confronted him with my viewpoint on montage as a collision. A view that from the collision of two given factors arises a concept.

  From my point of view, linkage is merely a possible special case.

  Recall what an infinite number of combinations is known in physics to be capable of arising from the impact (collision) of spheres. Depending on whether the spheres be resilient, non-resilient, or mingled. Amongst all these combinations there is one in which the impact is so weak that the collision is degraded to an even movement of both in the same direction.

  This is the one combination which would correspond with Pudovkin’s view.

  Not long ago we had another talk. Today he agrees with my point of view. True, during the interval he took the opportunity to acquaint himself with the series of lectures I gave during that period at the State Cinema Institute. . . .

  So, montage is conflict.