The Advent of Structuralism or, and was to äth the lividual ty> op-3s con-th their geois," ies and espised ind In-icism. usiasm led the i, they / asked Rocha ted on )posed s made .e "so-»rldists :stions could ofso-e pro-What sed or ndent within iisting dnter-inter-A^orld strat-most propriate? What was the relation between production methods d aesthetics? Should Third World cinema emulate the Hollywood ntinuity codes and production values to which Third World audi-ces had become accustomed? Or should it make a radical break th Hollywood aesthetics in favor of a radically discontinuous and ti-populist aesthetic such as the "aesthetic of hunger" or the "aes-;tic of garbage?" To what extent should cinema incorporate in-venous popular cultural forms? To what extent should films be nti-illusionistic, anti-narrative, anti-spectacular, and avant-garde? 1 his last question was also being asked by the First World avant-rde.) What was the relation between Third World filmmakers rgely middle-class intellectuals) and the "people" whom they pureed to represent? Should they be a cultural vanguard speaking ■ the people by proxy? Should they be the celebratory mouth-xes of popular culture, or the unrelenting critics of its alienations? Unfortunately, perhaps because of an assumption that Third World ellectuals could only express "local" concerns, or because their ays were so overtly political and programmatic, this body of work : s rarely seen as forming part of the history of "universal" - read .rocentric - film theory. The Advent of Structuralism The intellectual movement called structuralism was not without relation to these Third World stirrings. Both structuralism and third_ worldism had their long-term historical origins in a series of events that undermined the confidence of European modernity: the Holo-' caust (and in France the Vichy collaboration with the Nazis), and the postwar disintegration of the last European empires. Although the exalted term "theory" was rarely linked to Third World Cinema theorizing, third worldist thinking had an undeniable impact on First World theory. The structuralists codified, on some levels, what anti-colonial thinkers had been saying for some time. The subversive_work of "dgjiaiuriiliaatiQň^xxŕhrined by what one might call the left wing 102 The Advent of Structuralism ofsemiotics - for_example, Roland Barthes's famous analysis of the colonialist implications of the Paris Match cover showing a black soldier saluting the French flag - had everything to do with the external critique of European master-narratives performed by Third World Francophone decolonizers like Aimé Cesaire {Discourse on Colonialism^ 1955) and Frantz Fanon (The Wretched of the Earthy 1961). In the wake of the Holocaust, decolonization, and Third World revolution, Europe started to lose its privileged position as model for the world. Levi-Strauss's crucial turn from biological to linguistic models^for a new anthropology, for example, was motivated by his visceral'aversion to a biological anthropology deeply fp tainted by anti-semitic and colonialist racism. Indeed, it was in the context of decolonization that UNESCO asked Levi-Strauss to undertake the research which culminated in his "Race and History" (1952), where the French anthropologist rejected any esscntialist. hierarchy of civilizations. Both the structuralist and the poststructuralist movements, in this sense, coincide with the moment of self-criticism, a veritable legitimation crisis, within Europe itself. Derrida's decentering of Europe as "normative culture of reference," for example, was clearly indebted to Fanon's earlier decentering of Europe in The Wretched of the Earth. Many of the source thinkers of structuralism and poststructuralism,' furthermore, were biographically linked to what came to be called the Third World: Lévi-Strauss did anthropology in Brazil; Foucault taught in Tunisia; Althusser, Cixous, and Derrida were all born in Algeria, where Bourdieu also did his anthropological field work. In terms of film, the adoption of the methods of the human sciences constituted a challenge to what were seen as the irriDressionis-tic, subjective methods of earlier schools of film criticism. In this period, film semiotics and its prolongations, later ^lledj'screen ^theory" or .simply. "j^m..tlieor^,"^ame to the center of the analytic enterprise. In a first stage ,vSaussure an structuraninguisu^provided the dominant theoreticaLmodel. Understanding the causes of this paradigmatic shift requires a brief detour into the origins or :he structuralist movement. Although language had been an obje:: of philosophical reflexion for millennia, it was only in the twenct:.- century 103 The Advent of Structuralism that it came to constitute a fundamental paradigm, a virtual "key"' to the mind, to artistic and social praxis, and indeed to human existence generally. Central to the project of a wide spectrum of twentieth-century thinkers - Peirce, Wittgenstein, Sapir, Whorf, Cassirer Heidegger, Bakhtin, Merleau-Ponty, and Derrida - is a concern with the crucial importance of language in shaping human life and thought. ' As the methodological success story of the twentieth century, structural linguistics generated a rich proliferation of structuralisms prem-ís'ed on the principles of Saussurean linguistics. The overarching y- -'^- Tneta-discipline of .semiotics^ in this sense, can be seen as a local manifestation of a more widespread "linguistic turn," an attempt, in Fredric Jameson's words, to "rethink everything through again in terms of linguistics."1 Film semiotics must be seen as symptomatic not only of the general language-consciousness of contemporary thought but also of its penchant for methodological self-consciousness, its "metalinguistic" tendency to demand critical scrutiny of its own terms and procedures. The two source thinkers of contemporary semiotics were the American pragmatic philosopher Charles Sanders Peirce (1839-1914) and the Swiss linguist Ferdinand de Saussure (1857-1913). Roughly simultaneously, but without each other's knowledge, Saussure founded the science of "semiology" and Peirce the science of "semi-" otics." In A Course in General Linguistics (1916) Saussure called for a "science that studies the life of signs," a science that "would show" ' what constitutes signs, what laws govern them." Peirce's philosophical, investigations, meanwhile, led him in the direction of what he called "semiotics," specifically through a concern with symbols, which he regarded as the "woof and warp" of all thought and scientific research. (That there are two words for the semiotic enterprise, "semiotics" and "semiology," largely has to do with its dual origins in these two intellectual traditions). It is Saussure, however, who constitutes the founding figure for F. n en p ř i n <; tr 11 r f i t r n 1 i i nv^-iMuCthuiľJnFlrn "i rľT o F "fíľm "šť'm i ri Kčs T Saussure's Course in General Linguistics ushered in a kind of "Co-, pernican Revolution" injinguistic thought by seeing language not as a mere adjunctj:o our grasp of reality but rather äs formative of it. 104 The Advent of Structuralism Saussurean linguistics forms part of a general shift away from the nineteenth-century preoccupation with the temporal and the his-torical - as evidenced by Hegel's historical dialectic, Marx's dialect- ~^kl materialism, and Dzrwiďs_^evolution of the species" ^io~the~ contemporary concern with the spatial, the systematic, and the structural. Saušsure argued that linguistics must move away from the historical (diachronic) orientation of traditional linguisTjcT rowarrPa synchronic approach which studies language as a functional totality ^at a"giveň point in time. In fact, however, it is virtually impossible to separate out the synchronic from the diachronic. Indeed, many of the aporias of structuralism derive from its failure to recognize that history and languagejare mutually imbricated. For the structuralists themselves, however, the qualifiers "synchronic" and "diachronic," then, were seen as applying less to the phenomena themselves, therefore, than to the perspective adopted by the linguist. What matters is the shift: in emphasis from a historical approach preoccupied with the origins and evolution of language^to a structural emphasison language as a functional system. i More a method than a doctrine, structuralism was concerned with the immanent relations constituting language and all discursive systems. Common to most varieties of structuralism and semiotics was an emphasis on the underlying rules and conventions of language rather than on the surface configurations of speech exchange. In language, Saussure rámously argued,~ctthere arě^ólTl^čIiSerences." Rather than a static inventor}' of names designating things, persons, and events already given to human understanding, Saussure argued, language is nothingjrcojre than a ■series _of phonetic differences matched with a series of conceptual differences.. Concepts, therefore, are purely differential, defined not by their positive content, but rather by their diacritical relation with other terms of the system: "Their most precise characteristic is in being what the others are not." Within structuralism as a theoretical grid, then, behavior, institutions, and texts are seen as analyzable in terms.-of an under-Jyjngn£t\rorJi_ofj^^ elements which constitute the. network gain their meaning from the relations that hold between the elements. 105 The Advent of Structuralism Although structuralism developed out of Saussure's •oundbreaking work on language, it was not untiľthé 1960s that it »gmejyi.dcly disseminated^ The process" by"which structuralism me to form a dominant paradigm is retrospectively clear. The sci-ltific advance represented by Saussure's Course was transferred to erary study initially by the Russian Formalists and later by the Prague nguistic Circle, which formally instituted the movement in the Theses" presented in Prague in 1929. The Prague School pho-ologists, notably Troubetskoy and Jakobson, demonstrated the >ncrete fruitrulness of looking at language from a Saussurean per-lective and thus provided the paradigm for the rise of structuralism the social sciences and the humanities. Levi-Strauss then used the lussurean method with great intellectual audacity in anthropology id thereby founded structuralism as a movement. By seeing kin-ip relations as a "language" susceptible to the kinds of analysis rmerly applied to questions of phonology, Lévi-Strauss made it jssible to extend the same structural-linguistic logic to all social, ental, and artistic phenomena and structures. Lévi-Strauss extended e idea of binarism as the organizing principle of phonemic systems to human culture in general. The constituent elementsot myihJiiEfi. those of language, only acquire meaning in relation to other ele-meňtTšucn as myths, socialpractices, and cultural codes, compre-heňšiblě-only on the_basis__of structuring oppgsjtjgnj^^When Lévi-Strauss delivered his inaugural lecture in 1961 at the College de France, he situated his stmcjura[anjhirgcglCigy^1'fo1'n the broad tkld of semiology. Bytsearchmg foi constants within a multitude of variations, and by banishing all resort to a conscious speaking sub-lect, Lévi-Strauss laid the bases for structuralism. [n terms of film, the structural approach implied a move away_ from_any_evaluative criticism preoccupied with exalting the artistic status of the medium or of particular filmmakers or films. Auteur-structuralism in the late 1960s built on Levi-Strauss's concept of mvth to speak of genre and authorship. In terms of directors, semiology was less interested in the aesthetic ranking of directors than in how films in general are understood. Just as Lévi-Strauss was uninterested in the "authors" .of Amazonian .myths, so structuralism was 106 The (^ht-estion of Film Language not particularly interested in the artsmanshíp .qf individual auteurs. While auteurism valorized sp^^^ all filmmakers are artists and all films, are art, simply because film^s socially constructed status is that of art. _ The Question of Film Language The shift from the classical film theory of Kracauer and Bazin to film semiology mirrored larger changes in the history of thought in general. Film semiology also reflects changes in French cultural institutions: the expansion of higher education and the opening up of new departments and new forms of research; new publishing venues willing to publish trans-disciplinary books like Barthes's Mythologies^ new jnsjiuitioQS such as the École Pratique desTiäutes Etudes (where Barthes, Metz, Genette, and Greimas all taught); and ne\yJou^na]s_ such zs^Commumoa^ns. Indeed, issue 4 of Communications in 1964 presented the structural linguistic model as the program of the_ "Hiture, with Barthes's essay "Elements of Semiology" providing a blueprint for a broad research project. Issue 8, two years later, on "structural analysis of the ^^!L(stpry), framed a narratological project that would be carried out over decades. In the wake of the work of Lévi-Strauss a wide range of apparently non-linguistic domains came under the jurisdiction of structural Iin-guistics. Indeed, the 1960s and 1970s might be seen~aš the height of semiotic "impenajism," when the discipline annexed vast territories of cultural phenomena for exploration. Since the object of semi-otic research could be anything that could be construed as a system qfjigns organized according to cultural codes or signifying processes, semiotic analysis could easily be applied to areas previously considered either obviously non-unguis tic - fashion and cuisine, for example - or traditionally deemed beneath the dignity of literary or cultural studies, such as comic strips, photo-romans, James Bond novels, and the commercial entertainment film. The^re^f tjj.^ to.define, the status-o£.^ 107 'L í i The Qtiestion of Film Language film as a language. Fiimolinguistics, whose origins Metz attributed to the convergence of linguistics and cinéphiíia, explored such questions as: Is cinema a language system (langue) or merely an artístíc .| language''{langageyi (Métž's 1964 article "Cinema: langue or ^\ langageV was the founding essay within this.currcnt of inquiry.) Is it legitimate to use linguistics to study an "iconic" medium like film!' If it is, isthere any equivalent 'n rW. cinema to the linguistic sign? If therejsjj^cinematic sign, is the relation between signifier and signi-fied "motivated" or "arbitrary," like the linguistic sign? (For Saussure the relation between signifier and signified is "arbitrary," not only in the sense that individual signs exhibit no intrinsic link between signifier and signified, but also in the sense that each language, in order to make meaning, "arbitrarily" divides the continuum of both, sound and sense.) What is the cinema's "matter of expression?" Is the cinematic sign, to use Peircian terminology, iconic, symbolic, or, indexical, or some combination of the three? Does the cinema offer anyequivalent to langué's "double articulation" (i.e. that between phonemes as the minimal units of sound and morphgmes as the minimal units of sense)? What are the analogies to Saussurean oppositions such as paradigm and syntagm? Is there a normative grammar for the cinema? What are the equivalents of "shifters" and other marks of enunciation? What is the equivalent of punctuation in the cinema? How do films produce meaning? How are films understood? In the background lurked a methodological issue. Rather than an essentialist, ontological approach - what is the cinema? - attention shifted to questions of discipline and method. Quite apart from the question of whether film was a language (or like a language), there was the much broader question of whether filmic systems could fee illuminated through the methods of structural linguistics, or any other linguistics for that matter. ^Metz exemplified a new kind of film theorist, one who came to the field already "armed" with the analytic instruments of a specific discipline, who was unapologetically academic and unconnected to the world of film criticism. Eschewing the traditional evaluative language of film criticism, Metz favored a technical vocabulary drawn fromlinguisties and narratology (diegesis. paradigm, syntagma). 108 í i í i The Qtiestion of Film Language With Metz we move from what Casseti (1999) calls .the "onto- -[logical paradigm" ä la Bazin to the "methodological paradigmy Although Metz clearly built on the antecedent work of the RussiaiT Formalists, along with that of Marcel Martin (1955) and Francois" Chevassu (1963) and especially Jean Mitry (1963,19j65)_, he brought a new degree of disciplinary rigor to the field. ""* Within a few years a number of important studies were published on the language of film, notably Metz's Essais sur la signification au cinema (1968; translated as Film Language \w 1974); Metz's Langage et cinema (1971; translated as Language and Cinema in 1974)-Pasolini's Empirismo Eretico (translated into French as VExperience heretique: langue et cinema in 1971 and into English as Heretical Empiricism in 1988); Eco's La Struttura Assente (The Absent Structure); Emilio Garroni's Semiotica ed Estetica (Semiotics and Aesthetics, 1968); Gianfranco Bettetini's Cinema: Lingua e Scrittura (The Language and Technique of Film, 1968); and PeterWoIleüls Signs and Meaning in the Cinema (19 69), all of which addressed on some level the issues raised by Metz. (The Italian work, as Giuliana Müscio and Roberto Zemignan point out, has generally been filtered though French channels.)1 Of these, Mcte^&ljnLangjiage^^ Metz's chief purpose, as he himself defined it, was to "get to the bottom of the linguistic metaphor" by testing it against the most advanced concepts of contemporary linguistics. In the background of Metz's discussion was Saussure's founding methodological question regarding the "object" of linguistic study. Thus Metz looked for the counterpart, in film theory, to the conceptual role played by lanjfue in the Saussurean schema. And much as Saussurexoncluded that the wO£ purpose of linguistic investigation was to disengage from the cha-Jític plurality of parole (speech) the abstract .signifying system of a Janguage, i.e. its.key units and their rules of combination at a given point in time, so Metz concluded that the object of cině-semiology was to disengage from the heterogeneity of meanings of the cinema its basic signifying procedures, its combinatory rules, in order to see to_wjiax_ejeen fil- . Metz's >ttpm of dvanced f Metz's regard -Le coun-'■ngue in_ that the :he cha-em of a a given..„ mology_ cinema^ •r to see :ed dia- Thc Qtiestion of Film Language for Me tz, jfoo^i ne rngjis the cinematic institution taken in its broaH-,. it sense as a^ multidimensional soci o-cultural fact which includes pEe-filmicjEvents (the economic infrastructure, the studio system ■i-rhnology). jx>st-filmic events (distribution,exhibition, and the social orjpolitical impact of film), anda-fihmcjyeßts(the decor of clieTheater, the social ritual of moviegoing).("Film,^) meanwhile ufeřs to ä localizable discourse/a text; not the physical object con-__ X r»- r lined in a can, but rather the signifying text. At the same time > [etz points out, the cinematic institution also enters into the multi-dimensionality of films themselves as bounded discourses concen-. ating an intense charge of social, cultural, and psychological teaning. Metz thus rejntroducesj^^^^ and i ntrna within the category^filnT," now isolated as the specific and r-oper "object" of film semiology. In this sense, "the cjneman^^ u presents not the industry but rather the totality of films. As a novel i""to~literature, or as a statue is to šculptureTl^tžargulŠ^šo "išTum £_> cinema... Thgjformerjrej^ irjgr refers to an ideaj^n^ejnhk^the-totalitv of films and their traits V rítHín the filmic, then, one encounters the cinematic. Thus Metz closes in on"the object^f^ejm^ ^ mrses^ftejítasjather than of the cinema in the broad institutional "•-nse, an entity much too rnultifaceted to cor^tute.i^iepj;o£expb-N ctjDf filrr^oHnguis^ an ob- l ct too multiform to form the proper object of linguistic science. 1 he question which oriented Metz's early work was whether the ^ rien^jva^^^g^g.(language_syg^em) or langage (langu^geTTMetz' ;gins by discarding the imprecise notion of "film language" that had predominated up to that time.~ít~7š~in this context that' Metz explores the comparison, familiar from the earliest days of film theory between shot and word, and sequence and sentence. Vgr Metz, important differences render such an analogy problematic: 1 ^^^JILJ^njXCityil' r£!?£Iii1-n .1' kej^oxdsj^ce the lexico njs Jn._ (-, principle finite )J?jut_like statements, aninfinitvof which can be consjmctedxmthe basis, of alimitecT number of words. 2 £hnrg_arf»_tfre rrearions of fhp fílmrniVprJ -nnlilrŕ-; W0rd^Xwn^cn no 1 : ä -—-—-■s The Qttestion of Film Language preexist in lexicons) but again like statements. The shnrj^rnyjdesjanJnorrjjjTan'- anmíinr of i n form n Hon and semi-__ otic wealthy The shot is an_actualized unit, imlike the_wojTJ_which is a purely^ virtual lexical unit to be used as the speaker wishes. Theword^ "dog" can designate_anyjy]^ejDfdo^ , wítfTäny accent or intonation, whereas a filmic shot of a dog tells _ us, at the very minimum, that we are seeing a certain kind of dog of a certain size and appearance, shot from a specific angle with a specific kind of lens. While it is true that filmmakers might "virtualize" the image of a dog through backlighting, soft-focus, or decontextualization, Metz^sjnore general point is that the cinematic shot more closely resembles ah utterance or a state- * _ment ("here is the backlit silhouetted image of what appears to be a large dog") than a word. Shots, unlike words, do not gain meaning by paradigmatic con- ■> jrast vviBTotneFihots that might have occurred in the same place jDn the syntagmatic chain. In the cmejma^hc^tsJ^n^art.o£a^ Raiágjlgm so open as to be meaningless. (Signs, within the Saussurean schema, enter into two kinds of relationship: paradigmatic, having to do with choices from a virtual, "vertical" set of "comparable possibilities" - e.g. a set of pronouns in a sentence - and syntagmatic, having to do with horizontal, sequential arrangement into a signifying whole. Paradigmatic operations have to do with selecting, while syntagmatic operations have to do with combining in sequence.) To these disanalogies between shots and words, Metz adds a further díšäňSIógy concerning the medium in generál: the cinema does not" constitute a languageLY^deTy^ä^iiile _a^^jcode^ All speakers of English of a certain ageliave Iňästered the code of English - they are able to produce sentences - but the ability to produce filmic utterances depends on talent, training, and access. Tospeak a language^ in other words, is simply to use it, while to "speak" cinematic lan-.guage_is^!wayjLtQAcexta^ argue, of course, that this asymmetry is itself historically determined; one can 111 The Qtte'stion of Film Language .Jrypothcsize a future society where all citizens will have access to the -1 code of filmmaking.. But in society as we know it, Metz's point must stand. There is, furthermore, a fundamental difference in the diachrony of natural as opposed to cinematic language. Cinematic language can be suddenly prodded in a new direction by innovatory aesthetic procedures (those introduced by a film such as Citizen Kane^ for example) or those made possible by a new technology such as the zoom or the steadicam. Natural language, however, shows a more powerful inertia and is less open to individual initiative and creativity. The analogy is less between cinema and natural language than between cinema and other arts like painting or literature, which can also be suddenly inflected by the revolutionary aesthetic procedures of a Picasso or a Joyce. í Metz concluded tha_t_the cinema was not a language system but that it was a language. Although film texts cannot bexonceived-as-jj generatedbvan underlying language system - since the cinema, lacks— í t^^rbitrary_iign,jninima) units, andjdouble articulation "they do nevertheless manifest a language-like systematicity. Although film language^hag no<&. priori lexicon or syntax, it is nevertheless a language. One might call "language^Metz argues, any unity defined.. j in terms of its "matter of expression" - a Hjelmslevian term that designates the materiäTin which signification manifests itself- or in 'Těřmšofwhat Barthes in Elementsof'Semiology'calls its ^typical sign." Literary language, for .example, is the set of messages whose matter of expression is writing; cinematic language is-the set of messages i whose matter of cxpressionyonsists of five tracks or channels: moy-_ ying photographic' i.nža39^J^cordcd T^Q.netig^ourid, recorded noises, recorded musical sound., and writing (credits, intertitles. written I materials injhejhot)._Cinema is a language, in sum, not only in a broadly metaphorical sense but also as a set of messages gröühdedin a given matter ofexpression, and as an artistic language, a discourse "ól:"sighi"fyrhg7]practice charactenzedJ>y_^pexÄ ^denng^rč^cluřěs^ Much of the early debate centered around the question of minimal units and their articulation in the sense of André Martinet's notion of the "double articulation" of minimal units of sound (pho- _ 112 The Qimtion of Film Language nemes) and minimal units of sense (morphemes). Inresponse tn J^etz's ar^menj:^hatJUmJ[acked double articulatior^Pier Paolo ^Pasolini argued that cinema did form a^ngj^£jo£re^lity" with its si pwn double articulation of Jlgineme^j[by^naJ^^ < and "im-signs" (by analogy tomorphemes). The minimal unit of cineňiatic language, for Pasolini, is^fojp^dbythe divers? real-wr^pf signifying objects in th'e'sfibt. The language of im-signs, for Pasolini was extremely subjective and extremely objective at the same time. He postulated minimal units of film, i.e. cinemes, the objects de. picted in a filmic shot, but which unlikephonem.es were infinite in number. The cinema explores and reappropriates the signs of reality. Eco argued that objects cannot be elements of a second articulation since they already constitute meaningful elements. Both Eco and Emilio Garroni criticized Pasolini's "semiotic naivete" for confusing cultural artifact with natural reality. But a / number of recent analysts have argued that Pasolini was far from naive; in fact he was actually in advance of his contemporaries. For Teresa de Lauretis Pasolini was not naive but rather prophetic, anticipating the role of cinema in "the production of social reality'^ (ibid., pp. 48-9). As Patrick Rumble and Bart Testa point out, Pasolini saw structuralism as only one interlocutor, along with Bakhtin, Medvedev, and others. For Giuliana Bruno, Pasolini is not the naive reflectionist portrayed by Eco; rather, he sees both reality and its filmic representation as discursive, contradictory. The relation between film and the world is one of translation. Reality is ai "discourse of things" which film translates into a discourse of im- \ ages, what Pasolini called "the written language of reality." Like i Bakhtin and Voloshinov, Pasolini was more interested in parole than i in langue (see Bruno, in Rumble and Testa, 1994). Pasolini was also interested in the issue of the analogies and disanalogies between cinema and literature. Just as written reworked oral discourse, the cinema reworked the common patrimony of human gestures and actions. Pasolini favored a "cinema of poetry" over a "cinema of prose." The former evoked an imaginative, oneiric, subjective cinema of experimental form where author and character blend, while the latter evoked a cinema founded on classical con- 113 spouse^ ier Pac :" with Tonern^ al unit ofjl 1-world i ■Pasoli"-' ime tirr jjects.c Infinite, is of real-l 1 articula-j semioticf ty^But aj far from; aries. Forj letic, an-| 1_ realityji Dint out,! 5ng withl lini is ni'i th reali The reL -:ality is se of im-1 ty." Like tro/^tho" gies ar ■. *ework< mony ■ fpoetr :, oneiri charact • ical co] i 3 The Qtiestion of Film Language motions of spatiotemporal continuity. In Empirismo Eretico Pasolini Iso discussed his notions of "free, indirect discourse" in the cin-tna. In literature "le style indirect libre" referred to the managing f subjectivity in a writer like Flaubert, whereby mediated represen-ition conveyed through pronouns like "Emma thought" modu-ted into a direct presentation "How wonderful to be in Spain!" In ie cinema it referred to the stylistic contagion whereby authorial ersonality would blend ambiguously with that of the character 'here a character's subjectivity would become the trampoline for ylistic virtuosity and experiment. Umberto Eeo, whose work on the cinema was part of his work h languaged articulations in general, rejected a double articulation >r the cinema in favor of a triple articulation.:, first, iconic figures; 5cond, iconic figures combined into semes; and third, semes com- ined in "kinemorphes." Garroni, meanwhile, argued that Metz had >ked the wrong question; the right question concerned the consti-itive heterogeneity of the filmic/artistic message. Bettetini preferred double articulation based on the cinematic "sentence" on the one and, and technical units (the frame, the shot) on the other. He ?oke of the "iconeme" as the privileged unit of film language. In Tndice del Realismo (The Index of Realism) he applied Peirce's Í£kô£2!£y ^_^^inen3.a-.as_deploying all three dimensions of the. gn: the indexical, the iconic, and the. symbolic. Bettetini argued lat the minimal signifying unit of film, the "cineme" or "iconeme," the filmic image and this -"orrcsponds not to the word but to the .ntence. Peter Wollen too, in Signs and Meaning in the Cinema L 969) foun31>aiissureAn^ sign overly rigid for a me-_ turn whose "aesthetic niches" derived from a computer and unsta-lejfcpjpyrnent-of alUhese type--of-signs*—.. - Film became a discourse, Metz argued, by organizing itself as arrative and thus producing a body of signifying procedures. As barren Buckland points out, it is as if the "arbitrary" relation of íussure's signifier/signified was transferred to another register, i.e. ot the arbitrariness of the single image but rather the arbitrariness fa plot, the sequential pattern imposed on raw events. Here we rid an echo of the Sartrean idea that life does not tell stories. The 114 ■\ i The Qtiestion of Film Language _truejmalogv between film and hingugge, for Metz, consisted in their com m on syntagma ti c nature- By irinving^rnnw^ i™*;»» \r, n*^, nTm becomes language. Both language and film produce_djsc^urse.__. through paradigmatic and syntagmatic operarjpjis^La^f^jage^eJejcts. — and combines phonemes and morphemesto form sentencesj_film__ sefects and combines images and sounds to form "syntagmas," i.e. ° ■* L units of narrative autonomy in which elements interact semantically^., While nómíagě~erKtrelv• resembles another image, most narrative films resemble one another in their prindpaLsyntagnxajicJIgu^i,"!^ their orderings of spatial and temporal relations. ' ~yTj^~Gráň^'"SjÍBJ^ the prin-_ cipal syntagmatic figures "or the spa tio temp oral orderings of ríär" rative cinema. It was proposed as a response to the question "How does film constitute itself as narrrative discourse?" against the backdrop of the notorious imprecision ofllm terminology^much ofwhicH . had been based on theater rather than on the specifically cinematic signifiers of image and sound, shots and montage. Terms like "scene" and "sequence" had been used more-or-less interchangeably, and were based on the most heterogenous criteria. The classification was at times based on a posited unity of depicted action ("the farewell scene") or of place ("the courtroom sequence") with litde attention to the precise articulations of the filmic discourse, and ignoring the fact that the same action (e.g. a wedding scene) might be rendered by a diversity of syntagmatic approaches. Ajctz used the paradigm/syntagma distinction, along with the largčrKíi5ryeirhsr or yfŕeth'od - "a shot is continuous or it is not" - to construct his Grande Syntagmatique. The Grande Syntagmatique constitutes a typology of the diverse ways that time and space can be jprdered through editing within the segments rája jia^ratjve^lrn^ Using a binary method ofrnmmiit^ti^(rorn^]jatJ^njgjtjJi^ye to do with discovering^w|^her^Aange.on,the_Ieyel of the signifier entails a change on the level of the signified), Metz generated a total of six types of syntagma (in the version published in Communications in 1966), subsequently increased to eighr (in the version in-^ eluded in Essais sur la signification au cinema in 1968 and also in j Film Language). The eight syntagmasji£_as-followsr----- 115 í r* The Question of Film Language 1 Th^j£^w/,w/wťj^^^ turn subdivided into (a) the single-shot sequence^ and (b) four kinds of y - inserts: the non-Megeticinsert f3l single shot which presents ob-^ 7ects exterior to the fictional world of the action); the displaced dietetic insert {"real" diegetíc images but temporally or spatially out of context); the subjective insert (memories, fears); and the explanatory insert (single shots which clarify events for the spectator). 2 The parallel syntagma: rwa-akernaring motifs without clear spatial or temporal relationship, such as rich and poor, town and country. 3 The bracket syntagma; brief'scenes given as typical examples of a certain order of reality but without temporal sequence, often organized around a "concept." 4 The descri^tivejynPagma: objects shown successively suggesting spatialjcoexistence; used, for example, to situate the action. 5 ' The alternating syntagma: narrative cross-cutting implying tem- poral simultaneity, such. asa chase alternating pursuer and pur-, sued. 6 The scene: spat|otemporal continuity perceived as being without flaws or breaks, in which the signified (the implied diegesis) is continuous as in the theatrical scene, but where the signifier is fragmented into diverse shots. 7 The episodic sequence: a symbolic summary of stages in an implied chronological development, usually entailing a compression of time. 8 The ordinary sequence: action treated elliptically so as to eliminate unimportant detail, with jumps in time and space masked by continuity editing. This is not the place to inventor}' the innumerable theoretical problems with the Grande Syntagmatique (for a sustained critique see Stam et al., 1992). Suffice it to say that while some of Metz's syntagmas are conventional and well established - the altexnating 4ynjagma, for .example, refers}c^\^^^v^^ox^iS^As^^'.. ÜY£_cross-cutting^others are more innovative. The bracketsyntagma, 116 The Qiiestion of Film Language for example, provides typical samples^o£a.giyxrLojder_ofrejility with-oliViinkiňg tnem chronologically .-The audiovisual logoTwElcrTöpen ^řlěvisíon sitcoms (tor example, the initial montage-segmgntshpw-Ing the typical aenvmes ot„^ '"pärTTyler Moore Show) mightbe seen as bracket syntagmas. Simi-Iflt-yjthe fragmented shots of two lovers in bed that open Godard's JL l/larried Woman provide a typical sample of "contemporary adul-tgry;" indeed, the sequence's lack of teleology and climax form part 0f a Brechtian strategy of de-eroticization, a "bracketing" of eroticism- Many of the films featuring significant numbers of bracket syntagmas can be characterized, not coincidentally, as Brechtian, precisely because the bracket syntagma is especially ^well-equipped for representing the socially "typical," GodarďsBrechtian fable about war, Les Carabiniers, mobilizes bracket syntagmas as part of the film's systematic decöhstructioh from within of the dominant cinema's traditional approach to dramatic conflict. TheJ>racket svfitagma^ffn.-phasison the typical - hen^tJie_behaviora^^ - is -v/ "■r eminently suited to the social and generalizing_intentions ofpolitK. qzed directors■====—. Äs a kind of illustration of his method Metz performed a syntagmatic breakdown of the film Adieu Phillipine into 83 autonomous segments. But given Metz's methodological restrictions, his syntagmatic analysis did not address many of the most interesting features of the film: its portrayal of the TV milieu; the chronotropic implications of the frequent TV monitors in the shot; the working-class attitudes and accents of the characters; the war in Algeria (in which the protagonist enlists); gender roles and flirtation in 1960s France. Once the linguistic analysis is finished, almost everything else remains to be said, whence the need for a Bakhtinían translinguistic analysis of the film as historically situated utterance. But Metz offered the Grande Syntagmatique in a more modest spirit than was often granted by his detractors, as a first step toward establishing the main types of image orderings. To the objection that "everything remains to be said" it might first be answered that it is in the nature of science to choose a principle of pertinence. To speak of the Grand Canyon in terms of geological strata, or of Hamlet in 117 :y with-^Tôpen Lshow-~Q£Lthe L.Simi-odard's yadul- rnvpart f eroti- bracket chtian, uipped e about e film's' La's tra^ I'seny ^ar - is ^poIitK med a utono->ns, his resting [Otopic )rkir.g-M'ia ^in 1960s ything tinian :rance. ;t spirit estab-■n that Lat it is ) speak mlet in The Question of Film Language is of syntactic functions, hardly exhausts the interest or significa- of experiencing the Grand Canyon or reading Hamlety yet that , not mean that geology and linguistics are useless. Second, the «11 k of addressing all levels of signification in a film is the task of ^ Lial analysis, not film theory. ] i Lanjfuajje and Cinema Metz redefined the Grande Syntag-n i qug_^s rarely a subcode of editing within a historically delim-^7j body of films, i.e. the mainstream narrative tradition from thp CTm .ohdation of the sound film in the 1930s through the crisis of The'studio aesthetic and the emergence of the diverse New Waves^ inthe_l£ť>Qs_i_^Ietz,s schema, clearly the most sophisticated developed up to that point, was subsequently applied (in myriad textual analyses) and was-later reconfigured by Michel Colin from the Chomskian perspective of transformational grammar (see Colin, in, "ßuckland, 1995). Film theory could still use a more sophisticated approach to the questions raised by the Grande Syntagmatique, one that would synthesize Metz's work with other currents: Bakhtin's suggestive notion of the chronotope as "the intrinsic connectedness of temporal and spatial relationships" in artistic texts; Noel Burch's work on spatial and temporal articulations between shots; Bordwell's work on classical cinema; and Genette's narratology insofar as it is transposable to film. Metz was subsequently criticized for surreptitiously privileging the j rřiáíhstřeam narrative film and marginalizing such forms as documen-[ tary arid the' avant-ga'rde. Ä Palditiniahtfänslingüistic formulation might ha/e-saved" cihě:semiologists in the Saussurean tradition a good deal of trouble by rejecting from the outset the very notion of i unitary (cinematic) language. Anticipating contemporary socio-linguistics Bakhtin argued that all languages are characterized by the dialecUcaLinterpjay_bet\ve^n_c^tri^^^P££ÍÍ^^JQwar^ normativ-. izationjmonogiossja) jjid^centrifugal energies favoring dialectal dryersifca^iyjiejejfog^ approach provides a valuable frame- work for seeing the classical dominant cinema as a kind of standard language backed and "underwritten" by institutional power, and.thus exercising hegemony over a number of divergent "dialects" such as che documentary, the militant film, and the.avant-garde cinema. A 118 -í i ■í ■I ''é í Cinematic Specificity Revisited translinguistic approach would be more relativistic and pluralistic about these diverse filmic languages, privileging the peripheral and the marginal as opposed to the central and the dominant. Cinematic Specificity Revisited In their attempts to legitimate film as art, as we have seen, theorists made conflicting claims about the "essence" of film. The 1920s Impressionists like Epstein and Delluc had earlier embarked on a quasi-mystical search for the photogenic quintessence of film. For theorists such as Arnheim, meanwhile, the artistic essence of cinema was linked to its strictly visual nature', and thus to its "lacks" (the limiting frame, the lack of a third dimension, etc.) that marked it as art. Others, such as Kracauer and Bazin, rooted film's "vocation for realism" in its origins in photography. -Film semiology, too, was concerned with this perennial issue. For Metz, the question "Is film a language?" was inseparable from the question "What is specific to the cinema?" The pertinent sensorial traits of film language help us distinguish the cinema from other artistic languages; in changing one of the traits, one changes the language. For example, film has a higher coefficient of konicity than does a natural language like French or English (although one could argue that ideographic or hieroglyphic languages are highly iconic). Films are composed of multiple images, unlike photography and painting which (usually) produce single images. Films are kinetic, unlike newspaper cartoons which are static. Metz's approach, then, involved teasing out the specific signifying procedures of film language. Some of the specific materials of expression of the cinema are shared with other arts (but always in new configurations) and some are unique to itself. The cinema has its own material means of cinematic expression (camera, film, lights, tracks, sound studios), its own audiovisual procedures. This question of "materials of expression" also brings up the issue of evolving technologies. Is an IMAX spectacle, or a CD-ROM narrative, or video art still a film? 119 Cinematic Specificity Revisited Metz's most thoroughgoing exercise in filmolinguistics was Langage et cinema, first published in French in 1971 and translated (disastrously) into English in 1974.1 Here Metz substituted the broad concept of "code," a concept thankfully free of specifically linguistic baggage, for both langueznd langage. For Metz, the cinema is necessarily a "pluri-codic" medium, one which interweaves (1) "specifically cinematic codes," i.e. codes that appear only in the cinema, and (2) "non-specific codes," i.e. codes that are shared with languages other than the cinema. Cinematic language is the totality of cinematic codes and subcodes insofar as the differences separating these various codes are provisionally set aside in order to treat the whole as a unitary system. Metz describes the configuration of specific and non-specific codes as a set of concentric circles, with a differential approach to cinematic specificity. The codes range from the very specific (the inner circle; for example, those linked to film's definition as deploying moving, multiple images - codes of camera movement, continuity editing, etc.), through codes which are shared with other arts (e.g. generally shared narrative codes), to codes which are widely disseminated in the culture and in no way dependent on the specific modalities of the medium or even on the arts in general (for example, the codes of gender roles). Rather than an absolute specificity or non-specificity, then, it is more accurate to speak of degrees of specificity. Examples of specifically cinematic codes are camera movement (or lack of it), lighting, and montage; they are attributes of all films in the sense that all films involve cameras, all films must be fit, and all films must be edited, even if the editing is minimal. The distinction between specifically cinematic and non-cinematic codes, obviously, is often a tenuous and shirting one While the phenomenon of color belongs to all the arts, the particu laríties of 1950s technicolor belong specifically to film. Even nonspecific elements, moreover, can be "cinematized" via filmic simultaneity, by their neighboring and coexisting with the other elements featured on other "tracks" at the same moment in the filmic-discursive chain. Within each particular cinematic code, cinematic subcodes repre- 120 Cinematic Specificity Revisited seht specific usages of the general code. Expressionist lighting, for example is a subcode of lighting, as is naturalistic lighting, eisensteinian montage is a subcode of editing, which can be contrasted in its typical usage with a Bazinian mise-en-scěne that would jtiinimize spatial and temporal fragmentation. According to Metz codes do not compete, but subcodes do. While all films must be lit and edited, not all films need to deploy Eisensteinian montage. Metz notes, however, that certain filmmakers such as Glauber Rocha at times mingle contradictory subcodes in a "feverish anthological procedure" by which Eisensteinian montage, Bazinian mise-en-scene, and cinema verité coexist in tension within the same sequence. The diverse subcodes can also be made to play against one another, for example by using Expressionist lighting in a musical, or a jazz score in a western. For Metz, the code is a logical calculus of possible permutations; the subcode is a specific and concrete use of these possibilities, which yet remains within a conventionalized system. There is a tension in Language and Cinema between an additive, taxonomie approach to codes, developed in the first half of the book, and a more activist "writerly" deployment of the codes, developed at the end of the book. A history of the cinema, for Metz, would trace the play of competition, incorporations, and exclusions of the various subcodes. In his essay "Textual Analysis etc.," David Bordwell points out some of the problems with Metz's analysis, arguing that Metz's characterization of subcodes shows covert dependency on received ideas about film history and the "evolution of film language," ideas which provide the unstated grounding for the recognition of subcodes. Bordwell therefore calls for the historicization of the study of cinematic subcodes.2 The invaluable historicization suggested by Bordwell is limited to the institutional and the art-historical; it does not include what Bakhtin would call the "deep-generating series" of both life and art, i.e. history in a larger sense as it impacts on film. Metz inherited the question of langue/langage from Saussure and the question of cinematic specificity from the Russian Formalists, with their emphasis on literary specificity or literatumost. Metz, in this sense, inherits the combined blindspots of Saussurean lin- 121 ing, for jhting. :>e con-: would o Metz st be lit s. Metz Dcha at :al pro--scene, :e. The íer, for z score ■ossible f these ;ystem. Iditive, : book, eloped ompe-. In his >me of acteri-about h pro-codes. :>f cin-:ed by t does ies" of film. assure >rmal-Metz, n lin- Cinematic Specificity Revisited ntistics (which "brackets the referent" and thus severs text from listory) and of aesthetic formalism (which sees only the autotelic lutonomous object of art). If Metz, like the Formalists, could be ,aid to have brought great "sharpness and principle to the problem )f specification," he was somewhat less adept, given these inherited ilindspots, at linking the specific and the non-specific, the social ind the cinematic, the textual and the contextual. In this sense the 3akhtin School critique of Formalism is pertinent to Metz's notions >f the "specifically cinematic," and, as I suggest later (p. 188), to he "neo-Formalism" of Kristin Thompson and David Bordwell. What is perhaps more promising in Metz's work is his attempt to listinguish film from other media in terms of its means of expres-ion. Metz distinguishes between film and theater, for example, by he physical presence of the actor in the theater versus, the deferred bsence of the performer in the cinema, a "missed rendezvous" that »aradoxically makes film spectators more likely to "believe" in the mage. In subsequent work Metz stressed that it is precisely the "im-ginary" nature of the filmic signifier that makes it so powerful a atalyst of projections and emotions (Marshall McLuhan implied omething similar in his contrast between "hot" and "cool" media). ŕíetz also compares film to television, concluding that despite tech-lological differences (photographic versus electronic), differences i social status (cinema by now a consecrated medium, television till deplored as a wasteland), differences in reception (domestic small creen versus theatrical large screen, distracted versus concentrated ttention), the two media, constitute virtually the same language. liey share important linguistic procedures (scale, sound off and on, redits, sound effects, camera movement, etc.). Thus they are two losely neighboring systems; the specific codes which also belong to be other are much more numerous and important than those which o not belong to it; and, inversely, those which separate them are íuch less numerous and important than those which separate them, i common, from other languages (Metz, 1974). Although one might rgue with Metz's conclusions here (for example, one might say lat technologies and reception conditions have evolved since the 970s), what is important is the differential, diacritical method: con- -1 ■í í L 122 Interrogating Authorship and Genre structing or discerning film's specificity by exploring the analogies and disanalogies between it and other media. Interrogating Authorship and Genre Linguistically oriented semiotics had the effect of displacing auteurism, since filmolinguistics had little interest in film as the expression of the creative will of individual auteurs. At the same time, auteurism had introduced a kind of system - one based on the constructing of an authorial personality out of surface clues and symptoms - which made it reconcilable with a certain kind of structuralism, resulting in a marriage of convenience called auteur structuralism. Undermining the cult of personality endemic to both the Cahiers and the Sarris models, auteur-structuralism saw the individual author as the orchestrator of trans-individual codes (myth, iconography, locales). As Stephen Crofts points out, auteur-structuralism emerged out of a precise cultural formation in the late 1960s, that of the structuralist-influenced left in London, and specifically of the film-cultural work of the British Film Institute's Education Department. Auteur-structuralism was exemplified by Geoffrey Nowell-Smith's study Visconti (1967), Peter Wollen's Signs and Meaning in. ■the Cinema (1969), and Jim Kitses' Horizons West (1969). The auteur-struetúralists highlighted the idea of an auteur as a critical construct rather than a flesh-and-blood person. They looked for hidden structuring oppositions which subtended the thematic leitmotifs and recurrent stylistic figures typical of certain directors as the key to their deeper meaning. For Peter Wollen, the apparent diversity of John Ford's oeuvre, for example, hid fundamental structural patterns and contrasts based on culture/nature binaries: garden/ wilderness; settler/nomad; civilized/savage; married/single. Auteur-structuralism had little to say on the issue of cinematic specificity, since many of these motifs and binary structures were not specific to the cinema but were, rather, broadly disseminated in culture and the arts. 123