Institutional Mode Of Representation 68106z

Film's Institutional Mode of Representation and the Soviet Response Author(s): Noël Burch Source: October, Vol. 11, Essays in Honor of Jay Leyda (Winter, 1979), pp. 77-96 Published by: The MIT Press Stable URL: http://www.jstor.org/stable/778236 Accessed: 28-04-2015 08:51 UTC

Your use of the JSTOR archive indicates your acceptance of the & Conditions of Use, available at http://www.jstor.org/page/info/about/policies/.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please [email protected].

The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to October.

http://www.jstor.org

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representationand the Soviet Response

NOEL BURCH

It is temptingto regard the systemof representationat work in the vast majorityof filmsproducedduringcinema's earliestperiod (which we may situate between1892and 1906) as an authenticallyworking-classsystem,in opposition to not only thebourgeoisnovel, theater,and paintingof thenineteenthcentury,but also an institutionalmode of representation as it was to develop after1906. In the countrieswherethefilmindustryfirstdeveloped,notonly was theaudience of this cinema largely proletarian,but in many respectsthe systemof representation which we may identifyas specificallyof this period derives little'from the characteristically bourgeois art formsof the eighteenthand nineteenthcenturies and almost everythingfrompopular art formsdescendentfromthe Middle Ages and before. However,much of the othernessof thefilmsof thisera is patentlyoverdetermined, oftendue to the contradictionbetween the aspirations-conscious and unconscious-of middle-classinventorsand entrepreneurs on the one hand and the influenceof such plebeian or otherwise"alien" art formsas the circus, the carnival sideshow, thepicturepostcard,or thelanternshow on theother.1In any of theworking case, one mustregardas highlyproblematicanydirectintervention classes,whose tastecould have directlyaffected only thesubstanceof thefilmsthey saw (in and England, especially); while the deepest aspirations of the workingclass were sometimescateredto symbolically,thesefilmscertainlynever reflected revolutionaryideology.In thisprivilegedrelationshipbetweenan essentiallypopulist cinema and the workingclasses lasted practicallyuntil the introductionof sound. In the United States,however,whereeven in the era of a wholly proletarianaudience the substanceof the filmsmostlyreflectedthe lives and ideals of theirpetit-bourgeoismakers,the industryquickly came to see that theconditionforitscommercialdevelopmentwas thecreationofa mass audience, that is, one which also included the various strataof thebourgeoisie,less fragile 1. Morerecentresearchhas shown thatothercontradictions, economicand psychological,playeda major role in this processof overdetermination.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

78

OCTOBER

economically and possessing more leisure time than the immigrantworking classes. It is importantto realize that the extraordinary expansion of the American cinema and its rise to world dominance afterWorld War I was a directconsequence of the creationof thataudience during theperiod 1905-15.In ,on the other hand, the industryremained content to exploit the early mode of representationfor nearly three decades, catering to a small domestic audience which was almost exclusivelyworking class, and counting on the skills of its thehuge international cameramenand actorsto continueto captivateindefinitely marketwhich it had conqueredearlyin thecentury.The corollaryof thissituation was thatthe Frenchbourgeoisiewas not to come to thecinema in any appreciable numbersuntil thescreenfinallyacquired a voice, thatcrucial elementof presence which would at last place it on a par with the legitimate--whichis to say, bourgeois--stage. Earlycinema was marked,in theeyesof theinternationalbourgeoisie,by the absence of the persona, of nearly all the signs of characterindividualization capable of satisfyingexpectationscreatedby naturalistictheaterand novels,and of theSubject.The voice steepedin theprimacyof theindividual,in thecentrality was indeed thebiggestlack, hence theconstantbut only veryrelativelysuccessful effortsto invent a sync-soundsystem,from Edison's Kinetophonographto the Gaumont Chronophone. However,thepersona was lackingon thevisual plane as well. One of the foundingvisual models forthe earlyperiod as a whole was the shot as exemplifiedin La Sortiedes Usines Lumiere and also in filmssuch as long L'Assassinat de Marat,which Hatot directedfortheLumiere Companyduringthe early months of its production. Films like the latter-and therewere many of them-illustrate in spectacular fashion the gap between early cinema and the bourgeoistheater:thecoextensionof prosceniumarchand filmframeproducesan effect ofdistancethroughsmallnessand low definitionwhichis verydifferent from the effectof presenceindissociable fromthe bourgeoisstageand producedby the and the eye's facultyof "sync" voice, by "natural" color, three-dimensionality, focusing in space, often with the aid of opera glasses. As the various sociothis ideological pressuresto make thecinema "morerespectable"became stronger, long shot came to be perceivedas an obstacle. We may take as both metaphorand illustrationof this a filmproduced in 1903,Edwin S. Porter'sGreat Train Robbery.As in so manyfilmsof theperiod,it would be impossible even to distinguishbetween,forexample, the outlaws and the posse were it not forthe bandannas worn by the former,so wide are all the shots of the action proper. For this film,however,Porter-who seems to have experienced with unusual acuity the contradictionsof that transitionalerasupplied exhibitorswith a small, separateroll of filmconsistingof a single shot, which theywere freeto splice onto the head or tail of the film,whicheverthey chose. This mobilityitselfstressestheothernessof an era in which filmswerenot

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

79

1903. EdwinS. Porter.The GreatTrainRobbery.

yetclosed objects.As is well known,thisshot showeda close,head-onview ofone of theoutlaws shootinginto thecamera,an image whichgivesclear,almostbrutal expression to the need then being felt to reduce distance. . at all costs. This "close-up," which seems to hover on the fringeof a diegesis which cannot assimilateit,is indeedthesign ofsomethingalreadysensedas lackingat thattime. is thatthelack was felt But sensedbywhom?All thatwe can saywithanycertainty by commentators, producers,exhibitors,directors,cameramen;what the mass of filmviewersmay have feltis an altogetherdifferent matter,which forthemoment can only be leftto conjecture. who singlehandedlysolved the Despite legend, it certainlywas not Griffith of facial of the legibility,thatsine qua non forthe interpolatedclose-up, problem institutionof the persona. In fact,while Griffith, during his richlyinnovative careerat Biograph,graduallymoved thecameracloserto all of his tableaux,true close-ups pictureobjects farmore oftenthan theydo people. Moreover,Griffith whichfor was one of thelast directorsto relenton thematterof actors'anonymity, a varietyof reasons had been a universallyrespectedrule duringtheearlyperiod. This belated adhesion to the starsystem-thecounterpartof the close-up in the constitutionof the filmiersona-is undoubtedlyboth cause and effectof the paucityof truefacialclose-upsin theBiographfilms,whichwerein otherrespects so forward-looking. Anotheraspect of earlycinema which did not fulfillexpectationscreatedby modes of representationdominant at the turnof the centuryresultedfromthe experiencedby early filmmakersin reproducing,under certain great difficulty circumstances(especially for indoor scenes), the depth cues long essential in

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

GeorgesMblies.L'Hommeaila tatede caoutchouc. 1901. Westernimagery,whetherin easel painting or on the prosceniumstage. A great many filmsmade during this era are characterizedby the relativeperceptual flatnessof their(interior)imagery.L'Assassinat de Marat and the other"theatrical" filmsproduced by the Lumiere brothersat the startof theirundertakingare examples of this. More significantstill, perhaps, are the many remarkable instancesfound in what I call the matureearlyera. Consider The Life of Charles Peace, the narrativeof a celebratedVictorian murdererfilmed by a remarkableartisan of working-classorigins, William Haggar.2 Most of the filmconsistsof a seriesof single-shot,richlyorchestrated tableaux,but it culminatesin a multishotchase sequence filmedon location. All of the stylizedinteriorscenes are shot against two-dimensionalbackdropsfrom which all illusion of haptic space seemsto have been cunninglyexcluded,and in frontof which actorsplay according to a strictlylateral blocking scheme.This traitis common to nearlyall scenesshotin thestudio until at least 1910 and was brilliantlyillustratedby the greatMelies, forwhom the "essence" of cinema was preciselyits capacityforrenderingthree-dimensional space and movementin two

This filmis a fineexample of the populist traditionin the earlyBritishcinema,as referred 2. to above. Peace is treatedas a kind of folkhero. 3. I should exclude the veryprecocious Danish cinema fromthis statement,however.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

81

dimensions (see in particular the trickeffectin L'Homme a la t2tede caoutchouc).4 and of most This tendencycontinuedto make itselffeltin thefilmsof Griffith of his contemporaries.This was due to the persistenceof two factorsthat had determinedits presencefromthe start.One was filmingin daylight,in studios withglass roofsor in theopen air, which gave an even,"flat"lightingthattended on thesame plane. The otherwas thestationingof thecamera, to place everything stillresolutelyfrontal,withthelens axis rigorouslyparallel to thefloorand always at the height of a standingman. Consequently,until about 1915 or even latera characterwould occupy full screenheightonly if he or she were standingin the foreground.When the actor was seated in a chair,crouchingon the ground,or standingin thebackground,his head onlyreachedthemiddleof thescreen,which familiarto graphicartists,in which thebackgroundeffect, produceda flattening set or landscape-seems to be looming overhead, ready to topple into the foreground,as it were. At the same time,however,otherfactorshad already been workingin the opposite direction. The generalizationof electriclighting made it possible to obtain more subtlemodelingand chiaroscuroeffects. Color had long been used by the French,including Mdlies himself,to counteracttheflatnessof certainimages (with the introduction,in particular,of artificialeffectsof aerial perspective). Around 1914 several directorsand technicians began to avoid placing their cameras at a ninety-degree angle to therearwall, as had been customary.Finally, there was the introduction-possibly by DeMille in The Cheat (1915)-of a systematic,slightlydownward tilt of the camera, which meant that characters would occupy thewhole heightof thescreeneven when theywereat theback ofa moderatelydeep set, and which furthermoreaccentuated the obliqueness of horizontallines. Together,all of theseprocedureswere graduallyto bringabout the creationof a full-blownhaptic pictorial space "in" which the diegeticeffect would be able to reach full development. However,thechiefproblemforthemajor pioneers,fromPorterand theearly Britishfilmmakers(Smith,Williamson, Hepworth.. . ) to Barkerand Feuillade, was, on the one hand, what I call the linearizationof the iconographic signifier and, on theother,theconstructionof a linearizeddiegeticcontinuum.Let us now brieflyexamine thesetwo closely linked issues. The panoramic tableau of the mostcharacteristic earlyfilmsofferstwo basic traitswhich may also be seenas complementary-forwe mustnot lose sightof the factthatall of these"inadequacies," as well as thestrategieswhich ultimatelyled to theirreduction,interpenetrate in complex fashion. First thereis the relative 4. In thisfilma magician-scientist pumps his head up to huge proportionswitha bellows. As is shown in Franju's filmLe Grand Mlies, the effectwas obtained by pulling Melies up an inclined plane on an invisible trolleytowardsthecamera. For Mlies, close-ups werealways "giant faces": the screen,he felt,was the only plane a filmcould contain.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

82

OCTOBER

rarenessofany of theindexesof individualization-differentiation alluded to above; then thereis a tendencyto confrontthe spectator'sgaze with an entiresurfaceto scan, at timesalong a relativelycontrolledtrajectory (but which generallytookin mostof thescreen'ssurface).At othertimesthegaze is undirected,consideringthe absence of most of the ordering procedures-strategies of isolation or signalization-which would graduallymake it possible to normalizethebehavior of the spectator'seye. One verystrikingexample of thetypically"chaotic" tableau is theopening shotof a Biograph filmof 1905, Tom Tom thePiper's Son, known to us todaythroughKen Jacobs'senlighteningrehandlingof it. The shotshowsa crowded marketplacedistractedlydominated by a woman tightropewalker in white. But she has no role in the narrative(in fact,she is theonly characternever to be seen again). On theotherhand, what is meant to be thecentralaction-the preliminariesleading up to the theftof thepig, the theftitself,and thestartof the chase as the thiefescapes-is nearly invisible for the modern spectatorat first viewing. For he is accustomed to having each shot in a filmcarefullyorganized around a single signifyingcenterand to the linearizationof all the iconographic signifiersthroughcomposition,lighting,and/or editing.5And as we know, the firststep in overcomingthis "handicap" was the dissectionof the tableau into successivefragments(closer shots),each governedbya single signifier, so thateach frame would be immediatelydecipherable (at least in accordance with certain normsof legibility)at firstviewing. However, in orderthat thesesuccessiveimages not bringabout thedislocation of the"original" profilmicspace-the space of thesingletableau,thespace,if one prefers,of theproscenium-a long evolutionwas necessary.Startingfromthe firstpremisesof thealternatingshotsin theworkof Porterand theBritish,and the earliest contiguitymatches (matches of direction and eyeline), this evolution, through the increasing ubiquity of the camera, was ultimatelyto succeed in establishing the conviction that all the successive separate shots on the screen referredto the same diegetic continuum. In otherwords, the time spans representedwerelinked togetherby relationsof immediatesuccession,simultaneity, or a more distant anteriorityor posteriority;the spaces pictured communicated directlyor at one or more removes;and above all the whole constituteda milieu into which thespectatormightpenetrateas an invisible,immaterialobserver,yet one who not only saw but also "experienced" all that transpiredthere. The camera's ubiquity and the strategieswhich led to the spectator'sidentification with thecamera's viewpoint,togetherwith thesystemof orientationalmatchesby

5. It should be noted that as oftenas not a contemporarypresentationof this or any otherfilm would have been accompanied by a "lecture,"the task of which was to centertheseacentricimages. it is my Independentlyof thealien natureof thistypicallyprimitivesplittingof thenarrativesignifier, contentionthat an audience which had been watchingsuch filmsforas many as ten yearsmay well have been sufficiently "on its toes," even without the help of a lecturer,to conduct spontaneouslya slightlymore topological reading than we are normallycapable of today.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

83

which the right/leftrelationships of the spectator's own body organized his apprehension of all contiguous spatial relationshipson the screenfromshot to shot, reinforcedthe sense of spatial integrity.These two acquisitions were set-up,6destinedto ultimatelyto convergein thefigureknownas thereverse-angle become the keystoneof the entireedificeat the level of visual signification. While the full head-on reverse-angledid not become generalized until the and while of course thesyncvoice was nothearduntil theend of the mid-twenties, decade, the systemthusconstitutedas a visual entityhad become fullyoperational in the United Statesbeforetheend of World War I and in WesternEurope by the early 1920s. FritzLang's Mabuse diptych(1922) is an earlyexample of thesystem masteredto a perfectionthat has perhaps never been sured. And it is not without interestthat Eisensteinhad the opportunityof studyingclosely such a supreme example of the systemwhose emergenceI have brieflysketchedhere, having been involved-in whatcapacityhas not,I believe,been clearlyestablished as yet-with the editingof the Soviet versionof Mabuse. Several years before the firstprojections at the Grand Cafe, Edison was already dreaming of filmingand recordingoperas, and in this his enterpriseis antitheticalto that of Louis Lumiere. Not only did the team working under Edison's auspices (W. K. L. Dickson and his associates) invent the first"sound movies" with theirKinetophonograph,whose eyepieceand earphonesprefigure, at thescale of theindividualspectator,thedark,womblikeisolation of themodern movie palace, but theyalso shot some of theearliestclose-ups. And all of thiswas done in theBlack Maria, thatprecursorof themodernsound stage.If thecompany was soon forcedby thecompetitionfromLumiere to give up theattemptat sound and to copy themoretypicalearlyEuropean models,theseearlyexperimentsattest to the existenceof a need,ideologicallydeterminedin part,but only in part,that would ultimatelygive rise to an institutionalmode of representation. We also find,as early as the firstLumibre films,and throughoutthe early periodof Frenchcinema up to themasterpiecesofJasset,Perret,Feuillade, and the emigre Fasnier, firstin scenes shot on location, later in increasinglyelaborate studio sets,a verythoroughexploitationof thepossibilitiesofdeep-focusmise-enscene. In fact,we are dealing here with an increasinglysharperprefiguration of that pseudomontage within a single take (except in the work of Feuillade, intrasequentialeditingwas still rarein beforeWorld War I) which would ultimatelybe capable of reproducingthe structuresof classical montage. This approach, which among French directorscontinued to serve as a vehicle for strictlyprimitiveelements,such as theinsistentglancingat thecamerawhichone still findsin Feuillade as late as 1916, would reappear twentyyears later in the canned theaterof the earlysound years,when it was simultaneouslytheorized by

6.

Also called shot-reverse-shot or shot-countershot in tributeto theFrenchchamp-contre-champ.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

84

OCTOBER

none other than Eisenstein (the paradox is only apparent) in his classes at the Moscow Institute.7 The dominance of theWesternmode offilmicrepresentation was determined neitherby ideological factorsalone nor by sheereconomic opportunism.Rather, it correspondsbroadlyto the mode of constitutionof the Subject in our culture, and it developed into an ideological vehicle of unprecedentedpower. However massive its political and social consequences, it was the resultof an overdetermined convergenceand not simplya class strategy.

At the timewhen thecivil war was recedingwithintheyoung Soviet Union and the great period of artisticexperimentationwas beginning, the systemof representation developingin thefilmindustriesof thecapitalistworldwas notyet consolidated. We have already noted that syncsound constituteda serious fully lack which, it is clear today,the intertitlenevercompletelyfilled.To theveryend of the silent era, it retained(and indeed still retains) a "distancing" potential which directorslike Gance and L'Herbier had sought to exploit to aestheticends. A filmindustryas culturallyimportantas thatof remainedverystrongly dominatedbyearlypractice,withfrontality stilldominantas late as 1925,withthe rules of orientationstill verypoorly assimilated.And in all countriesthevarious "punctuating" opticals made possible by recenttechnologywere as yetscarcely encodedand wereoftenused-and not only in avant-gardefilms-to contributeto a freelydecorativestyle.This "unfinished"statein which thesystemfound itself, especially in Europe, played a decisive role in the orientationsof the most importantSoviet directorswho, with only one exception,were otherwisequite preparedto accept the system'sclaim to a privilegedstatus. It is no doubt this twofold circumstancewhich determinedthe earliest options of Lev Kuleshov and his troupe. It was this which led themto the first theorizationof the systemof orientationmatching.Their most famous experiment consistedof a seriesof montage fragmentslinked by actors' entrancesand exits,so thatvarious partsof Petrogradwereseen as contiguous,whereasanyone familiarwiththecityknewthattheyweremilesapart.This experimentwas in fact nothing more than the rational formulationof the contiguitymatch long since masteredat the practicallevel by D. W. Griffith. In The Musketeersof Pig Alley (1912), forexample, a whole "imaginary" neighborhoodis similarlyconstructed by laying end-to-endfragmentsof settingswhich are broughttogetheronlyby the successiveframeexits and entrancesof the actors. Following these laboratoryexperiments,the films that came out of the See VladimirNizhny,Lessons withEisenstein,trans.and ed. JayLeyda and Ivor Montagu,New 7. York, Hill and Wang, 1962.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

85

Kuleshov workshopattestto anotherconcern,not unrelatedto thefirst:studying and appropriatingthe codes governingthe major genres of the capitalist film industry-the spy serial,as in The Death Ray (1925); thecomedy,The ExtraordinaryAdventuresofMr. Westzn theLand oftheBolsheviks(1924); the "Far North" adventuredrama, By the Law (1926). The guiding principle behind all these productions was that the institutionalmode of representation,the genres and other coded systemsfounded upon it, offeredideal vehicles in the ideological strugglebecause of the privilegedrelationshipswhich theyalreadyenjoyedwith mass audiences. Afterthe Kuleshov group disbanded in 1925, its ambitions were no doubt best achieved in Miss Mend (1926), directedby an ex-discipleof Kuleshov, Boris Barnett(in collaborationwithFyodorOtsep). In thisfilmtheprincipleof political didacticismthroughpasticheis maintained,but withone fundamentaldifference: this monumental "serial" (three parts, over four hours long) frequentlyshifts abruptlyfromone popular genreto another.Spy thriller,sentimentalmelodrama, romantic comedy, slapstick farce follow each other in quick succession. The intention is clearly to undercut the escapist and alienating absorption of the popular genres. I have no wish to establish,in thecontextof thisinventory, any hierarchical order whatsoever.The wide range of Soviet attitudesand options, which runs fromKuleshov's pastiche to Dziga Vertov's"deconstruction,"correspondedto a theveryconcreteand pluralism indispensableto thesocialistethic.It also reflected needs of Sovietsociety,cominginto existenceundernotoriously highlydiversified conditions. Kuleshov's undertakingthus appears doubly complex and difficult justified.The urban masses werealreadyquite familiarwith thecurrentmode of representationand formsof expression,and it was obvious that one important way of reachingthemconsistedin acquiring the theoreticalmasteryof thatmode and in appropriatingits formsof expression.Furthermore, although thebulk of thepeasantrydid not come to know thecinema until aftertherevolution,it takes theoptimismofa Vertovto becomeconvincedthatlinearexpectationswithregard to thecinema would only be producedbypreviousfilm-goingexperience,and that thesepeasant masses were consequently"unspoiled." V. I. Pudovkin also came out of the Kuleshov workshop.His approach was not fundamentallydifferent fromthatof his mentor,althoughhis methodologyand, of course, his stylistics,which are not the subject of this essay-is quite different and his ambition,in a sense,fargreater.Pudovkin was strivingprincipally to extend the possibilities of the existing system,while maintaining its essential principles. This undertakinghas undeniably enriched our cultural heritage,withsuch remarkablefilmsas The End ofSaint Petersburg(1927) or The Deserter(1933), but it was certainlynot devoid of contradictions.Significantly enough, theseactuallyrepeated,at a higherlevel ofelaboration,thecontradictions experiencedby the pioneersof the earlyand formativeperiods. Some yearsago, in a programmaticessay which has not surprisinglyfallen

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

86

OCTOBER

into neglect,two Britishcritics,Michael Orromand RaymondWilliams,voiceda number of criticismsconcerning the Pudovkin method in connection with a sequence fromStorm Over Asia (1928).8They argued that the directordisrupted too radically the cohesion of the spatiotemporal continuum, that essential And guaranteeof verisimilitude,in otherwords,of thefull-blowndiegeticeffect. theywent on to compare this outmoded,disjunctivestyle,too analyticfortheir taste, with the techniques of modern cinema, illustratedby a sequence from George Stevens's Shane. Here, they demonstrate,continuityis ensured by the presentto the scene,theirordering juxtaposition in long shotsof all thesignifiers of theirbeing linked togetheronly assured instead being by picturecomposition and as was so oftenthe case in Pudovkin's screen-direction by eyeline matching, films. Despite the ingenuous character of their demonstration,these writers pointed to a fundamentalcontradiction,one which is of considerableinterestto basic researchin thisfield. Pudovkin's writingsand his polemic with Eisensteinclearlybear out the evidence of his films:his chiefconcern was to draw the ultimateconsequences fromthathistoricalprocessof linearizationof theiconographicsignifiers to which I have already referred.Let us consider the sequence at the beginning of The Motherin which thefather,tryingto takethehousehold clock down fromthewall to exchange it forvodka, is confrontedbyhis son and wife.He accidentallybreaks the clock and leaves the house carryingoffthe laundryiron which servedas the clock's counterweight.This scene is a perfectillustrationof Pudovkin's method. The scene is broken down into a series of key fragments,big close-ups whose meaning is whollyunequivocal and which,while respectingand renderingquite the continuityof the action, primarilyserveto spell out thataction satisfactorily in a seriesofelementary, carefullydifferentiated signs,in a simple,causal chain. A face grows tense,an arm is raised,a wheel of theclockworkrolls across thefloor. There is no room forthegaze to roam unguided (or evenguided) about theimage forso much as an instant.The director'sconstantconcernis, on thecontrary,to regulatethe "flowof signs" as closelyas possible. Moreover,in a sequence such as this one-and it is here that Pudovkin adds a new dimension to an approach which is otherwisefundamentallyGriffithian-acceleration of tempo and strong rhythmiatternsin certaineditingfragments generatethepathos of theclose-up, to use an Eisensteinian term which seems perfectlyapt in this context. The scansion of thesignifiers no longerhas as its sole aim to conferorder fragmentary on denotative signs. It also serves to control the underlyingproduction of meaning,theconnotativedimensionof thefilmicdiscourse-what has oftenbeen called the "emotion" of thescene-by means of thedynamicsof the successionof the montage fragments.Moreover,the connotativeproduction is also used by Pudovkin in a reiterativemanner, in particular to suggest sound effects.One 8.

Raymond Williams and Michael Orrom,Prefaceto Film, London, Film Drama, 1954.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

87

VsevolodPudovkin.The Mother.1926. thinks of the bar scene in The Mother, which immediatelyfollows the one describedabove. The livelyatmosphere,the throbof the music are suggestedby theverydynamicsof thesuccessionofdetails,no longersubjectto causal orderbut ratherswirlingabout in an impressionistdescription,a verticalequivalentof the horizontaltransparencyof Griffithian linearization. It is a fact, however, that in many ages, especially those involving confrontationsbetweenmore than two characters-hereI have in mind a scene involvingthe mother,the son, and thetsaristsoldiers,or anotherin which "The Heir to Genghis Khan" confrontstheEnglish furtraders-Pudovkin'sanalytical penchant, his concern to make each picture into a "brick" as elementaryas whichhe can controlas closelyas possible,does possible in a chain ofsignification indeed lead him to weaken the verisimilitudeof the diegeticspatial continuum. And yet this verisimilitudewas a foundinghistoricalcondition of the system which subtendshis whole endeavor.Wishingto carryto itsextremeconsequences the logic of linearizationthroughediting,Pudovkin comes up against the same obstacle encounteredby the pioneerswhen theywerecastingabout formethods capable of overcoming the unfortunate"dissociative" effectwhich the first interpolatedclose-ups had upon the unityof filmsthat still depended almost exclusivelyon the layout of the primitivetableau. In both cases, thisdisintegration, as it were,was the price thathad to be paid foran increasein "expressive-

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

88

OCTOBER

ness," in otherwords,a greatercontrolovertheproductionof meaning. Striving to remain within the bounds of fundamentallinearityand to strengthenthat of thesystemis linearity,Pudovkin fails to see thattheenunciationcharacteristic not simplya successionof signs,as decomposedas possible, but thatit is founded on a dialectic between such "stripped-down" images and a more complex spatialityofferingcomplementaryguarantees. The close-up, as integratedinto filmicdiscourseby Griffith, Barker,theInce brothers,etc., drew an importantshare of its significationfromthe widershots thatproceededand followedit and fromwhich it was in a sense excerpted.It was through this alternation of long-shot and close-up that the stronglydiegetic cinema was to attain maximum effect(and in this sense a filmlike Shane is certainlyan example). Intoxicated,as it were,by the possibilitiesrevealedin the new-foundmasteryof orientationalmatchingprocedures(and in particularthe eyeline match), Pudovkin sought to reconstitutea given profilmicspace in its entiretysolely throughthe successivepresentationof its details. He attemptedto renderthe full presenceof characters,objects,and indeed thediegesis itselfsolely throughthis "nearsighted"approach. Yet in so doing he sethis workat odds with a whole dimension of the systemhe was seeking to improve, since in many sequences of his silentfilmsdiegeticspace is reducedto such an abstractionthat such as theillusion of thepresenceofcharactersto each otherare importanteffects weakened. considerably Paradoxically,one of thefinestmomentsin The Motheris thatin which the method described above is abandoned completelyand Pudovkin returns,for reasonsofstylisticand dramaticcontrast,to a space which is much closerto thatof the filmsof Louis Lumiere. The firstpart of the irable scene showing the in the factory confrontationbetween revolutionaryworkersand strikebreakers remind us how in shots suitedis the is filmed which fixed, aptly wide-angle yard primitivetableau to scenes of mass struggle.This demonstrationwill be confirmedagain and again throughoutthe Soviet cinema's silent period and well beyondit. It was Dreyer,in The ion of Joan of Arc,who went on to derivefrom what we might call the Pudovkin contradictiona coherent dialectic based preciselyupon that diegeticdissolution, assumed as such, of profilmictopography. However,it was the Ukrainian masterAlexanderDovzhenko,in theopening sequence of his greatestfilm,Earth (1930), who went furthestin putting that contradictionto work, designatingit as such, showing how it was possible to construct,withclassical spatiality,an ambiguous diegeticspace, in thesense that it is essentiallyand disturbinglyuncertain. This celebrated sequence deals with an old man dying and a dialogue betweenhim and friendsand relativesstandingor sittingaround him. But is it so certainthattheyare actuallyaround him?Some shots,especiallythoseof thebaby playing, seem to involve a relationshipwhich has nothing to do with ordinary contiguity; they seem more like elements of "attraction" in the manner of

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

89

Eisenstein.The charactersare neverseen togetherin thesame shot;theyare linked only by theireyelineexchanges.A close reading of this sequence shows a whole seriesofdiscrepancieswhichactuallyrenderimpossiblea readingofdiegeticspace in keeping with the traditionalsystemof orientation:"in the place where" the orientationof a glance fromthe old man had enabled us to situatethisor that character,we now encounteranother;in the course of anotherseriesof apparent shots,we encounterstillanothercharacter"in thesame place," and reverse-angle yet,as far as we are able to judge, the hieraticstillnessof the scene has been preservedthroughout.At othermomentsa shot of a fieldof wheat seems to be located in an "impossible" space with respectto the eyelinedirectionsof those who see it. And thisopening sceneoffersonly one of various strategiesemployed by Dovzhenko: oftenan articulationbetweensequences will leave fundamental doubts about the precise momentwhen the spatial or temporalhiatus actually occurred.Along another axis certainshots, though more closely related to an "emblematic" space/timethan to the diegetic space/timeproper,nevertheless continue to entertainsubsidiarylinks with the latter(witnessthe seriesof shots indicatingthe ing of the seasons, the quasisymbolicsequence of "the lovers' night,"or theshotsofa youngwoman standingbya sunflower).Conversely,other momentswhich are firmlyanchored in the primarydiegesis (such as Vassili's famous dance-and-deathor his father'snight of mourning) seem to partakein turnof thoseemblematicshots,tendingto suspend themovementof thediegesis. It is throughsuch ambiguitiesas these,such derogationsfromtheseamlessnessof

AlexenderDovzhenko. Earth. 1930.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

90

OCTOBER

therepresentationalfabric,such attackson themetonymicintegrity of thediegesis ofa metaphoricdiscourse,thatDovzhenkocomes bytheintroductionof fragments closestto an importantaspect of Eisenstein'sgreatadventure.But in factthistype of constructionin one way or another was a major concern of nearly all the important Soviet directors.Yet while it has often been identifiedwith them exclusively,it should be pointedout thatthesetechniquesgrewout ofan objective in Intolerance encounterbetweencrosscuttingof the type perfectedby Griffith in of which are merely and the turn (1916) metaphoriccutaway Gance, procedures inserts and earliest of the crosscutting. extrapolations The finalsequence of Eisenstein'sStrike(1925),consistingof an alternation betweenshotsof butchersat workin a slaughterhouse(shotswhich absolutelydo not belong to theprincipal diegeticspace/time)and images of thepersecutionof the strikingworkersby mounted Cossacks, providesthe earliestexample of this typeof figurein narrativeSoviet cinema. Here the relationshipbetweendiegetic and metaphoric space/time (which in this instance involves its own strongly diegetic effect)still derivesfroma linear concept which is perfectlycompatible with the Griffith approach. In fact,one mightcite severalmainstreamfilmsof the sound era which have incorporatedthis technique of parallel and extended metaphor(WalterGraumann's Lady in a Cage comes to mind ... but do not the shots of buildings and citystreetsin Muriel functionin a similar way?). On the otherhand, Eisenstein'sdevelopmentsof thisstrategyin The General Line (1929) and above all in October(1928) may be said to be fundamentallyat odds withthis linearity. The mechanical peacock in October,which appears fragmentarily within the montagepiece associated with theopening of thedoor as Kerenskyentersthe great room which is to shelterhis precarious power, is of course a symbol of Kerensky'sfatuouscharacter.But it is so tightlymeshedinto themovementof the door itselfthat it resistsany reductionto a single signifyingfunction.A naive reading, predicatedon the inviolabilityof diegeticspace/time,might conclude thatthis is an automaton set in motion by machinerywhich connectsit with the door. This is but one (perfectly"legitimate") aspect of a complex productionof meaning irreducibleto any linear model. I cannot draw herea completepictureof Eisenstein'scontributionto thefarreaching investigationof the establishedrepresentationalmode, undertakenin factby all the most advanced of the Soviet school. One would have to discuss typage,thatimportantreconsiderationof the cinematiersona, and the complex relations which it entertainedwith the stereotypedcasting of the capitalistcinema. One would have to discuss as well the concept of themass-ashero and therevaluingof the long-shotassociatedwithit, as well as themixtures of styleand genrein Strike,October,and The General Line, and of course such ambitiousattemptsto extendthedirector'srangeas "tonal montage"or "intellectual montage." However, it seems to me that Eisenstein presents his most stimulating

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

91

challenge when, firstin his films,later in his teaching,he strivesto found a dialecticsof orientationalmatchingwhich, thoughhe saw it as a complementto the systemthathad risenover theprevioustwentyyears,also tendsto undermine the veryfoundationsof that system. I have alreadysuggestedthatit was preciselybecause oftheunfinishedstatus of the representationalsystemthat Eisensteinand his fellow filmmakersfound themselvesin a relativelyprivilegedsituationfora rethinkingof filmpractice.At that timein Europe, eyelineand directionmatching,forwant of any universally acceptedcodification,forwantofa "continuitygirl,"was no more thana working hypothesis,one which seemsto have enjoyedfavor,it is true,but whichremained only one possible option among others(witnessall the mismatchedeyelinesin Frenchand Germanfilmsof everycategoryas late as the mid-twenties).9 In this connection Strikecontains an extremelysignificantsequence. The spyingforemanis knockedoffhis feetbya clout froma huge steelwheel swinging on a crane driven by a group of mischievous revolutionaryworkers.In this sequence, perhaps forthe firsttimein filmhistory,we see illustratedtheproposition that "correct"directionmatching,the logic of which correspondsto thatof the right/left orientationsof a real or imaginary"establishingshot," could very well coexist withothersystems,and thatalthough thelattermightcontradictthe logic of the former,togethertheycould constitutea single compositespace/time characterizedby its unnaturalness(i.e., its rejectionof left/right body logic). For indeed, in the successive shots showing the foremanbeing knocked over, the swinging wheel changes screen directionat each shot change, and yet all the diegeticevidence(and our own common sense) tellsus thatin realitythedirection of the wheel remainsconstant. Of course the intentionhere and, at one level,theeffect produced consistin an exteriorization,throughthis "violation" of representation, of the latentclass violence behind this relativelyharmlessincident.In his of Eisenstein's teachings,Vladimir Nizhny tells how the mastertheorizedhis doctrineof the "montage unit," which advocates dividing up a given sequence into subsequences definedby successivecrossingsof the 1800 line. These "bad" position/ directionmatchesare ofcoursemeantto emphasizeprivilegedmomentsof tension in the narrativeflow.Indeed, wheneverEisenstein provided a rationale forhis innovations-invariably after the fact-he invoked criteria derived from the And thedramaturgyat workin thesequences thatare ideology of representation. most representativeof his dialectics of matching provides confirmationof this "expressionist"outlook and of thecorrelationbetweensuch experimentsas these and Eisenstein'squest forthe effectwhich he called pathos. However,it seems to me no less truethatthereis a preciousparallel statementin thisstrategy, foritalso

9. In Lang's Metropolis,L'Herbier's L'Argent,Raymond Bernard'sLe Miracle des loups.... are clearlynot talkingabout the mistakesof amateurs.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

We

92

OCTOBER

involvesa jeopardizingofthesystem'sgreatest"secret":thefactthata filmis made up of fragmentsof montage,thatit is not bynaturebutbyartificethattheclassical d&coupageproduces an effectof continuity. We find one particularlyvivid illustration of this in the Odessa steps sequence in Potemkin(1925). Here the extremediscontinuityof the editinggoes farbeyondmereimpressionisticsubjectivism,and theprincipleof montageunits intervenesspectacularlyto organizetheclimax of theepisode. In thisinstancethe expressiveintentionis accompanied by a programmaticstatementof no small endowed withrelative importance:thata secondaryorganizationof thesignifiers, autonomy,can give filmicdiscourse an entirelynew dimension, irreducibleto linear expressiveness.The sequence is constructedaround two broad montage units,of which the second intervenesonly when the nursefirstappears with her baby carriageand is thenassociated with the carriageas it rolls alone down the are intercutwithshotsof the steps.However,theimagesof thisdramatictrajectory these are and filmed from angles which belong to the first continuing massacre, theend of itsrun,to "fall back" unit. the toward carriageseems, Finally, montage into thefirstunit (in otherwords,into theinitial right/left relationship),and after this "dissonant" period consistingof cuts back and forthbetweenthe two units, the sequence ends entirelyin the first.It is throughsuch constructionsas thesewe mightalso cite thesecond sectionof Potemkin,"Drama on the Quarterdeck," thecream-separator sequence in The General Line, or theraising of thebridgein October-that Eisenstein became the firstto succeed in relativizing certain fundamental norms of the institutional mode of representation.This mode would, of course,reintegratetheminto a subsystemderivedfromit,but which at the same time contained the premisesof a more fundamentalcontestation.We natureof this may, I believe,sum up both the progressiveand the contradictory work with the following well-knownobservationtaken fromNotes of a Film Director: The strengthof montagelies in thefactthattheemotionsand mindsof the spectatorsare included in the creativeprocess. The spectatornot only sees those elementsof the work which are capable of being seen but also experiencesthe dynamirocessof theemergenceand formation of the image just as it was experienced by the author. This probably is the highestpossible degree of approximation to visually conveying the author's sensations and conception in the greatest possible completeness,to conveyingthemwith "that almost physical tangibility"withwhich theyarose beforetheauthorduringthecreative process,at the momentsof his creativevision.10 Under close scrutiny,this text may be seen to reveal with great precision 10.

S. M. Eisenstein,Notes of a Film Director,New York, Dover Press, 1970,pp. 77-8.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

93

Eisenstein'scomplex attitude.On the one hand, he rejectseverythingin therepresentational system which causes the spectator to see only "those elements This is thecredo, capable of being seen," in otherwords,he rejectstransparency. as it were,thatunderlieshis "dialecticization"of thematchingsystemand all the other "illusionist" strategies;his goal is to make theworkof thesignifiervisible. into a spectacleof theclassical type, Yet at the same timethisworkis reintegrated one which is certainlyon a "higher plane" than the other, but one which neverthelessmustin the last analysis submitto thesame linear,we mightevensay totalitarian,model: what the spectatoris supposed to grasp at the end of the process, whateverwork he or she may have been called upon to perform,is assumed to be what the authorput into it. We findourselvesface to facewith the old illusion thatholds theworkofartto be a mediator,a means ofcommunication between two sensibilities.This will perhaps also help us to understandwhy Eisensteinneversought (not even in Strike,despiteall claims to thecontrary)to oppose thesystemby thenestablishedwithany notion of a tabula rasa. In spiteof theirdifferences, in spite of theirdisputes,he sharedwithPudovkinand Kuleshov thedeep convictionthatthe "language" with which thename of Griffith was then so closely associated was tantamountto a basic language whose fundamental who proclaimedhis attachmentto componentswereintangible.Even a filmmaker dialectical and historical materialismand who felthis task was to enrich that systemthroughcriticalreappraisal was bound to remain within the conceptual frameworkwhich it defined.This is the nervecenterof his polemic with Vertov. Needless to say, it would in my estimationbe foolish to reproachhim forthis. Among the Soviet masters,Dziga Vertovalone advocatedan uncompromistabula rasa. In the USSR of the 1920s, such a position also involved ing contradictionswhich are far fromnegligible. The fact remains, however,that Vertovwas the firstfilmmakerand theoreticianto have produced-in ways that were at timescrude, at othersdeceptivelypolemical-a critical definitionof the nature of cinematic representation,and to have undertaken,in his masterwork The Man with a Movie Camera, a practicalcritique of it. Reading certainVertovtextsoverlyliterally,commentatorshave oftenmade of him the irrepressiblechampion of documentaryagainst the fictionfilm. However, what this readingof his careerfails to reveal is that the reason Vertov seemed to be combatingfictionper se was thathe perceivedin the fictionfilmof thatera thehegemonyofa deeplyalienatingsystemofrepresentation. This was in part because of the ideological substancewhich in capitalist countriesit almost invariablypurveyed-explicitlyor implicitly-and in part because of theive attitudethatit requiredof thespectator.And if he attackedEisenstein,seemingto confuse him with the mastersof Hollywood, it was because he feltthat in the revolutionarycontexta tabula rasa strategywas indispensableto clean theeyesof themasses,as he mighthave put it. Reading his texts,seeinghis films,it is hard to believe that he did not realize that The Man witha Movie Camera (or Kino-Glaz, forthatmatter)was as much, or as little,a fictionas Potemkinor The Mother.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

94

OCTOBER

We know that Vertov'sprojectdid not merelyconcern the perceptionand reading of images. Vertovhad a deep political commitment,and he even had the presentimentthatexercisesin the decoding of images could provide trainingfor thedecoding of reality.This projectstillholds promisetoday,and we have had, in In the Soviet Union of the filmsof socialist Cuba, a glimmerof its fulfillment. the 1920s, however,such amalgamations could easily lead to serious political illusions. It is also truethatin Vertov'scase, it producedmasterpieces.Through recentliteraturewe are beginningto have a betterappreciationof thetruebreadth of this film,long regardedas a simple display of cinematic fireworks."This classical response, so common among viewerseven today,is a symptomof the almost totalillegibilityof thisfilmforseveraldecades,theveritablecrisiswhich it causes within filmicrepresentationas a whole.., .and all the light which, at a second level, it sheds upon it. I can only sketchthe broad outlines of the work accomplished in this immensefilm,and I muststartwith theobservationthatits chief target is the fundamentallinearityof filmic representation,a linearity contestedin all its aspects,and no longer simplyin thatof syntax,as was chiefly the case with Eisenstein. This filmis not made to be viewedonly once. It is impossible foranyone to assimilate its workin a single viewing.Far more than any filmbyEisenstein,The Man with a Movie Camera demands that the spectatortake an active role as deciphererof its images. To refusethatrole is to leave the theateror escape into revery.For the relationships proposed between these images are seldom selfmovesbackwards,denyingour evident;oftenthe logic of successivesignifications usual senseof chronology,and evenmoreoftenit will takeus along an axis which is no longer syntagmatic,but paradigmaticof the film'sveryproduction(frozen frames,photograms,editingscenes,shooting scenes,screeningof the filmbefore an audience). Here again, however,the trajectoryfollowedis not determinedby any simple chronologyof productionbut is theresultof the multipleinteraction of other structures-thecycle of the workingday, the cycleof life and death, a reflectionon the new society,on the changing situationsof women withinit, on the vestigesof bourgeois life, on povertyunder socialism, and so on. Further associated with all this is a reflectionon filmic representationitself,on the constitutionof haptic space, theillusion of movement,and so on. One maysafely say that thereis not a single shot in this entirefilmwhose place in the editing schemeis not overdetermined chains ofsignification, bya whole setofintertwined and that it is impossible to decipher fully the film'sdiscourse until one has a completelytopological grasp of the filmas a whole, in otherwords,afterseveral viewings.12Resolutely reflexive,this filmwas the most radical gesturethat the silent cinema had known-in the Soviet Union or elsewhere. See forexample AnnetteMichelson, "'The Man with the Movie Camera': From Magician to 11. vol. 10, no. 7 (March 1972) 60-72. Epistemologist,"Artforum, In my work the concept of the univocalityof theinstitutionalmode of representation 12. refersof

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

Film's InstitutionalMode of Representation

95

iiOV? 4-1::::i:

.:::::i:-low

The Man witha MovieCamera.1929. Dziga Vertov. Vertovwas, however,a communist;as long as he was permittedto do so, he stroveto involve his work in the concreteconstructionof socialism. At thesame time,his analyses-writtenand filmic-were some thirtyor fortyyearsahead of theirtime.Not until the 1950sdid theyoung Stan Brakhageproduce a critiqueas penetrating,albeit writtenfromthe opposite ideological position; not until the mid-1960sdid European Marxist criticsreintegrateVertovinto Left aesthetics. Small wonder, then, that Vertov should have fallen prey to the pedagogical illusion, that he should have imagined that filmswhich have probably only become legible in thepast tenyearsor so (and even thenonly throughmuchhard work), could spontaneously"educate the senses" of the illiteratepeasant masses or, forthatmatter,of theurban masses,howeverhighlydeveloped theirpolitical consciousness.For theirexpectationshad long since been programmedby their experienceof dominantfilmpractice. Nothing will ever excuse or justifythe persecutionsto which this great master was subjected during the latter part of his life, when he was given courseto a relationshipbetweenthefilmsand thespectators-mostspectators-who have been written into theinstitutionby society.The others-a fewscholars,critics,filmmakers-willoftenperceivethe veryreal polysemicdimensionofjust about any filmtext.However,thisreadingis not onlyconducted fromoutside the institution(whose vocation,as ChristianMetz remindsus, is "to filltheaters,not to emptythem"); it is ultimatelyirrelevantto our understandingof the institutionas a single text.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions

96

OCTOBER

we mustmake no mistakeabout practicallyno opportunityto work.Nevertheless, it: if theworkof Vertovstill containsan immensetheoreticalpotential,ifit helps us to understandthe systemwhich still governs 99% of the world's filmand televisionproduction,if it helps us to reflecton the possibilitiesof eventually developing-within a political and social contextcomparable,at theveryleast,to Vertov's--methodsof audio-visual education and propaganda which might fromthebasic normsof cinematicrepresentation, he invented departsignificantly no magic recipes.In particular,it is clearlya delusion to imagine thatreflexiveness has automatic pedagogical value. The key to educating the senses of the masses, an education that would enable themto read the filmicsystem-to read themselvesinside it ratherthan simplybeingwritteninto it again and again-lies in changes a good deal more far-reaching.Even at the strictly audio-visual level, the education of the senses must through the schools of Kuleshov and Eisensteinbeforethatof Vertov,mustmove,in otherwords,in an ascendingorder of contradiction.

This content ed from 150.244.109.223 on Tue, 28 Apr 2015 08:51:22 UTC All use subject to JSTOR and Conditions