What I shall attempt here is a critical reading of the body of work that has served as the foundation of the proposed semiology of the cinema—the writings of Christian Metz.41 My reading of Metz, like Althusser’s reading of Marx, will be a “symptomatic reading” in which I shall be interested not only in what the author says but also in what he does not say; and I shall direct my attention not only to what Metz highlights but also to what he prefers to leave in darkness.
Such a reading is said to be a symptomatic reading simply because one of its basic operations is to consider what is said as “symptoms” of what is not said, of what is repressed. A symptomatic reading of Metz is particularly appropriate and particularly revealing, for there is one vast and crucial area of conceptualization that is conspicuous in its absence in Metz’s writings. And in light of Althusser’s insistence that it is ultimately ideology that determines the “structuring absences” in a text, it is striking to note that what is taboo in Metz’s writings, what violently shapes and disfigures his theoretical model by its structuring absence, is precisely any consideration of ideology itself.
This is strange, to say the least, for by 1968 the ideological character of the cinema was so hotly debated in France that it was very much a potentially explosive issue. A number of things contributed to this: the growing radicalization of Godard; the leftward turn of the old Bazin-oriented Cahiers du Cinéma; the solidarity of French film-makers in the face of government attempts to depose Henri Langlois as director of the Cinémathèque; the disruption (by Godard and, of all people, Truffaut) of the Cannes Film Festival in the spring of 1968; and, finally, the May Events themselves and the participation of film-makers—indeed of a large segment of the French film community—in the rethinking from “zero” of French cinema in Les Etats Généraux du Cinema as well as in the tumultuous debates over education and culture in the Sorbonne and the Theâtre de l’Odéon, or simply in the streets. All of these phenomena (and many more, such as the debates over television programming and government interference with news broadcasts on the state-owned television network, the ORTF) had catapulted the cinema (and television) into the forefront of the debates over the function of art in society. And the debate continues.
Old assumptions now are being challenged; and many people are discovering for the first time the profoundly ideological character of art. In the cinema two notions in particular have been challenged and largely discredited for their ideological ramifications. The primacy of the narrative, the use of the cinema to tell stories, has been denounced for playing up to people’s escapist tendencies and for leading the passive spectator down the path of least resistance, thereby inculcating a painless, even pleasurable but nonetheless authoritarian relationship between the cinema and the audience. And, second, the notion that the cinematic image was a faithful reflection of reality, that the camera doesn’t lie, was seen to fit in all too well with the ruling class’s need to present the status quo in the guise of some natural, immutable reality.
Within this historical and political conjuncture—and with the cinema clearly the focus of ideological struggle on the cultural front—Christian Metz’s Essais sur la signification au cinéma inexplicably became the latest fad in academic circles, at first in France, then in England and America. The reasons for Metz’s popularity among academics are not at all easy to understand. After all, his Essais are extremely difficult and tedious reading, and his overall argument is impenetrable to most readers and only barely discernible to a diligent few.
One has to understand the intellectual market, however, and the central position within that market of the latest intellectual fads from Paris. With interest in the cinema mushrooming on university campuses—especially in the U.S.A. but also in England and France—Metz’s book was timely. Moreover, Metz’s recourse to structural linguistics enabled the book to ride the already swelling wave of popularity of structuralism. At the same time, Metz’s semiology could be—and was—touted in the intellectual press as a “new and improved structuralism.” (An even better product than our previous model! )
Finally, by a strange twist in the workings of ideology, Metz’s writings also benefited, in a sense, from the political debates over the cinema, and this is true in spite of the fact that Metz systematically ignores the political debates. In the arts, however, there is often a confusion (which the promoters of culture are quick to take advantage of) between the artistic avant-garde and the political avant-garde. Generally, we can expect the artistic avant-garde, no matter how commercial or decadent—in fact, the more commercial and decadent the better—to be promoted as somehow subversive. (We have the Dadaists and Surrealists to thank for this fetishization and commercialization of scandal.)
In the case of Metz’s writings, however (and let us note that the first volume of Essais sur la signification au cinéma was published in 1968), the fact that the political turmoil had been felt in the cinema as well; and that people were talking about rethinking the cinema from “zero”; not only provided a nice market for a book which put forth a new approach to the cinema, it also endowed such a book with connotations of political avant-gardism. Regardless of its content (and in this sense the highly dense, impenetrable style of Metz’s writings helped to mask the content), a book on the cinema, featuring a new and improved structuralism, and coming from Paris in 1968, could very easily pass for a work of the political as well as the artistic avant-garde.
But let us look at where Metz stands on the very issues which were—and still are—hotly debated by those of us who are concerned about the ideological function of art. What do we find? Well, it is hard to find where Metz stands. So dense is Metz’s elaboration of his semiology that the ideological ramifications of his theoretical model almost pass unnoticed. Few readers of Metz penerate sufficiently the web of his rhetoric to recognize that at a time when the cinema’s famous “reflection of reality” argument has been discredited for its ideological complicity with the status quo, Metz blithely revalorizes both the argument and the complicity. Few recognize that at a time when the narrative film has become suspect for its authoritarian pacification of the audience, Metz simply equates the cinema with narrativity, and leaves it at that. Finally, few recognize that at a time when the exploration of the relation between cinema and ideology has come so urgently to the forefront of our critical and theoretical debates, Metz archly refuses to confront the existence of ideology, drawing back inside his methodological shell each time he finds ideology looming in front of him.
But just what is Metz’s system? What is Metz’s semiology? It is very difficult to answer these questions satisfactorily, for in spite of Metz’s obsessive preoccupation with the notion of “science” (what kind of science we shall see a bit later), there is regrettably little clarity either of organization or of expository style in the convoluted ruminations of Metz’s two volumes of Essais sur la signification au cinéma or of the later Langage et Cinéma.
And where there is a bit of clarity in Metz’s work—as in his ability to sift through possible confusion over terminology and to make distinctions between, for example, the terms “cinema” (and “cinematic”) and “film” (and “filmic”)—one cannot avoid the observation that Metz really doesn’t tell us anything new but merely gives the academic’s seal of approval, after pages and pages of explanation and qualification, to the common usage of these terms. (Metz laboriously spells out, for example, that we don’t come out of the theater after seeing Renoir’s La Grande Illusion and say to a friend we happen to meet, “I just saw a marvelous cinema”; therefore, he concludes, the terms “film” and “cinema” are not interchangeable in both directions; although, in some cases, for example, when speaking in English of the “art of film” or of “film history,” Metz acknowledges that the term “film” in this sense is interchangeable with the term “cinema” although Metz prefers the latter.)
Moreover, one cannot avoid the additional observation that such distinctions are by no means arrived at through some special methodology of semiology. (Significantly, however, Metz will later use these distinctions as a justification for ignoring, in this particular semiology of cinematic language, the study of individual films.) Rather, Metz’s clarification of the terminology is simply the product of a rhetorician’s respect for words (as well as for simple logical consistency) and an academic’s patience to plod tediously through the many different uses and misuses of these terms in the literature of the cinema. In short, Metz’s clarification— if it is that —is really nothing more and nothing less than an example of pedantry. One can argue that it is pedantry at its very best; but it is definitely pedantry, and not semiology, unless one wants to argue that these two terms are interchangeable.
Let us concentrate, however, on the elaboration of Metz’s semiology. Let us attempt to discover on what foundations it is laid. As Stephen Heath put it, “In fact, the development of Metz’s semiology itself might be grasped in connection with its posing and exploration of a central issue; it is given as a systematic appeal to the methods and concepts of modern linguistic theory in order to examine rigorously the prevalent metaphor in writing on film of cinematic language.”42
Thus, in attempting to use the methods and concepts of modern linguistics in order to determine whether the term “cinematic language” has any validity, Metz is faced at the outset with the question of whether cinema ought properly to be considered what French linguists, after Saussure, call a langue or a langage? Taking Saussure’s definition of langue as “a system of signs for the purpose of intercommunication,” Metz finds cinema lacking the status of a langue because, in his view, (1) cinema is a one-way communication with no return channel for the response of the viewer-listener; (2) cinema is only partly a system of signs; and (3) cinema has very few true signs (ESC, I, p. 79).
Now in some ways all three of these reasons for denying cinema the status of a langue are debatable. Even within the framework of linguistic theory, as we shall see in a moment, the latter two conclusions by Metz are extremely questionable. And where the first objection is concerned, one can certainly argue, following Godard, that the real dynamics of the cinema are not on the screen (or even in the relations between images and sounds) but in the relations between the images and sounds, on one hand, and the viewer-listener in the audience.
The fact that the general practice of (bourgeois) cinema systematically deemphasizes the possibility of any active intervention by the viewer-listener—even mental intervention—thus indicates simply that intercommunication takes place in an indirect and unequal way in the cinema because intercommunication takes place in an indirect and unequal way in society as a whole. And the ruling class (which controls, among other things, the cinema) makes sure things stay that way.
Thus, in Godard’s far-reaching view, the cinema serves to inculcate the bourgeois concept of representation, of our acquiescing in letting other people represent us, speak for us, give us images of the way we ought to be. There is intercommunication in the cinema all right, but the stage on which it takes place is not the stage (or the screen) of the movie theater itself, as if it existed in a vacuum, but rather the stage of social practice, of which the cinema, like everything else, is merely a part.
Even comfortably seated in the movie theater in our habitual role as passive consumers, we are vaguely aware that “behind” the film “someone is talking to us.” Who is talking? Well, we occasionally say that such-and-such a film got produced because so-and-so’s money was “behind” the film. Perhaps, then, in the logic of our unconscious formulations, we already know who is talking. Money talks.
And far too often we communicate back to the people who have the money in just the way they want us to—by our passivity, by our silence, by our compulsive consumption, by our readiness to let others represent us, both in the cinema and outside. But it doesn’t have to be that way, either in the cinema or outside. We can change our ways of responding to films that seek to pacify us or to brutalize us. (We can even stop going to see them.) We can change the kinds of films that get made. (Critics can educate the public to support films other than those that pacify or brutalize us.) And we can change society. In fact, the struggle to change society is a struggle on many fronts: one of them is the cinema.
In any case, Metz’s bland assertion that the cinema does not provide for intercommunication should not simply be taken as a fact and left unquestioned. Moreover, what is at stake is certainly more than merely finding out whether the cinema fits into the category langue or the category langage. In itself, the question of whether cinema is a langue or a langage is an academic question. But we should recognize that academic questions don’t just fall from the sky. And we should recognize that it is precisely the function—the ideological function—of petit bourgeois intellectuals to defuse potentially explosive issues by transforming them into academic questions in which the masses of ordinary people can’t see the issues clearly and couldn’t care less because everything is worded in such an abstract and convoluted way.
But just what constitutes a methodology that has permitted Metz to define for himself such a severely—and to my mind, falsely—delimited object of research? Moreover, just what are the foundations of Metz’s semiology? Even in the light of structural linguistics the issues are by no means as cut-and-dried as Metz makes them out to be.
As Stephen Heath points out,
It is the idea of the image that represents the blind spot of Metz’s initial formulations, the point at which the articulation of significance collapses in the face of analogy. It is the “pure analogy” of the image (ESC, I, p. 51) that determines the absence of signs; where the linguistic sign is arbitrary, the image of a dog resembles a dog; the distance between signifiant and signifié is minimum, there being a “quasi-fusion” of the two. The image is envisaged as a duplication of reality; “cinema has for its basic material a set of fragments of the real world mediated by their mechanical reproductions” (ESC, II, p. 49). Hence cinema lacks any equivalent to the double articulation of linguistic langue, its very economy, the combination of systematically defined units of a lower level (phonemes) to form units of a higher level (monemes) : instead of articulation, duplication; instead of economy, an infinity of analogical resemblances.43
In other words, Metz’s semiology rests on the presupposition that the cinematic image is a “pure analogy” of the “real world”; and that, therefore, the image is not a sign. However, even within the realm of semiological research, this position is flatly rejected by Umberto Eco, among others, who argues that images, as well as verbal signs, are cultural rather than natural; that they are arbitrary and subject to conventions in a systematic way. 44
Thus right at the foundational starting-point of Metz’s investigations—the langue/ langage opposition—we find that there are presuppositions that underlie Metz’s thinking (here, for example, the tendency to consider the cinematic image as a duplication of reality) which, although cryptically stated or simply implicit in his writings, are generally not identified by Metz, nor recognized by the majority of his readers, for what they really are, namely, presuppositions which one has the obligation to question and upon whose theoretical foundations the entire edifice of Metz’s semiology must ultimately stand or fall. Moreover, as we shall see shortly, these presuppositions carry with them ideological ramifications of the gravest consequence—ramifications which Metz feels entitled to ignore precisely because he has methodologically ruled them out, naively or duplicitously, right from the beginning.
On the question of the cinematic image’s relation to reality, as on most questions, however, Metz is careful to cover his tracks. As Heath points out,
In fact, Metz’s early statements concerning the image almost always add the condition that the image is not reality, that it is a mediation, that it is a “deformation” (ESC, I, p. 111). . . . As generally in this kind of discussion, however, the qualifying condition tends to be overshadowed and the impression of reality the cinema may produce is seized directly as “reality,” foreclosing the thinking of the production of that impression of reality ... ; the explication serves, in fact, less to analyse the production of that impression than to confirm its reality; it becoming an established and unchallengeable fact; cinema equals impression of reality.45
Another of Metz’s presuppositions—which does not come into play until he has anchored cinema on the side of langage rather than langue —is his equation of the cinema with narrativity. Granted, Metz recognizes, as anyone must, that there are films which simply cannot be characterized as narrative films. But these Metz simply sets aside, acknowledging that such films are outside his area of competence; while at the same time he defines his own area of competence as “cinematic language” and unabashedly equates the cinema with narrativity. (It is characteristic of Metz’s rhetorical style that even in reducing his claims to competence— and seeming to make them more “precisely and scientifically delimited”—he really makes them infinitely more vast. )
Where Metz does address himself to the question of why he identifies the cinema with narrativity, his arguments (as Michel Cegarra, writing in Cinéthique, has pointed out) are circular, specious, and often so naïve as to be possibly disingenuous. For example, observing that the cinema took the path of narrativity quite early in its history, Metz makes the following pronouncement: “It was necessary that the very nature of the cinema made this evolution if not inevitable, at least possible, perhaps even probable” (ESC, I, p. 52). As Cegarra cautions in a footnote, “This would give us a tautology to infinity: the cinema is narration because it is cinema, therefore narration, therefore cinema, etc....” 46
Or, for another example, there is Metz’s explanation of how demand is to be taken into account as one of the foundations for equating the cinema with narrativity: “There were the spectator’s needs, in short; the demand” (ESC, I, p. 52). And this pressure is, Metz acknowledges, “perfectly capable of exerting an influence ... on the formula of the spectacle” (ESC, I, p. 52). Therefore, Metz tells us, “the basic formula which has never changed consists in calling a ‘film’ a large unit which tells us a story; and to ‘go to the cinema means to go and see that story” (ESC, I, p. 52). And the reason why that formula has never changed, Metz informs us, is because “it has been accepted ... it is a rather pleasing one.” As Cegarra, justifiably taken aback by such reasoning, points out, “The naivety of such an analysis is astounding. One wonders whether Metz has ever heard of ideology , or whether he knows that film is also a commodity which can be sold and makes profits”47
Having laid such shaky foundations (that, as we have seen, ultimately rest on implicit presuppositions which themselves are extremely shaky), Metz then proceeds to use another set of concepts and methods borrowed from linguistic theory. What he does, however, is to use the breakdown of linguistic units into those which sustain paradigmatic relations and those which sustain syntagmatic relations, in such a way as to appear to justify “systematically” what in fact he has simply built into the “system” right from the start, namely, the apparent primacy of narrativity.
Because Metz has already denied the cinematic image the linguistic status of a sign (the validity of which would entail its being systematic along linguistically formal lines), he can now argue that since, in his view, the image is “pure analogy” (the validity of which is mere “duplication of reality”), it is thus extremely difficult if not downright impossible to speak of the cinematic image sustaining anything but syntagmatic relations. In other words, images can be combined, but they cannot in any linguistically systematic way, he argues, be substituted for one another.
( Incidentally, denying the image the status of a sign, of course, also denies it the status of a word. This forms another of Metz’s original justifications for denying cinema the status of a langue. And, consequently, Metz can now assert that the cinema could not a priori possess significative paradigmatic oppositions—between words—and distinctive paradigmatic oppositions—between phonemes.)
Thus, applying the linguists’ commutation test to the cinematic image (in order to determine which relations are paradigmatic and which are syntagmatic ), Metz “finds”—because he has stacked the deck—that where paradigmatic relations would be expected to appear there is in fact no commutation, merely an infinite set of possible images. “In the cinema . . . the number of realisable images is indefinite. Indefinite many times over. . . . For profilmic scenes are themselves unlimited in number” (ESC, I, pp. 102-103).
Drawing from this commutation test the following conclusion, “It seems thus that the paradigmatic of film is condemned to remain partial and fragmentary, at least if one looks for it at the level of the image [Metz’s italics]” (ESC, I, p. 103), Metz then attempts (at a later date and chiefly through additional notes and corrections to the second version of the text) to reintroduce the paradigmatic—but primarily as a paradigmatics of the syntagmatic order. Recognizing that “the syntagmatic and the paradigmatic are by definition correlatives of each other” (ESC, I, p. 73, note), and acknowledging that “this is precisely why I am less sceptical today on the question of the paradigmatic of the film than I was when I wrote that article” (ESC, I, p. 75, note), Metz nonetheless persists in privileging the syntagmatic and conceives of the paradigmatic in the only way his particular bias—for the cinema as narrative —can account for, namely, as a choice between ways of combining images, as paradigms of syntagms.
Once again, however, Metz attempts to cover himself on all sides, acknowledging that “the existence of different kinds of paradigmatic associations between the images themselves is not excluded, since different social groups all have cultural ‘symbolisms’ which also relate to iconography. The point is simply that these paradigms are not specifically linked to cinematic language” (ESC, I, p. 73, note).
But in this seemingly offhand disclaimer there is still another Metzian presupposition, or, more precisely, another bias ensuing from his initial presuppositions regarding the nature of the image and his equating the cinema with narrativity. In casting out the paradigmatic relations between images themselves from the sacred realm of cinematic language, Metz happens to be conveniently ridding himself of something that could cause him trouble. (For one thing, these “cultural symbolisms” he alludes to might force him to deal with ideology.) In any case, he betrays here not only his bias for the syntagmatic relations at the expense of the paradigmatic relations, but also his bias for denotation at the expense of connotation.
While, characteristically, Metz is careful to acknowledge that “the semiology of the cinema may be conceived either as a semiology of connotation, or as a semiology of denotation” (ESC, I, P. 99)י his own exclusive bias toward the latter and his inability to conceive of connotation as anything but an ornament to denotation, that is, to narrative, have been exposed quite clearly by Cegarra.48 Moreover, at the point where Metz denies the paradigmatic associations between the images themselves the right to enter the kingdom of cinematic language—conveniently pushing them somewhere “out there” in the mundane world of the cultural —Metz seems to prefer not to consider that, as Cegarra insists, “the paradigm of the image never exists in itself for it is also (and above all) a paradigm of connotation.” 49
It is clear, however, that the object of study awaiting Metz’s longwinded and circuitous foundational justifications has all along been nothing other than the denotation of the narrative. Thus, his “discovery” of the “surprising poverty of the cinema’s paradigmatic resources” (ESC, I, p. 72) hardly comes as a surprise. On the contrary, he now has exactly what he has been looking for—suficient excuse for turning his attentions exclusively to the syntagmatic relations, to the combination of images into sequences, to, in fact, narrative.
Defining his own practice of semiology ever more narrowly— but at the same time imputing ever grander significance to its findings—Metz now focuses his attention on what he calls the Grande Syntagmatique of the narrative film—or, still more “modestly”—of cinema. (For Metz, of course, the two are synonomous.)
The Grande Syntagmatique, as it is worked out in two different versions, is developed as a model for analyzing a film into the possible sets of narrative “figures” or “autonomous segments.” In its later—but, as Metz cautions, not necessarily definitive—version, the model consists of eight types of combinations of images into autonomous segments of the narrative. And the model is then applied descriptively to Jacques Rozier’s film Adieu Philippine, which is laboriously broken down into a total of eighty-three autonomous segments of the narrative, each of which is classified as one or another of the eight basic types.
Much has been made of this Grande Syntagmatique, and particularly of its application to an individual film. If one was hoping —and it is a legitimate expectation—that theory, when finally applied, even if descriptively, would enrich our understanding of the film by providing us with some food for thought, one comes away malnourished. The table set for us by maitre Metz is spare indeed. Metz’s “Table of the ‘Autonomous Segments’ of Jacques Rozier’s film Adieu Philippine” (ESC, I, pp. 151-75) turns out to be a desiccated and ossified taxonomy of the banal. All it can offer us as food for thought are the bare bones of the story—picked clean, of course, with scientific rigor.
Having persevered in following Metz’s laborious and circuitous wanderings in linguistic theory to arrive at this frugal repast is too heavy a price to pay for too little—a price one pays in those most precious of commodities, time (lost) and energies (misdirected). Moreover, aside from the meager pickings at the end of the journey, there is also the risk of having swallowed along the way some of those half-baked presuppositions that Metz has tried to slip in unnoticed right from the start. And to swallow these, we have insisted, is to swallow a dose of ideological poison.
Granted, Metz (unlike all the journalistic reviewers whose film criticism consists of retelling the story) concentrates less on what the story is about than on how it is told in images. (Incidentally, Metz’s Grande Syntagmatique is unable to explore any relations between images and sounds. It is designed to apply only to the images.) Nonetheless, the story is still preeminent, even if Metz, unlike the journalistic reviewers, approaches the story from a certain (false) notion of science rather than shooting from the hip with his subjective impressions. (But, as Ernest Callenbach insists, those journalistic reviewers are able to tell us just as much, if not more, about how a film is told as Christian Metz does . . . and in an infinitely more interesting way!)
In any case, as the editors of Cinéthique point out in their critique of Metz’s Langage et Cinèma ,50° this shift from the impressionistic to the scientific merely marks the passage from one form of idealism (subjective) to another (objective). And the latter is perhaps more dangerous precisely because its practitioners, unlike the subjective idealists, can’t be accused of simply ignoring the realm of ideology; instead they rule it off limits “methodologically.”
Thus, “the idealist discourses of the new type find an accomplice in semiology: like them, semiology leaves the class struggle at the door of its laboratory, never for a moment examining what it is that always gives signification (its object) a class character, namely, the inscription of films into the ideological struggle.”51
Moving from Metz’s two volumes of Essais sur la signification au cinéma to his later Langage et Cinéma, one becomes gradually aware of three things. First, in the later work there is less singleminded concentration on narrative, and in its place a rather belabored development of the notion of cinematic language conceived as a schema of codes and sub-codes (about which more in a moment). Second, far from bringing about any radical rethinking of his earlier positions, plus ςα change plus cest la mème chose. And, finally, Metz’s shift of emphasis in the later book seems motivated by an attempt to buttress his earlier positions by arriving at them from a slightly different, although equally circuitous, path. Interestingly enough, in the course of Langage et Cinéma this defensiveness on Metz’s part becomes increasingly focused on warding off criticisms from the left, particularly the kind advanced by Cegarra in Cinéthique.
Let us examine, however, Metz’s development in Langage et Cinéma of the schema of codes and sub-codes that in his view constitute cinematic language. Drawing further on structural linguistics, Metz defines “code” as follows: “What is called a code is a logical entity constructed in order to explain and elucidate the functioning of paradigmatic relations in the texts and to explain and elucidate the functioning of syntagmatic relations in these same texts. The code carries in it the intelligibility of the syntagm as well as that of the paradigm, without itself being either paradigm or syntagm” (LC, p. 122).
We’ll leave Metz the responsibility for demonstrating, and not merely asserting, the intelligibility of the syntagm as well as the paradigm; but where intelligibility is concerned we can’t help but point out that such a definition of “code” does not go beyond the original elaboration—without the term “code”—of the paradigmatic and the syntagmatic in his earlier Essais. In short, Metz is introducing, or constructing, a logical entity that will be used to justify, and thereby salvage, his earlier positions.
And, in fact, it quickly becomes clear that what is at stake, ultimately, in Metz’s formulation of the schema of codes and subcodes is the original langue / langage problem and all of its ramifications. Speaking of Metz’s Langage et Cinema, Stephen Heath observes,
Cinematic language is now defined as the totality of cinematic codes and sub-codes in so far as the differences separating them are provisionally set aside so that the various codes may be treated as one unitary system (LC, pp. 51 and 98). . . . The advantages of the term “cinematic language” lies simply in the possibility it allows at certain levels of the analysis of thinking of the various cinematic codes in one block together ... in order to arrive at propositions of the type, “There is nothing in cinematic language that corresponds to the word in linguistic language.”52
The reader interested in films will most likely greet this type of proposition with a resounding “so what?”—and he is hardly likely to find much of an advantage in such terminology. But, of course, Metz is not methodologically interested in films but in cinema; and, in this respect, the notion of the code provides Metz with material for defending his definition of cinema as a langage without a langue. (Incidentally, it is not my intention, in criticizing Metz, to reverse his position on the langue I langage question and to assert that cinema is indeed a langue. Whether it is one or the other or neither is, as I stated earlier, an academic question. The ramifications of asking such a question about the cinema are of far greater consequence than the answers Metz, or anyone else, may come up with. But even in terms of Metz’s answers, the ramifications need to be pointed out, for Metz carefully avoids doing so.)
Contrary to langue (which in Saussurean terms is constituted by a single code— le code de la langue ), language is not a single code but a combination of several codes. And Metz now lays stress on what he calls the “pluricodic character of cinema,” or the “heterogeneity of cinematic language” (LC, p. 143). In the face of this “pluricodic character of cinema,” then, it will become Metz’s central focus in Langage et Cinema to disengage those codes which are “specific” to cinema—the “cinematic codes.”
In Metz’s terminology the codes qualified as “specific” are distinguished from those qualified as “nonspecific” in that the nonspecific codes (for example, the code of dress or the code of speech), although they may be present in films, are present elsewhere as well; thus, they are not specific to cinema. Whereas the cinematic codes appear only in films and are thus specific to cinema.
We’ll come back in a moment to the relations between the specific and the nonspecific codes; but within the specific codes themselves, Metz makes a further distinction between what he calls at one point general and particular cinematic codes, or, as he later chooses to call them, between cinematic codes and cinematic sub-codes. General cinematic codes (or “codes”) are those that are not simply specific to films but are effectively or potentially common to all films—for example (Metz’s), the general code of cinematic punctuation. On the other hand, particular cinematic codes (or “sub-codes”) are those which, while specific to films, are only found in certain particular classes of films—for example (Metz’s), the particular codes of cinematic punctuation peculiar to certain schools or genres of films.
So far so good, although it is difficult to say where all this terminology leads. And once we get some idea of where it leads, there is still the question one always has to ask with Metz of whether the journey is productive. For example, let’s pursue the elaboration of the schema of codes and sub-codes a bit further. Stephen Heath, echoing Metz, tells us that
it is important . . . not to confuse codes and sub-codes. Codes are not in competition with one another in the sense that they do not intervene at the same point of filmic process; there is no choice between, say, lighting and montage. Choice arises between the various sub-codes of a code, they being in a relation of mutual exclusion; there is no possibility of choosing at one and the same time the montage developed in Russian films of the twenties and that associated with the theories of Bazin. Codes and sub-codes of different codes are in a (syntagmatic) relation of addition and combination; sub-codes of the same code are in a (paradigmatic) relation of substitution.53
If we stop and reflect on this summary of Metz’s position for a moment, it is rather revealing. Not that it tells us anything new about cinema, but it does reveal something we might not have seen clearly before about Metz. Heath’s paraphrase of Metz is not only accurate (Heath is a very diligent exegete), it also has the advantage (believe it or not) of being more concise and coherent than the original. And yet Heath, absorbed as he is in the unenviable task of faithfully explicating Metz, seems not to realize that in rendering Metz’s exposition somewhat clearer he is also making it easier for the discerning reader to grasp the ludicrously pedantic and superfluous character of so much of Metz’s work.
Who needs to be told—much less in such a convoluted way— that “there is no choice between lighting and montage”? Can anyone imagine a film-maker laboring under the delusion that lighting and montage represent an either/or proposition? And certainly no critic or theorist who wasn’t wearing the methodological blinders of the semiologist would deem it necessary or the slightest bit enlightening to work out a structural linguistic ‘proof’ of such a self-evident fact. Once again, it is clear, Metz’s semiology boils down to little more than a tedious taxonomy of the banal.
Moreover, precisely because it restates in a pedantic and uninteresting and unproductive way the commonplaces of cinematic practice, Metz’s semiology ignores or is admittedly unable to deal with films that challenge and break down these commonplaces. (Metz acknowledges, for example, that even within his chosen area of competence, narrativity, there is a sequence in Godard’s Pierrot le fou—the escape down the side of the building from the Paris apartment—which, although clearly belonging to the narrative, could not be fitted into any of the eight combinations of shots that comprise his Grande Syntagmatique. )
When Metz (and, following him, Heath) pedantically spell out that “there is no possibility of choosing at one and the same time the montage developed in Russian films of the twenties and that associated with the theories of Bazin,” they seem to ignore the fact that in several of his films Godard has explored these seeming oppositions dialectically; and that in One Plus One Godard has demonstrated the possibility of dialectically employing, “at one and the same time,” Eisenstein’s principles of montage and Bazin’s principles of mise-en-scène. Moreover, it is also worth pointing out here that Godard’s dialectical explorations in One Plus One go against the grain of another of Metz’s basic assumptions, for contrary to Metz’s emphasis on the primacy of denotation and his relegating connotation to a mere ornament of the narrative; Godard demonstrates in One Plus One the richness and primacy of connotation in contrast to the crudeness and poverty of denotation —while narrative itself, in this film, is revealed as a perfect vehicle (perhaps the perfect linguistic vehicle) for fascism. (See Chapter 5, which is devoted to Godard’s dialectical explorations in One Plus One. )
Nonetheless, there is one way in which Metz’s elaboration in Langage et Cinéma of the schema of specific and nonspecific codes might have a limited use—although, characteristically, Metz himself does not do so, for if he did he would finally be obliged to deal with ideology. On the question of the relations between specific and nonspecific codes, Metz, whose preoccupation is largely, almost exclusively, the specific codes, nonetheless acknowledges that “a code which is more specific than another is not necessarily more important; it is merely a code which the language under consideration shares with fewer others (and which therefore characterises that language to a greater degree)” (LC, p. 183). But the whole thrust of Metz’s work, emphasizing as it does the narrative properties of cinematic language, persistently privileges the codes that are specific to cinema’s ways of telling stories.
But, as the editors of Cinéthique point out, “If the analysis takes into account not just the primary object of semiology (the study of cinematic language ), but its secondary object as well (the study of films as systems ), it is possible to argue that the specific codes are not necessarily the most important codes of a film system. The nonspecific codes also play a role in the establishment of that system, and the question then is whether this role is not perhaps always the primary one.” 54
For Cinéthique even to pose such a question amounts, as they are the first to acknowledge, to a rigorous rethinking and autocritique of their own earlier dogmatic position emphasizing cinematic specificity. (Of course, Cinéthique emphasized specificity for very different reasons than Metz.) Moreover, so thorough is Cinéthique’ s self-criticism that they are now willing to characterize as a “fanaticism” their own (past) and Metz’s (continued) “fetishism” of specificity.
In the past Cinéthique’s own fanaticism regarding specificity manifested itself in their dogmatic refusal to consider any film that did not operate self-reflexively in an explicitly political selfquestioning of its own cinematic elements. Of course, such a self-reflexively political “meta-film” is still highly valued; but it is hoped that Cinéthique’s insistence on this kind of specificity will no longer be invoked so dogmatically. (In the past, even such a film as La Hora de los Hornos was dismissed by Cinéthique because it was found to lean a little too heavily on the cinema’s notorious “impression of reality.”)
In any case, where my friends and comrades at Cinéthique are concerned, it is my hope that once having relaxed their rigidity on the matter of specificity—one of the few points on which they and I ever disagreed during our many discussions (and even on this issue the disagreements were only on degree of emphasis)— the editors of Cinéthique may now open up the magazine to some of the productive work that urgently is needed in examining just how ideology is inscribed (within what Metz would call the combination of heterogeneous codes) in various types of films with varying degrees of self-refiexiveness about their specificity.
On the question Metz poses (but does not resolve) of the relations between the specific and nonspecific codes within one filmic system, Cinéthique quotes with qualified approval Metz’s assertion that “the system of the text is the instance which displaces the codes, deforming each of them through the presence of the others, contaminating the ones by the others . . . and placing each code in a determined position in the overall structure. The process of displacement thus results in a putting into place, which is itself destined to be displaced by another text” (Metz, LC, p. 78 ... as quoted by Cinéthique, p. 199).
But the qualification advanced by Cinéthique makes far more sense than Metz’s original statement—and in my opinion can simply supersede the former, since it poses an anterior question, indeed, the fundamental question that Metz has so systematically avoided. In any case, Cinéthique’ s qualification is of crucial importance: “We could agree with this structural view of roles if the question of ivhat produces this structure and its systems were posed with equal explicitness. In a sense, Metz does pose the question, but fails to follow it through to the end, invoking reasons of a methodological order, whereas we see the evasion as a concrete result of the limits within which semiology is enclosed by its ideological presupposition” (Cinéthique, p. 199).
Here, finally, is the real crux of Metz’s elaboration of the code systems in films. And, as we have seen again and again, what is methodologically ruled off-limits by semiology’s ideological presuppositions is precisely any consideration of ideology itself. Moreover, to pose the question of the relations between the nonspecific and specific codes in a filmic text without considering them individually and in combination in terms of ideology is simply to leave out the only conceptual element that can bring clarity out of confusion and make the theoretical model at all useful. (Its use, however, even with ideology taken into consideration, would be very limited, for all the really productive work would have to be accomplished outside of the framework of semiology, that is, within the framework of the Marxist analysis of class struggle and its censorship in ideology. In other words, all semiology offers is an uninteresting and unproductive way of mapping, after the fact, the productive explorations carried out by another methodology, in this case Marxist methodology. )
Cinéthique spells it out quite clearly in the following passage, in which they pick up, and transform, Metz’s terms:
Translating this into Marxist terms, a film is therefore a set of contradictions between two types of heterogeneous elements— the specific and the nonspecific codes—and of contradictions within the specific and within the nonspecific codes. These are the contradictions which can be distinguished from the standpoint of ideology. The contradictions which traverse the nonspecific codes are in fact those which principally characterise the conflict between the proletarian and bourgeois ideologies. . . . The contradictions which structure the specific codes are those which assign a greater or lesser role to some element from the standpoint of its ability to conserve or reproduce a particular ideology. In this context, the codes of iconic analogy are certainly among the most ideologically loaded. (Metz says that analogy is itself coded, which is of course true, but it should also be said that it is coded by the conflict of ideologies [what some would call “culture”].) (Cinéthique, pp. 198-99.)
For his part, Metz may lay out a topographic map of codes classified according to their varying degrees of specificity; and he may even call for an exploration of the relations between specific and nonspecific codes—whose relations, however, he can only characterize in terms of a vague “balance of forces.” But Metz himself draws back timorously from such exploration, and he can only repeat in summary “that the ‘balance of forces’ between the cinematic contributions and the external contributions is extremely variable from one filmic system to another. But these considerations engage the psycho-sociology of cineastic ‘creations’ (and of the spectatorial receptivities), as well as diverse problems of general epistemology, rather than the structural analysis of films themselves” (LC, pp. 83-84).
Once again Metz seeks refuge in cinematic specificity, conveniently thrusting everything else “out there” into “the psychosociology of cineastic ‘creations1’,” “ whatever that may be. Moreover, Metz may argue that his theory is descriptive and not normative; but that is actually a false issue. The fact is that Metz’s description is incomplete; and it leaves off at precisely the point where, because ideology would finally have to be confronted, the theory would inescapably become normative, i.e., for one class’s ideology or another, against one class’s ideology or another. (Actually, of course, Metz’s theory, like all theory, has been normative all along, only without seeming to be. And that camouflaging of its normative character is eminently ideological in itself, and is what has made Metz’s semiology such an attractive ideological tool for the bourgeois intellectuals of our academic institutions.)
Significantly, it is precisely at this point in Langage et Cinéma where, having set out—but not having followed through—on the exploration of the balance of forces, Metz makes a backhanded allusion to the overtly normative character of Cinéthique’ s theoretical positions. Metz acknowledges the existence of what he calls “a new type of ‘political’ film—such as those that Cinéthique would like to see the multiplication of.” But Metz brings up the subject in a way that implies that ideology is only the determinant factor in a few aberrant cases, in the overtly political films that appeal to the lunatic fringe of the far left. And this transparent maneuver makes it all too easy for Metz to invoke the familiar defense mechanisms by which he attempts to preserve his peacefui sanctum of specificity and to dismiss the unruly children with a stern admonition to “go play elsewhere”—in those realms of “the psycho-sociology of cineastic ‘creations’ ”—if they insist on saying that bad word ideology.
It has not clearly enough been recognized just how much Metz’s semiology has developed not just alongside of but in opposition to the simultaneous development in the 1960’s and 1970’s of the Marxist theory of ideology and its application to film theory and criticism. But Metz’s work, which goes through the most excruciatingly painful methodological contortions to avoid having to deal with ideology, is literally haunted by a specter, the specter of Marxism and the Marxist conception of ideology; and this specter hovers constantlv over the pages of Metz’s circuitous detours and semiological deadends. Moreover, like all taboos, the Marxist concept of ideology is constantly present in its repressed state in Metz’s writings—a presence-as-absence, or, in Althusser’s terms, a “structuring absence” whose very repression must violently shape and form/deform the text in which it is “absent.”
One reason, of course, why this fact has not been widely recognized is simply that it is so difficult to pin down just what Metz does say, much less what he represses, covers up, and does not say. But, to utilize the currently popular metaphor of an author’s strategies, one might say that Metz’s density and abstruseness almost seem part of an overall diversionary strategy. And if this were the case, it ought to be clear from our foregoing analysis of the lengths Metz will go in order to avoid ideology, just what it is that Metz’s diversionary strategy would divert us from. (Incidentally, one does not even have to impute such a conscious strategy to Metz personally, it being one of the curious properties of ideology to work its strategies through individuals whether they are conscious of its workings or not. And in this sense one does not have to say that “Metz speaks bourgeois ideology” but rather that “bourgeois ideology speaks Metz.”)
In any case, on the few occasions where Metz cannot avoid at least some allusion to Marxism, he is capable of being extremely coy and even downright duplicitous. One case in point is Metz’s treatment, in Essais sur la signification au cinéma, of Eisenstein, whose “genius” he is careful to acknowledge at the same time that he repeatedly rebukes this “great artist” for his refusal to avoid the cinematic “heresy” of manipulation. At times Metz sounds just as caught up in the transcendentalist position of idealist metaphysics as was André Bazin, who, of course, was also highly critical of Eisenstein’s “manipulation.” And in a confusing and ultimately self-contradictory argument, Metz ultimately winds up reproaching Eisenstein for not acknowledging “the ‘natural’ sense of beings and things” (ESC, I, p. 45 ), whatever that might be; and he laments Eisenstein’s refusal to show us “the course of the world, but always, as he himself says, the course of the world refracted through ‘an ideological viewpoint’..( ESC, I, p. 44).
Ultimately, of course, it’s the same old argument left over from Bazin’s realist aesthetics. While protesting that “it’s not a matter of politics,” Metz, like Bazin (whom he praises, significantly, for being “more subtle than those who reproach Eisenstein for being a Communist”), seeks to cast aspersions on any film-maker who openly acknowledges the ideological character of a film’s discourse. (And is it mere coincidence that Metz, like Bazin, singles out as the main example of such an ideological discourse the work of a Communist?)
Here it is clear that Metz has learned absolutely nothing from all the critiques that have been leveled at Bazin. Metz here betrays a belief in the cinema’s ontological realism just as naive as the one Bazin once championed. And, like Bazin, Metz thinks that the only films which function ideologically are the films which do so openly and consciously, like the films of the Soviet silent masters, or the “new type of political’ film . . . that Cinéthique would like to see the multiplication of.”
Needless to say, there couldn’t be a grosser distortion of ideology and how it works. Is Metz really unable to see that a filmic discourse which purports to show us “the course of the world” in its so-called “natural” state (we’ll leave it to Metz, and God, to define that one for us) really performs an immensely more effective and an immensely more manipulative ideological function than the openly ideological discourse of an Eisenstein, precisely because this bourgeois filmic discourse covers up its own signifying practice and seeks to pass off the status quo—”the course of the world” under the thumb of the capitalist mode of production— as reality ... as “the ‘natural’ state of beings and things”?!?
And this brings us to another instance of Metz’s disingenuous mistreatment of Marxist notions. As the editors of Cinéthique have pointed out in their critique of Langage et Cinéma, the word “ideology” (used as little as possible by Metz) nonetheless “never appears without pejorative connotations . . . and is often associated as a synonym or equivalent with: ‘stereotype’ . . . ‘propaganda’ . . . ‘phantasm’ . . . ‘banality’ . . . ‘constraint’ . . . ‘fanaticism’. . . .”
As if this weren’t bad enough, however, Metz twists the opposition ideology/science (first elaborated by Althusser) to his own uses, distorting and devaluing the concept of ideology while at the same time valorizing his own very narrow and idealist notion of science.
In fact, nothing could be farther apart than the positions of Althusser and Metz, for where Althusser elaborates the foundations (which he finds in Marx’s Capital) of a new science that will, among other things, enable us to understand and account for ideology, Metz merely discredits and devalues ideology in order not to have to deal with it. Moreover, in his very notion of science, as Cinéthique pointedly observes,
Metz remains true to the general project of the “human sciences.” These “sciences” segregate their objects in order to make them autonomous and to cut them off from one another and from what links them, namely, the class struggle. For the basic exclusion on which these sciences are established is clearly that of historical materialism. Metzian semiology is no exception in this respect. . . . The multiplication of the human sciences is in fact very convenient for those who want to stay locked up in their laboratories—they can always tell a caller to try next door, look in another drawer or a different cupboard. Thus Metz can say: “These considerations involve the psycho-sociology of cineastic ‘creations’ (and of the spectatorial receptivities) as well as diverse problems of general epistemology, rather than the structural analysis of films themselves.” ... In this way, the problem which expresses the ideological struggle in a condensed form is no sooner posed than it is thrust somewhere “prior to” film or beyond it, as if the ideological struggle did not intervene decisively in the formation of every filmic system. (Cinéthique, p. 205.)
However, Metz’s avoidance of ideology and his cloak of scientificity are bound to be attractive to academics who are threatened by the intrusion of politics into their sanctuaries. With Metz, though, they have nothing to fear. No unruly disturbances by proletarian and Third World youths here . . . just the reverential murmurings of the mumbo-jumbo of semiological jargon. The academics of film studies, uptight about their intellectual respectability, get a whiff of salvation when they encounter semiology. They rediscover the religious vocation of the pedagogue. At last they have some densely obscure mysteries to impart to the uninitiated.
The distressingly egalitarian, even plebeian, aspect of the study of films gives way to a hierarchically structured ritual presided over by a glib priestly elite, which, in the name of Metz, excuses itself from the need to bother with the critical analysis of individual films. The path to salvation lies in the semiology of the cinema. But like all paths to salvation this one is poorly marked, obscure, and difficult to follow. Quite a lot must be taken on faith. Meanwhile, the uninitiated are asked to await The Word. An article on Metz in the American review Cinema announced the publication in English, at that time still forthcoming, of Metz’s Essais sur la signification au cinéma in terms that unabashedly equated this event with the Second Coming.
One would hope, however, the fact that Metz’s writings are now available in English will have the effect of deflating rather than expanding the cult of semiology. In fact, the bubble of the semiology fad may very quickly burst as readers of the English translation of Essais sur la signification au cinema discover for themselves how much hot air is contained in Metz’s rhetoric.
In any case, where intellectual fads imported from abroad are concerned, there is generally an inverse ratio between availability of the works in English and their position on the ladder of intellectual snobbery. As long as Metz’s writings had not yet been translated into English, there was a certain mystique about them. And the mystique, of course, was played up by the academic impresarios who can be counted on to hop on the bandwagon in a rush to import the latest intellectual acts from Paris. Now that Metz’s work is available to English readers, however, the mystique is fading quickly as the obscurantist mystification that went into the mystique is more clearly discernible.