Metz and film semiotics:
opening the field

by Sam Rohdie

from Jump Cut, no. 7, 1975, pp. 22-24
copyright Jump Cut: A Review of Contemporary Media, 1975, 2004

“There are said to be certain Buddhists whose ascetic practices enable them to see a whole landscape in a bean. Precisely what the first analysts of narrative were attempting: to see all the world’s stories ... within a single structure: we shall, they thought, extract from each tale its model, then out of these models we shall make a great narrative structure, which we shall reapply (for verification) to any one narrative: a task as exhausting ... as it is ultimately undesirable, for the text thereby loses its difference.”—Roland Barthes(1)

A year ago, at a seminar at the British Film Institute centering on the work of Christian Metz and on cinesemiotics, a well-known critic spoke. Robin Wood found Metz insufficiently concrete, too concerned with abstract proposals, instead of with the actual analysis of “real” films. Wood himself has repeatedly stressed the organic quality of specific films—the singularity of the body of the text. “What does all this have to do with films?” Wood wanted to know. The suspicion was that he knew already in the very insistence of his demand   to know what the “pay-off” was, what the “returns” of this theorizing were. Could it be “used,” “exploited” in filmic analysis?

Robin Wood, in his own work, had ceaselessly struggled to grasp the full meaning, the essential nature of individual films, their essence. Whatever that was for any given film, it related to the systematic structure of the film, the way it cohered, in short its coherence, the working out of elements and their finalization, closure, and resolution. The subject author, creator of the unique film, always stood as the fundament of this critic’s criticism. The film, in “the final analysis,” was the expression of the author and the author’s world. (The film as world. The film as representation. The representation of the world in film. The expression of the “truth” of the world. The “truth” of the film.) Behind the film, waiting to be recognized, waiting to tear the mask of the film and to appear, was the filmmaker. Robin Wood’s analysis was a reconstitution of the subjectivity of the author, and of the reader, the author’s subjectivity, and ours. Robin Wood was in good company, and on familiar terrain.

It was right that Wood employed economic terms for his interrogation, criticism and pose of feigned incomprehension of Christian Metz and of cinesemiotics. It revealed a conception of the film as commodity, as product, or, in a different but related system, as fetish, as object of desire. The work of criticism was to possess, own, have, love, admire, worship, grasp that object. And it was the film’s coherence which permitted such unseemly conduct, coherence which was the core and handle which desire might seize. If the film did not present in its structure such a defined hold, center (and many films do not, are incoherent, or at least de-centered), then a negative evaluation was applied. There appeared almost a terror, a sense of loss, at an inability to answer the query, “what is this film about?” as there was terror at the posing of this kind of question. What if it could not be possessed? What if it had no meaning?

This kind of criticism is a perversion. The model for it is the body. To call it organic is simply to point to the sexuality which informs it. The film text reconstituted by this critic is the body of the text, the Copy of the body, behind which, for ultimate possession, is a Hawks or a Hitchcock, a somebody indeed.

Curiously, a close reading of Christian Metz causes some wonder at the virulence with which Metz was rejected that evening. True, Metz fails to talk about “real,” “living,” “existing,” “actual” films. But he does indicate, in the movement of his thought towards structure and the notion of the singularity of the system of the film text, assumptions not so far removed from a traditional organicism.

What is it that was so disturbing, that caused such impoliteness, such nastiness, those social signs of repressed English middle-class violence? In part, one purpose of this essay is to attempt to answer that question in a reading of Christian Metz’ cinesemiotics, to name the threat posed by his work to traditional views despite the traditionalism of his own approach. But to anticipate ever so slightly something of the problematic involved, one relating not only to cinesemiotics, but to semiotics generally.

The problem for the organic critic (let us name some things for what they are) is the fact of the commonness of the body, and its uniqueness, film as both replica and reduplication, the copy of the model, and a singular entity. To posit a singular entity for a given film and to found that singularity in its coherence and system is to run the risk of sacrificing precisely the stake in the game. The unique film becomes common by the very fact of declaring its coherence—a common attribute. System becomes an essential feature of films. In that very gesture of proclaiming a singular uniqueness of the system, the uniqueness is destroyed. To rationalize this problematic of criticism too far is to question the very project of that criticism. A contradiction is exposed at the very center of it, and the force of the contradiction further exposes the force of a desire. It is Metz’ merit and of semiotics to make manifest this desire. Exhibitionism is one thing, to be exposed, to be caught with one’s pants down, is quite another. It was not that this critic did not understand Metz, he understood only too well.

Christian Metz’ cinesemiotics has been fully and sympathetically explicated.(2) No repetition of that exegesis is required, even less so with the appearance in English translation of most of Metz’ major writings.(3) The response to these so far in the United States has been either piecemeal, over-detailed, basically irrelevant, or, it has been academic and scientific, that is, without value.(4) In France, semiotics has been subjected to serious rethinking and criticism by semioticians, whereas in the United States the semiotics which is the object of that criticism can hardly be said to have been established. This fact makes the task of analysis of cinesemiotics doubly difficult, for the terms of such an analysis presuppose the knowledge and effect of the semiotic activity.(5)

To characterize that activity here would only be to distort it. And it might too compromise an understanding of it by condensing and making abstract a work, a genuine activity, practice, production into what it is not—dogma, formula, Law.(6)

In the circumstances, a single aspect of Metz’ cinesemiotics will be treated, that which most clearly relates to problems in traditional criticism, the point of the disturbance Metz causes to that criticism, and which at the same time opens, out into the areas of a critical semiotics. That point is the distinction opposition Metz poses of “text/textual system.”(7)

The “text” is the actual film, “an object ‘of the real world.'.. a given.” The text is what the filmmaker (cineaste) creates. It is defined by its source. The text is fully manifest as real presence, attested process. The text is in the can, it really exists.

The “textual system” is the intelligibility of that reality (of “the text”). The textual system is not the text itself but its system, its structure. The textual system is constructed by the analyst from the text. The textual system is an abstract entity defined not by its source, but by its destination. It is distinct from the text as the text is distinct from the system of that text.

.”.. the text, as text, is distinct from any system, and even from the unique system of which it is the only text. And the system, even if unique, is distinct from any text, including its own.”

“What the cineaste constructs is the text, while the analyst constructs the system.”(8)

The text is an unique, real, singular object. The textual system is an unique, abstract, singular entity. The textual system is distinct from its own text and it is distinct from other textual systems. The text is the material real of the textual system, that which permits the assertion of the textual system as “the singular textual system.”

The opposition “text“/“textual system” contrasts a difference which informs nearly the entirety of Language and Cinema. Indeed, that writing can be organized in terns of oppositions of the kind “attested process“/“constructed intelligibility.”(9) “Film“/“cinema,” the rounding paired opposition of Language and Cinema marks the distinction between actual films, what we see when we go to “the cinema,” and the system derived by analysis from that actuality.(10) That system is designated elsewhere as “the cinematic language system,” which is formally rather similar to the textual system, but writ large. And, as for the textual system, named as “singular textual system,” Metz takes considerable pains to delimit the specificity of the cinematic language system, i.e., its singularity, from other language systems, and above all from language itself. Just as he seeks to establish the specificity of the textual system, so he seeks to establish the specificity of the cinematic language system. Both the textual system and the cinematic language system are abstractions of a process into a system. In establishing specificity, Metz reconverts the textual systems into a process from which he then further abstracts the cinematic language system.

The paired oppositions Metz operates are not only similar in what they oppose, but logically connected. They call to one another in a circle of overall systematicity. The cinematic language system is not derived from texts, but from textual systems, the result of a prior analytic operation. At the same time, the textual systems can only be fully comprehended if the elements of the cinematic language system (codes and subcodes), which it organizes, are known and identified. On the one hand, the textual system, in its very uniqueness, establishes many of the codes of the cinematic language system. On the other hand, it manipulates the codes of the cinematic language system as in part established by other unique textual systems.

Certain important implications derive from this grid of structured oppositions. The end term of one pair becomes the first term of another. If “text” stands as attested process in relation to the constructed intelligibility of “textual system,” the second term stands as attested process in relation to “cinematic language system.” Stated differently, the opposition is one of “heterogeneity/homogeneity.” At each move towards a larger compass, towards the notion of “the cinematic language system,” greater homogeneity is evident between ever greater units. In effect, there is an increase in system.

From this point of view “attested process”—at whatever level it is conceived—poses itself as a threat to “constructed intelligibility,” in other words as a threat to system. System is ever poised to absorb the vagaries, the heterogeneity, the free play in process, to generalize and connect up elements, in short, to systematize, to encompass the free movement of a signifying practice (the film) into the rigidity of the sign (meaning), or, more precisely, the place of the sign within a structured totality (the cinematic language system).

What is meant by this notion of system? It appears to mean nothing other than the dominant established codes of the culture. What is systematic is in effect what has been established, what is conventional and therefore what is communicable. The systematic is what is replete with meaning, what is confirmed by social practices and serves those social practices.

It appears that Metz’ project is a project of repressing that which is not yet sign, which is presignifying, which is a work and an activity and might be termed “signifying practice,” into an ordered system of signs.(11) The text, or rather certain texts, are a danger to system and threaten it, particularly if they move against the dominant codes, refuse any attempt to be absorbed, in short, refuse the attempt to be dominated.(12) The sign of domination is system itself, yet to establish systems is the central concern of Metz (and of traditional semiotics), and that establishment is always directed against the work of signifying, the work of the text, in the name of the structured sign, in the name of a full social meaning.

Metz’ writings throughout employ models of analysis and of structure derived from structural linguistics. At the same time, Metz was forced to point to the difference between linguistic phenomena and aesthetic ones. A task of his cinesemiotics became the definition of differences between language and cinema (“cinematic specificity”) and yet those definitions depended on linguistic notions (“the cinematic language system”). This possible contradiction, or at the very least, this problem, seemed to have been overcome by shifting the relative weight contained in the opposition in linguistics between “language“/“speech.” For the most part, linguists are concerned with the first term of the opposition, and not the second, with the structure of language, rather than the individual “real” utterances of the language. Metz gives greater weight to the “speech” of cinema, to actual films, as the site of the construction of codes which then become part of the cinematic language system, to be used in other texts, now as a fully coded element. The “all at once” character of language—it must have appeared completely formed, otherwise communication would not have been possible—contrasts with the development (evolution?) of the cinematic language system. This would apply to all areas of the cinema, most obviously the syntax of cinematic punctuation, or serious subcodes of lighting, or spatial dispositions.

An analogy from literature might help to further clarify this problematic with which Metz deals and which derives from his close relation to linguistic models. The denotative level of literature is language. Literature is expressed through language, yet it subverts and transgresses language, defines itself as other than language by its use of the fully-fledged signs of the linguistic system as only first term signifiers of another system—its own as text and the larger one of Literature.(13) Literary texts are in excess over the language, supplement, “over-determine” the linguistic sign. Like the other arts, Literature is connotative.

All literary texts perform this work, but some go beyond it in a subversion of Literature itself. If Literature is not a language, it nevertheless, like Cinema, has laws, conventions, rules. These are Literature. To refuse these rules, to negate them, is to refuse the domination of the dominant codes (Literature). It is to literally produce a writing that which is not yet read. To produce within the codes of Literature is virtually to produce a product which has already been read. i.e. conforms to the dominant codes, and is therefore fully developed sign. A writing on the other hand is pre-sign, pre-system. not-yet Literature, a practice and production of signifying (which has not yet been read, not yet systematized) and which by that very fact contests the dominant Literary and other codes of the culture.

It should be clear, at least minimally, that the prime thrust of Metz’ semiotics is on the side of the codes, the system, in effect on the side of the established, the ideologically dominant. His project is to delimit Cinema, codes of the social practice and not the presignifying work which might contest these. On the contrary, it is precisely such work and activity that Metz himself appears to contest,

It might be helpful at this point to return to Metz’ opposition “text/textual system,” in particular the way he specifies the textual system and its workings. It is only a mention in his writing and can be quoted in full:

.”.. a singular system is a combination of several codes ... Saying that a filmic system is a combination of several codes implies that it consists essentially in a displacement ... no code plays a central role in the overall structure of a given text, not even those ‘mobilized’ by the text. What makes the system of a film is passage from one code to another; each film takes shape with various codes, and it is in this ‘with’ that the importance lies.”

“With diverse codes, but also against them. In this sense, each film is built upon the destruction of its own codes. It is not enough to prove that in a filmic system each code is inessential because it is only the combination of codes that is essential ... the proper task of the filmic system is to actively underplay each of these codes by asserting its own particular logic and because it asserts it, an assertion which is necessarily accomplished through the negation of that which is not itself i.e. codes ... In each filmic system the ... codes are both present and absent: present because the system is built upon them (on the basis of them, with/against them), absent because the system is only a system to the extent that it is something other than the message of a code (or a series of these messages), i.e., because it begins to exist only when (and where) these codes begin to cease to exist in the form of codes, because it is this very moment of negation, destruction construction.”

.”.. the movement of the code seems to us to retain all of its importance, not only because the study of codes, outside of any filmic system, is, for semiotic research, an end in itself (although not the only one), but because filmic systems themselves, as active processes of displacement, are only intelligible if one has a sure idea of what it is that has been displaced.”(14)

The deployment of the notion of codes in this statement marks an important advance in Metz’ cinesemiotics. In particular it is an advance over positions taken in Film Language where the only code identified in the cinema was a code of narrative connectives—the grand syntagmatic. That code was given a privileged place as the defining characteristic code of cinema, the most pertinent, most homogeneous, most ordered, most dominant. It limited Metz’ semiotics to a narrow linguistic model and by highlighting narrative committed his semiotics to an extreme realistic defense of the cinema. The essence of cinema, its pertinent feature, was made to rest in a strict logical-temporal order. If there was a cinematic language system that system was, above all else, a system of narrative.(15)

Here in Language and Cinema the text is conceived as pluricodic, no single code being dominant. The grand syntagmatic becomes one of a multiplicity of codes while more general codes of narrative are not even given the status of being specifically “cinematic” since shared by other artistic systems, particularly literary ones. Metz’ notion of codes creates a flexibility in the concept of a cinematic language system and introduces quite new relations between that system and the system of texts.

Despite this advance Metz’ cinesemiotics still seems locked into realist assumptions, most notably at the precise point at which he discusses the pluricodic structure of the film text. A plurality of codes is a permanent given of all artistic texts and it was proper for Metz to have stressed this in relation to the cinema. However, all texts are not plural texts. Some, as Barthes has pointed out, have only a meager plural, are parsimoniously plural, as in the case with the classic realist text.(16)

The parsimoniously plural represses the pluricodicity of the text by establishing a hierarchy of codes, a systematized domination of one set of codes over others. For example, in the classic realist text narrative codes dominate codes of the symbolic, or the cultural, and it is these narrative codes which bestow sense, direction and meaning to the text. That Metz established narrative codes in Film Language to the exclusion of others in the cinema in part pointed to the de facto dominance of narrative in most films. Films, as it were, repressed, kept hidden their pluricodicity through the dominance of narrative. That was the only code which Metz in the first instance could disentangle because it was the only code which immediately presented itself to view, taking the stage and blocking out others.

A double ideological pressure seemed to have operated. The face of realism presented a single view, while Metz already predisposed to that view, already ideologically committed to a classic realism, could not see behind it. A glance was sufficient.

The repression of other codes by dominant codes arrests the activity and free play of these codes by literally assigning them a “place” in a system in which the dominant codes are central, are dominant. In that sense the dominant codes “read” the text, “read” the other codes, establish a place of knowledge and sense, define entrances and exits from which the text may be approached and understood.

The parsimoniously plural of classic realism is a text already played out before play begins. It is that fact which makes clear the notion that such texts are “already read.” Not only do they set a reading, but set a reading in accord with the dominant codes of the ideology, fixed in Literature, or Cinema, or Painting. That ideology, under which we already live, is simply restated, recycled and confirmed. It is an already-known, an already-read.

What is the function of the codic domination of classic realism? It functions not only to close off other codes, arrest their play and their differences, but to establish the unique singularity of the text.

That the text is already singular as a material object naturalizes a process accomplished by the textual system. The textual system actually “ends” in the classic text by appearing to resolve the suspensions, contradictions, differences, enigmas, puzzles which the text has set in motion. It is like a sentence fully predicated, finished, finalized. The fullness of sense and meaning is a function of the closure of the textual codes, the final denial of their plurality.

The multiple voices (codes)in the texts end in a superior voice, a singular voice which finally silences all other voices.(17) That voice is no less than the voice of the subject-author fully constituted by the singular unique textual system. In that sense the dominant code “speaks” as the code of unique individual expression and creativity. The text literally constructs a transcendental subject as the source of the text. To whom does it speak? It speaks to me, the reader-subject. It establishes a social exchange, a communication of meaning between subjects, author and reader, or more precisely, producer end consumer. The text has become complete, replete, in short, a product. The textual system becomes the expression of the author. The unique expression of the author is the unique textual system.

The coincidence of the ending of the material text with the finalization of the textual system makes the latter appear as self-evident and natural, in fact, as “real.” “The end” is truly the end. All is fixed, glued together, with the glue of system, logic, connectives and subject.

When Meta refers to the “singular textual system” the reference is to an ordered systematicity and wholeness of the text, what marks it off, constitutes it. It is not difficult to see why, in the circumstances, his writing so insists on the real materiality of the text. It is that materiality which helps to license the notion of the singularity of the textual system—doubly singular because it is uniquely systematic and because it is real, actualized.

It is here that Metz’ cinesemiotics comes close to the organicism of traditional criticism and to its realist assumptions. What possibly could establish the singularity of the text outside of an unique creative subject-author for which the text is the expression and the sign? What else could bestow that quality of singularity, but a hierarchy of codes which fully establishes the totality of the text, its integrity and its meaning?

It is in Metz’ notion of codes that assumptions of classic realism and the criticism which supports it are challenged. It is where he “fixes” that notion, effaces the play of codes and their play of differences in a presumed overwrought systematicity that he resumes the assumptions which elsewhere his writings contradict.

The gap between a critical semiotics of the cinema and traditional but serious analyses of the cinema will never be closed, or never established as an area of real struggle between conflicting ideologies, unless that semiotics self-critically substitutes its obsession with system, and hence its commitment to the dominant codes of social practice and social cohesion, by a practice which concentrates on the activity of signifying, the work of the text, an area which is pre-signifying and which challenges the dominance of the sign, dominance tout court.

That work can only succeed by a greater attention to texts. To say, as some do, and Metz among them, that such work cannot yet proceed before codes are fully enumerated appears to be the most abstract of empiricisms. It is not a list of signs and systems that is required but rather a study of the practice of signifying, a practice only carried out by texts against systems and against the sign.

The tools are available for such analyses and semiotics has provided them. What blocks its operation is not “not enough facts” or not enough method but too much ideology which pervades a narrow semiotics as it does a wider criticism.

The obsession of traditional semiotics with the sign, with system, with center, with the expressing subject must be critically examined and exposed. That is a work of theory by semiotics on semiotics, a theorizing on the work of the text, a work of signifying, and one which parallels that textual operation in its investigation of the sign, revealing its gaps, absences, fissures, and repressions. The work in fact which pre-cedes the sign and its placing within established dominant discourses.

Robin Wood in his demand for greater attentiveness to the actual film text, however correct, could not have been aware, nor could have suspected the full implications of that demand, in particular for his own critical project.

“The end,” the “pay-off,” the closure of theorizing is not the text as coherent product. Theory is not a “ready-made,” a suitably tailored commodity, a thing to be applied where most profitable. It is instead a production, a signifying practice. And the specific object of that practice, the practice of a critical semiotics, is the signifying work of the text, not its system, its wholeness, its totality, but its production, its generation. Critical semiotics entails a perpetual work on the work of signifying, its own and that of the text.

The cinesemiotics of Christian Metz marks a crucial beginning for a critical semiotics of the cinema. The gaps, the contradictions in his writings point to the necessity for a genuine return to the text, not as site of finalization, as ultimate goal, but as opening, initiation, commencement into an infinitude of difference, play and pleasure.(18)


1. Roland Barthes, S/Z (New York,1974). p. 3.

2. See Screen 14:1/2 and in particular the two articles by Stephen Heath, “Film/Cinetext/Text” and “Metz’s Semiology: A Short Glossary.” And also by Heath, “The Work of Christian Metz” in Screen 14:3.

3. Christian Metz, Film Language (New York, 1974) and Language and Cinema (The Hague, 1974).

4. For an example of the former see Noel Carroll’s review of Metz’ writings in Film Comment 10:6, and for the latter Paul Sandro’s account of Metz in Diacritics, Fall, 1974.

5. See in particular Julia Kristeva. “The Semiotic Activity,” Screen 11:1/2 and her “The System and the Speaking Subject,” in The Times Literary Supplement, 12 October 1973. Also Jacques Derrida, “Structure: Sign, and Play in the Discourse of the Human Sciences,” in R. Macksey and E. Donato, The Languages of Criticism and the Sciences of Man (Baltimore, 1970 .

6. This is done in particular by the “academic” approach to Metz taken in Diacritics while it makes of Metz’ cinesemiotics yet another academic formalism.

7. See the glossary by Heath mentioned above. The entire glossary is structured in terms of paired, related sets of oppositions.

8. Note, Language and Cinema, p. 73.

9. See Heath’s glossary.

10. Ibid.

11. Kristeva, op cit.

12. For Metz, Godard simply elaborates and enriches the narrative core of cinema rather than effecting a subversion of that core, it’s the same again, but better, “richer.”

13. This concept is detailed in Roland Barthes, Elements of Semioloogy, (tr. NY, 1967) and Mythologies (tr. NY, 1972).

14. Metz, op. cit., pp. 101-102.

15. See in particular Michel Cegarra, “Cinema and Semiology.” in Screen 14:1/2.

16. Barthes, op. cit., p. 6.

17. We could say that in the Cahiers du cinéma’s analysis of Ford’s YOUNG MR. LINCOLN, the Cahiers critics refuse to allow the dominant codes to “do their own thing” by going beyond them to pick up other codes which go against the grain (Screen 13:3 Autumn 72).

Charles Eckert’s “The Anatomy of a Proletarian Film: Warner’s MARKED WOMAN” (Film Quarterly 17:2, Winter 73-74) elaborates the concept of overdetermination in film.

18. I am grateful to Ben Brewster for having read this article and for his comments and criticism. I reproduce much of that criticism below. The points I think are correct and important, and I offer them to the reader as a critique of some of the positions expressed in my article or rather as a pointing to some of the dangers which such positions could give rise to:

.”.. there are dangers in making anti-organicness, anti-systematicity a critical value in itself, dangers of a kind of Croceanism, the true work of art evading the fixation of any system and thus attaining a pure, uncoded expressivity. It’s important to specify the way current notions of organicness, systematicity help unify and thus underpin the currently dominant ideologies and thus to be able to give strategic priorities to any deconstruction project. Metz’s current interest in psychoanalysis, the first major statement of which should appear in the next issue of Communications which is on psychoanalysis and the cinema, stems from this problem I think (or rather, as his interest in radical conceptions of art is a bit modish and derivative, I suspect, from the problem of a unification of the textual system based on something less nebulous than the cinematic language or cinematic specificity). Moreover, such a generalized notion of deconstruction is vulnerable on the grounds Paul Willemen raised that the reader is always capable of reading any text offered to him as a system because of its material closure—he construes each gap as an ellipsis, each deviation from a code as a metabole, and so on. For deconstruction to have a political effect, there must be structures that resist this process of ad hoc redefinition of the conventions, which is what I take Kristeva to be getting at in her theory of psychic Bahnungen and social pressures that constrain the processes of signification from outside—once again, Freud being introduced to handle the aporiae of semiotics. “