I think that henceforth it is no longer possible to begin a discourse on cinema as language without taking into account at least the terminology of semiotics. Indeed the problem, if one wishes to set it forth briefly, appears in the following way: whereas literary languages found their poetic inventions on the institutional basis of an instrumental language, quite common to all who speak, cinematic languages seem not to be founded on anything like this. For their real basis, they do not have a language whose primary objective is communication. Thus literary languages appear immediately as distinct, in their practice, from the pure and simple instrument which serves to communicate; while communication by means of cinema would seem arbitrary and devious, without such an instrumental basis used normally at all.
Men communicate with words, not with images; this is why a specific language of images would appear as a pure and artificial abstraction.
If this reasoning were correct, as it seems to be, cinema could not materially exist; or at the very least it would be only a monstrosity, a series of insignificant signs, Semiotics envisages sign-systems indifferently: it speaks, for example, of “systems of linguistic sign,” because these exist, but in fact this in no way excludes the theoretical possibility of other sign-systems, for example a system of signs by gestures, the more so as a complement to the spoken language. Indeed, a word (lin-segno, i.e. linguistic sign) pronounced with a certain facial expression takes a certain meaning, pronounced with another expression it takes another, perhaps even the opposite (especially if the speaker is from Naples). A word followed by a gesture has one meaning, followed by another gesture it has another, etc.
This system of signs by gestures which, in practice, accompanies the system of linguistic signs as its complement, can be isolated as an autonomous system and become the object of a study.
One can even suppose, by abstract hypothesis, the existence of a unique system of signs by gestures as unique instrument of communication for man (in sum: deaf and dumb Neapolitans): it is from such a hypothetical system of visual signs that language derives the foundation of its existence and the possibility of allowing the formation of a series of naturally communicative archetypes.
Of course, this would still not amount to much. But we must immediately add the intended recipient of the cinematic product is equally accustomed to visually “read” reality, that is to keep up a dialogue with the reality which surrounds him and which is used as the environment of a collectivity which can be felt even in the pure and simple manifestation of its acts, of its habits. The fact of walking alone in the street, even with our ears stopped up, constitutes a continual dialog between ourselves and an environment which expresses itself by the mediation of the images which compose it: the physiognomy of the passersby, their gestures, their signs, their actions, their silences, their expressions, their collective reactions (people waiting at red lights, a crowd around a street-accident or around a monument), besides, traffic signs, indicators, counterclockwise rotaries are in sum objects charged with meanings and which utter a brute “speech” by their very presence.
But there is more: in man, an entire world is expressed by means of significant images shall we therefore propose, by analogy, the term “im-signs” (imsegni, i.e. image-signs). This is the world of memory and of dreams.
Every attempt at memorization is a series of im-signs, that is primarily a cinema sequence. (“Where have I seen this person? Wait… I think it was at Zagora” — image of Zagora with its green palms against the pink soil — .walking with Abd Kader… — image of Abd El Kader and of the person in question walking past the encampment of the French outposts — etc.). And thus, all dreams are a series of im-signs which have all the characteristics of the cinematic sequence: close-ups, long shots, etc.
In sum, there is a whole complex world of significant images — formed as much of gestures and of all sorts of signs coming from the environment, as of memories or of dreams which is proposed as the “instrumental” foundation of cinematic communication, and prefigures it.
Here, we must immediately make a marginal observation: whereas the instruments of poetic or philosophical communication are already extremely perfected, truly form a historically complex system which has reached its maturity, those of the visual communication which is at the basis of cinematic language are altogether brute, instinctive. Indeed, gestures, the surrounding reality, as much as dreams and the mechanisms of memory, are of a virtually pre-human order, or at least at the limit of humanity in any case pre-grammatical and even premorphological (dreams are unconscious phenomena, as are mnemonic mechanisms; the gesture is an altogether elementary sign, etc.).
The linguistic instrument on which cinema is founded is thus of an irrational type. This explains the profoundly oniric nature of cinema, as also its absolutely and inevitably concrete nature, let us say its objective status.
Every language is recorded in a dictionary, incomplete but perfect, of the sign-system of his surroundings and of his country. The work of the writer consists in taking, from this dictionary, words, like objects arranged in a drawer, and in making a particular use of them particular insofar as it is a function both of the writer’s historical situation and of the history of these words. The result is an increase of historicity for the word, that is a growth of meaning. If this writer passes into posterity, his “particular use of the word” will figure in future dictionaries, as another possible use of the word.
The expression, the invention of the writer adds, thus, to the historicity, that is to the reality, of the language: he makes use of the language and serves it both as a linguistic system and as a cultural tradition. But this act, toponymically described, is one: it is a new elaboration of the meaning of a sign which was found classified in the dictionary, ready for use.
In return, the act of the filmmaker, although fundamentally similar, is nonetheless much more complex.
A dictionary of images does not exist. There are no images classified and ready for use. If by chance we wanted to imagine a dictionary of images, we would have to imagine an infinite dictionary, just as the dictionary of possible words remains infinite.
The cinema author has no dictionary but infinite possibilities. He does not take his signs, his im-signs, from some drawer or from some bag, but from chaos, where an automatic or oniric communication is only found in the state of possibility, of shadow. Thus, toponymically described, the act of the filmmaker is not one but double. He must first draw the im-sign from chaos, make it possible and consider it as classified in a dictionary of im-signs (gestures, environment, dreams, memory); he must then accomplish the very work of the writer, that is, enrich this purely morphological im-sign with his personal expression. While the writer’s work is esthetic invention, that of the filmmaker is first linguistic invention, then esthetic.
It is true that after some fifty years of cinema,, a sort of cinematic dictionary has been established, or rather a convention, which has this curiosity it is stylistic before being grammatical.
Let us take the image of train wheels rolling amid clouds of steam. This is not a syntagma, but a styleme. [Styleme = a unit of style. Tr.] This allows us to suppose that, from all evidence, cinema will never attain a true grammatical normativity which would be proper to it, but rather, so to speak, a stylistic grammar each time a filmmaker makes a film, he has to repeat the double operation of which I spoke and, as a rule, be content with a certain quantity of uncounted means of expression, which, born as stylemes, have become syntagmas.
In compensation, the filmmaker does not have to deal with a centuries-old stylistic tradition, but only with a decades-old one: he has practically no conventions to be contradicted at the risk of an excessive scandal. His “historical contribution” to the im-sign is brought to a quite short-lived im-sign.
Hence, perhaps, the feeling of a certain fragility of cinema: its grammatical signs are part of a world each time chronologically exhausted. The clothes of the ’30s, the cars of the ’40s, etc., are so many “things” without etymology, or at least whose etymology exists only in the corresponding system of words.
The meaning of words fits the evolution which presides over the creative fashion of clothes or of the lines of cars. Objects, in return, are impenetrable to it: they do not allow modification and say by themselves only what they are at that moment. The imaginary dictionary in which the filmmaker classifies them in the course of the primary stage of his work is not sufficient to give them a historical background, significant for all, now and forever. One thus notes a certain determinism in the object which becomes a cinematic image. It is natural that this should be so, for the word (linguistic sign) used by the writer is rich with a whole cultural, popular and grammatical history, whereas the filmmaker who is using an im-sign has just isolated it, at that very moment, from the mute chaos of things by referring to the hypothetical dictionary of a community which communicates by means of images.
More precisely: if the images or im-signs are not classified in a dictionary and if they are not ordered by a grammar, they nevertheless constitute a common heritage. We have all seen personally the locomotive in question, with its wheels and push-rods. It belongs to our visual memory and to our dreams. If we see it in reality, it “tells us something.” Its appearance in a desert land tells us. for example how moving is the labor of man and how great is the power of industrial society — hence of capitalism to annex in this way new exploitable territories; and, at the same time, it tells some of us that the engineer is an exploited man who, in spite of everything, accomplishes his work with dignity, to the profit of a society which is what it is, even if its beneficiaries identify themselves with it, etc. The locomotive object as a possible cinematic symbol can tell all this by communicating directly with us, and indirectly — as to the common visual heritage — with others.
Thus in reality, “brute objects” do not exist: all are meaningful enough by nature to become symbolic signs. This is why the work of the filmmaker is legitimate in its primary stage. The filmmaker chooses a series of objects, things, landscape or persons as syntagmas (signs of a symbolic language) which, if they have a grammatical history that is conferred on them at this precise moment — as in a sort of happening ordered by a choice and a montage — have nonetheless a pre-grammatical history which is already long and intense.
In sum, just as; in the poet’s style, free rights belong to what is pre-grammatical in the spoken signs, so in the filmmaker’s style, free rights will belong to what is pre-grammatical in the objects. This is another way of saying what I have already said, namely that cinema is fundamentally oniric by reason of the elementary character of its archetypes (that is, once again, habitual and consequently unconscious observation of environment gestures, memory, dreams) and of the fundamental pre-eminence of the pre-grammatical character of objects as symbols of the visual language.
We must add that, in the course of his preliminary and fundamental work, which is the constituting of a dictionary, the filmmaker will never be able to gather abstract terms.
This is perhaps the principal difference between the literary work and the cinematic work. The linguistic and grammatical domain of the filmmaker is constituted by images. Now images are always concrete (only by a foresight embracing millennia could one conceive image-symbols which would know an evolution similar to that of words or at least roots, originally concrete, which, with use, have become abstract.) This is why cinema is, today, an artistic and not philosophical language. It can be a parable, but never a directly conceptual expression This is the third way of affirming the profoundly artistic nature of cinema, its expressive force, its power to embody the dream, that is its essentially metaphoric character.
In conclusion all this should suggest that the language of cinema is fundamentally a “language of poetry.”
Quite on the contrary, historically, in practice, after several abortive attempts, the cinematic tradition which was formed seems to be that of a “language of prose,” or, at least, of a “language of narrative prose.”
But in fact, as we shall see, this is an altogether peculiar and ambiguous prose, insofar as the irrational component of cinema cannot be eliminated. In truth, at the very moment when it was established as a new “technique” or “genre” of expression, cinema was also proposed as a new technique or genre of escape-spectacle, profiting from a number of consumers unimaginable for any other medium of expression. This means that cinema has undergone a violation which was moreover rather foreseeable and unavoidable: everything in it that was irrational, oniric, elementary and barbarous has been kept this side of consciousness, has been exploited as an unconscious factor of shock and glamour, and upon this naturally hypnotic monstrum which a film always is, there was quickly constructed a whole narrative convention which has authorized useless and fallaciously critical comparisons with the theatre and the novel. There is no doubt that this narrative convention refers by analogy to the language of written prose communication, but it has in common with this language only an exterior aspect: illustrative and logical methods — whereas it lacks one of the fundamental element, of the “language of prose”: the rational. This narrative convention relies upon a mystic and embryonic film, a “sub-film “which, from the very nature of cinema, unwinds behind every commercial film, even a decent one, even a socially and esthetically rather adult one.
However — as we shall see below — art-films themselves have adopted for their specific language this “language of prose,” this narrative convention deprived of expressive accent, neither impressionistic nor expressionistic. But one can also assert that the tradition of the cinematic language, which dates from these last decades, has a tendency toward naturalism and objectivity. There is a contradiction here, unusual enough to require careful observation in its reasons and profound connotations.
To resume, let us say that the linguistic archetypes of im-signs are the images of memory and dream, that is, the images of communication with oneself (and of only indirect communication with others, in the sense that the image which another person has of a thing about which I am speaking constitutes a common reference). These archetypes consequently give an immediate basic of “subjectivity” to the im-signs, the mark of belonging totally to the poetic. So that the tendency of the cinematic language should be expressly subjective and lyrical. But the im-signs, as we have seen, also have other archetypes: the integration of gestures into the spoken language and also realization such as we see it with its signs that have only the value of signals. Such archetypes are profoundly different from those of memory and dreams, namely, they are brutally objective, they belong to a type of “communication with others” common to all and strictly functional, so that the tendency which they stamp upon the language of im-signs is rather flatly informative. Moreover, the primary work of the filmmaker — the choice of the im-signs of a true common and instituted vocabulary like that of words. A subjective intervention thus comes into play as of this primary stage, insofar as this primary choice of possible images is therefore quite necessarily subjective.
But this too is subject to contradiction. The brief history of cinema (due to the limitations of expression imposed by the very large number of intended viewers of the film) has been such that the systems which immediately became cinematic syntagmas — and thus form a part of the linguistic institution are few and, at bottom, crude (remember the example of the locomotive wheels: the infinite series of close-ups all like that…). All this underscores the elementary, objective and conventional character of the language of im-signs.
In sum cinema, or the language of im-signs, has a double nature. It is at the same time extremely subjective and extremely objective (an objectivity which, ultimately, is an insurmountable vocation of naturalism). These two essential aspects are closely bound together, to the point of being inseparable, even for the needs of an analysis. The literary function also is double by nature: but its two faces are discernible: there is a “language of poetry” and a “language of prose” so differentiated that they are diachronical and have two divergent histories.
With words, I can proceed with two different operations and thus end up either with a “poem” or with a “narrative.” With images, I can only at least to date create cinema (whose more or less poetic or prosaic character is merely a matter of nuances. This in theory. In practice, as we have seen, a tradition of a “language of narrative cinematic prose” was quickly constituted).
There are of course extreme cases, where the poetic character of cinema is altogether evident. The Andalusian Dog, for example, is flagrantly obedient to a will to pure expression; but to get there, Bunuel had to have recourse to the descriptive panoply of surrealism and one must say that, as a surrealist product, it is of the first order. Few of the other literary works or paintings of this movement can be compared to it insofar as their poetic quality is corrupted by a naive hypertrophy of the content appropriate to the poetics of surrealism, which harms the expressive purity of the words or colors. On the contrary, the purity of cinematic images is no longer thwarted but exalted by a surrealist content. Because it is the true oniric nature of dreams and of unconscious memory which surrealism finds in cinema…
Cinema as I said before, because of its lack of a vocabulary of concepts, is directly metaphorical. However, each metaphor intended in particular inevitably includes something crude and conventional witness those fights of excited or peaceful doves that are supposed to render a character’s torment or joy.
In sum,, the nuanced metaphor, scarcely perceptible, that subtle poetic halo which separates, by a breath and a chasm, the language of Leopardi’s “A Sylvia” from the classical petrarcho-archaic language this metaphor would not be possible in cinema. The most poetic cinematic metaphor possible is always closely bound to the other nature of cinema, the strictly communicative one of prose, which has prevailed in the short tradition of cinema history, spanning in a single linguistic convention art-films and escape-films, masterpieces and adventure serials.
And yet, the tendency of the most recent cinema from Rossellini, compared with Socrates, to the “new wave” and to the production of the last few years, of the last few months (including, I suppose, the majority of the films presented at the Pesaro festival!)  is towards a “cinema of poetry.”
The question which arises is thus the following: how can the “language of poetry” be theoretically explainable and practically possible in cinema? I would like to answer this question by exceeding the strict domain of cinema, by widening the issue and profiting from the liberty which my particular position between cinema and literature assures me. I will therefore, for the moment, transform the question: “Is the ‘language of poetry’ possible in cinema?” into this one: “Is the technique of free indirect discourse possible in cinema?” Indeed, we shall see below how the birth of a technical tradition of the “language of poetry” in cinema is bound to a particular form of free indirect cinematic discourse. But first I must specify what I mean by “free indirect discourse.”
It is simply this: the author penetrates entirely into the spirit of his character, of whom he thus adopts not only the psychology but also the language.
Examples of free indirect discourse have always been numerous in literature. Thus, Dante employs a sort of free indirect discourse when he uses, mimetically, terms which one hardly imagines were familiar to him, and which belong to the vocabulary of his characters’ social milieu: expressions from the courtly language and love-novels of the age for Paolo and Francesca, crude words for the town loafers… Naturally, the use of free indirect discourse blossomed first with naturalism, such as that poetic and archaistic of Verga, then with intimist and twilight literature, i.e. that of the nineteenth century, essentially composed of re-lived discourses.
The characteristic of all re-lived discourses is that the author cannot abstract from them a certain sociological consciousness of the milieu he is evoking: the social condition of a character determines his language (specialized languages, dialect, jargon, dialectal language).
We must also distinguish the interior monolog from free indirect discourse: the interior monolog is a discourse relived by the author through a character who is, at least ideally, of the same class and generation. The language can therefore be the same for the character and for the author psychological and objective characterization is in this case not a fact of language, but of style. Whereas free indirect discourse is more naturalistic, for it is really a direct discourse without quotation marks and which thus necessitates the use of the character’s language. In bourgeois literature without class consciousness (that is, in which there is identification with all humanity), free indirect discourse is most often a pretext. The author constructs a character speaking, if need be, an invented language — which allows him to express his particular interpretation of the world. In this indirect discourse, which, for good or bad reasons, is only a pretext, one can find a narration studded with many borrowings from the “language of poetry.”
In cinema, direct discourse corresponds to the “subjective” shot. In direct discourse the author puts himself aside and allows his character to speak, in quotation marks:
And saying: ‘Come now you see the meridian Touched by the sun and on the bank The foot of night already covers Morocco.’
By direct discourse Dante relates, as spoken, the words of his master. When a screenwriter writes, as seen with the eyes of Accattone: “Stella runs through the vacant lot,” or else “Close-up of Cabiria looking around; she sees, farther off, through the acacia, some youngsters who dance by, playing instruments,” he is outlining the scheme of what, during shooting and even more during editing, will become “subjectives.”
There is no lack of famous “subjectives,” if only for their extravagance: remember, in Dreyer’s Vampyr the “subjective” shot which sees the world as the corpse sees it, as we can see it if laid out in a coffin that is, looking up from below, and moving.
Just as writers do not always have a precise technical awareness of an operation such as that of free indirect discourse, so directors have, up to now, created the stylistic conditions of this operation totally unconsciously, or with very approximate awareness.
Yet it is certain that a free indirect discourse is possible in cinema all the same. Let us call this operation (which, compared to its literary analog, can be infinitely less flexible and complex) “free indirect subjectivity.” And, since we have established a difference between “free indirect discourse” and “interior monolog,” we shall have to see to which of these two methods “free indirect discourse” is more closely related.
It cannot be a true “interior monolog” insofar as cinema does not have the faculty of interiorization and abstraction which the word has: it is an “interior monolog” in images, and that’s all. Thus it lacks a whole abstract and theoretical dimension, evidently present in the monolog, which is an evocative and cognitive act. Thus the lack of an element (concepts of literature) prevents the “free indirect subjective” from corresponding perfectly to what the free indirect monolog is in literature. I would not be capable of citing any instances of total interiorization of the author into a character in the history of cinema up to the ’60s: I do not believe any film exists which is an entire “free indirect subjective.” in which the story is told through the character, and in an absolute interiorization of the system of allusions belonging to the author.
If the “free indirect subjective” does not altogether correspond to the “interior monolog,” it corresponds still less to the true free indirect discourse.”
When a writer “re-lives the discourse” of one of his characters, he steeps himself in his psychology, but also in his language: “free indirect discourse” is therefore always linguistically differentiated from the language of the writer.
If he is able to reproduce, by reliving them, the different languages of the various social categories, it is because they exist. Every linguistic reality is a whole composed of differentiated and socially differentiating languages; and the writer who employs free indirect discourse must above all be aware of this: it is an aspect of class consciousness.
But, as we have seen, the “institutional language of cinema” is only hypothetical: or if it exists, it is infinite, for the author must always create his own vocabulary. But, even this particular vocabulary ends up a universal language; for everybody has eyes. It is not a question here of taking into consideration special languages, sub-languages, jargons, social differentiations, for if there are any, they are completely uncatalogable and unusable.
It is evident that the “look” directed by a peasant (the more so if he comes from an underdeveloped region) and by a cultivated bourgeois upon the same object embraces two different realities: not only do the two men perceive two different “series” of things, but also, the same thing offers two different “faces” to the two “looks.” But this too is only inductive and escapes all codification.
Practically, then, on a possible common linguistic level based on these “looks,” the difference which a director can encounter between himself and his character is psychological and social, but it is not linguistic. Which completely prevents any naturalistic mimesis between the filmmaker’s language and the language, the hypothetical “look” directed by another upon reality.
If the filmmaker assimilates himself to his character and, through him, tells a story, or represents the world, he cannot have recourse to that formidable instrument of differentiation which is language. His operation cannot be linguistic, but stylistic.
Besides, even the writer who re-lives the discourse of a character socially identical to him cannot characterize his psychology thanks to language which is his own — but thanks to style and practically thanks to certain turns belonging to the “language of poetry.”
The fundamental characteristic of the “free indirect subjective” is therefore not of a linguistic nature, but of a stylistic one. It can be defined as an interior monologue without its conceptual and philosophic element, which as such is abstract.
This implies, theoretically at least, that the “free indirect subjective” in cinema is endowed with a very flexible stylistic possibility; that it also liberates the expressive possibilities stifled by traditional narrative conventions, by a sort of return to their origins, which extends even to rediscovering in the technical means of cinema their original oniric, barbaric, irregular, aggressive, visionary qualities. It is the “free indirect subjective” which establishes the possible tradition of a “technical language of poetry” in cinema.
To take concrete examples of all this, I shall have to make Antonioni, Bertolucci, and Godard undergo the test of analysis. (But I could choose, too, authors in Brazil; Rocha, in Czechoslovakia and probably among a good number of those represented at Pesaro.)
As for Antonioni (Red Desert), I would not want to stop at points which are universally recognizable as poetic and which are numerous in this film. For example those two or three violet flowers which are in the foreground, out of focus, in the shot in which the two characters go into the neurotic worker’s house and which, a little later, reappear in the background of the shot, no longer out of focus but fiercely sharp, when they come back out. Or else the dream sequence, which after so much refinement in the colors, is filmed very simply in the most natural technicolor (to imitate, or better: to relive through a “free indirect subjective” a child’s idea which comes from the comic strips of tropical beaches). Or, again, the scene of the preparation for the voyage to Patagonia: the workers who are listening and that stupefying close-up of a worker from Emilia, strikingly truthful, followed by a crazy vertical pan along an electric-blue stripe on the whitewashed wall of the warehouse. All this bears witness to a profound, mysterious, and at times extreme intensity in what illuminates Antonioni’s imagination: the formal idea.
But, in order to demonstrate that the basis of the film is essentially this formalism, I would like to examine two aspects of a particular stylistic operation (the same which I shall examine in Bertolucci and Godard) an extremely significant one. The two moments of this operation are: (1) The close follow-up of two viewpoints, scarcely different from each other, upon the same object: that is, the succession of two shots which frame the same portion of reality first from close in, then from a little farther away; or else first head-on, then a little obliquely; or else, finally, quite simply, on the same alignment but with two different lenses. From this arises an insistence which becomes obsessive, as myth of the pure and anguishing autonomous beauty of things. (2) The technique which consists in having characters enter and leave the frame, so that, in a sometimes obsessive way, the montage is the succession of a series of pictures — which I shall call informal into which the characters enter; so that the world appears as ordered by the myth of a pure pictorial beauty, which the characters invade, it is true, but while submitting to the rule of this beauty instead of profaning it by their presence.
The inner law of the film, that of “obsessive framing,” thus shows clearly the preponderance of a formalism as a myth finally liberated and hence poetic (the fact that I use the term formalism does not imply any value judgment; I am well aware that an authentic and sincere formalist inspiration does exist: the poetry of language).
But how has this liberation been possible for Antonioni? Quite simply thanks to the creation of a “stylistic condition” by a “free indirect subjective” which coincides with the entire film.
In Red Desert, Antonioni no longer applies, by a somewhat awkward contamination as in his previous films, his own formalist vision of the world to an engaged content (the problem of the neurosis of alienation); but he looks at the world at one with his neurotic heroine, re-living it through the “look” of this woman (who is, not for nothing, this time beyond the clinical stage, suicide having already been attempted). Thanks to this stylistic mechanism, Antonioni has given us his most authentic work. He has finally succeeded in representing the world seen through his own eyes because he has substituted, wholly, the world-view of a sick woman for his own vision, which is delirious with estheticism: a substitution justified by the possible analogy of the two visions. But even if some part of arbitrariness entered into this substitution, one could make no objection. It is clear that the “free indirect subjective” is a pretext which Antonioni has, perhaps quite arbitrarily, used in order to obtain the greatest poetic liberty a liberty which, precisely, borders (and this is why it is intoxicating) upon the arbitrary.
Obsessive still shots are also characteristic of Bertolucci’s film, Before the Revolution. However, they have a different meaning than for Antonioni. The world-fragment, imprisoned in the frame and transformed by it into a fragment of autonomous beauty which refers only to itself, does not interest Bertolucci as it interests, in return, Antonioni. Bertolucci’s formalism is infinitely less pictorial: his frame does not intervene metaphorically upon reality, sectioning it into so many mysteriously autonomous places, like pictures. Bertolucci’s frame adheres to reality, according to the canon of a certain realist manner (according to a technique of poetic language, followed by the classics from Charlie Chaplin to Bergman): the stillness of a shot upon a portion of reality (the river, Parma, the streets of Parma, etc.) reveals the grace of a profound and confused love precisely for that portion of reality.
Practically, the whole stylistic of Before the Revolution is a long “free indirect subjective” based on the dominant state of mind of the protagonist, the neurotic young aunt. Whereas there was, in Antonioni, a whole substitution of the sick woman’s vision for that (of febrile formalism) of the author, in Bertolucci such a substitution does not take place. What there has been is a contamination between the vision the neurotic woman has of the world and that of the author, which are inevitably analogous, but difficult to perceive, being closely intermixed, having the same style.
The intense moments of expression in the film are, precisely, those “insistences” of the framing and the montage-rhythms, whose structural realism (derived from Rossellinian neo-realism and the mythic realism of some younger master) is charged, throughout the uncommon duration of a shot or a montage-rhythm, till it explodes in a sort of technical scandal. Such an insistence on details, particularly on certain details in the digressions, is a deviation in relation to the system of the film: it is the temptation to make another film. It is, in sum, the presence of the author, who, in a measureless liberty, goes beyond the film and threatens continually to abandon it for the sake of an unforeseen inspiration which is that latent of the author’s love for the poetic world of his own life-experiences. A moment of a naked and raw subjectivity, entirely natural, in a film in which as in Antonioni’s subjectivity is mystified by a method of false objectivism, the result of a pretextual “free indirect subjective.”
Beneath the style generated by the disoriented, disorganized, beset-by-details state of mind of the protagonist, is the level of the world as seen by an author no less neurotic, dominated by an elegiac, elegant, but never “classicist” spirit.
In the world-view of Godard, there is, on the contrary, something rough and perhaps even slightly vulgar. For him, elegy is inconceivable. Perhaps because he lives in Paris, he cannot be touched by such a provincial and rustic sentiment. For the same reason the classicist formalism of Antonioni is also foreign to him. He is altogether post-impressionist, he has none of the old sensuality which still impregnates conservative lands and which is marginal, Paduan-Roman, even when it is very Europeanized, as in Antonioni. Godard has set himself no moral imperative: he feels neither the need of a Marxist engagement (that’s ancient history), nor academic mad conscience (that’s all right for the provinces). His vitality knows neither restraints, nor modesties, nor scruples. It is a force which reconstitutes the world to its measure, which is cynical towards itself. Godard’s poetics is ontological; its name is cinema. His formalism is thus of a technical character, poetic by its very nature. Everything that is moving and is fixed by the camera is beautiful: this is the technical and therefore poetic restitution of reality. Godard too, naturally, plays the usual game: he too needs a “dominant state of mind” of the protagonist to establish his technical liberty. A neurotic and scandalous dominant state in his relationship with reality. The heroes of his films too are therefore sick exquisite flowers of the bourgeois class, but they are not in treatment. They are gravely affected, but full of life this side of the brink of pathology: they simply embody the norm of a new anthropological type. Even their obsession is characteristic of their relationship with the world: the obsessive attachment to a detail or a gesture (and this is where cinematic technique comes in; even better than literary technique, it can push such situations to the extreme). But this insistence on a single object does not exceed a bearable duration; in Godard, there is no cult of the object as form (as in Antonioni) nor a cult of the object as symbol of a lost world (as in Bertolucci): Godard has no cult and puts everything on a level of equality. His “free indirect discourse” is the systematic alignment of a thousand details of the world, which follow one another, undifferentiated, without continuity-solution, arranged in sequence with the cold and almost satisfied obsession (typical of his amoral characters) of a disintegration reunified in an unarticulated language. Godard is a complete stranger to classicism otherwise one could speak in his case of neo-cubism but we could very well speak of an atonal neo-cubism. Behind the narrative of his films, behind the long “free indirect subjectives” which imitate his characters’ state of mind, there always unwinds a mechanical and asymmetrical film, made for the pure pleasure of restoring a reality broken by technique and reconstructed by a vulgar Braque.
The “cinema of poetry” as it appears several years after its birth characteristically produces films of a double nature. The film which one sees and receives normally is a “free indirect subjective” which is sometimes irregular and approximate in short, very free. This comes from the fact that the author uses the “dominant state of mind in the film,” which is that of a sick character, to make a continual mimesis of it, which allows him a great stylistic liberty, unusual and provocative. Behind such a film unwinds the other film the one the author would have made even without the pretext of visual mimesis with the protagonist; a totally and freely expressive, even expressionist, film.
Obsessive framings and montage-rhythms testify to the existence of this underlying, unrealized film. Such an obsessive force contradicts not only the rules of the common cinematic language following a different and perhaps more authentic inspiration liberates itself from its function and appears as “language in itself,” style.
The “cinema of poetry” is therefore in reality essentially based on the stylistic exercise as inspiration, which is, in the majority of cases, sincerely poetic. This removes all suspicion of mystification as to the role of the pretext which is that of the free indirect subjective.”
What then does all this mean?
It means that a common technico-stylistic tradition is in process of being formed: that is, a cinema language of poetry. This language tends to appear henceforth as diachronical in relation to narrative cinema language: a diachronism which is destined to be emphasized increasingly, as happens in literary systems.
This emerging tradition is based on the collection of cinematic stylemes which have been constituted almost naturally in function of the irregular psychological characteristics of the characters chosen as pretexts, or, better: in function of a primarily formalist world-view of the author (informal in Antonioni, elegiac in Bertolucci, technical in Godard). Expressing such an inner vision necessarily requires a special language, with its technical and stylistic formulas simultaneously serving the inspiration, which, as it is precisely formalist, finds in them at once its instrument and its object. The “cinematic stylemes” which have thus appeared and been classified in a tradition barely established and still without norms unless intuitive, pragmatic ones all coincide with typical procedures of cinematic expression. They are linguistic facts, which therefore require specific linguistic expressions. Enumerating them amounts to outlining a possible “prosody,” not yet codified, in gestation, but whose rules already exist in potential (from Paris to Rome and from Prague to Brasilia).
The primordial characteristic of these indications of a tradition of the cinema of poetry consists in a phenomenon which technicians define normally and tritely as “making the camera felt.” In sum, the maxim of wise filmmakers in force up till the ’60s “Never let the camera’s presence be felt” has been replaced by its opposite.
These two opposite points, gnosiological and gnomic, indiscussibly define the existence of two different ways of making films: of two different cinematic languages…
But then it is necessary to say that in the great cinematic poems of Charlie Chaplin, of Mizoguchi or of Bergman, the common character was that “you didn’t feel the camera”: they were not filmed, therefore, according to the laws of “the language of the cinema of poetry.”
Their poetry resided elsewhere than in the language considered as linguistic technique. The tact that one did not feel the camera in them means that the language was adhering to the meanings by putting itself in their service: it was transparent to perfection, did not superimpose itself upon the facts, did not do violence to them with mad semantic deformations the very ones which are due to a language which is present as incessant technico-stylistic awareness.
Let us recall the boxing sequence in City Lights, between Charlie Chaplin and a champion who is, as usual, much stronger than he is. The astonishing comedy of Charlie’s dance, his little steps taken a bit here and there, symmetrical, useless, overwhelming and irresistibly ridiculous, well, here, the camera was still and took just any long shot. One didn’t feel it. Or again let us recall one of the last products of the classic cinema of poetry: The Devil’s Eye, by Bergman, when Don Juan and Pablo leave Hell after three centuries and see the world again: the appearance of the world — such an extraordinary thing — is filmed with a shot of the two heroes against a background of somewhat wild springtime country, one or two very common close-ups and a long shot of a Swedish panorama, overwhelmingly beautiful in its crystalline and humble insignificance. The camera was still, it framed these images in an absolutely normal way. One didn’t feel it.
The poetic character of the classic films was therefore not the fact of a specifically poetic language.
This means that these films were not poetry, but narratives. Classic cinema was and is narrative, its language is that of prose. Its poetry is an inner poetry, as, for example, in the narratives of Chekhov or Melville.
Thus one feels the camera, and for good reasons. The alternation of different lenses, a 25 or a 300 on the same face, the abuse of the zoom with its long focuses which stick to things and dilate them like quick-rising loaves, the continual counterpoints fallaciously left to chance, the kicks in the lens, the tremblings of the hand-held camera, the exasperated tracking-shots, the breaking of continuity for expressive reasons, the irritating linkages, the shots that remain interminably on the same image, this whole technical code was born almost of an intolerance of the rules, of the need of unusual and provocative liberty, a diversely authentic and pleasant taste for anarchy, but it immediately became law, a prosodic and linguistic heritage which concerns all the cinemas in the world at the same time.
Of what use is it to have identified and, in a way, baptized this recent technico-stylistic tradition “cinema of poetry?” A simple terminological convenience, evidently, and which is senseless unless one then proceeds to a comparative examination of this phenomenon in relation to a larger political, social and cultural situation.
Cinema, probably since 1936 — the year Modern Times was released — has always been in advance of literature. Or at least, it has catalyzed, with an opportuneness that made it chronologically anterior, the profound sociopolitical reasons which were to characterize literature a bit later.
Cinematic neo-realism (Open City) prefigured all the neo-realism in Italian literature in the post-war years and part of the ’50s; the neo-decadent or neoformalist films of Fellini or Antonioni prefigured the revival of the Italian neo-avant-garde and the extinction of neo-realism; the “new wave” anticipated the “school of the Look” in brilliantly publicizing its first symptoms; the new cinema of some of the socialist republics is the primordial and most remarkable datum of a reawakening of interest in these countries for a formalism of Western origin, as an interrupted twentieth-century motif, etc. In a general framework, this formation of a tradition of a “language of poetry in cinema” appears as the hope for a strong and general resumption of formalism as typical and average production of neo-capitalism. (Naturally, there remains the reserve, due to my Marxist moralism, of a possible alternative: i.e., of a renewal of the writer’s mandate, which for the moment appears to have expired.)
Indeed, to conclude:
(1) The technico-stylistic tradition of a cinema of poetry originates in the climate of neo-formalist researches, corresponding to the stylistic and linguistic inspiration which has again become current in literary production.
(2) The use of the “free indirect subjective” in the cinema of poetry is only a pretext enabling the author to speak indirectly — through some narrative alibi — in the first person; thus the language used for the interior monologs of the character-pretexts is the language of a “first person” who sees the world according to an essentially irrational inspiration and who, to express himself, must therefore have recourse to the most brilliant means of expression in the “language of poetry.”
(3) The character-pretexts can only be chosen from the author’s own cultural circle: therefore analogous to him by their culture, language and psychology: “exquisite flowers of the bourgeois class.” If they happen to belong to another social world, they are always sweetened and assimilated via the categories of anomaly, neurosis or hypersensitivity. The bourgeois class itself, in sum, even in cinema, identifies itself, again, with all humanity, in an irrational interclassism.
All this belongs to the general movement of recuperation, by bourgeois culture, of the territory it had lost in the battle with Marxism and its possible revolution. And this is a part of the somehow grandiose movement of the evolution — we shall call it anthropological of the bourgeoisie, along the lines of an “internal revolution” of capitalism, i.e. of a neo-capitalism, which questions and modifies its own structures and which, in the case which concerns us, re-attributes to the poets a pseudohumanistic function: myth and the technical awareness of form.
[This text was read in Italian by Pier Paolo Pasolini in June 1965 at the first New Cinema Festival at Pesaro. The present version is from the French translation by Marianne de Vettimo and Jacques Bontemps which appeared in Cahiers du Cinéma No. 171, October 1965.]
In Movies and Methods. Vol. 1. Ed. Bill Nichols. Berkeley: University of California Press, 1976. 542-558.