Excerpted from D. N. Rodowick, An Elegy for Theory (Cambridge: Harvard University Press, forthcoming).
A CARE FOR THE CLAIMS OF THEORY
D. N. Rodowick (Harvard University)
Those who know Metz from the three perspectives of writer, teacher, and friend are always struck by this paradox, which is only apparent: of a radical demand for precision and clarity, yet born from a free tone, like a dreamer, and I would almost say, as if intoxicated. (Didn’t Baudelaire turn H. into the source of an unheard of precision?) There reigns a furious exactitude.
— Roland Barthes, “To Learn and to Teach”
One sees reborn everywhere, after a long eclipse, the interest for theoretical discussion.
— Christian Metz, “On Classical Theories of Cinema”
Often considered to be the discursive founder of the structuralist enterprise in film, revisiting Metz’s earliest publications reveals a more complex and often surprising picture. In a group of texts published between 1964 and 1972, Metz marks out a conflicted conceptual space within structuralism — between a precedent aesthetic discourse in film theory and an emergent discourse of signification, between phenomenology and semiology, between semiology and film, and between sign and image — whose stakes are played out in the imagination and construction of “theory” as a concept whose rarity before the 1960s cannot be underestimated. Indeed the early Metz takes on two projects in the early sixties whose scales are enormously ambitious. Having become associated with the École Pratique des Hautes Études (EPHE) from 1963 under Roland Barthes’s tutelage (and in 1966 elected a directeur d’études), Metz takes on one of the central obstacles to expanding linguistics into a general semiology of culture, that is, to show that the methods and concepts of structural linguistics and the study of speech or langue are applicable to non-spoken phenomena; in short photography and film. As is clear even in Barthes’s early essays on photography, the image is viewed here as both an object of fascination and an obstacle to a general science of signs, which can only demonstrate its universality if it can master the image in signification. The enunciative a priori or implied defining question of the aesthetic discourse from the 1910s through the 1930s was “In what ways can film be considered an art?” And in repeatedly returning to this question, debating it, worrying it, probing it from different angles and from a variety of conceptual frames, the discourse fractured and eroded the concept of the aesthetic itself in a way commensurate with the larger project of modernism in the arts. The enunciative a priori of the discourse of signification, raised by Barthes in “The Rhetoric of the Image,” is “How does meaning get into the image?,” as if the image itself, in its analogical plenitude, is opaque to meaning. Semiology can only lay claim to founding a general science of signs if it can demonstrate that the image is surrounded by meaning, crossed with or shot through with signification, bathed in sense. However, and in a way analogous to the aesthetic discourse, semiology founders in its confrontations with the image; or, as Barthes’s encounters with the image makes clear from the beginning, from a semiological perspective there is something traumatic, anxious, or imponderable in the image that semiology feels compelled to master, and in many respects fails to master. Barthes will finally embrace the idea of an unmasterable core of non-meaning in the image in his return to “phenomenology” in Camera Lucida.
Therefore, one central concern of Metz’s earliest essays is to make a contribution to a general semiology of culture by working within the context of the EPHE in a specialized domain — the cinema. Alternatively, out of this project unfolds another one, less remarked upon yet equally ambitious. More than Barthes, I think, Metz quickly became keenly aware of the difficulties, not of the image, but of renovating the concepts of structural linguistics to extend them to non-linguistic expressions. At the same time, if the semiological program was to include film one also needed to take into account a historical discourse on cinema reaching as far back as the 1920s to show how these writings were already approaching, if often in conceptually imprecise and non-systematic ways, the problem of film as discourse. After Guido Aristarco’s pioneering Storia delle teoriche del film  Metz is one of the first important figures to place the aesthetic discourse in an historical frame, to consider it in all its disparity and dispersion across continents, languages, and decades as a special genre of discourse, distinguishable from both film history and criticism, and one that has a history seeking conceptual unity. Like Aristarco, Metz is constructing an archive (which will be recognized retrospectively as the first canon of classical film theory) but a directed one — selecting texts, identifying predecessors, locating where conceptual foundations have been laid.
This project is not without its ironies and paradoxes. On one hand, Metz is entirely a product of his discursive context. In excavating and refashioning the aesthetic discourse in the early 1960s he is guided ineluctably by a retrojecting framework that revisits and unavoidably rediscovers in the first fifty years of writing on film a preoccupation with language and signification commensurate with, if only incompletely and in a fragmentary way, the larger discourse of structuralism. On the other hand, through his cinephilism, his commitment to phenomenology, and his attachment to postwar French film culture, Metz is at odds with structuralism. The twinned project of contributing to a new cine-semiology, and to recovering and paying homage to a special literature on film, does not necessarily lead to building a general science of culture through linguistics. Metz desires to be rigorous, conceptually precise, and methodologically systematic, but he refrains from making this into a desire for science or for philosophy — it is, rather, a desire for theory.
Emerging out of a series of overlapping yet conflicting discursive formations —phenomenology, filmology, structuralism, classical film aesthetics, and cinephilism — in a series of important texts of the 1960s, Metz finds his way in theory, and in so doing, begins to construct an enunciative position or perspective that can finally be recognized as theoretical. Metz builds a map and a picture of the history of film theory through the discursive formations of structuralism and semiology. Contrary to the usual conception of the early Metz as the founder of a certain discourse and of a method — cine-semiology and the structural analysis of film — Metz here becomes a fairly unique figure within the larger discourse of signification in its era of methodological passion. Metz’s particular conception of theory is directed by a kind of ethical searching at odds with the discursive context that produced him, one that questions a whole mode of existence (in structuralism, in film study, in theory) through the conceptual will to forge a new form of life in thought around the cinema. A closer look at his essays of the 1960s, gradually uncovers the will to locate a position or perspective expressed in the form of a certain moral reasoning. An inheritor of the institutional and academic discourse of filmology, as well as the phenomenology of André Bazin, and inhabiting discourses that are simultaneously cinephilic, philosophical, and ethical, in these essays Metz positions himself as the conciliator between several postwar discourses traversing film and the human sciences, as if to find a new place for film in the human sciences through theory.
Metz’s construction of a place for theory — its positions of address, its points of intersection and conflict with other forms of discourse, its epistemological extensions and limits — unfolds on a sinuous path that moves forward by looping back on itself at frequent intervals in a recurrent process of revision and refashioning, moving in uneven lines across several essays. Undoubtedly, the most fascinating and most complex account occurs in the first half of Metz’s first professional article, “Cinéma: langue ou langage?” published in 1964 in an issue of Communications devoted to “Semiological Research.” In short order, Metz takes up the problem of history and theory again in his review of the first volume of Jean Mitry’s Aesthetic and Psychology of Film, “Une étape dans la réflexion sur le cinéma” (“A Stage in Reflection on the Cinema”). The line continues in a 1967 review of Mitry’s second volume, “Problèmes actuels de théorie du cinéma” (“Current Problems in Cinema Theory”) before another phase of methodological reflection and revision occurs in parallel: first in the opening chapter of Language and Cinema, and then in the republication of the two essays on Mitry in Essais sur la signification au cinéma, II, which are grouped together with a new prologue in a section entitled, “On Classical Theories of Cinema.” Among his many significant contributions, then, Metz was one of the first key figures to adopt a metatheoretical perspective in film study — a reflection on the components and conceptual standards of theory construction, as well as a historical view of the development of film theory. Metz is also one of the first main figures after Aristarco to make present and perspicuous a new concept of theory by constructing theory as an object, examining its history, and testing its present and potential claims to generate knowledge.
That Metz moves, as if searching out stepping stones to cross an unruly stream, from a stage in reflection, to current problems of theory, and then to the assertion of an antecedent and historically locatable period of film theorizing is significant, as we shall soon see, and all the more so in that the canon of film theory so familiar to us today was still fragmentary, incomplete, imperfectly translated, and hardly known. Still, one finds throughout the sixties the emergence of a certain historical consciousness in the form of a desire to revisit, recollect, reorganize and systematize thought about the cinema, especially as represented in Kracauer’s Theory of Film (1960) and Mitry’s great books, preceded by Jay Leyda’s pioneering translations of Eisenstein’s Film Sense (1942) and Film Form (1949). Nonetheless, up until the 1970s a great number of key theoretical texts were unavailable in French, and indeed, in many other languages: Eisenstein and Pudovkin’s work appeared only in scattered fragments and excerpts, Vertov was hardly known, and key texts by Arnheim and Balázs were available only in German. The French genealogy scattered across the diverse texts of Canudo, Delluc, Dulac, Moussinac, Faure, Epstein, Gance, Clair, Cocteau, Feuillade, L’Herbier, or the Surrealists, was dispersed in often hard to find publications. The fiftieth anniversary of the invention of cinema inspired the publication of two important collections in 1946, Marcel Lapierre’s Anthologie du cinéma : retrospective par les texts de l’art muet qui devint parlant (Paris: La Nouvelle Édition) and Marcel L’Herbier’s Intelligence du cinématographe (Paris: Éditions Corrêa), but valuable as they were these volumes were hardly more than a mélange of testimony by directors, actors, and inventors interspersed with selections from aesthetic writings assembled under rubrics that revealed no special concept of “theory.” Still, in France as in Italy, postwar film culture did have a sense of a canon for the aesthetic discourse, as represented by Henri Agel’s little pedagogical volume for the Que sais-je? series, Esthétique du cinéma (Presses Universitaires de France, 1957), which refers to and closely follows Aristarco’s canonization of Balázs, Pudovkin, Eisenstein, Arnheim, and Spottiswoode though without reproducing any of their texts. The first collection of Eisenstein’s texts in French, Réflexions d’un cinéaste, appeared only in 1958.
Throughout this period of recovery, collection, and anthologization an historical perception emerges of there being a corpus of film theory that is relatively delimited and self-contained if only one could assemble all the texts in an orderly way. This desire to discover or construct a canon is fueled both by the rarity of sustained studies of film aesthetics in the classical period and by the cultural and academic marginality of film and film studies. Even in Metz’s case, this perception of rarity and marginality leads to a tendency to think of the history of film theory as a series of monuments: Balázs, Arnheim, Eisenstein, Kracauer, Bazin, Mitry, all major figures who could anchor a field or mark out its borders. (And one believed this territory could in principle be taken in from a single field of vision — even in the early seventies, the devoted student of cinema could still dream of reading every published work in film theory, in English or in French, as the books would hardly fill one shelf.)
Metz’s expert command of German and English, and his institutional placement as an academic researcher in a field which as such did not yet exist, no doubt abets and fuels a drive to assemble, organize, and arrange, methodically and systematically, the available “research” on cinema, as if to reassure himself of a certain place in the history of thought about cinema, or even to show that this thought exists and has a history. No doubt he is also inspired by Mitry’s own drive to organize systematically a certain thought about cinema, to ratify it and to show that it has methodological unity and value. At the same time, it is not clear that Metz viewed the initial phase of his work as contributing to a (semiological) theory of film, so much as appealing to film as a problem in the transition from linguistics to a general science of signs. Metz will thus regroup and reconfigure the canon of film theory as constituted by Aristarco and others to include film semiology as a necessary stage toward developing a “scientific” problem and attendant vocabulary in which film is only a part.
To better understand Metz’s construction of theory, along with the epistemological stakes and perspectives invested in that term, it may be best to begin at the point where Metz concludes the first phase of his thinking: the Introduction to his magisterial thèse d’État, Language and Cinema. Nearly ten years after filing a proposal to study “filmolinguistics” at the Centre nationale de recherches scientifiques, the connection to filmology had not been forgotten. In hindsight it is clear that Metz conceived both “Cinéma: langue ou langage?” and Language and Cinema as functioning in ways analogous to Gilbert Cohen-Séat’s foundational Essai sur les principes d’une philosophie du cinéma (Paris: Presses Universitaires de France, 1946), that is, as setting out a methodological foundation as a kind of conceptual grid: imposing conceptual order, reducing the problem to a manageable scale, defining and aiming at certain problems while excluding others. Language and Cinema is a sort of reconception and rewriting of the Essai but from the standpoint of the discourse of signification, which in 1971 has fully bloomed, meaning also that is has begun to fade. Four years later, with the publication of yet another deeply influential methodological statement in Communications, “The Imaginary Signifier,” Metz would help found again a new discourse, that of the subject and ideology.
In a strong sense, the central question of the Introduction to Language and Cinema is how to bring theory to cinema? Or in other words, how to filter, reduce, or circumscribe the object of investigation to make it the proper object of a theory? The cinema in its largest possible conception, Metz argues, is a total social fact in Marcel Mauss’ sense. As a multidimensional whole it does not lend itself to a unified and rigorous examination, but rather, only to “a heteroclite mass of remarks implicating multiple and various points of view.” As a possible object of theory, this is another way in which “cinema” is analogous to “language,” for language in its largest sense also confronted Saussure as a global, variegate, and multidimensional social whole whose scale and complexity escaped any theoretical purchase. A theory, then, requires a principle of pertinence, a sort of filter or grid that sets the conceptual perimeters of a theoretical object and establishes the lines of latitude and longitude guiding its systematic study. The cinema as such, like language as such, is too vast to be a possible object of knowledge. Saussure laid the foundations for a theory of signs — semiology — in defining langue as a system of signification underlying language more generally, and therein lies a possible opening into film theory. In examining the system of signification, semiology refinds language in another sense, and finds other senses in language. A theory of film, rather than a theory of cinema, will have to perform a similar reduction, isolating only those components of the filmic fact that are discursive or textual.
Metz continues by observing that although narrative film began to emerge about the same time as Saussure was giving his course on general linguistics, theory was a long time coming to film, or at least the components of a theory wherein one could clearly establish criteria for defining filmic and cinematographic facts. That the history of film theory has unfolded, higgledy-piggledy, in the accumulation of heteroclite and syncretic observations and texts is a result of its relative youth as an art form and lack of institutional setting. The history of cinema has not wanted for “theorists,” Metz observes, though it has until recently lacked the constituents of a theory. To make film a possible object of knowledge means reducing the scale of investigation, plotting out recognizable property lines, flattening and shaping the landscape, giving it an architectural design. For Metz the profile of the classical “film theorist” echoes the eclecticism of the writings themselves. In the early decades of writing on the cinema, Metz observes, “What one most often called a ‘cinema theorist’ was a sort of one-man-band [l’homme orchestre] who ideally held an encyclopedic knowledge and a quasi-universal methodological formation.” One needed to be a historian, with complete knowledge of world film production, as well as an economist who could understand the industrial circumstances of production. To define film as art one also needed to be an aesthetician, and if one wished to comprehend film as a meaningful discourse, one was also a semiologist. Finally, to the extent that one wanted to excavate in the content of particular films various psychological, psychoanalytic, social, political, or ideological facts, “nothing less than a total anthropological knowledge was virtually required.”
In short, the classical era risked producing little more than “a heteroclite mass of remarks implicating multiple and various points of view.” What is surprising, nonetheless, is the conceptual richness and precision of early contributions to understanding film (here Metz draws clearly his canon) in the texts of Balázs, Arnheim, or Albert Laffay, in the writings of Eisenstein and the Russian Formalists, or later, Edgar Morin and Gilbert Cohen-Séat where, as Metz notes, the choice of principles of pertinence is already more self-consciously made. For Metz, these names represent phases, stations, or stages on the way to theory, or a theory yet to come. The classical period is thus not a total but only a partial eclipse — light peers through, and it is waxing. If the space opened between Aristarco in 1951 to Metz in 1964 defines a period in which film theory will gradually achieve historical consciousness of itself, in the period between 1964 and 1971 film theory not only acquires a name, it also takes on a form and acquires a method and epistemology — it becomes a genre of discourse.
1964 is not only the date of publication of Metz’s seminal and foundational essay, “Cinéma: langue ou langage?” It also falls between the publication of Jean Mitry’s two volumes of Aesthetic and Psychology of Cinema (1963 and 1965). No doubt, a figure like Mitry embodies more than any other the image of an homme orchestre that Metz sketches on the first page of Language and Cinema. Metz’s deep appreciation of Mitry’s arguments and his accomplishments — fully set out in his two critical reviews on Mitry in 1965 and 1967 respectively, and his frequent citations of Mitry’s magisterial if flawed work — are sincere and his praise fulsome. Nonetheless this praise is attenuated by the curious place reserved for Mitry in Metz’s genealogy of theory. Metz praises Mitry’s books as the synthesis and the outcome of an entire era of “reflection on film,” reflection, however, and not theory. For as Metz will soon make clear, from the standpoint of a possible film semiology Mitry’s work is the apogee, but also the denouement and conclusion, of a certain way of thinking about film. The question before Metz here is “theory”: what counts as a theory of film, what are its conceptual components and its characteristic activities, and who can lay claim to being a subject of theory, its author or enunciator? In posing these questions in a series of works between 1964 and 1971, and sketching out historical markers and directions, in fact, in raising theory’s history as a theoretical question, Metz not only invents film theory but also becomes the first exponent of what I have called the metatheoretical attitude. In these seven short years, for film studies at least, Metz becomes “discursive” in Foucault’s sense. Not just the author of film theories but the focal point of a new system of address, which emits from a new institutional context with its own rhetorical style and sense of place in history, setting out a new conceptual framework defined by precise principles of pertinence and implicit criteria of inclusion and exclusion for the practice of theory.
In looking back retrospectively at the first phase of general reflection on film, Metz observes that in fact there are two kinds of “theories” proposed. (The quotation marks are Metz’s.) On one hand, in everyday language the word “theoretician” still “frequently designates an author whose writings are above all normative and whose principle aim is to exert influence on films to come, indeed, to prescribe a preferential choice of subject for these films.” But another path has been forged through the aesthetic discourse, above all by the authors that occupy Metz’s preferred canon. These are writers who “have devoted all or an important part of their cinematographic efforts to analyzing films such as they exist, and who appear as so many precursors of a description of film, in the sense given this work in the human sciences and notably in linguistics.” These authors are precursors, then, of a descriptive rather than prescriptive form of analysis that attends to films as they are rather than some possible future ideal film yet to be created. There are two sides or dimensions of this pre-theoretical reflection then:
one on the side of the work to come, thought in terms of influence, which does not hesitate to advise or prescribe, which wants to respond directly to the working problems of an ‘artist creator,’ and which only has sense in this perspective, and one on the side of filmic discourses already given, and which seeks to analyze them as facts.
An analogous situation exists in aesthetics, Metz suggests. But the significant point here is Metz’s preference for a descriptive theory of cinema whose main outlines are prefigured, though in a scattered and disunified way, in the most important authors of the discourse of aesthetics. These writers, however, lacked principles of pertinence that could ground and unify their observations about the state of film language. As such, they could follow only furrows they had already plowed, circling endlessly back to the aesthetic a prioris guiding their thought.
However, the first epoch of general reflection on film has now come to an end. One can no longer be satisfied with a variety of heteroclite observations but must clearly choose a principle of pertinence; in other words, theory must rally around a method, which can unify synthetically from a singular perspective the data and knowledge gathered within its domain. What was previously called “film theory” included observations concerning filmic and cinematographic facts but often without differentiating them. Though often illuminating, these approaches were eclectic and syncretic, drawing on a variety of methods without applying any one in a consistent or even self-conscious way. The discourse of aesthetics was not yet a theory of film. The discourse of structure and signification signals another mutation in this history, then, as the opening of a new phase, which Francesco Casetti has quite rightly characterized as “methodological.” In this transitional moment, Metz argues that methodological pluralism is a necessary though nonetheless provisional exigency. One sees here both a defense of filmology, its persistence as a fellow traveler supporting the discourse of signification in film, as well as the flowering of a “theory of the filmic fact” derived from the methods of a linguistically informed semiology. Most striking throughout this chapter is Metz’s implication that semiology is somehow provisional or less stable than sister disciplines in the human sciences, and that theory has not yet arrived here in the form of a singular and unifying method. A striking commonality, then, between the discourses of aesthetics and signification, despite all the characteristics and criteria that divide them, is the sense that theory is yet to come, always ahead of us as a third possibility, envisageable but so far unattained.
Metz’s concern with method in the Introduction to Language and Cinema is already on full display in “Cinéma: langue ou langage?” Throughout the sixties, it is fascinating how Metz seems so concerned with mapping out and clarifying the variety of epistemological frameworks within which film study takes place, as if in his first published essay he needs to create a new mode of existence, in film and in theory. The essay is thus a manifesto and methodological statement, dividing and ascribing tasks, probing and defining concepts, and laying out positions of address. More importantly, it wants to explore the conditions of possibility wherein a synthetic and unified theory of film might be constructed, and as such it is both a prelude and pendant to the Introduction to Language and Cinema. That such a global and unified approach to film might be possible is the lesson Metz learns from Mitry’s Aesthetic, and that a global and unified approach to the problem of signification as such is possible is the very air Metz breathes throughout the sixties. This idea directs, after all, the project for semiological research outlined in Communications 4, especially in Barthes’s “Elements of Semiology,” with all its methodological passion. What remains to be understood is the place of a possible film theory in this discursive universe — now already somewhat ahead of Mitry’s summing up and closing off of classical film aesthetics, but also somewhat behind in making its own positive contributions to a general semiology. Theory as such is yet to arrive in academic film study.
The title of the essay is significant: can the sense of film be studied from within the concepts and methods of linguistics, whose object is langue? Or if film is a language (how could it not be since it conveys meaning) what kind of language is it, or by what rights do we refer to it as a language? The essay aims not only at rendering more precisely an object of study but also at creating and evaluating a perspective from which that object can be known, and in many respects, valued. Already, this is a somewhat strange position to occupy within the context of a “scientific” structuralism. Be that as it may, if theory is a problem searching for an explanation, Metz here redraws a fairly cloudy picture in sharp outline. In so doing, he shifts the discursive landscape and remaps the entire territory of the aesthetic discourse onto the discourse of signification. Where before the persistent problem was “Is film an art, or has it transformed the concept of art?”, now the problem is: “How do images convey meaning, or in what ways can images be considered as signs?” This question lies at the heart of the semiological enterprise and is the key to its aspirations to become a general science. If linguistics is only a subdomain of a more general semiology, then the conceptual domain of speech, and the scientific foundation of linguistics, must be extendible to images, and especially, moving images. This turning of the question shifts all the centers of gravity of the earlier discourse; it displaces elements in their orbits and creates new sources of illumination, lighting up new features of the landscape and throwing shadows over previously prominent landmarks. With what would soon be recognized as Metz’s characteristic precision and attention to detail, the very long prologue to the essay works back through the history of film theory as it was known at the time but with a specific agenda in mind. The prologue focalizes a persistent question of earlier writings on film, though running in the background, as it were, and brings it forward. Again, one outcome of this move is to recast retroactively this discourse as “film theory,” indeed to see in a variety of otherwise eclectic accounts the problem of language and signification in film, and to assess them as false starts or incomplete movements waiting for the proper general concepts and methods to place them in a framework where they can be articulated and resolved, moving forward in a genuinely dialectical fashion.
Here key differences become apparent. More often than not the aesthetic discourse proceeds through an immanent analysis. It begins with the idea that filmic expression has a specific identity anchored in materials, processes, or automatisms that belong only to film. Semiology extends these medium specificity arguments for a certain time only finally to renounce them in the second semiology, whose turning point is Metz’s Language and Cinema. However, Metz’s earlier essay produces another, more violent mutation of perspective, and one that accounted for the resistance to semiology by more aesthetically inclined thinkers. In a very real sense, film as such was no longer the object of theory (and in Language and Cinema that object will entirely disappear into a conceptual, virtual space). Rather, the discourse of signification begins from a general yet precise methodological perspective — that of the “science of signs” — of which film or photography will only be a part of the universe of cultural signification. In the context of the EPHE, this science was forged in the commitment to linguistics and marked by Saussure’s unaccomplished dream of creating a general theory of signs. In this respect, semiological film theory was initially considered as only one component or sub-domain of a general account of signs. However, if photography or film were of special interest to both Barthes and Metz in the early sixties, this is because they posed a special, and in many respects intractable problem for a general and inclusive theory of signs, at least from a Saussurean perspective.
As I have remarked in several contexts, the aesthetic discourse inherited from the philosophy of art a system of categories that divided and ranked art forms according to criteria of spatial or temporal expression. Among the many disorienting features of film was to present itself as an uncanny hybrid of space and time, thus producing the need for new concepts and categories, and in some cases, unsettling and remapping the idea of the aesthetic itself. Being forged in the history of linguistics (running parallel in a curious coincidence with the history of film), semiology confronted in film another intractable division, that of speech and image. Through its commutation tests and concepts of double articulation, syntagmatic and paradigmatic analysis, denotation and connotation, messages and codes, semiology was born in a scientific context confident that its analysis of speech or natural languages was extendible into anthropological and literary structures of expression. The open question in the heroic era of structuralism was whether these concepts and categories would prove pertinent or even applicable to more general forms of expression, especially analogical and pictorial images. Or even, and this is the question that Metz’s essay both wants to answer and finds nearly impossible to answer, is the very notion of “film language,” so prominent among the Soviet theorists and in the fad for grammars of film in the 1950s, a legitimate formulation, or is it in fact an oxymoron? If the image cannot be considered a sign, and if narrative film cannot be analyzed as a language or aesthetic discourse, then the scientific project of a general semiology, a complete theoretical account of signifying phenomena, was an impossible fantasy. This is the project that would preoccupy Metz throughout the sixties, which would bring him into conflict and debate with Umberto Eco and Pier Paolo Pasolini, and which would in fact create the discursive genre of film theory within the context of the larger episteme laid in place by the more general history of structuralism.
Metz’s essay is thus the launching pad for a new sense of theory, marked by the adoption of a vast new range of concepts, a shift in rhetoric and positions of address, and new institutional contexts. Film becomes an academic enterprise, subject to scholarly debate in university seminars and colloquia by trained researchers, in ways that presuppose a common methodological background or framework, even if that framework is open to revision. But here there is another important point to emphasize. Before the discourse of signification there is no “film theory”; there are only aesthetic writings on film. Aristarco’s rhetorical move is ratified thirteen years later by the discourse of signification; or rather, by the early sixties the invention of theory as a discourse in the context of structuralism has fully and invisibly accomplished a retrojection, both carving out and bridging over an epistemic breach, wherein theory enters the ordinary language of academic discourse as if it were always there, as if, from the time of Canudo’s earliest essays, we were and had always been “theorists.”
We find ourselves again beginning with an ending. The conclusion to “Cinéma: langue ou langage?” comes round again to the opening to underscore the stakes of Metz’s arguments. (It also anticipates in interesting ways the Introduction to Language and Cinema.) It is certainly the case that the essay remains a foundational text, laying out the elements for a semiology of cinema, performing for film studies the work that Barthes’s “Elements of Semiology” performed for the study of literature and of culture in general. Metz is concerned not only with working through and critiquing metaphorical uses of the concept of language in relation to film form and narration, but also with making more conceptually precise how one may speak of filmic meaning within the conceptual vocabulary of linguistics and semiology, and finally, with how film both challenges and enlarges the prospects for achieving a general semiology of culture.
These accomplishments would have been enough to assure Metz a place in the history of modern film theory, and this with his first professional academic essay at the age of 33. But half of the text is fully devoted to another question, and one not often discussed: the specificity of theory as a concept. Just as Metz is clearing the ground and making more precise how and under what conditions the concept of language can be applied to the study of meaning in film, he is also concerned with mapping precisely appropriate uses of the term “theory.” Here Metz is equally convinced that there is a literature or language of theory, and that not all writings on film are theoretical; thus, his implicit desire to establish the parameters of theory as a discursive genre. Recall that, with the exception of Aristarco, the term as such has up till now, 1964, been deployed only infrequently, irregularly, and inconsistently; no one embraces it, or if they do, they equivocate even in the larger context of structuralism. Through the discourse of signification, Metz draws the contours of the concept, gives it form, shape, and appearance through a nominative process. Hereafter, vernacular uses of the term will become less habitual as theory comes to denominate a specific kind of practice and a more or less well defined genre of (academic) discourse.
Metz concludes his essay then asserting that up until 1964 there have been four ways of approaching film study: film criticism, cinema history, filmology, and “theories of cinema.” (The scare quotes are Metz’s.) While the history and criticism of film must certainly contribute to a complete understanding of the cinematographic institution, they are not the central focus of Metz’s interest. Nevertheless, what Metz calls the “theory of cinema” is less a present discourse than a historical one (if one is past, another new one must be emerging), whose great exponents were Eisenstein, Balázs, and Bazin. Metz characterizes this approach as “a fundamental reflection (on the cinema or on film, depending on the case) whose originality, interest, significance and, in sum, whose very definition is tied to the fact that it was also made from within the world of cinema: ‘theorists’ were either cineastes, enthusiastic amateurs, or critics”. In contrast, filmology approached the cinema from the outside, carrying out research on cinematographic facts through the domains of psychology, psychiatry, aesthetics, sociology, and biology, whose fundamental figures are Gilbert Cohen-Séat and Edgar Morin. No doubt, many of the concerns of film theory and filmology are complementary as represented by what Metz calls the border cases of Rudolf Arnheim, Jean Epstein, and Albert Laffay. Both approaches are indispensable to the territory of activities that Metz wishes to mark out, a synthesis no doubt possible since it is nearly accomplished in the first volume of Jean Mitry’s Aesthetic and Psychology of Cinema. But there is something missing in this story. Despite the variety and repetitiveness of the appeals to the idea of language in theoretical writing on film, and given the fact that no less a figure that Cohen-Séat underlined the importance of the study of the filmic fact as discourse, there have been few points of contact between linguistics and semiology, and the study of film. That linguistics has ignored film is not unreasonable. But here Metz has a more daring move in mind. The time has come to bring together in a synthetic way the work of the principle theoreticians of film, filmological research, and the vocabulary and methods of linguistics as a way of finally realizing
in the domain of cinema the great Saussurian project of a study of the mechanisms through which individuals transmit human significations in human societies. The master of Geneva did not live long enough to witness the importance that cinema would have for our world. No one contests this importance. We have to make a semiology of cinema [Il faut faire la sémiologie du cinéma].
Curiously, the specificity of the study of film would seem to disappear in the accomplishment of a general semiology; at the same time, the project of semiology cannot move forward without a passage through the problem of how meaning is transmitted through images.
This is a thorny problem that requires some tricky conceptual gymnastics in the essay. We will eventually find our way back to them. But for now let us return to the idea that Metz is trying to survey a vast landscape, in both film study and linguistics, to lay out the perimeters of a new and more contained conceptual space. For the moment, he is less certain of what it is than what it is not. It borders on history and criticism and draws support from them but at the same time it is spatially distinct from them. It appears to be temporally distinct from “film theory” as a historical discourse; at the same time, coming from outside the cinematographic world, filmology is also not “film theory.” What is, in fact, the discursive position that Metz is trying to construct for himself and for the academic study of film?
This question in fact functions as a sort of enunciative a priori, structuring the conceptual and rhetorical space that links “Cinéma: langue ou langage?”, “On the Classical Theory of Film,” and the Introduction to Language and Cinema into a common discursive network. In each iteration of the question, in pursuing a drive towards theory, Metz recurrently finds himself equally confronting the idea that film theory does not yet exist; rather, we find ourselves in a middle period where at best we are only on the way to theory, and that in most respects what will be finally accomplished is not a “film theory” but rather an incorporation or subsumption of the filmic fact into the general domain of a semiology of culture.
This untimeliness of theory as a conceptual and rhetorical position — always to come and always past, never fully present as an epistemological perspective — is on full display in Metz’s writings on Mitry. The interest of these essays lies primarily neither in Metz’s clear and useful account of Mitry’s books, nor in his criticisms of certain of Mitry’s concepts, but rather in Metz’s attentiveness, striking in its perspicuity, to a certain concept of theory. Through Metz, film theory achieves a certain presence, stature, or standing. There is confidence here that film theory has a structure and a history, that it develops and evolves according to a definable arc, and that it seeks a form, which it has not yet attained. For Metz, Mitry’s books are thus a stage or stepping stone in this progressive arc of film theory. They have an intermediate status — summing up and concluding one phase and opening out to another — and an uncertain temporality. They have deep roots in the past, and thus belong conceptually in most respects to classical film theory, yet in their drive towards building a global and synthetic account of meaning and the moving image, Mitry’s work anticipates a theory yet to come. (It is significant that Mitry produces an “aesthetic”; Metz calls this work a “theory.”) Thirteen years after Aristarco’s pioneering book, film theory gels, thickens, and begins to appear in clear outline as the possibility of a systematic and unifying conceptual framework for the study of cinema.
In “On the Classical Theory of Cinema,” Metz also outlines a historiography of theory: that theory is a way of thinking about film that has a history, that it has had a “classic” phase, which is coming to a close in Mitry’s work, a future that can contribute to a global account of the social life of signs, and a present though intermediate phase, which is laying the conceptual foundation for a possible general semiology of the cinema, though in a fragmentary and piecemeal fashion. (Though Metz himself does not say so, this vision of theory does not arise, actually, from the history and discursive structure of aesthetic writing on film, but rather from a larger discursive territory — that of the history of structuralism, already anticipated in Russian Formalism, and especially Eikhenbaum’s “Theory of the ‘Formal’ Method.”) Metz’s 1971 presentation of the two texts on Mitry, contemporaneous with the writing and publication of Language and Cinema, is striking in this respect. In a few short paragraphs, Metz takes pains to lay down definitive historical markers, so many stages in the theory of film marked by discursive fissures and breaks that overlap in uneven strata. The first section of Metz’s 1972 collection — on the classical theory of cinema, and in particular, the works of Jean Mitry — is meant to give an account of how problems of theory were posed in the years of publication of Mitry’s two volumes, 1963 and 1966. Metz wants to put into perspective the “classical” period of film theory (the parentheses are his own, a doubt or hesitation concerning the temporality and conceptual cohesion of such a concept), of which Mitry’s books are at once the apogee and closing gesture, and from which they draw their conceptual and historical significance. The books thus define a precise historical segment in the stations of theory:
It was before the theoretical renewal of 1968-69; just before and in another sense, well before. It was well after the great theoretical era of silent film. It was just after the Bazinian wave. As for filmology, one no longer spoke of it. A hollow period [période creuse] [...]: there was not enough interest in theory to know who was already part of it, and who was then passing into a vast forgetting.
The lack of interest in Mitry’s important books, Metz argues, is caused by their uncertain historical position — they bear witness to the importance of a past tradition that had reached its point of culmination, and having thus exhausted itself had also outlived its audience.
Metz puts the “classical” period within quotation marks not only to signal its temporal uncertainty (How far into the present has it dilated? How deep into the past has it contracted?) but also to clear a space for a new discursive terrain. Through Mitry, the classical discourse has reached its point of culmination in the present but it is not part of the present; it cannot find a resting place within the modern or actual discourse, the discourse of signification, but must remain disjunct from it on several levels. The deep irony of this disjunctiveness is Metz’s recognition of the many points of contact between Mitry’s work and the emerging discourse of signification.
This hole or hollow in the progress of film theory (“période creuse”) would not long remain empty. Metz quickly notes that his own first steps in conceiving the project of a film semiology, “Le cinéma: langue ou langage,” was published in 1964 in between Mitry’s two volumes. (“Une étape” is contemporaneous with that essay, as I have already noted.) But despite the novelty of semiology and the possibility it presents for real theoretical advancement, Metz reiterates his sentiment that it cannot be considered as an absolute beginning for film theory. In its inaugural moment, semiology must take into account, reconsider, and reevaluate what preceded it and made it possible. This task is neither an afterthought nor a supplement, Metz emphasizes, but rather engages directly the value of theory itself.
A single page, then, and apart from a foreword the first page of Metz’s book, but one can already begin to see clearly his conception of the place of semiology in the broader historical perspective of film theory. What is not so clear is how the gesture of placement itself constructs a history of theory with divisions, continuities and discontinuities, way stations and mile markers, retrospective glances and retrojecting movements. Classical writers were on the way to theory, as it were, but could take it only so far. Writing in 1971, Metz believes he sees a future for theory, a renewal and setting of new directions. In between falls a period of transition, a time of taking stock, clearing terrain, and of clearly establishing principles of pertinence that can make real theoretical work possible. Among the other hopes placed in it, film semiology was thus charged with the task of finally building the foundations of a film theory that would contribute to the larger project of constructing a general science of signs.
But what in fact are the criteria defining theory in this sense? How is it different from previous writing on film, and how does it anticipate its place in the general, critical semiology to come?
Mitry’s conceptual concerns here overlap with those of the younger Metz and of semiology in other interesting ways, above all with respect to questions of analogy, representation, the “coefficient of reality” attributed to film, and film’s phenomenological character. In fact, these are all qualities of photography and film that would rub up against and resist the incorporation of mechanically produced images within a linguistically inspired account of signs in both Barthes’s and Metz’s texts of the early sixties. Metz remarks upon this as a problem for the “first semiology,” which constructed an intractable opposition between the analogical and the coded. As Metz relates,
The first semiology could not conceive that analogy itself might result from certain codes, whose proper action is to produce the impression of their absence. And further, today still, if one wishes to critique the illusion of reality, is it not necessary to take the fullest account of the reality of that illusion? Thus a gap still resides between arbitrary codes and analogical codes, even if the latter, precisely, are at present conceived of as codes.
In retrospect, one of the most striking aspects of Metz’s first text on Mitry, “Une étape dans la réflexion sur le cinéma,” is not only his suggestion of a clear historical transition between two ways of thinking about the cinema but also his sense that this thought distributes itself historically in distinct if sometimes overlapping and interpenetrating genres. Metz writes of Mitry’s book that “This work, taken on its own terms, represents the most serious effort of general synthesis to date of which cinema has been the object.” In its breadth, ambition, and logical structure, one imagines it suggests for the first time the real possibility of a general and synthetic theory of film.
If Mitry’s book embodies both a point of culmination and a distinct division, how is it alike or different from other texts that historically considered themselves, or were considered, “theories of cinema”? Metz sets aside journalistic or anecdotal accounts as well as film history to first describe as theory general accounts of film itself divided onto two lines:
The first emerges from what one calls the “theory of cinema”: written by cineastes or critics, or by enthusiastic amateurs, they place themselves in any case within the cinematographic institution and consider the cinema first as an art. The others, of more recent appearance, adopt the “filmological” perspective: approached from the outside, the cinema is grasped as a fact with psychological, sociological, and physiological dimensions, and — more rarely — aesthetic dimensions.
Whereas they might have complemented one another, theory and filmology have, more often than not, experienced tense relations. Perhaps they are two sides or dimensions of a single theoretical approach? They are alike in their generality, Metz offers, as well as in their distinctiveness from what Metz calls “differential studies” of individual filmmakers, genres, or national cinemas. “How can one understand the cinema without being a bit of a ‘filmologist’,” Metz asks, “since film puts to work phenomena that go well beyond it? And how to understand it without being a bit of a ‘theoretician’ because the cinema is nothing without the cineastes who make it? Among Mitry’s great achievements is that he brings these two dimensions together in a single work by a thinker who is also a maker. Moreover, in its great synthetic arc, Mitry’s book establishes a line of thought and a network of filiation and common concerns that reasserts, once again, the emerging canon of classical film theory: Balázs, Arnheim, Jean Epstein, Eisenstein, Bazin, Albert Laffay, Gilbert Cohen-Séat, and Edgar Morin. One finds conjoined within Mitry, then, the aesthetic or “theoretical” line of classical film theory and the scientific or “filmological” line that is a sort of precursor to modern film theory.
Later in the review, Metz characterizes the classical period as a time of violent polemics and blind combat, of too general analysis and contradictory claims for the metaphysical essence of cinema. Although Metz would later revise this opinion, Eisenstein and the Soviets come in for particular criticism for their lack of rigorous terminology, approximative and inexact analysis, and avant-garde enthusiasms rendered in an “artistic” style. In contrast, Metz offers that Mitry’s book marks the passing of this era and the emergence of a new phase of reflection on film, opening
an epoch of precise research, which even if its objectives are general, will no longer be vague or uncertain in its methodological reasoning […]. This book has brilliantly concluded an epoch that was sometimes brilliant but which risked aging badly if prolonged immoderately. Aesthetics and Psychology of Cinema opens a reflection on film to the perspectives of a new epoch, which will have the face of those who make it.
This new era, of course, is the era of signs and meaning, and if Mitry marks the point of termination of one line of thought, moving towards theory, perhaps Metz marks the beginning of another.
We are finally approaching the beginning of “Cinéma: langue ou langage?” The essay is divided into two, almost equal halves: the implicit concern of the first half is to review the history of film theory and to construct an idea of what it means to have a theory; the second half works through methodological problems of applying linguistic concepts to film. It is revealing that most glosses on this foundational essay ignore the first twenty-five pages as if there were something there that was inassimilable or perturbing to the project of the second half, which lays down the ground work for a semiology of film. There are perhaps two reasons why the first half of the essay seems so out of place, or perhaps out of time, a long delay or digression before Metz gets on to the presumed semiological heart of his argument. To understand the first reason means comprehending that Metz himself does not know or has not yet found the place or position from which a theory can be articulated. It is as if one were trying to speak without yet knowing the grammatical rules of a language or even its pronominal functions. Metz is searching, trying to find his place in theory without yet being certain of what defines the epistemological stakes and value of theory construction. The ground continually shifts beneath his feet as he seeks out a stable foundation on which to build a new epistemological perspective (the semiological) alongside an ethical analysis. In fact, it is this ethical dimension of Metz’s questing for theory that seems indigestible though in hindsight it may be the most original and fascinating line of thought in his argument. The reflexivity of these pages is dizzying as Metz tries to put in place a vision or concept of theory that does not yet exist as such, and at the same time also reflects continually on the value of theory as an enterprise. Though Metz is no Nietzschean, one sees him here in almost a Zarathustrian mode, asking, “What does the ‘theorist’ want, and what does he will in wanting it?”
The second reason derives from the place the essay itself occupies in the history of film theory: not only does “theory” as such not yet exist as a concept (we almost literally see it here in a process of discursive emergence), one also cannot yet place it in a history. It is as if the concept cannot emerge without having a certain historical consciousness of itself, heretofore lacking. Theory’s archive does not yet exist. It must be reassembled and evaluated from scattered texts in multiple languages; one must make of it a corpus, defining within it salient questions, problems, and debates with their own internal regularities and zones of classification.
This historical self-consciousness of theory, and the desire to assemble critically an archive from which the potential for theory construction can be adjudicated, is a fairly unique accomplishment for the period. By the same token, this sense of a history of (film) theory could only occur under two conditions. It requires, first, that there is a sense of a canon of aesthetic writing on film as a sort of prelude to theory. Filmology by no means provided this canon nor is there yet textual evidence that Metz was aware of Aristarco’s Storia. However, both polyglot and polymath, and an intensely curious and exacting researcher, Metz constructs his own canon as it were, from German and English as well as French sources. Metz’s canon conforms in interesting though coincidental ways with the first canons of Daniel Talbot and Richard Dyer MacCann, though with an exception: Metz is refining the definition of theory and who is capable of constructing theories; his principle of selection is guided by a concept of theory where earlier collections are not. Second, this canon must define a certain kind of historical space, where there is not only “theory” but competing theories and ideas, grouped together historically. Francesco Casetti has commented astutely that theories in the classical period were local formations contained in distinct social and national communities that were rarely in direct contact with one another. In the post-war period, a new discursive environment occurs, where not only is a new idea of theory coming into existence, but where there is also the awareness of an international history of film theory comprised from an archive whose fundamental texts are now co-present, spatially and historically, and in dialogue with another. Moreover, here the syncretism and eclecticism of the classical era is defined retrospectively from the point of view of an epistemological space where structuralism follows on the heels of filmology, and where a unified and globally applicable theory in the human sciences seems possible. In constructing a space for theory, Metz is clearing the grounds, shifting back through the history of writing about film to sculpt a concept with precision, to review its possible senses, and to reorganize it in a unified field held together with well formed and consensually accepted principles of pertinence.
We have finally arrived, through a series of loops and digressions, though important ones, at the first pages of “Cinéma: langue ou langage?” Most astonishing in retrospect is how Metz begins emphatically with an implied ethical question: from what place does theory speak? In an essay that wants to explore what a theory of language can offer film, the stakes first unfold in a critical evaluation of the language of theory and what theory values in taking film as an object of knowledge. In this respect, it is odd that so much of 1970’s theory opposed Metz to André Bazin, for in the opening paragraph of the essay the cards of the argument are fully stacked in Bazin’s favor. Citing a 1959 interview with Roberto Rossellini in Cahiers du cinéma, Metz observes that at the very turning point of modern cinema in Europe, Rossellini speaks of the great silent age of Soviet montage and the idea of editing as an all powerful manipulation of meaning as things of the past. The era of montage was an indispensable phase of cinematic creation but now it is giving way to other strategies, and other aesthetic approaches, to reality. Here, Rossellini (and Metz) might as well be quoting chapter and verse from Bazin’s “Evolution of the Language of Cinema.” Montage was also thought a theory, Metz suggests, not only because it was one of the first sustained concepts of cinema but also because of its scientific pretensions. Trained as an engineer, the young Eisenstein came to believe in the possibility of engineering reality and subjectivity through the reconstruction of film language. And in this respect, a certain concept of montage became co-extensive with the cinema itself in a long line of influential writers: not only Eisenstein but also Pudovkin, Alexandrov, Dziga Vertov, Kuleshov, Balázs, Renato May, Rudolf Arhneim, Raymond J. Spottiswoode, André Levinson, Abel Gance, and Jean Epstein. Pudovkin introduces yet another variant in the discussion — of the relation between shot and montage, where the shot is only an element of montage whose sense is found in the whole of the construction, not in the content of an individual part. Metz calls this a fanaticism for montage, whose adherents refuse doggedly and categorically any form of descriptive realism to the cinema. Two problems thus arise about the status and location of language in cinema, especially in relation to the shot and to the referential status of profilmic space. Eisenstein’s process is one of fragmentation and reconstitution. That an uninterrupted segment would have its own sense and beauty is unthinkable. In the early Eisenstein, the profilmic space is a raw material to be dissected and reconfigured into a new series whose meaning is unambiguous. Thus for Metz, “Eisenstein does not miss any opportunity to devalue, to the profit of concern for sequential arrangement, any art that would invest itself in the modeling of the segments themselves.”
Metz thus characterizes the era of montage as being dominated by a spirit of manipulation and of engineering the spirit. The theme of the ethical dimension of theory starts to emerge along these lines, and very soon it will be clear that Metz is contrasting two forms of life or modes of existence characteristic of his modernity — the structural and the phenomenological — in order to explore how an aesthetic semiology comes to designate a third path inspired by the phenomenological aesthetics of Mikel Dufrenne, and to a certain extent, the early Barthes. In the opening pages of this essay, an unquestioned foundational text in the history of film theory, what we find then is rather a strong ethical statement, which continues into the second section. The question of film language has hardly yet been asked. The central problem here seems to be the value of the shot of whatever duration in relation to the sequence, and then the question of where meaning is expressed in the composed film? What is most striking in the second section is how the ethical question, rather than the theoretical one, advances; or yet more complexly, how the ethical and the theoretical advance in turns like two strands that weave one around the other. The engineering spirit of sovereign montage has not fallen into the past except in the cinema, Metz asserts; instead, it finds itself reborn in the new cultural attitudes of the human sciences. Where one would think that Metz’s ambit is to present the value of structural linguistics for the study of film, one finds instead a heartfelt plea to soften the structuralist activity by bringing it into contact with modern film, that is, with art. What links the Soviet obsession with decoupage and montage to a certain modern attitude in the human sciences is a passion for manipulating elements through dismantling and reconstructing them — Metz calls this a “jeu de mecano,” playing with Erector Sets, a childhood preoccupation that carries forward into the more adult activities of “engineers, cyberneticians, indeed ethnographers or linguists”.
So here, slowly and subtly, before it is even apparent that Metz is addressing the question of langue or langage, the problem of linguistics, and of the multiple and confusing overlapping senses of “language,” weave themselves into his text. Film should be confronted as a language, but what kind of language, with what sort of linguistics, and from what perspective? Indeed what languages of theory must be spoken or rewritten to examine the possible senses of language in relation cinematographic art? With undisguised irony, Metz associates information theory and distributional analysis with playing with model trains: disassembling, classifying, and reassembling always interchangeable parts — rails, straight, curving, and forked — into ever renewable configurations. Though himself trained in structural linguistics, what Metz is straining towards slowly is a deep criticism of modern linguistics for denaturing and de-aestheticizing language. No doubt, like boxes of rails and connectors in a model train set, ordinary language may be characterized by fairly strict kinds of paradigmatic choices that yield richly varied syntagmatic chains, all of which are open to modelization. (This, in point of fact, is close to what Saussure referred to as langue, an implicit and restricted set of invariant operations underlying mechanisms of sense in ordinary language.) But there is still something in language that resists modelization and the engineering of meaning, something that remains open and ambiguous, only ever partially and incompletely coded, and something also that sticks to the world of experience and is not so easily reduced to a virtual system. Information theory wants to reduce the thickness of language to a message, because language
pulls along too much “substance” within itself, it is not totally organizable. Its double substantiality, phonic and semantic (that is to say, two times human, by the body and the mind) resists complete pigeon-holing [résiste à l’exhaustivité de la mise en grilles]. Furthermore, has the language that we speak become — quite paradoxically when one thinks of it — what these American logicians call “natural” or “ordinary” language, whereas in their eyes no adjective is required when they speak of their machine languages, more perfectly binary than Jakobson’s best analyses. The machine has stripped human language of its bones, sliced it up into neat sections where no flesh adheres. These “binary digits,” perfect segments, now only need to be assembled [montés] (programmed) in the required order. The perfection of the code is triumphantly achieved in the transmission of the message. This is the great celebration of the syntagmatic mind.
In case one misses his meaning, Metz continues by focalizing in the “linguistic machine” a variety of modern preoccupations with automatization, commodification, and the over-processing of raw nature into denatured products where finally, “The prosthesis is to the leg what the cybernetic message is to the human sentence.”
In the opening sections of his essay, then, Metz is objecting to two kinds of theory, in film and linguistics, which are connected by a preoccupation with “engineering” and a way of construing language. What Metz is searching for now is a theoretical alternative both to montage “theory” and to hard structuralist analysis. In hard structuralism language is treated as a product, Metz asserts, or more clearly, a raw material that must be refined in a well-defined process: one analyses by isolating constitutive elements of paradigms, then these elements are redistributed into isofunctional categories (“straight rails to one side, curved rails to the other”). However, the moment that one anticipates in theory,
which one thought of from the beginning, is the syntagmatic moment. One reconstitutes a double of the first object, a double totally thinkable since it is a pure product of thought: the intelligibility of the object has become itself an object.
And one has not in the least considered that the natural object has served as model. Quite to the contrary, the constructed object is the object-model; the natural object has only to hold up to it. Thus the linguist tries to apply the givens of information theory to human language, and what the ethnographer will call “model” is not in the least the reality examined but rather the formalization established from it.
Reality has disappeared into its simulacrum.
Published in 1964 in the rapidly ascendant arc of structuralism, and in the flagship journal of the semiological enterprise in France, this paragraph must have been stunning, even bewildering to some readers, for Metz continues by linking information theory to French structuralism itself. No less a figure than Levi-Strauss is chided for “passifying the real as ‘non-pertinent’.” This theory of abstracting and modeling the real is then linked to the structuralist activity as defined by Roland Barthes, Metz’s mentor at the École Pratique, who is himself criticized because his aim is not to represent the real, but to simulate it. The structuralist activity
does not try to imitate the concrete face of the initial object, it is not “poesis” or “pseudo-physis”; it is a product of simulation, a product of “techne.” In sum, the result of a manipulation. Structural skeleton of the object erected into a second object, always a sort of prosthesis.
Metz, soon to be considered the godfather of cine-structuralism, has here retreated from the core concepts of structuralism. Or perhaps he is trying to imagine another kind of structuralism, and another path to theory, one where the hard structuralism of Levi-Strauss can be softened in the passage through aesthetics in general and film in particular?
After Levi-Strauss and Barthes, the next link in Metz’s chain of argumentation is Eisenstein, considered as a hard structuralist avant la lettre. And in a similar fashion, film theory must seek still another path, not in a return to the filmic past, to the engineering or manipulative attitude that now, ironically, replicates itself in hard structuralism, but rather one in relation to modern cinema, which presents an ethos alternative to the machinic mind. Rossellini is again the avatar of a new way of thinking. “To Rossellini who exclaimed: ‘Things are there. Why manipulate them?’,” Metz writes,
the Soviet might have responded, “Things are there. They must be manipulated.” Eisenstein never shows the course of the world, but always, as he himself said, the course of the world refracted through an “ideological point of view,” entirely thought and signifying in each of its parts. Meaning does not suffice; one had to add signification to it [Le sens ne suffit pas, il faut que s’y ajoute la signification].
This is not a political contrast, as Metz makes very clear, but it is an ethical one, and one with theoretical consequences. If Eisenstein veers too far towards the materialist side of modernity, the scientific and engineering mentality, on the phenomenological side, Bazin’s desire for a direct contact with things through film is too idealist. At stake in this contrast is how one approaches the concept of sense or meaning in relation to signification. At this very moment in the text, semiology makes a surprise appearance as an intermediary possibility, perhaps bridging the materialist and the phenomenological attitudes, or in fact, softening structuralism with phenomenology. Rather than a direct contact of consciousness with things, or a deconstruction and remaking of meaning in a simulacrum, semiology, Metz argues, is concerned with something else:
what I call the “sense” of the event narrated by the cineaste would be, in any case, a meaning for someone (no others exist). But from the point of view of expressive mechanisms, one can distinguish deliberate signification from the “natural” meaning of things and beings (continuous, global, without a distinct signifier, thus the joy read on a child’s face). The latter would be inconceivable if we did not already live in a world of meaning, but it is also only conceivable as a distinctive organizing act through which meaning is redistributed: signification loves to cut up precisely discontinuous signifieds that correspond to as many discrete signifiers.
In this Eisenstein goes too far, not aesthetically but theoretically. Referring to the magnificent segment of the stone lions rising up in protest in Battleship Potemkin (1925), Metz argues that “It wasn’t enough for Eisenstein to have composed a splendid sequence, he intended in addition that this be a fact of language [langue].” How far can the passion for construction go, Metz protests? One variation on the imagination of the sign would be a cybernetic art finally reconciled with science, a vision of poetry programmable by machines. This is an extreme example of a certain orientation of modernity, one of its possible paths, where whether carried forward into aesthetic creation or into cybernetics or structural science, leads to dubious results.
There is a genealogical line, then, that Metz draws from the modernity of sovereign montage to that of Barthes’s vision of “structural man.” Along this line, it must be said, there are many points of attraction for Metz. Both cinéphile chevronné and structural linguist, admirer of Eisenstein (in theory and practice) no less than Rossellini, adept at phenomenology no less than semiology, how to counter-balance all these opposing forces? And how to do so in theory and through language? Indeed, how to seek out in language — both a theoretical conception of language and in a certain conception of theoretical language — a place that reconciles these interests? How to find one’s distinct place in theory? In implicitly asking these questions, Metz is forging for himself here a new form of life in theory.
But to return to my reading, here Metz notes two reservations with respect to his criticisms of structural man or the “syntagmatic mind.” The historical existence of Constructivism in film and film theory waxes and declines well before the emergence of structural man, who appears after the Liberation in France. In fact, the historical situation is yet more complex. The emergence of a Formalist or structuralist attitude is contemporaneous with the triumphant period of Soviet cinema and aesthetics. The two evolve in tandem and in close contact with one another, especially in the pages of Lef and through the work of Eikenbaum, Osip Brik, Victor Shlovsky, and Roman Jakobson in Moscow and St. Petersburg. Moreover, even if the period of sovereign montage is thought to be concluded, structuralism in the thirties was just entering a period of gestation before arriving with Levi-Strauss, Jakobson, and André Martinet in France, all fresh from their encounters in New York. This does not detract from Metz’s main point, however. In the historical moment when a certain mentality (call this from our perspective, a certain form of life in language and in theory, but what Metz calls an “intellect-agent”) becomes conscious of itself and gains confidence in itself, it deserts the cinema, where a new form of modernism is asserting itself in neorealism and the French New Wave. Moreover, the cinematic domain is too small; structuralism needed to deploy its forces on larger territories. It is thus understandable that at the beginning structuralism would have to feel its way slowly toward a field so rich and complex as film.
But here Metz’s second reservation arises. Metz finds it paradoxical that the cinema would be considered such a rich domain for the early syntagmatic mind of the 1920s, for it seems to be in conflict with the analogical power of the film image as well as its phenomenological sensitivity for the real — what Metz calls a continuous and global image without a distinct signifier, which is resistant, in fact, to strict codification. Even from a semiological perspective, Metz’s bets for a new film theory, indeed, for modern theory as such, are placed on the real, or at least, a certain image of the real:
Is it not the peculiarity of the camera to restore to us the object in its perceptual quasi-literality, even if what one gives it to film is only a fragment pre-selected from a global situation? The close-up itself, the absolute weapon of the montage theorists in their struggle against visual naturalism, is it not at the smallest scale just as much respectful of the face of the object as a wide shot? Is not the cinema the triumph of this “pseudo-physis” that the manipulative mind precisely refuses? Is it not based completely on the famous “impression of reality” that no one contests, which many have studied, and to which it owes simultaneously its “realist” tendencies and its aptitude for staging the fabulous?
And here is the dilemma in which Metz finds himself, the double bind that requires a solution in theory — what is most modern in theory, structuralism, finds itself in conflict with what is most modern in cinema, Rossellini or Bazin’s phenomenology of the real. And indeed film (or more precisely, the analogical image) — which might be thought as marginal to the larger enterprise of structuralism whose concern is with all of culture and all of language — will soon become the focal point through which semiology must distinguish itself from linguistics. The image is in conflict with language, and what is most advanced in theory is at odds with the most powerful aesthetic concepts of modern cinema. In league in many respects with Barthes’s writing on photography in the same period, Metz must now remodel a concept of language to find a new way of approaching semiology — not a science (filmology), but something methodologically rigorous and conceptually precise; not a hard structuralism, but a soft one.
From a semiological perspective, film theory could only have a paradoxical status in its current state. Given Metz’s view that the cinema does not lend itself well to manipulation or to the engineering mind, why did it generate so much enthusiasm for certain “theorists of construction” [“théoriciens de l’agencement”] like Eisenstein and the Russian Formalists? The great attraction of film for Constructivism was based on a fundamental conceptual error in Metz’s opinion. Like a language, film seemed to have fundamental and distinct levels of articulation — from the photograms on the film strip, to shots, to sequences, and to larger structural parts — that could be broken down, reconfigured, and rearticulated. Why should one not see a meaningful system of articulations there? Metz continues in observing that
the error was tempting: seen from a certain angle, the cinema has all the appearances of what it is not. It seems to be a kind of language; one saw there a langue. It authorized and even required decoupage and montage: one believed that its organization, so manifestly syntagmatic, could only proceed from a prerequisite code, even if presented as not yet fully conscious of itself. The film is too clearly a message for one to suppose it without code.
This is perhaps the moment to follow Metz in a short digression. The problem of the essay — cinema, langue ou ¬langage? — so limpidly posed in French has always presented obstacles to English readers, above all in translating the term langue. Langue is not exactly speech nor is it language. In a footnote to these paragraphs, Metz explains the basic conceptual distinction where for Saussurianism langue is a highly organized code, while language covers a zone of interest more amorphous and more vast:
Saussure said that language is the sum of langue and speech. Charles Bally or Émile Benveniste’s notion of the “language fact” goes in the same direction. If one wants to define things and not words, one would say that language, in its most extensive reality, appears every time that something is said with the intention of saying [...]. No doubt, the distinction between verbal language (language properly speaking) and other “semes” (sometimes referred to as “language in the figurative sense”) imposes itself on the mind and must not be mixed. But it is [also] normal that semiology would take an interest in all “languages” without prejudging from the beginning the extension and limits of the semic domain. Semiology can and must draw important support from linguistics, but the two cannot be confused.
Two problems arise from this terminological digression. On the side of code, langue is neither speech nor language, nor is anything gained from opposing natural and aesthetic languages. Metz needs something more here than Formalism’s main principle of pertinence, the distinction between practical and poetic language. Secondly, semiology must deal with a vast range of meaningful phenomena (semes), many of which are not linguistic in nature. Yet, as a science of meaning linguistics has not been surpassed, and must still nourish the concepts and methods of semiology. The contrast between langue and parole, or code and message, is not only a key principle of pertinence for Saussure’s linguistics, it is also essential to his imagination of a more general semiology. Message, speech, language, and seme are all actualized instances of meaning, but the langue underlying them is virtual. Where langue is so close in French to “tongue,” or “national language,” here it is more like a virtual force, nowhere present in any instance of signification, yet at the same time underlying all meaning as the structured system of differences from which an expression gains and transmits sense.
Herein lies a conceptual confusion where all the various “grammars” of film and treatises on “film language” have come to grief. Because films are understood, and are repeatedly understood, one searches in them for a conventional syntax. Yet, at best one will find only fragile and partially coded elements torn from reality, like
a great river whose always moving branches deposit here and there its bed, in the form of an archipelago, shaped from the disjointed elements of at least a partial code. Perhaps these small islands, hardly distinct from the watery mass, are too fragile and scattered to resist the external forces of the currents that gave birth to them, and to which in return they remain always vulnerable.
Metz later continues this line of thought in a significant passage:
In the cinema, everything happens as if the signifying richness of the code and that of the message were connected together [unies entre elles] — or rather, disconnected — by the obscurely rigorous relation of a kind of inverse proportionality: the code, when it exists, is coarse. Those who believed in it, when they were great cineastes, did so in spite of themselves. When the message becomes more refined, it undermines the code — at any moment, the code can change or disappear; at any moment, the message can find a way to signify differently.
The impermanent, unstable, and even historical nature of code in aesthetic expressions already throws up a challenge to Saussure, who insisted that only a synchronic analysis could reconstruct the underlying system of a langue. All the (phenomenological) qualities of analogical artifacts, and indeed the historical variability and innovativeness of art, erect conceptual barriers to a theory of the code, at least in a strict sense.
The open question for theory, then, is how to remain sensitive to the open and complex processes through which films have, gain, or give the appearance of intelligibility? On one hand, Constructivist or Formalist writing on film goes too far in taking shots for words and sequences for phrases, thereby finding the structure of langue, speech, and other forms of “pseudo-syntax” within the filmic message. Sovereign montage dismantles the sense interior to the image to slice it up into simple signs exploitable at will. On the other, without montage, or rather, the extreme forms of montage, modern cinema unveils another kind of expressivity, and therefore a kind of “language” immanent to the analogical image itself in its phenomenological density and richness. Metz calls this another or alternative kind of organization [agencement], where “the signifier is coextensive with the whole of the signified, a spectacle that signifies itself, short-circuiting the sign properly speaking.” Following Merleau-Ponty’s lecture on “Cinema and the New Psychology,” and indeed a whole line of post-war reflection on the phenomenology of the image, Metz finds film to be the phenomenological art par excellence, where the moving image,
like a spectacle of life, carries its meaning within itself, the signifier only uneasily distinct from the signified. “It is the felicity of art to show how a thing begins to signify, not by reference to ideas that are already formed or acquired, but by the temporal and spatial arrangement of elements.
The film image short circuits the linguistic sign, but at the same time it is not life itself but rather a composed, complex, heteroclite image; not a langue, but nonetheless a language, and again following Rossellini, a “poetic language.”
Thus the title of the essay already gestures towards Metz’s key dilemma in theory. The problem of meaning in film must navigate carefully between, on one hand, the domain of langue and the conceptual precisions of structural linguistics, and on the other, language, or the phenomenological richness of the analogical and aesthetic image. This dilemma organizes all the great rhetorical poles of the essay, including the recurrent contrast between Rossellini and Eisenstein in the realm of poetics, and the historical distinction between the “classical theorists” of film and the broader, more synthetic semiology to come. At the same time, these are also ethical choices, laying out approaches to life and to thought as the odd introduction to the essay makes clear. As an alternative to structural linguistics, Metz searches out an aesthetic or poetic semiology to forge a compromise where the search for a place in theory might define a domain that is both conceptually precise and aesthetically rich. Even more striking is the way that for Metz the new, modern cinema already anticipates, reconciles, and transcends these oppositions in its very forms; it is ahead of or anticipates theory in this respect. The modern cinema includes both montage and sequence-shot in its creative repertoire, and here Metz agrees completely with Mitry that there is no film without montage, or rather, editing. The analogical power of the image, the near fusion of signifier and signified, cannot define the whole of the film image but only one of its most important components — the photographic image. The image is not reducible to the photographic alone. The shot enters into many kinds of combinations and on various scales or degrees: “A film is made of many images, which take their sense, one in relation to the others, in a play of reciprocal implications”. The signifier and the signified are thus separated in a way that indeed makes “language” possible. Therefore, through their interest in aesthetic or poetic language, even the Bazinians and Left Bank filmmakers have the merit of having conceived a sort of spontaneous and intuitive semiology that refuses any consideration of cinema as a langue.
Finally, there is yet another polarity that must be reconciled in Metz’s essay, and this polarity poses two obstacles to the kind of aesthetic semiology Metz is searching for. Within the historical space of “classical theories,” which Metz no doubt considers the precursors to a more modern approach signaled by semiology, there are two possibilities or pathways on the way to theory: one which veers too closely to language, the other of which strays too far from it. On one hand there is Formalism or Constructivism, what Metz calls the adherents of “cine-langue”; on the other, there are the “aestheticians,” such as Balázs and Arnheim. In each instance, it seems always to be the case that theory has not yet arrived: one constructs the components of a theory, but then there occur the false starts, detours, digressions, and cul de sacs where in the aesthetic discourse either one veers towards Constructivism and cine-langue or towards art and expression — theory must reconcile the two. The second obstacle is that the conceptual genealogy of cine-semiology descends directly from the Formalists (in the broadest sense), who, Metz implies, may have posed the problem for film in a limited or inadequate manner. And this observation turns round to complicate the first problem. In 1964 a linguistically inspired semiology passing through structural anthropology aims high, hoping to construct a general and critical account of culture as language. But if a general semiology is to transcend linguistics to become a comprehensive account of the life of signs in society, of signifying culture, it must widen conceptually the province of language to include non-linguistic expressions. And here all the most intractable problems will pass through the analogical arts, primarily photography and cinema, “messages without codes” as Barthes put it at the time. The artistic domain, which at first glance seems tangential, now becomes the central obstacle to constructing a general theory. Suddenly, the minor art of film is a major concern for semiology. Moreover, to construct a theory by bringing the two domains in contact with one another, to produce a defendable epistemological perspective on the filmic fact that is equally attentive to the phenomenological experience of film, Metz needs a new concept of language, one which, like filmology, comes from outside the cinematographic institution but which also remains attentive to the expressive power and complexity of the works themselves.
To be on the way to theory, then, means returning to but also remapping the problem of speech or cine-langue in pre-war writing on film, and also, from the perspective of modern aesthetics and structural linguistics, to pass judgment on the first stage or phase of theory, which now implicitly, though in a scattered and disunified way, follows the Ariadne’s thread of the concept of cine-langue, and this, paradoxically, in the era of silent film. Metz is well aware of the irony: “No era was more verbose than that of silent film. So many manifestos, vociferations, invectives, proclamations, prophetic statements, and all against the same fantasmatic adversary: speech.” And all seeking purity of expression, as it were, in a moving visual image of universal power.
At the same time, the concept of cine-langue sought out something like a universal syntax in the silent image, something that made of images a “language” but a non-verbal one. In returning to and remapping the canon of aesthetic writing on film, Metz defines a two-fold project. On one hand, he identifies and defines a certain genre of writing on film — film theory — and gives it a conceptual valence distinguishable from history and criticism. Historically, this is both a backward looking and forward projecting gesture, which in each case launches itself from a space located within the discourse of signification. The objective of constructing a new idea of film theory is to make it part of a larger project — the general semiology to come as the foundation for the human sciences. At the same time, this rewriting or remapping is a retrojection, reformatting the aesthetic discourse in the structure of the discourse of signification, making of it the first or preliminary archaeological phase to which film semiology will be a second and intermediary step contributing to a general science of signs.
After stating his criticisms and hesitations concerning the status of the concept of cine-langue, Metz returns to them to examine what elements or characteristics bring them close to theory, or render them as stages or stepping stones, partial and fragmentary attempts to find a path towards theory. The seduction and the sin of early writings was to have been on the right road but going too fast in the wrong direction. Many found a path toward theory through the problems of meaning and language; nonetheless, they operated with an inexact, even mistaken, concept of signification and of language,
for at the moment when they defined the cinema as a non-verbal language, they still imagined confusedly that a pseudo-verbal mechanism was at work in the film [...]. A thorough review of theoretical writings of the period makes easily apparent a surprising convergence of conceptions: the image is like a word, the sequence is like a sentence, a sequence is constructed from images like a sentence from words, etc. In placing itself on this terrain, the cinema, proclaiming its superiority, condemned itself to an eternal inferiority. In comparison to a refined language (verbal language), it defined itself without knowing as a courser double.
This is what Metz calls the paradox of “talking cinema,” in expression and in theory. The key aestheticians of the silent period and the transition to sound had an unclear and even somewhat perverse understanding of the complex relationship of speech to image. They viewed this relationship as antagonism and rivalry, which blinded them in theory to the wealth of possible combinations and interactions between image and speech, each equally impure, each equally enriched, by their mutual interaction. Looking back at this period historically, like Bazin but for different reasons, Metz observes that for a certain cinema nothing changes during the transition to sound. In fact, not until a new modern cinema was born, perhaps with Citizen Kane (1941), did the image transform itself to welcome a new relationship with speech, and not any kind of speech, but rather, a modern aesthetic discourse. Modern cinema appears again in Metz as a sort of herald for theory — the proto-conceptual Theôry who announces a new relationship of image to language, which can only be finally understood in a new construction of “theory” where Metz’s aesthetic version of structuralism hopes to make a contribution. Here the modern cinema finally becomes a “talking” cinema that conceives itself as a supple aesthetic language, never fixed in advance, always open to transformation. Referring explicitly to Étienne Souriau (and implicitly to André Bazin), Metz writes that the long take has done more for talking cinema than the advent of sound, and that a technological innovation can never resolve an aesthetic problem — it can only present the problem before a second and properly artistic creation comes to suggest possible solutions, which can consequently be expressed in theory. In this manner, the modern cinema of Alain Resnais, Chris Marker, and Agnès Varda constructs a new conceptual relation of language to image, a complex yet “authentically ‘filmic’” discourse. In many respects, they present to semiology what is a stake in a film discourse.
We are now close to the end of the first half of the essay. After all of his criticisms of Constructivism, of cine-langue and erector set cinema, Metz then concludes the first half of “Cinéma: langue ou langage?” with an appreciation of cine-langue as theory, or perhaps pointing the way towards theory. Metz offers that these writings formed a whole body of theory (“La ciné-langue formait tout un corps théorique”), which must be evaluated as such. The open question here is what are the components and conceptual stakes of theory that appear in outline or in their initial steps in the 1920s and 30s, and which are more or less clearly distinguishable from criticism on one hand, and history on the other? And there is another term in this equation — art. Metz observes that there may have been an erector set cinema but not erector set films. “Cinema” here means an idea or a concept imagining, desiring, or proselytizing for a certain kind of film. But, pace Arnheim, the great films of Eisenstein or Pudovkin transcended their theories: “The common tendency of many films of this period were only hypostasized in the writings and manifestos. The tendency never realized itself completely in any particular film”. Aesthetic thinking through a filmic discourse, in this respect, always remained ahead of theoretical expression itself. This observation is related to Metz’s subsequent comment that from a historical perspective the cinema could only become conscious of itself, as film and as art, through excess or exaggeration; hence, the ecstatic tone of the period’s manifestos and various cris de coeurs. The period of cine-langue is thus important for two reasons. After 1920 or thereabouts, it coincides with the birth of an idea of cinema as art and thus represents the emergence of a kind of historical consciousness as well as an anticipation of theory through aesthetic practice. Secondly, Metz notes that his central question — cinema, langue or language? — could only begin to be presented at the moment when the first film theories were being conceived. The whole conception of cine-langue — though preliminary, incomplete, and excessive — nonetheless raises questions of both art and language. Though Metz does not say so directly (he says it everywhere indirectly), the path to theory is sign-posted here as passing through, and perhaps beyond, the domains of the aesthetic and the linguistic. The possibility of theory, however, had to wait for more modern approaches to both art and linguistics, and in this respect film, like every art, exhibited its proto-conceptual and anticipatory force. At the apogee of sovereign montage, Metz concludes, and without attendant theories or manifestos, directors like Stroheim and Murnau prefigured the modern cinema. This idea of cinematic modernity is, of course, Bazin’s. And at the same time, il faut faire la sémiologie du cinéma.
The theory to come — film theory as a stage or step towards a global and unified semiology — must pass through the linguistic and the concept of langue, and at the same time it must become “translinguistic” passing through the problems posed by non-verbal languages. The question of cinema has pride of place in this framework. And here, interestingly enough, Rossellini is evoked once again to establish that film is an art rather than a specific sign-vehicle, and must be treated as such semiologically. The simple conclusion and the profound irony for the discourse of signification is that while films are powerfully meaningful and expressive nothing can be gained for semiology by considering them as analogous to a langue. But just as a general semiology will only come into being by transcending and subsuming the domain of linguistics, film theory will become a sub-domain of semiology in recognizing concretely the ways in which cinema is a language without a langue. Testing the conceptual limits of langue in order to map out the possible and legitimate ways of treating filmic expression as language is the great technical task of the second half of Metz’s essay. That useful pedagogical task must be left aside here. The important point to conclude with is to account fully for the role played by the aesthetic, or a transformed idea of the aesthetic, in forging the discourse of signification. In one of the most remarkable sentences of the essay, Metz writes that “The ‘specificity’ of cinema is the presence of a language that wants to be made art, in the heart of an art that wants to become language. [La ‘spécificité’ du cinéma, c’est la présence d’un langage qui veut se faire art au cœur d’un art qui veut se faire langage].” There are two directions of “language,” then, neither of which is predisposed to being understood as a langue. On the one hand, there is what Metz calls an “imaged discourse” [“discours imagé”], that is, the moving photographic image as “an open system, difficultly codifiable, with its non-discrete fundamental unities (= images), its too natural intelligibility, and its lack of distance between the signifier and signified.” But there is also a “filmic discourse” that draws upon a variety of other elements to compose a film expressively, not only with moving images and montage, but also with dialogue, music, sound effects, written elements, structures of narration and patterns of spatial and temporal articulation both invented and borrowed from the other arts, which are only partially codifiable. “Art or language,” concludes Metz,
the composed film is a yet more open system [...]. The cinema that we know (there will perhaps be others [...]) is a “menu” with many pleasures: a lasting marriage of art and languages that constitutes a union where the powers of each tend to become interchangeable. It is a community of wealth, and in addition, love.
To construct a film theory while maintaining a love of cinema, to make this theory conceptually possible and terminologically current, now means knowing to what extent the vocabulary of linguistics advances or blocks the passage through film to a general semiology. For the possibility of semiology is also the path to having or possessing a theory, or to know that one thinks theoretically. This is why to become or be on the way to theory, the discourse of signification has to find itself pre-figured in the aesthetic discourse. Or to put it in a different way, theory is only the partial and intermediate transition point toward a more general science.
For all the pages so far written in this essay, and for all the twists and turns taken in Metz’s brilliant argument, the question still before him, then, is that if the cinema can in no way be considered a langue, then how to defend his conviction that a “filmolinguistics” is both possible and desirable, and that it must be solidly grounded in the vocabulary and method of linguistics? One of the founders of the discourse of signification, for Metz the path to a global semiology and a science of signs must pass through a linguistically inspired film theory. This conviction produced two consequences for his writings of the period. First, his retroactive historical reconstruction of a certain history of writing on film from the 1920s produces a canon where in fact to claim their status as theory means to have considered the problem of language in whatever form. A process of retrojection is at work here, where the highly variegate and contradictory aesthetic discourse is being (has been) transformed by the discourse of signification. The past canon of film theory is thus selectively formed to contribute to a debate in which filmolinguistics or cine-structuralism will be both the culmination and the passage to new, broader, and more synthetic forms of knowledge. Theory here becomes a theory of language and structure, inspired by Saussure, a process begun already by the Russian Formalists in the twenties and thirties. Tracking back for the moment from our restricted view on Metz’s first essays to include the prolific work of other writers of the period, including Umberto Eco, Pier Paolo Pasolini, Raymond Bellour, Noël Burch, Emilo Garroni, Yuri Lotman, Peter Wollen, Sol Worth, and many others, even including Jean Mitry, we can see that despite the will to forge a common method and conceptual vocabulary for (cine)semiology, the discourse of signification was itself a highly variegate and in some senses syncretic discourse. Nonetheless — and here filmology indicated a real and fundamental change — there was a sense common to almost everyone of a shared, international dialogue or debate within a more or less common set of problems and concepts, of moving forward through conceptual conflict to a more precise and unified approach defined by the problem of signs and meaning in images.
Marc Vernet has observed that Metz’s writings can be organized into three distinct phases, each with their particular style of writing, each of which defines its own particular conceptual and epistemological space distinct from the others: the collected essays of the 1960s, Language and Cinema, and finally, “The Imaginary Signifier.” These phases are all points of passage or transition in theory, moving from the problem of signification to that of the text, and finally, to psychoanalytic accounts of the signifier. In taking account of the variety of Metz’s contributions, and his extraordinary drive and commitment continually to revisit critically and to remap the stakes of theory, both epistemologically and evaluatively, we can better understand his unique contributions not only to building film studies as a modern university discipline but also to forging a discourse now often taken for granted: the theory of film. What drives Metz’s epistemological and ethical searching from the very beginning is his dual sense of both the fundamental necessity of theory as conceptual critique and innovation, and an idea that theory is always open and incomplete, not yet arrived and always to come. In the decades of semiology’s methodological passion, Metz was one of structuralism’s most powerful critics, and also one of its true believers, but by the early seventies the dream of a global and unified science of signs was rapidly fading — the discourse of signification was fraying and splitting into new formations; structuralism was turning into poststructuralism, and theory was becoming Theory. In this respect, it is interesting to return to the Introduction to Language and Cinema and its retroactive account of what Metz calls the three phases of “film theory.” In the first phase, what was referred to as the theory of film was eclectic and syncretic, and “called upon several methods without applying any of them in a consistent manner, and sometimes without being aware of doing so.” The semiology of the cinema, which preoccupied Metz throughout the sixties, and whose crowning achievement was Language and Cinema, is obviously here only an intermediate stepping stone — not yet a theory, but building the foundations of a methodology on the shoulders of filmology through a process of conceptual clarification and reorganization in the context of a general science of signs. Metz continues by anticipating a third phase to come,
where various methods would be reconciled in depth (which could imply the disappearance in common of their present forms), and film theory would then be a real synthesis, non-syncretic, capable of precisely determining the field of validity of different approaches, the articulation of various levels. Today, it may be that we have reached the beginning of the second phase, where one may define a provisional but necessary methodological pluralism, an indispensable course of treatment through division [une cure de morcellement]. The psychology of film, the semiology of film, etc., did not exist yesterday and may no longer exist tomorrow, but must be allowed to live today, true unifications never being brought about by dictate but only at the end of numerous studies.
It is a tribute to Metz’s influence on the field, and his own capacity for self-criticism and innovation, that Noël Carroll will echo this sentiment twenty-four years later in his own introduction to a collection co-edited with David Bordwell, Post-Theory. Moreover, Metz’s major turn to psychoanalysis only four years after Language and Cinema would force a wild shift in the discourse of signification and, at the same time, set in place a new discursive situation of increasing conceptual pluralism, opening the era of contemporary theory in film, media, and art. There is a certain irony here in noting Metz’s close agreement with Bordwell and Carroll about the prospects for theory and its incompleteness, that we have not yet entered a conceptual space where a theory of film is possible. At the same time, in what may have been his last interview, Metz characterizes this openness or incompleteness as a kind of ethics or modesty in theory. The interview with Marie and Vernet ends with Metz offering a tribute to Roland Barthes as his only real master. Metz describes this debt to Barthes as a care for the claims of theory, of thinking theoretically, while maintaining a certain flexibility or openness: to not be attached to a theory but to change positions according to need. In this, one better understands Metz’s rejection of the idea that the study of film could be the object of a science or Wissenschaft, and that in fact the serious or theoretical study of film would always take place through a methodological pluralism that was open-ended and irresolvable. But there is something else. “This practical philosophy, which [Barthes] transmitted to me rather than taught me,” Metz offers, “is a sort of ethic — the will to furnish, in the very movement of research, an amiable and open space [un espace amical et respirable].” Call this, theory as generosity.
 Roland Barthes, “The Rhetoric of the Image” in Image/Music/Text (Glasgow: Fontana Press, 1993), 32. Also see, in the same volume, “The Photographic Message” and “The Third Sense.”
 Guido Aristarco, Storia delle teoriche del film (Milan: Guido Einaudi, 1951).
 Christian Metz, “Cinéma: langue ou langage?”, Communications 4, ed. Roland Barthes (Paris : Seuil, 1964), 52–90. The article was reprinted in Metz’s first collection of essays, Essais sur la signification du cinéma (Paris: Éditions Klincksieck, 1968), 39-93. It appeared in translation as “Cinema: Language or Language System” in Film Language: A Semiotics of the Cinema , trans. Michael Taylor (Chicago: University of Chicago Press, 1991), 31-91. As I will discuss further on, the English title of the essay is misleading and the translation itself marred by many errors and infelicities. For this reason, I will refer to the essay with its French title. All citations from this essay are my own translation, with French page numbers from the Essais given first, and corresponding pages from the English translation given after in italics. For a detailed overview of Metz publications and curriculum vita see Christian Metz et la théorie du cinéma, special issue of iris (Paris: Éditions Klincksieck, 1990), 299-318.
 The first Mitry review essay was published in Critique 214 (March 1965): 227-245, and the second in Revue d’esthétique 20 (April-September 1967): 180-221. Dudley Andrew has recently stated that he has “always dated the advent of academic film studies at the moment when Metz leapfrogged over Mitry as he reviewed the latter’s Esthétique et psychologie du cinéma.” See his “The Core and Flow of Film Studies,” Critical Inquiry 35 (Summer 2009): 896.
 Christian Metz, Essais sur la signification au cinema, II (Paris: Éditions Klincksieck, 1972).
 Christian Metz, Language and Cinema, trans. Donna Jean Umiker-Sebeok (The Hague: Mouton, 1974): 9, 5 (trans. modified). Originally published as Langage et cinéma (Paris: Larousse, 1971). Wherever I have modified the translation, page numbers in the French original will be given in italics.
 Ibid., 10, 5.
 Ibid., 10, 6.
 Ibid., 11, 6.
 Ibid., 11, 7.
 See, in this respect, Cassetti’s authoritative account in Theories of Cinema: 1945-1995, trans. Francesca Chiostri and Elizabeth Gard Bartolini-Salimbeni, with Thomas Kelso (Austin: University of Texas Press, 1999).
 Christian Metz, “Cinéma: langue ou langage?” in Essais sur la signification au cinéma (Paris, Klincksieck, 1968), 92, 90. Trans. Michael Taylor as Film Language. I remind the reader that all English translations are my own, and that corresponding page numbers from Film Language are given in italics.
 Ibid., 93, 91.
 “Sur la théorie classique du cinéma: a propos des travaux de Jean Mitry” in Essais sur la signification du cinéma II (Paris: Éditions Klincksieck, 1972), 11 (my trans.).
 The question of whether the analogical image could be subdivided into smaller distinctive units, and thus to what degree one could identify codes interior to the image, as it were, is one of the key points of contention between Metz, Umberto Eco, and Pier Paolo Pasolini. Referring to this as a debate on the relative value of graded and coded signs, Peter Wollen provides an astute commentary in his Signs and Meaning in the Cinema  (London: British Film Institute, 1998). See especially the chapter on “The Semiology of the Cinema,” 79-118.
 Metz, “Sur la théorie classique,” 12.
 “Une étape dans la réflexion sur le cinema,” in Essais sur la signification du cinema II (Paris: Éditions Klincksieck, 1972), 13 (my trans.). Metz also refers to the book as the “first general treatment of cinema available in the world” (13), strangely ignoring the 1960 publication of Kracauer’s Theory of Film.
 Ibid., 14.
 Ibid., 34 (my trans.).
 Compare, for example, Daniel Talbot, ed., Film: An Anthology (New York: Simon and Schuster, 1959); reprinted (Berkeley: University of California Press, 1966), and Richard Dyer MacCann, ed., Film: A Montage of Theories (New York: E. P. Dutton and Co., Inc., 1966).
 Metz, “Cinéma,” 41, 33. In later essays and retrospective footnotes, Metz significantly softens and complicates his original assessment of Eisenstein. Metz also ignores, of course, Eisenstein’s fascinating experiments with the sequence-shot as described in Vladimir Nilsen’s The Cinema as a Graphic Art (London: Newnes, 1959). This is a selective ethical reading of Eisenstein, incomplete though not unjust, to make him an exemplar of “structuralist man.”
 Metz, “Cinéma,” 42, 34.
 Ibid., 43, 35.
 Ibid., 44, 36.
 Ibid. The interior citations are from Barthes’s essay, “The Structuralist Activity,” originally published Lettres modernes (February 1963), 71-81, and reprinted in Barthes’s Essais critiques (Paris: Éditions du Seuil, 1964).
 Metz, “Cinéma,” 44, 36-37.
 Ibid., 45, 37.
 Ibid., 46, 38.
 Ibid., 47-48, 40.
 Ibid., 47-48, n. 5, 40.
 Ibid., 48, 40-41.
 Ibid., 56, 48-49.
 Ibid., 50, 43.
 Ibid. The interior citation is from Merleau-Ponty’s “The Film and the New Psychology,” a lecture originally given at IDHEC in 1945 and reprinted in his Sense and Non-sense, trans. Hubert L. Dreyfus and Patricia Allen Dreyfus (Evanston: Northwestern University Press, 1964), 57-58 (trans. modified).
 Ibid., 51, 44.
 This observation draws out an interesting contrast between the early work of Metz and Raymond Bellour, who otherwise were so closely allied. Bellour understood early on that the primary testing ground for the structural analysis of film should target a certain classicism; in short, Hollywood film, especially Hitchcock. In contrast, to put his “large syntagmatic categories” to the test, Metz turned to a minor though important New Wave film, Adieu Philippine (1962) and investigates narration in Fellini’s 8 ½ (1963), both of which have essays devoted to them in Film Language. His frequent references to Left Bank filmmakers are not simply a matter of taste, I think, but rather are more generally representative of a feeling for the conceptual power and inventiveness of the new cinemas, and one that would be echoed later in Gilles Deleuze’s work on the time-image. The tight link between “modern theory” and “modern cinema” is also in full view in the panels on semiology and film incorporated into the Mostri del Nuovo Cinema in Pesaro in 1966 and 1967, which generated important critical discussions in which Barthes, Metz, Eco, and Pasolini all had roles to play. See for example Casetti’s Theories of Cinema, 135.
 Metz, “Cinéma,” 51, 43.
 This contrast was implicitly understood in the classical period. In a 1931 review of Granowky’s Das Lied vom Leben (1931), Arhneim complained about “the strange way in which Russian film artists ruin the chance of visualizing things through their penchant for theoretical constructions. The Russians are real fanatics of film theory. They have thought up almost cabbalistic systems; yet the application to the actual work of art is for the most part not very satisfactory.” The implication is that there should be some conceptual alternative to the Russian approach, and perhaps one that is more aesthetic and less theoretical. See “Granowsky probiert,” in Rudolf Arnheim: Kritiken und Aufsätze zum Film, ed. Helmut H. Diederichs (Frankfurt am Main: Fischer, 1979), 233. I was led to this fascinating quote by Sabine Hake’s The Cinema’s Third Machine: Writing on Film in Germany, 1907-1933 (Lincoln: University of Nebraska Press, 1993), 278.
 Metz, “Cinéma,” 56, 49.
 Ibid., 57, 50-51.
 Ibid., 62, 56.
 Briefly, Metz’s main objective is to trace out all the ways signification in film is unlike a langue but like a language according to the following criteria: within the image discourse there is no double articulation; filmic syntax is forged at the level of sequence composition, making film more like “speech” than langue; narrative film is characterized by strong syntagmatic organization with weak paradigms, or rather, commutations are only possible at the level of large units of organization; and that film, like other art forms, is less communication than an open system of expression. Linguistics, in other words, points the way to showing what film is not (langue) and what it is, a language or discourse of art.
 Metz, “Cinéma,” 65, 59.
 See Michel Marie and Marc Vernet, “Entretien avec Christian Metz,” iris 10 (April 1990): 276.
 Metz, Language and Cinema, 20, 13.
 Ibid., 21, 14.
 David Bordwell and Nöel Carroll, eds., Post-Theory: Reconstructing Film Studies (Madison: University of Wisconsin Press, 1996).
 Marie and Vernet, “Entretien avec Christian Metz,” 295.