The Scene Text 113 7 The Scene Text Format One of the arguments frequently advanced against the screenplay as; literary form is that it is obliged to follow rigidly defined rules^Lfarmaf that reveal its function as an industrial blueprint. The problem with trif blueprint metaphor has been addressed in Chapter 3, but it is undoubtl edly the case that, to a far great extent than with the superficially^ comparable stage play, it is required to demonstrate the mechanisms I . which it may be realised within its target medium in terms prescribed^ within the conventions of a more or less standard format. Some ot its conventions are rather arbitrary (a screenplay/usually begins wi r the words 'FADE IN', whether the writer actually envisages a fade or? not), and much of the language is purely functional, as in the foi u of the slug line (see below). Margins, layout, arid lineation are subject' to established convention, which are all specified in any compeTenT screenwrlting manuaj^JThe purpose of this aspect of format is parlly J| to enable individual membjrajgfjhe cast and crew (actors, locate i managers, lighting technicians, and so forth) rapidly to locate the f places irTthe scWpX'^t'call for their individual input. Neverthele within the 'master-scene' format reproduced here, this is less important initially than ease of reading for the target reader of either the selling^ or the published script. The present book follows Claudia Sternberg's separation of the scree i play into the 'scene text', considered in thjs chapter, and the__'dialogueJ -text' in the next. Essentially, the scene text is everything bar the di lo^ueJextj_th£latter of which includes notjusJJthe_words sjpokenjjy the *Jj characters but also indiH'tionrof' whether the .speech is voiceoverr^O.),-offscreen (O.S.), or continued (CONT.) after interruption by either a"** BbrjajL-Or-jan_jJe^ Parenthetical direction kerning the delivery of the line is widelx.feccBiraged on the grounds litis the job of the actor or director, but not the writer, to determine life should b~e delivered. Screenplays vary widely Th~th"e"Hegree to |yt|they conform to this and other prescriptions of this kind; if such jfrection is indicated, it will be centred below the speaker's name, -ip scene heading, unattractively but generally termed thejslug line', ffaitis three elements: an indication of whether the scene is interior or a^ridr, the locationTand time of day. This is frequently merely a state-■it of whether it is day or night, but for local reasons, a more specific |g|tidn of time will sometimes be given. The elements in the slug ^xontainjniormation that assists location managers, lighting crew, sjera operatives, and so on. The slug line also, of course, indicates -ie divisions. Most screenplays and films consist of a large number iort scenes, and it is usually argued that in classical narrative films ||e scenes are linked together into coherent sequences of cause and p|ct. Just as in cinematic montage the meaning of the individual shot Tfnly revealed in the succession of images, so the meaning of the lyidual scene is determined by its position in a sequence of scenes. Hthbugh screenplays take this form_jo^__collaborative industrial lons^jfiy^^^ to its fjjjnconst ruction: this jsjjiemost self-reflexive of textual genres. Not f|y does it continually identify itself as a fictional construct, as does the leation of poetry, for example; iFalso cqr^tantly "reminds the reader jyphe_jndustrial process that is its raison d'etre. Unlike poetry, then, ^conventional communicationbetween implied author and implied llfrer js_to>ken, and the non-professional reader is forcecTto recognise Btthe implied reader appears to be someone other than himself or fi^self. Once this is recognised, however, and once the conventions of tl ^format have become sufficiently familiar, there is no intrinsic reason reading screenplays should be any more alienating an experience ^th^ni reading any other kind of text. Moreover, many aspects of this > plat tend to be used quite flexibly by the most accomplished screen-Jers; arguably, the better (or at least more successful or prestigious) jitters have greater scope for experimenting with the form without ^jeopardising the commercial prospects of the script. lodes far side from the dialogue, the most prominent aspect of the screenplay tjfotis the prose narrative. As with theatre plays, this is written in the 112 114 The Screenplay ^ >hi«|iift»( :% HoU present tense, for the same reason: the script is a direction to a reador*: «H who is_imaginatively present at the performance. Sternberg helptuijy^Sj distinguishes between tr^e_^modes' in the prose narrative: descrirJ3| '. tion, comment, and report (she adds a fourth -^geech - but as this iS wj-ktt, s'imply the dialogue text she considers it separately). To illustrate th . J at I Sternberg discusses the Prologue sequence from the Third Revised F: n ,j Script of Citizen Kane (dated 16 July 1940 and incorporating revision^ from 19 July). This is the script published alongside Pauline Kael's >■ 'Raising Kane' in The Citizen Kane Book. "The mode of description is composed of detailed se(&0n$_ajbou0j| p^o^ction^desjgn in addition to ecOTiomicaJLslug:]ni^ Description generally combineslwc^^ unique hybrid's The first is the 'frozenness' of prose description: a prose writer Choi pauses j^anj^iectinj>rderj^ generally renders it inactive, and freezes_;fae_narrative^ The second, unique to the" screenplay, is"the frequent indication_of camera movement,1 'The report mode is typified by events and their temporal sequence an<$ generally centers on "the actions of human beings' (p. 72). This fociisl on human activity, combined with the movement of the camera lrf-j the &>scrlptic^ itsjcharact^tic^ialfi dynamic movement in time. i *~~The remaining mode, that of 'comment', which 'eWpJain[s], interpret|s)'] or add [s] to .tiie^ dearly vl^ibi^ (p. 73), is on the face of it the most problematic. As Stinib^r^nbleVscreenwriting mam£ als. tend to insist that a screenplay should omit cc^ment, because it^ cannot be translated into visual terms. We may add that the convention that one page equals a minute of screen time means that excessive com? ment will interfere with this temporal equivalence. While ail screenplay? are written substantially in the report, action, and dialogue mo 1 tlielFIscon^ rancemjngto^comrli t«1 modeTSteTn^eTpu^ miss the opportuu touse the mode of comment. It is in this mode of presentation that evd new forms and designs of screenwriting shall be revealed' (p. 74). Thisij certainly the case with the Prologue of Citizen Kane, which is replete wifjj comment, such as information about the past history of the locatiorij or the screenplay description of Kane's Xanadu as 'literally incrediS' > which by definition cannot be filmed. The Prologue is extraordinarily evocative, largely because of such comment. The mythical assoaali of ancient, dead kingdoms summoned up by 'Xanadu' are amplified byj the 'exaggerated tropical lushness, hanging limp and despairing - M ■ moss, moss. Angkor Wat, the night the last king died'.2 The Scene Text 115 Itotizen Kane is undoubtedly a remarkably rich text, offering a wholly afferent kind of experience either from other screenplays or from latching the film, and it repays careful analysis. Precisely because it offers perhaps the most extreme example in all of screenwriting of gsyery literary use of the comment mode, however, it does not well jHu i ate how that mode functions in screenplays more generally. The jlfrual qualities of a given screenplay are inseparable from the antici-Ited production context. WjjHesjwas ro^ he i^ji£jwas,t^dj£ecti anJ_therefo develop a style gpvas appropriate for^im_as his ownjeader. The sCTigtTof some jier directorsjsuch as Davio^Mamet) are e^epjtionally minimalistic aendt comes to tM.sjg.ne, text, but arguablxlor Sesame reason: lie ^^TtopjvsJioj^J^w^ts tp^film it, orheknows he will be rdyjng on fjflpoateibutiQnsu^ to help him &|sjdlsJ5Jj^«a^ screenplay Is WeletaLprecisely ijftiake it amenable. to.mul_tiplejealisations. Sternberg considers the ^rriodes^separately from the rest of the text, it seems, because they can be considered aspects of style ssfflier than of indjisMal form. This division, which TiaTT^ertaTn -liV i' > suggests a jlistin^tio^ jn.| indus^ia^aspectsjjf the^scxeenplay. Partly in consequence, she ^ndsTtogive less emphasis to the former, in keeping with the overall goethodology, which is that of a linguist and film scholar rather than jptt'of a literary critic. Nevertheless, this rigorous approach can prove tictive. Thej^jiiejUjs_pf_the_jc^ _andjemantk dementsJhajL^^ J^firc*ionl°Hffi* 01 sTttlngLforexamp_le) m^more^d^amic, less ll&cretejisjDej^^^ as narration and chara£terisa- gton.il frese_ar^ §$jpreover, tb£^tkic±i0ns4>efweej^ as ^^yjgjatJie^m^For example, action that is reported is also action that Eid. The frequent absence ofinodifiers in screenplays is hot an jgenceof description; it is a style of description, and one that could be Spaded as commenting on, as well as describing, a reported action. In lEfect, Sternberg offers a version of 'close reading', but unlike the close ading of poetry, hers reveals a text that must constantly refer outside fe^Si^^^i?-*1' in a kind oJf reflexive recojL bringjhe filmba^k yearly, most screenplays suggest that the material can be realised OH the screen; this is its raison d'etre. Equally, however, the majority of 116 The Screenplay The Scene Text 117 screenplays do not make substantial reference to many of the 'elements' Sternberg identifies, including colour, lighting, sound, and music, which"' are ordinarily regarded as the responsibility of other specialists workir on the film. The same is true to some extent of camera, montage, anosr- !^ic^-P^J^^8^^ttH»e and location directly designate cutsj^aii 11 kilijo ^SP^^^LS^^BS^"scripjjto^wfaicjveach cutjs predefined'. In tl ^idcllmg_scilpts, '[e]diting_marjiejt^^ 5 such as indicaUc^j)|jty£ejof^hot (pp. 209-10). Spatib-tempor.i I an|g is easily indicated withouTspecifying the precise technical mt"ans',isL transition (cuts or dissolves, for example), by the simple juxtaposition oHT images or scenes to create stylistic, narrational, or functional i :ts.31|| The relationship between^scn^en^^ is perl ■ m3jj persuasively^iscussed byTier Paolo Pasoljni, in an essay en in A 'Tim, Screenplay as a "Structure that Wants to Be Another Structuio" Herjlj Pasolini is concerned not with the screenplay as merely a ■ e in-|| creative process. Instead, he investigates 'the moment in which it ( in big considered an autonomous "technique", a work complete and linisnliT in itself.3 He gives the example of a script that is neither an adaptatiM of another work nor has been filmed itself, although it could hr arguS that in theory one should be able to consider any screenplay ■ Mdiffll to these criteria, since one can always encounter a screenpla\ ni a fiffiji one has never seen. N *f| Pasolini argues that the methodology of what he calls 'styi nc crinf cism' is inappropriate to the analysjs^ptthej«ejraplay. First, 1 creerS play is distinguished notlio much by thejiuances of textua ail to m^hTexpecTto analyse^n^poemofapjiece of prose fictiorV but instea|| by;---an^emenf"tMFi'rnolJh.ere, that is a-"desire for form"' (p. 54)1 Second, the screenplay demonstrates 'thej:g_npnums_allusion i<> ''jtewj|| gpingj:inematogmphic work' (p. S3), and_this compels the read-1 wjh||S Pasoliniregards as a. lancl of: collaborator with the screenwrit* i Wjjfll in images, reconstructing in his own head the film to which the scra-npiWM -s^tem ^analysis, in which meaning is^supplied by a post-factq_recognition_of ^or~sffucture, and a micro-level that attends to the specific ways in jích thFscreenplay negotiates between, or simultaneousfy keeps in [jgý,rTts verbaTandj/is.uaLsigrr-sysj:em^ Jically different, less precise, but perhaps still more suggestive , short essay by Sergei Eis^nstgin, which explores from a director's rlpective the consequences of working with these two distinct sign-stims. The difference between a written text and a film cannot simply erased. Eisenstein gives the example of a phrase uttered by one of |je~survivors of the Potemkin mutiny, which became the source of one the director's most celebrated films. The veteran said that 'A deathly Jence hung in the air'. Eisenstein saw no difficulty with a writer corporating these words in the script, which ^séts out the emotional requirements. The director provides his visual tréšolution. And the scriptwriter is right to present it in his own £ language.... Let the scriptwriter and the director expound this in ftheir different languages. The scriptwriter puts: "deathly silence". The director uses: still close-ups; the dark and silent pitching of the aattleship's bows; the unfurling of the St. Andrew's ensign; perhaps "a dolphin's leap; and the low flight of seagulls.4 >» |ually, of course, the emphases may be reversed: the script may Inscribe a setting or character in literary terms that apparently exceed S caTíiío^nSeTišč^etri^ languaglTbTfiimVbTít^ jp^píóřrrpOEe director's imagination into_providing a correlative image, moud, "of těxřuTí ' pme the scr^np^ayjswritte^^ what lie spectator is to imagine is happening on the screen at that moment. ÍĚ^C5Tí5É-E550eJlse;Jri_^most_ all piose_fictj[on. tends to^draw attention to narration, because the discourse demonstrably comments 'J* 118 The Screenplay The Scene Text 119 retrospectively on story elements that have occurredjjrior to the moraej,t in which they arVnamtedlleuse^f the present tense in the screenplay obsgn-.es this gap between story and discourse, as does its construction , a series of more or less brief ^ a short scene within the film! In other words, unlike the retrospection, [ conventional prose fiction, "the screenplay hovers between present-tense -narration and sjjadpwyanticipation of a future realisation in a different^! medium. The stage play similarly unfolds in the present tense, yet there is a sig^ nificant difference between theatre and screenwriting on the one handj| and cinema on the other, because the image on the screen is at best; approximate record of an event that can only have happened at somig point in the past. On the cinema screen, it is never now. The screenplay' reads in the present, but it is the past of the film. Two of Woody Allen? films make great comic play with exposing this mechanism. In T-Purple Rose of Cairo (1985), a character within a film steps out from t! screen and enters the auditorium to join one of the spectators who has/ fallen in love with his screen image. The joke lies not just in the erasue. of the distinction between the fictional world of the film and the 'real'i world of the spectator, but in removing the distinction between thej past-ness of the film world and the present of the spectator. A similar^ conceit is seen in Deconstructing Harry (1997), in which an actor who ujj out of focus when filmed remains so in the 'real' world. In each case tl J conceit plays upon the powerful illusion of present-ness in a mediur that is inescapably a record of the past. Convention holds that one script page represent^one_jrmute,"j_ screen "time," aridthat the writer will ordinarilyconstruct the_sc|' 1 in accordance with the rhythmic demands of this e'qiwtion. This Ji i. j important consequences: lengthy enumerations of the items in a rooi fof"^ recordedTTTnTm^ have the wealthy detail often fouridln the realistic novel; it is more akin to poeuy, the* short story, or the Chaucerian fabliau. " "James F. Boyle goes a little furtherlh positing a 'readingjime' that isijj considerably shorterJh^nJJie^rojejc^ A script page = Reading time = Projection time = Fictional time5 eleven inches approx 25 sec. 1 minute variable This is, of course, merely a hypothesis, a guess; different readers wilC read at different speeds, and some screenplays are harder to read than jjhers. Nevertheless, empirical observation of one's own reading habits ^ds to support this assumption, which is not surprising in view of &_dialQgue.-intemiye nature of jmany screenplays, the economy of pr descriptive modes, and the generous margins and line spacing 'Sanded by studio conventions. The e^e^nc^c^iiadlngTOcJ^t ay, then, should correspond rhythmically to the viewing of a film, but ^accelerated speed: accepting Boyle's approximation, the hundred pltwenty-page script for a two-hour film should take something Per one hour to read. ffipyle's 'fictional time' is what most film theorists would describe as g>ry time/: that is, the duration of events as they 'really' happened. |s islo be distinjguisj^jrom^ jp^raHrar^w^in^which the story events are narrated, and which, ihe_i^ssicaLjQarr^^^ and others, can bitort story time in three basicjw^^ ||jjby theuse"6T'flashbacks (analepsis) or flash forwards (prolepsis); it palter the duration (much easier to quantify in cinema, by the use of 0p motion for example, than in prose fiction); and it may change the ' agncy, as in Rashomon (Akira Kurosawa, 1950), in which the same v|jnt is shown on multiple occasions. The technological constraints Jthe earliest films, such as the Lumieres' Sortie d'Usine (1895), meant It there was no distinction between story and discourse time. A rare frnple of a later film that supposedly unfolds in 'real time' is High |n (Fred Zinneman, 1952), in which the discourse time purports to be Wt]Y equivalent to that of the story time, with many shots of clocks ■Bell the spectator exactly how long they will have to wait before the jjaactic arrival of the train at noon. As such, it is an illustration of pure p|chcockian suspense. fact, there are some slight distortions in this equivalence in High W&W, but it remains highly unusual, since almost all films condense Ory-time in the discourse through the use of cuts and other transitional |eiyices such as fades and dissolves. A much more radical experiment is 't Year at Marienbad (Alain Resnais, 1961). Its writer, Alain Robbe-llet, 'saw Resnais's work as an attempt to construct a purely mental [j&e and time - those of dreams, perhaps, or of memory, those of any Iwtive life - without excessive insistence on the traditional relations ause and effect, nor on an absolute time-sequence in narrative'. As j£ avant-garde novelist, Robbe-Grillet was interested in questioning Sjjptrbf existence the whole basis of narratology, which depends on the imption that there is a story in the past that can be recovered in gpresent discourse. Instead, 'our three characters ... had no names, uid ii i 120 The Screenplay no past, no links among themselves save those they created by then own gestures and voices, their own presence, their own imaginati n -Robbe-Grillet was so fascinated by the potential for fiction of Resnais's radical approach to time in cinema that he described his own scrip f T the film as a 'cine-novel'. Narration Narration has long posed a difficult problem for film theory, one 1\x its roots in the Aristotelian distinction between 'showing' and 'telling', The early actuality of waves breaking on a beach, trains entering a station, or leaves blowing in the wind had the appeal of appan it unmediated realism: for the first time, a technological apparatus < record the movement of natural forces that could I not be captured theatre. The camera therefore appeared to be 'shoeing' incident, r; than 'telling' or narrating it. \ A screenplay composed solely of Sternberg's fnodes of diali . description, and report, and lacking the mode of 'comment', is possibly;* the textual medium that comes closest to realising the ideal of 'show4 ing' without narration. With the dialogue and_report modes bemg7 simply a record of what fsJSd or se^ first-person or third person narrator of prose fiction. This is undoubt-~ edly part of what is'realTy WTdeolo^caral:gument against the usejrfl yoice-over that one frequently encounters in the same manuals that-counselagainstthe conament mode. The difficulty with this argument is that it presents screenplay and? film in impossible terms: as media that evade mediation. This lies ati the heart of the problem of cinematic narration, which needs to be| differentiated from narration in the screenplay. A comparison to the' beginning of a short story by Ernest Hemingway, who has a very une^ matic' style in the sense that it is often rigorously confined to the Tepcsg mode, the description mode, and dialogue, establishes this well: The door of Henry's lunch-room opened and two men came in. They4 sat down at the counter, 'What's yours?' George asked them. T don't know', one of the men said. 'What do you want to n Al?' T don't know', said Al. 'I don't know what I want to eat.' Outside it was getting dark. The street-light came on outside the| window. The two men at the counter read the menu. From the othef •* The Scene Text 121 lend of the counter Nick Adams watched them. He had been talking ^to George when they came in. J*. 'I'll have a roast pork tenderloin with apple sauce and mashed -potatoes', the first man said. 'It isn't ready yet.' 'What the hell do you put it on the card for?' 'That's the dinner', George explained. 'You can get that at six . clock.' George looked at the clock on the wall behind the counter. 'It's five o'clock.' 'The clock says twenty minutes past five', the second man said. 'It's twenty minutes fast.' 'Oh, to hell with the clock', the first man said. 'What have you got to eat?' 'I can give you any kind of sandwiches,' George said. 'You can have ham and eggs, bacon and eggs, liver and bacon, or a steak.' 'Give me chicken croquettes with green peas and cream sauce and mashed potatoes.' 'That's the dinner.' 'Everything we want's the dinner, eh? That's the way you work fc it.'8 superficially, such a style has the effect of minimising or even eliminating narration. It simply records a series of events as they |appened, and invites the reader to supply the connections that would jnti 'rate them within a coherent story. 6*TM Killers' is a very well-known text, but even on first encounter H|jtie style is likely to seem very contemporary to a reader today, partly because the set-up of the two voluble hit-men has undoubtedly influ-mced; directly or indirectly, such well-known works as Harold Pinter's Stage play The Dumb Waiter (1960), Quentin Tarantino's Academy ftward-winning Pulp Fiction (1994), and Martin McDonagh's In Bruges |500S), nominated for an Oscar in the original screenplay category. It is ^pt-coincidental that Pinter and McDonagh were acclaimed dramatists "|ffore turning to film, or that in Tarantino's screenplays there is such ff preponderance of dialogue that, in this respect, on the page they ■'ft- n bear a closer resemblance to stage plays. In all of these works the gialogue is both exceptionally prolix and remarkably vivid. V, s shall consider a comparable sequence of dialogue in Pulp Fiction •in Ui=; next chapter, but it is clear that part of the effect of the dialogue in I lie Killers' comes from its juxtaposition with the style of the prose 122 The Screenplay The Scene Text 123 description. The latter is syntactically simple and eschews modifli r enumeration, and metaphor. The same is largely true of the dialogue! except that the two men are extremely particular in detailing the itu ' they want from the menu. This could be read in a number of ways (j^ psychopathic need to order the world by naming things with precision^ as in the obsession with brand names in Bret Easton Ellis's America^ Psycho, or indeed Tarantino's dialogue; an attempt to intimidate Geergf by establishing linguistic mastery), but it clearly emerges as a distinctive idiolect, a style................................................................... The description is also, on closer examination, heavily stylised m ways that bear comparison with the modes ofreport and description », the screenplay. Compare the first scene of a ran4om example, Willi m Peter Blatty's The Exorcist (William Friedkin, 1973):-- An Old Man in khakis works at section of mound with excavaUngl pick. (In background there may be two Kurdish Assistants carefully^ packing the day's finds.) The Old Man now makes a find. He extrac it gingerly from the mound, begins to dust it off, then reacts witl dismay upon recognizing a green stone amulet in the figure of thTSp demon Pazuzu. Close shot. Perspiration pouring down Old Man's brow. Close shot. Old Man's hands. Trembling, they reach across a iude" wooden table and cup themselves around a steaming glass of hot tea, -j| as if for warmth.9 The series of shots specified or implied in this passage (and almost any screenplay would have worked as well or better to illustrate the pcinii resembles the prose of The Killers' in privileging the report m ■ * actions are described simply and in sequence. Most important is the ■ in each case, of parataxis: events are described without being connected"' by the use of conjunctions. This seemingly eliminates narrational commentary and plainly records events as they happen. Hemingway's use of parataxis, however, contributes to what is in rack a highly distinctive style that creates his masculine, existentialist w rl I view. 'Character' is action, as Aristotle - a ubiquitous authority in screen-, writing manuals - observes. In 'The Killers', all of the characters decide t&|j perform or not to perform certain actions (to give the men what they warir or not, to contradict them or not), and this sequence of actions builds^ towards what will turn out to be the story's major event: the decision. 5f ;01e Anderson, the man the killers are seeking, not to act on the Tbwledge that they have arrived in town. He does not explain this; it is feasion, revealed in action, that defines the situation and the character, prnilar effect is produced by the succession of actions in a screenplay. [Mause this sequence implicitly or explicitly anticipates its realisation ^cinematic editing, it is usually presented, as in the example from fit Exorcist, as a series of events without conjunction or comment. "fet this does not^t^lljneanjtiatjhjr^^ On the itrary, narration is supplied in at least two ways. First, thestyle is ^tQ^nyjnic: it is a^decfiOT^eve^Us^oj^objects consciously chosen from |jhinj:heji^ We are direrteTtoTook at the amulet, le^erspiration, the hands, and the glass of tea. Realist prose fiction aetimes attempts to conceal this process of selection by provid-excessive, redundant detail. By contrast, other forms, such as the Idieval fabliau, depend for their effect on the conventions of meto-fyjnic selection. In Chaucer's 'Miller's Tale', for example, every element at is introduced in the first part of the story will contribute to the filiations visited on the characters in its comic climax. Most screen-jjiiys therefore have someriiijrig_ofJhejm a joke. Because of the impressed nature of the form, jury objecttowhich itdirects attention Sablejobejh^wn to be a set-up, to have a particular significance that "'Jl only be revealed later on:"the child^s" fed coat in Don't Look Now, the ^fowsHaRef in Citizen Kane, the amulet in The Exorcist. le narrati^ iS then confirmed Wj*^., gra se^ond,._coro]laryjpj^^ shots are arranged into a ^ se*.*€i ^^ence^again in antJcipariOTTof^m edjting. AlthouglTtlnFcornbi- '" "Y a'tiqn of shots is paractactic (there will ordinarily be no comment to igplain exactly why the images follow in this particular sequence), the reader will ordinarily have no difficulty in inferring the explanation for SpParataxisjn the^creenplay therefore appears to havejhe opposite fa1"ft jNjjgct to parataxlsjri prose fiction: in the former, knowledge _pf. the ^onventi^ or Rational presence, yet in fiction, parataxis attempts to suppress the Effect of narra^tjon altogether. le resulting problem in film theory has involved the question , ■ rjwho or what is_Jojn^j^_n_airating. As Christian Metz observes, ^Tjie spectator, perceives images which have obviously been selected W5*i-4v'^ (tjiey could have been other images) and arranged (their order could A *li|ve been different). In a sense, he is leafing through an album of redetermined pictures, and it is not he who is turning the pages but toe "master of ceremonies", some "grand image-maker".'10 As the 124 The Screenplay The Scene Text 125 scare quotes suggest, the question of how to-describe the presence a 11 activities of this image-maker remains problematic, because in Edward Branigan's words 'the "person" whose voice is "heard" in a [film! text*ft may be a much more complex (invisible and inaudible) entity than a/"* voice-over narrator or someone being interviewed'.11 These complexities have been discussed at length in at least two majo^ studies, by Edward Branigan and David Bordwell, and the specificall. filmic_asDects_of narration are not necessarily relevant to narration;jn «a the screenplay. What is remarkable about the analysis of cinematic^ nrnraHo^nn the present context, however, is that the_screenplay ls$ almost never mentioned as its possible source. For example, Bordw^jT notes that in Eisenstein's films 'there is the sense that the text before: us, the play or the film, is the performance of a "prior" story', and is narrated by 'an invisible master of ceremonies who has staged this" action, chosen these camera positions, and edited the images in just this way', so that there is 'a continual awareness of the director's shaping hand'.12 This captures very well the ontological status of the film in| relation to its 'prior' sources, and as noted in Chapter 3, the relationsh ■ * between film and screenplay is of major importance in this respect. Th. difficulty in film theory appears to be prompted in part by the desir< * construct a single narrator (hence perhaps the status of the directoi j auteuf), even though Bordwell dismisses the 'implied author' of a fil-ii as 'an anthropomorphic fiction'.13 Bazin's paradoxical 'genius of the| system' appropriately suggests that the sense of a single centre of c sciousness may in fact be the result of extensive collaboration. Within the screenplay,_as opposed to the film, Sternberg distinguishes* betweenan impersonal narrative 'voice', which 'shows' by indications of editing, mise-en-scene, and overt or covert 'perspectivemes' (indications of perspective)^ and, the personal narrative voice, which 'speafa^f in voice-over, on-screen narration, or a written text (pp. 133-41). \ ■ if is difficult to concur that in the screenplay 'telling by a narrative agentf does not take place despite its high degree of prose. The text only anticUj| pates a narrative perspective in the target medium of film' (p. 157). Thisf sits uneasily with Sternberg's conclusion, in which she suggests thai the 'scene text' tends to 'narratize' for the blueprint reader, and '[t]ha screenwriter therefore becomes a hidden director1 (p. 231). *■ Character Superficially, character is a much more straightforward concept tharj narration; we all know what we mean by the characters in a film. Ev|T Ho, much of what we think of as a fürn character is supplied by the actor, ^ndftf»?. inust.be differentiatedtext. reenrMa^s^rejoften vaguewhenusing the descrijgüvejnode to portray «gc^acter,...partly because it is not the writer who wilLcastThe"actor, le comment mode is also widely regarded as an inappropriate means presenting character, as these comments cannot be filmed. As in the i\m cription of action, it appears that one is left with only the resources <öf dialogue and action, which consequently tend to construct the fetagonist in particular as a more or less existential being. Moreover, gbecause ^is general screenplay and describe gtaracTeis when they first appear',14 the text usuaTlyTacks theTeTouices ^available to the novelist of the accümüläffo'nrmodification, and even .contradiction of detail during the course of the narrative. Accordingly, |charactensati()n in the screenplay, in this sense at least, is skeletal. ^Before simply accepting this as fate, howeverfit is worth pausing to feisider the enormous number of highly acclaimed screenplays that * ano. heed to these strictures, and. d.escribe„the characters "in some-gnes highly novelistic ways. At the beginning of Taxi Driver (Martin Scorsese, 1976; written by Paul Schräder), before any slug line or action ponies a detailed physical description of Travis Bickle, interspersed with pi crisp, vivid dissection of his blasted past and inner life: |0]ne can see the ominous strains caused by a life of private fear, emp-||r-.tiness and loneliness. He seems to have wandered in from a land where y it: is always cold, a country where the inhabitants seldom speak. [...] He has the smell of sex about him: sick sex, repressed sex, lonely sex, bat sex none the less. He is a raw male force, driving forward; towards v- what, one cannot tell. Then one looks closer and sees the inevitable. 4- The clock spring cannot be wound continually tighter. As the earth moves towards the sun, TRAVIS BICKLE moves towards violence.15 IN objections that could be made to this passage in its entirety require g>0 elaboration: you can't film smell, you can't film the inevitable. Yet ^isyould be difficult to deny that Schräder has captured the essence of %e character as most spectators experience it; or, more accurately, that gcorsese and Robert De Niro have managed to film the 'unfilmable' jäements of the script, and that this is done in the manner suggested joy Eisenstein: the writer has one sign-system, the director another, and ffhileit may be the job of the writer to think in the visual terms of the äSEector, it is equally the director's job to find correlatives for the verbal |f within the cinematic system. The issue returns to Steven Maras's 126 The Screenplay Wie Scene Text 127 previously considered question of whether the film should be regarded as merely the execution of a prior conception detailed in the screenplay In any case, even if the script contains material that cannot be filrni it can still be read. Some scfeenplays go still further, and preface the script with descrip./ tions of the characters in a list of the mosy^ In the 18 October 1950 'final' draft for Hitchcock's Strangers on a Train) (credited to Raymond Chandler and Czenzi Ormonde^, two pages, are ^ devoted to paragraph-long descriptions of eight characters. The longest^ of course, are for Bruno Anthony and Guy Haines, with Bruno's porti t being particularly novelistic: About twenty-five. He wears his expensive clothes with the tweedy nonchalance of a young man who has always had the best. He has, the friendly eye of a stray puppy who wants to be liked, and the sai ■ wistful appeal for forgiveness when his impudence lands him in lhe_ doghouse. In the moments when his candor becomes shrewd call i . lation, it is all the more frightening because of his disarming charrrf-and cultured exterior. It is as if a beautifully finished door, carved of= the finest wood, were warping unnoticeably, and through the tiny" cracks one could only glimpse the crumbling chaos hidden insid -and even then, not believe it. (p 1) After the opening description of the shoes, our first view of Bruno" repeats the description from the first two sentences above; the samC pattern is repeated with Guy, introduced at the same moment (p. 2) Such descriptions can be viewed in several ways. They may, of com > be dismissed as merely the novelistic character sketches of a pr writer who has failed to realise the script in visual terms. Alternatively the writer may be doing the very opposite: rather than connnui 1 interrupting the narrative to indicate aspects of character, providi u a figurative insight intojfte^cjwa<*e£jn£ry^^ and the actor to draw on this conception in the coursejgfjhe^lfn."The metaphor that describes Bruno has a temporal dimension: the d ■-is 'warping unnoticeably'. The challenge to the director (and actor, and designer) is to translate this unfilmmable conceit into a cinematic equivalent, just as a similar challenge routinely confronts a screenwr-i adapting a source novel. The same script furnishes one of the most memorable introduclii i to a pair of characters in all of cinema: the feet - or, more precisely, i ■ IJjoes - that serve to characterise Bruno and Guy. Chandler and Hitchcock apidly fell out, and it has been widely accepted that Hitchcock simply ^abandoned Chandler's work and substituted Ormonde, an inexperi-tin.;d and compliant writer, after which Chandler tried unsuccessfully )-have his name removed from the credits.16 However, Bill Krohn ||ports that, after previously submitting a short treatment on 18 July p50, Chandler then wrote a second that anticipates the film's memo-ifble opening: •J; [Chandler's] next treatment, written between 29 and 12 August, " begins with the image of the feet walking, although here and in all subsequent versions of the screenplay there are three tracking shots f the feet before they touch, rather than an alternating montage s in the film.... It is possible that Chandler ... misunderstood the feidea of the feet, if it was in fact Hitchcock's, or else came up with it |§ himself, but in a less 'cutty' form which Hitchcock simply never took E-time to change in the script. w Hside from one crucial scene 'where it looks as if Guy is going to kill ||uno's father, which Chandler [ironically] found absurd', 'all Hitchcock Rept from [Chandler's] draft were the feet at the beginning.'17 pit may be that this is all of Chandler that survives, but if so, there are jther moments in the screenplay that follow a similar method. Towards |Ke end, Bruno scratches around frantically for the incriminating fparette lighter, which has fallen into a drain. Warners put out a press J|lease to the effect that Hitchcock 'spent the afternoon directing jjRobert Walker's hand. At the end of the day the actor was exhausted, |g>ut Hitchcock was satisfied with his "performance"'.18 The emphasis on £the hand is anticipated in the script. ||These are but two examples of a method of characterization that is peculiar to the screenplay among_textuanfc^^ put, 'fijn contrasFto thTtfieatre7 whichmust present the performer on J|age as physically "whole", film is able to fragment space and objects as ij|ell as the human bod_vl_(p. 115). Samuel Beckett is radically different £p:om almost all other playwrights in the frequency with which he Joes present the onstage body in a state of fragmentation: Nell and pjagg confined to dustbins with only their heads occasionally visible Endgame, Winnie buried up to her neck in Happy Days, the isolated '|fouth in Not I. As we shall see, there is also a cinematic quality to the |se.of voice-over in Rockaby and Footfalls. But Beckett is very much the stception that proves the rule. 128 The Screenplay The Scene Text 129 In contrast to the excess of descriptive information in the realist nuvpls most screenplays indicate character with minimal recourse to mj\e\y fiers. For example, The Usual Suspects (Bryan Singer, 1996, screenplay by Christopher McQuarrie), introduces two of its characters as I j,j Hockney, a dark, portly man in his thirties', and 'Fred Fenster, a tall, thiri man in his thirties'.19 McQuarrie's screenplay is deservedly regarded as a masterpiece, but the characterization in this particular respect is pure Agatha Christie. The obvious alternative for a writer aiming to write 'cin-r ematically' is to make a virtue of visual fragmentation by selecting a sali-ent metonymic feature tojndicate character. A E^<^Jth^bodystand$ forjthej^olebqdy, or;is^selected asaj^artlcularly memorable featu r^, j0 tJaat it^imultaneously signifies something of the inner self while n r . ducin^_aJcind_cJ-sjio to the individual There is a superficial resemblance to what E. M. Forster described as 'flat^characters_in a novel, those who possess a single repeated quality that is not in contradiction with others. Some of Dickens's crialnfctcrs' are represented by a dominajitphysical characjenstic, such a I proto-detective Mr. Bucket in Bleak House: Mr. Bucket and his fat forefinger are much in consultation togethe under existing circumstances. When Mr. Bucket has a matter of thi pressing interest under his consideration, the fat forefinger seems tn rise, to the dignity of a familiar demon. He puts it to his ears, and it whispers information; he puts it to his lips, and it enjoins him to secrecy; he rubs it over his nose, and it sharpens his scent; h shakes it before a guilty man, and it charms him to his destruction The Augurs of the Detective Temple invariably predict that when Mr. Bucket and that finger are in much conference, a terrible avenge-will be heard of before long. Otherwise mildly studious in his observation of human nature, ol the whole a benignant philosopher not disposed to be severe upon the follies of mankind, Mr. Bucket pervades a vast number of house and strolls about an infinity of streets, to outward appearance rather languishing for want of an object. He is in the friendliest condition towards his species and will drink with most of them. He is free with his money, affable in his manners, innocent in his conversation - but through the placid stream of his life there glides an under-current cr forefinger.20 What is noticeable, however, is how unnatural and Hwcinematic thi seems: one cannot film 'an undercurrent of forefinger'. Dickens's memo! fllre complex than at first appears, since the forefinger has become ^Klturalised, and used to signify qualities that cannot easily be recon-||d_to the signifier itself. Another reason why it appears stranger than ■pawninglv comparable method of the screenplay, however, is simply Tffllse in the latter, bodily fragmentation is so ubiquitous as to have Blcfrie naturalised, whereas in prose fiction it represents a conscious Tp|;eemingly perverse choice on the part of the author. WE. concentration on the eyes is a staple of film theory and criticism: Bm|§the commonplace observation that the eyes are 'the windows of ■Sfil>ui\ and therefore especially revealing of character, to the devel-Illlnent of the 'eyeline match' and the need to avoid the direct look gfjt&the camera as principles of continuity editing, to more theoretical jglfebraBons of the ways in which the 'eye' of the camera dramatises or llpfbilises the interaction of spectator and screen. Hitchcock's films give offered particularly fruitful illustrations: one thinks, for example, f~°6f the extraordinary crane shot that closes in on the eyes of the killer ffiVmtng and Innocent (1937), the dead eye of Marion Crane on the !.jl'..wer floor in Psycho (I960), or Norman Bates's unnerving stare into sahe rcamera at the end of the same film. 3 S-T While the eyes may have a privileged status, the fragmentation of th ■ body in general became almost a necessary conditipji^^Lcinema ■" ,'riie technological advances and innovations in~editing in the early ^900s had allowed directors to dispense with the theatrical framing of febe body in long-shot as the usual means of shooting character. Today, ^ntire genres - the horror film, pornography, any post-watershed cop ilshow with a wisecracking pathologist - exist partly-to display the body Ktajneces. These are particular illustrations of the general ontology of B^nlm. Movies are almost compelled to cut up the body via close-ups and editing, although some will do so more self-consciously than others, lip-some will use the body part as a persistent signifier of character or |||tWation: the hands in Pickpocket (Robert Bresson, 1959), the nose in Wftjiatown (Roman Polanski, 1974). Bin-exploring the relationship of the fragment to the whole, this Ipcreenwriting method recalls the technirafand psychoanalytical analy-■ps of 'suture' - the stitching together of disparate shots in the continu-Kty system to create an ideological effect of seamlessness - which has jpngbeen a staple of post-structuralist and Lacanian film theory. From pijs perspective, the film text can be deconstructed into its constituent Pfrhents to show that cinema, while offering an illusion of wholeness, [never entirely succeeds in repressing its scandalous revelation that the j§fpi'an subject is not individual, indivisible, complete, but instead ft 130 The Screenplay The Scene Text 131 decentred, incomplete, lacking. Cinema, then, may be a representati n of the'mirror stage', that moment when a mother holds a baby before a mirror and pronounces its name. For the child, this is a profouft |:» ambivalent event: the previously involuntary motor functions of hands"' and feet now appear to be the movements of a complete, individual-self, its identity confirmed by its possession of a name] yet that self is" revealed to be separate from the mother who confers the name, End the figure in the mirror is itself illusory, a representation of the self that * perceives it. And this is, perhaps, what 'character' and 'identificati i mean in the cinema: the spectator temporarily loses the sense of s |. possession, and becomes caught up - or 'stitched up', via the effects of suture - in the world of a protagonist who problematically repress t the viewer without being identical to him or her. This process is unique to the experience of cinema spectatorship, and has no direct analogy in the screenplay> which nevertheless seems ' to prefigure it through the process of bodilyjfragmentation. Instead -gaining an illusion of wholeness, the reader oscillates between expe- i riencing the visible character as an accumulation of body parts air as a rough sketch of a figul:encontafnlng minimal signifying detail This sense"oT'fra^mentation need not be confined to the visible. The* multiply authored screenplay, depending on its stage of development, will often include contributions from writers brought in to change OT^dd jo an individual role, rjgrhaps to accommodate the wishes -requirements of a particularfactor. For this reason or otherwise, 11 is ' not difficult to think of roles that have been supplied with what mi; 11 be termed 'personality' rather .thjnj^hjirjcter'. Tersojiajity^jconl a sense of individuality by^means of non-essfintial attributes (Nice ■ Cage's Beatles obsession in The Rock [Michael Bay, 1996], for example)," instead of subordinating the character to its structuraLfcmctionvwithin fJlEpiQt^1 It is with character that creative writing classes and screenwriting manuals on the one hand, and literary criticism on the other, dive ■ most sharply. The former tend to promote 'naive' thinking: that is, I. -practical purposes they encourage the reader and writer to think of the. s characters and the story world as 'real'. This has been outmoded irr literary criticism at least since the 1920s, and some of the most important screenwriters (such as Mamet) and screenwriting gurus (Rob * McKee) have explicitly rejected it in favour of seeing character as_huth ^ a textual construct, and a concept that is meaningjul only when i he individual^character is seen in relation to the structure of the screen--play as a whole. It is what enables Mamentoirfgue''tiiat '[t]herc is no laracter. There are only lines upon a page'.22 Whatever the creative ■[vantages of naive thinking, then, the analysis of screenplays as texts should insist on the critical distinction between writing a screenplay ^ reading it. Ijarnet describes the task of the writer as beginning with the creation tjf^ 'logical structure', after which 'the ego of the structuralist hands the outline to the id, who will write the dialogue'.23 From this point of view, fo speak of a 'character' as an individual would be misleading, because a properly structuralist analysis the character has no essence -uj'positive' terms - but gains its meaning only from how it is posi-loned within a set of relationships. A less purist approach might see creative contradiction in screenplays such as Taxi Driver that make a Ebeavy investment in individuality. On the one hand, the character is 1o be seen as an autonomous person with the capacity for choice: life j roal-oriented, and redemption is available. On the other hand, he is "a function of the structure of the screenplay, which maps out his life |por him. r. However we view this question, character is inseparable from the Picturing role that is generally argued to be the screenplay's primary iction. More broadly, then, and to borrow Rick Altaian's terms in his analysis of film genres, we may see Sternberg's 'elements' as local, seman-H'propcrties of the screenplay text, but to understand fully how screen-Says operate we have to understand their syntactic organisation.24 g^ructure and structuralism %The Art of the Moving Picture (1915, rev. 1922), and The Photoplay: ^Psychological Study (1916), Vachel Lindsay and Hugo Munsterberg, Ilspectively, argued in different ways that film was a visual medium, p^iereas literature and drama are linguistic. Therefore, language should |lay no part in the ideal film, and a scenario must be 'entirely imperfect rand becomes a complete work of art only through the actions of the 3irector]'.25 Such arguments imply that the literary writer is concerned 0ly with the aesthetic effects of words in combination, and that Prama is merely dialogue. This overlooks the structuring force both of llpe dramatic text, and of the scenario in silent film in particular, and in "Kuiema in general. More perceptive in this regard is Victor O. Freeburg's // Art of Photoplay Making (1918), which explores film as a synaesthetic iri' lium and recognises the effects of time, fluidity, and arrangement, j^b &fl-of which imply the writer's structuring role. Freeburg thereby Spticipates some of the discoveries of Soviet montage, and it is perhaps 132 The Screenplay The Scene Text 133 significant that both Vsevolod Pudovkin (in The Film Scenario and.h Theory [1928]) and Sergei Eisenstein were to write trendiaritiy"on function of the scenario. Eisenstein puts it simply: 'the basic and ch, task of the shooting-script is in forming that compositional spine albn$» which must move the development of the action, the composition the episodes and the arrangement of their elements'.26 While this is arguably the major function of the screenplay, it pro."; vides one more explanation for its critical marginalisation, since n favours story structure over enunciation (the particular qualities ari^ choice of words j^hjitjiir^^ literary texts). Moreover, a reader cannot simply point tostrucUue but, instead, hailo infer it or construct* it, usually retrospectively, since it is often only at the end of a screenplays that its shape becomes entirely clear. In this way the screenplay exern-* plifies at a purely structural level the temporal dynamics of anticipatir n re-evaluation, and retrospection emphasised jn literary reader-respornr theory. Structuralism has always been most effective when used to anabse^ a large corpus of texts, especially those which are 'unliterary'. Liteťaní criticism, by contrast, tends to privilege the individual, the different, the.-unique; indeed, it is arguably precisely these qualities, often combine^ with ideas of stylistic complexity and self-reflexivity, that constitu -literature itself. It is no accident that one of the most influent I structuralist analyses, Roland Barthes' 'Introduction to the Structui I Analysis of Narrative', used the James Bond novels to illustrate a stru turalist methodology.27 Barthes's predecessors include Vladimir Propp's" Morphology of the Folktale (1928) and Joseph Campbell's The Hero with" a Thousand Faces (1949), studies that seek to uncover the pattern - the'* 'monomyth', in Campbell's revealing word - connecting an enormous]? range of fairy tales and myths, respectively. It is not implausible to regard the story departments of the maiotf MtJ Hollywood studig_ias possessing an acutely structuralist sensibility Ion-' before even Propp's investigations. From the beginning, Hollywo ' «^ was developing a story-gathering organisation and analysing the results-generically, and soon began the systematic combing of the world story; 6^ market. America was producing two thousand five hundred films a year", by 1910, six thousand five hundred by 1915, and with the Weste i"" European powers crippled by war, 'by 1917 the American industry^ was making nearly all the world's motion pictures'.28 The producer Dore Schary reported that in the_1940s the readers at Loew's offices hit NewJ&nJcJJaris and LonckmjaiDjjldJb£tween themjpjoyidg synopses almost 25.000 items per year; of these, just thirty to fifty would go intc gjion.29 From all those synopses the producers werej.ooking for just _things: First of all a story must be'for us': it must fit our program, permit practical casting, and generally be ready to go. But it must always have wide appeal to all kinds of people, it must be adaptable to visual felling, contain fresh pictorial elements to satisfy the audience eye, must be built around strong and intriguing characters (preferably with a-good part for one of our contractors), permit telling on the screen [n not much more than ninety minutes, be non-topical enough not to 'date' before we get our investment back. And it must sparkle with enough of that intangible showmanship.30 ollywood also shows parallels with structuralist thinking in its approach story development. Michael Hauge's popular screehwriting manual Hfgues that a 'story jdea ... can be expressed in a single sentence: It is a Jorv about a-[character] who-[action]'.31 On reason for this is crisply explained by one of the Hollywood producers in Mamet's stage satire Speed-the-Plow: 'You can't tell it to me in 5-ine sentence, they can't put it in TV Guide'.32 Yet the idea that a text, jjgr body_of texts,, is structured like a language is classically structuralist, gauge's sentence has both a linear (in structuralist terms, syntagmatic) teds, and a vertical (paradigmatic) axis. The linear axis provides the story development; the vertical axis allows for the substitution of different Sclaracters and actions. Such a model can very rapidly generate enormous ipnbers of 'different' stories. his analysis of the recurrent structural forms of the folk tale, Propp not speak of character in the ways that a traditional, humanist liter-critic would; instead he speaks of a common structure to the tales, ph of which consists of a selection of thirty-one possible 'functions', per-|p;med in an invariable sequence by the dramatis personae, who occupy Ken 'spheres of action' (villain, donor, helper, princess, dispatcher, hero, pd false hero). In an early example of the practical application of this pdel to a cinematic genre, Will Wright offered a 'liberalized version' of Sqpp's methodology. He incorporated 'attributes' as well as 'functions' into his analysis of the Western, noted the distinction between simple |pd collectively retold folk tales and the complex individual film text, Ijjjidifound 'unnecessarily restricting' Propp's insistence on an unvarying pjuence of actions.33 pThe structuralist model has certain advantages as an analytical tool ttCthe present context. It is very clear; applicable to both adapted and 5 134 The Screenplay 'tupluj 'Ví T if ^ original screenplays; helps to account for the recurrence of narratiyp* paradigms across different periods, cultures, and media; suggests i , .1 even most art-hóuše Iflms operate accordingly more codified genmo demands than is the case with "literary" fiction; and shows_Jaow_the-* individualscreenplay isintertextually related to a larg^_number~of-^ others. SenšTiTvely applied^ it can provide a particularly conviňcřngt demonstration of the internal structuring mechanisms of the individual-' screenplay. And it also helps to differentiate the screenplay from ;|n film text: it is the latter that challenges the system of the screenplay by inescapably introducing the structurally redundant signifies ijf^ the actor's appearance and performance, for example, and the gei n r tj serendipity of production. The primary theoretical weakness of classical? structuralism is that it has an unwarranted confidence in the stability of^j the system, as if stories were chess games that may have infinite number -but that all obey the rules of a game confined to sixty-four squares *\s~ an analytical tool, it is universally applicable - any narrative film can be expressed within Hauge's sentence or Campbell's monomyth - yet fotil this reason, lacks discriminatory power. Most important from the present perspective, however, is that exam-ining the screenplay as a self-reflexive structure problematises analysis! that breaks it djoym jntgjtsjx^nsti^ The meaning óTěach'^ aspect of the text is bound up with other, answering sjgns^JRosibuď 9hiffigesjts_me^mgr-.-t^ mentary on the nature^ of the_character, an individual scene acquires its jneaning through its position within farger sequences, and so on, "t| Consequerrtíyjth^a^m^ make its own sense within its own stoicmre, even though this verbal text will also be read in relationjo an external,.cinematic sign-system, so that its fragmentation into disciete elements_suitable for reading by individual professionals in no way prohibits the reading of it as alext like any other. "he Dialogue Text pulogue in film has received very little attention in comparison to Ithe technical and theoretical sophistication of image-based studies of uiierna. Those attempting to establish the credentials of film as an jFjit form have tended to emphasise its medium-specific qualities: in particular, the expressive possibilities unleashed by the editorial juxtaposition of moving images in a linear sequence. From this perspective, £Íhé>introduction of sound in the late 1920s represents a retrograde step ecause it arrested the camera's freedom of movement and compro-lised the integrity of the medium, although Busby Berkeley's work for Warner Brothers amply demonstrates that the technical difficulties of ^marrying sound to the moving camera had largely been eliminated by 1933. Moreover, 'silent' movies had almost always had some form of laural accompaniment, from the commentary of early exhibitors to the ^lear-ubiquitous use of a musical score, improvised or otherwise. \s Mary Deveraux observes, '[t]he first sound film, The Jazz Singer, ^brought not sound but a new kind of sound ... [t]he real change brought labput by synchronization was speech'.1 It is dialogue specificalj^rather Cfhan sound in general, thatjpreoccupjedmuch subsequent analysis of the ^médium. Uevereaux surveyša rangeoínEFiěbreticians and přačlíítloněrs, íífřořn Alexander Dovzhenko to René Clair to Charlie Chaplin, to show -that there was a 'split conception of sound' in which the ideal was 'a _wordless cinema, not a soundless one'.2 For example, the theoreticians of Soviet montage, including Eisenstein and Pudovkin, were excited bvihe possibilities of counterpointing sound and image; but the problem with the voice specifically, as far as Eisenstein was concerned, was rat it presented a kind of rhythmic tautology, since (in Devereaux's liimmation) 'the sound of human speech exactly correspond [ed] to a phot of a man talking'.3 135 136 The Screenplay The Dialogue Text 137 The theoretical foundations of this position are perhaps mosf^H influentially expressed in Rudolf Arnheim's Film as Art, first publish^ r in German in 1933, significantly just after sound had eliminate '',51 silent film production. Devereaux shows that Arnheim's objections u dialogue derive from an aesthetic and philosophical essentialism, whicjC holds that artistic value is inextricable from the materials peculiar to^S each medium. He is therefore obliged to enforce the boundaries that"! separate film from other arts, one consequence being that the sound-fL film, which utilises a form of speech with theatrical antecedents, rrm i be dismissed as (in Devereaux's word) a 'mongrel'.4 Hence Arnheirr insistence that pantomime of the Chaplin variety was preferable 1 ■ speech as a medium of human communication in cinema. In tl slightly more liberal and equally influential view of Siegfried KracatKr, '[a] 11 the successful attempts at an integration of the spoken word ha*i one characteristic in common: they play down dialogue with a view i i reinstating the visuals'.5 Although this hierarchical conception of film is still dominant-irt many areas of film study, Devereaux's conclusion that Arnheim 'refusfe j to see film as a continually evolving art form' and 'elevates the practic of a particular moment in film history to the principles of film art expresses an increasingly widespread view. As Noel Carroll observes^! , object to sound films on the basis that they are theatrical is illogical: th specificity thesis itself shows that they are distinct. On the other hand, *?l if one believes that the one can contaminate the other, then neither can" in fact be unique and self-contained, and the specificity thesis falls. Th plain conclusion is that art forms tend to be both more hybrid and mo varied in their applicability than the 'specificity thesis' can concede The only result to be expected from creating a hierarchy of channels i communication within a medium as synaesthetic as cinema is a cane i in which certain films will be excluded purely because they fail to meet a narrowly restrictive set of criteria. As Devereaux, Claudia Sternber and Sarah Kozloff all point out, certain genres are almost unthinkab without dialogue, while many others possess distinctively genre-specif ■ modes of speech, as the second half of Kozloff's Overhearing Bl Dialogue demonstrates in its analysis of westerns, screwball comedie gangster films, and melodramas. Although the prominence^ofjjie specificity thesis in film_jtadi; helps to^explain the scant critical attention t&^ci&ejrwriting dialogu-even those scholars who~have~aTt:empted to establish the screenpl; as_a serk^^far'nTbrJwfjting have tended (^iffier to accord dialogue . relarivel_y_jnajr^nal sta^uV~oT~ttrnave diSinguishedlt insufficient Mfim stage dialogue, everyday conversation, or the film actor's vocal *Jpvery. Kevin Boon's chapter on 'dialogue as action' in Script Culture jfltf the American Screenplay, for example, is inexplicably devoted to Kgcene from Glengarry Glen Ross, which David Mamet's screenplay ^joduces almost verbatim from the same writer's original play for the Efcge. Consequently, Boon's analysis of the screenplay dialogue might oprequal effect be applied to Mamet's published play text, and Boon nature of their conversational interaction, the language peculiar to indi/| vidual speakers, tfie~use of foreign languages^jdialects,and jargon, andS the-pattetns of dialegoe-wl^Mi-«idMduat-fllms. The critical distinctions outlined above offer a range of possiblg approaches to the analysis of screenplay dialogue. We shall now turrf» fbncrete examples, beginning with two kinds of dialogue commonly flysed in theatre plays. cis and offstage space fj£ of the most important functions of stage dialogue is deixis. This She set of signs that indicates relationships between speakers and igveen the speaker and the surrounding, on-stage space. It includes fjsonal pronouns such as 'I' and 'you', adverbs of place and time such here' and 'now', and demonstrative pronouns such as 'this' and at'. By implication, too, the definition of the _spatial limitations of ^"onstage space helpsto define its own relationship to the offstage forld. Because deixis is almost unavoidable in the playwright's task of jfablishing relationships of space and time in the theatre, it is argu-j|y 'the most significant linguistic feature - hnth stgtisjjr^y_anrl Itetionallv - in the drama'. It has been argued, for instance, that even Such a highly poetic and conceptual play as Hamlet, more than 5,000 §b'of 29,000 words are deictic.28 IS-well as establishing these on-stage relationships, dramatic dia-gue ordinarily does far more work than film dialoguejn creating an ^agmatTveJink~b^t^e^n the sCeTTe" tnat is pTesentedlo the spectator, rip'offstage or off-screen space. To cite only the most obvious example, ff£0H.^ gff, gtheatre audience of a play by Harold Pinter is wholly reliant on the >U% Saracters for information about the world beyond the room. As Pfister r^.^* marks, '[t]his sort of semantic interpretation of the contrast between ) ftenor and exterior space is particularly common in.-modem dramas jjjtten under the intellectual auspices of existentialism'.29 There is an iistent pressure on the Pinter character to justify his or her existence iithe dramatic here and now; appeals to whatever may be happen-gor may have happened outside the room, in the past, are to be eared with suspicion. When Pinter adapted The Caretaker for the 1963 version directed by Clive Donner, he created several new exte-|>l scenes that, in cinematic fashion, 'opened out' the action. While Igely wordless, these exterior scenes significantly alter the ontological pus of the interior episodes. Combined with a number of cuts to the |gthier monologues, and some additional new writing, they make | screenplay of The Caretaker a substantially different text to the stage r'|sion.30 "It was argued early in the history of film criticism that the restricts of time and place confronting the dramatist make writing for the latre a more exacting discipline than writing for the screen.31 This is 142 The Screenplay The Dialogue Text 143 possibly another contributory factor in the general evaluation of thi two forms, since the essentially functional and expository demands <5a deictic dialogue are significantly reduced in film. Indications of camera movement, close-ups, establishing shots, and easy cutting betweeflT locations separate in space and time - for example in the now clichQjl use of the expository montage sequence - are merely some of the iriosli obvious illustrations of the screenplay's capacity to provide alternati^hj to deixis...and verbal pjejentation_of off-screen space. Moreover, thdJH comparative brevity of sc_enes-jfinables the writer and directC£jo_rgftlril atjwill to situations that in drama must be developed contin at greater_length. The differences between screenplays, films, and stage plays in theirs treatment of deixis and space, and the potential for confusion betweeju them, are well illustrated by a consideration of the screenplays for two? different versionsof Shakespeare's Henry V, In the introduction to'theS published screenplay of the 1944 version, Laurence Olivier described^ his Henry V as 'perhaps, the first serious attempt to make a dulW Shakespearian film'. In this, Olivier felt that he was simply exploiting,® notable quality of the plays themselves. 'Shakespeare, in a way, "wiofi3p for the films'" by 'splitting up ... the action into a multitude of sma'** scenes', while 'more than one of his plays seems to chafe against cramping restrictions of the stage'.32 Certainly, in Henry V Olivier exploits the space-time fluidity of filrj He at first attempts a reproduction of Shakespearean staging, by havir the Chorus speak within the confines of the Globe theatre. Then, bej1 ning with the Prologue to Act II, the camera dissolves the stage walls by moving from the Globe to an obviously theatrical-looking ship that nevertheless is not contained within the confines of the stage, belori moving to scenes that are clearly not to be regarded as being played front of the theatre audience seen at the beginning. Yet it is not quits accurate to say with Olivier that '[fjrom the very beginning the plays gests a film'.33 On the contrary, the play is unique in the Shakespear in canon in the degree to which it insists from the beginning that this i play and nothing but a play, as the Prologue explicates with exceptie n il richness the deictic problem ofjising_s^g^space_to represent scene that are irnaginatively present yet physically absent. ^Kenneth Branagh also saw the play as 'tremendously "filmic"', biS his version was in part constructed in conscious opposition to Olivier^ 'nationalistic and militaristic' wartime production. Instead, Bramgh was excited by the prospect of using 'close-ups and low-level dialo to draw the audience deep into the human side of this distant medu i Sid'. His original intention was to have the Chorus begin to speak §a?disused theatre before 'throwing open scenery doors to allow the ?n;era to travel outside and into the "real" world of our film'.34 Early £tfre writing process, however, the decision was taken to situate lelGhorus 'in a deserted film studio' with 'a semi-constructed set'.35 Kssibly to accentuate the cinematic effect Branagh cut lines 19-27, in 10 the Chorus, conceding the necessary limitations of stage repre-|n*ation, asks the audience to 'Piece out our imperfections with your flights' by imagining that the 'monarchies' of England and France pxitained 'within the girdle of these walls'. Like Olivier, Branagh jilt ttiat the Prologue 'can be interpreted as alluding to the mystery and pagination employed in the medium of film'.36 Set however impressive the respective films are in accommodating ke Shakespearean text to the demands of cinema, they nevertheless liain bound by the essentially theatrical deixis. Russell Jackson has leently noted, in a discussion of Shakespeare on film, that Elizabethan 1 p s may resemble cinematic adaptations of theatrical texts in the ways ^hiich tTif^'open ou^ and time'_are aurally indicated simply by 'stateme!SsTn~the dialogue'.37 Henry V Eipels both Branagh and Olivier foTrnd a space - a theatre or a film "{equivalent to that in which the Chorus speaks, and to preserve, j|j only very minor cuts, the rousing words that establish a spatial ppll as temporal connection between the Chorus and the audience, issin turn is provoked by a desire to preserve a kind of authenticity Drrall the radical cutting of the text later in the screenplay, Branagh rted the film to remain 'Shakespearean in spirit'38) doubly prompted ^traditional notions of adaptation and by the pre-eminent place of (cespeare within the literary canon. A much more radical approach, ncult to visualise in film but perhaps attempted by Peter Greenaway i*lkospero's Books (1991), might have been an attempt to realise in ■> Striatic terms the insight of director Peter Brook, who once declared h »i i he power of a Shakespeare play on stage stems from the fact that i ippens "nowhere"'.39 Is; wlech acts bureaux and Kozloff reject any assumption that speech stands Fposed to action; speech itself is action. Drawing on the work of lour Chatman and other theorists of literary narrative, Kozloff i* -that dialogue itself can itself often be a kev story event, as in ^disclosure of a secret or a decIaraticui^fJLose^She also enumerates 144 The Screenplay several different kinds of conversational interaction, noting that lr/« ^ct;.w each case the effect depends on the dramatic context, and th^ decree: to which the speaker is successful injecunnjjjhejm^ onscreen lfsTielieTanaTKFoS^g For example, j?Hir>tica£ dTilo^uTmay signal to the audience that the characters are intimate* and part of our interactive engagement with the film will lie in truno to penetrate or decode a private language. Alternatively, the charai r may misunderstand one another, leading to 'dialogues of the d it The progress or interruption of dialogue can also reveal or change the-nature of a relationship, as in overlappirig dialogue, the deployment" of 'tag questions', or the silencing of a character, including by the use of 'toppers' (killer lines that attempt tosh^jdown a conversation and ?' often o^Kluc(e]a^«rie)r ~~ Studying dialogue in this way naturally leads both Kozloff j Devereaux tomention speech-act theory, a philosophy of languages ?f f^ developed by J. L. Austin. Austin began by proposing a distinction-* ^fo"'3 between 'constative' (proposition-bearing) utterances, and 'perfoim -tive' utterances in which 'the issuing of the utterance is the performing^ of an action'. He eventually concluded that the opposition was false,-! since 'stating is performing an act ... It is essential to realize that "true^. andj^fjliel^likej^'free^nd "unfree", do_not standj^ at all; but only for a generaTdimenlsioj^^ to say asc^p"^ ence, for these purposes and with these intentions'.40 it is easy to see why speech-act theory has proved to be a productive™ method of analysing drama. As Andrew K. Kennedy observes, 'the names given by Austin and other philosophers to "the speech act", and 2 to "performative" utterances points to their relevance to both converse"]* tion and to dramatic/theatrical performance'.41 Austin E. Quiglev I- r example, brilliantly clarifies the dialogue of Pinter's plays by recognis-~ ing that apparent contradictions and uncertainties about facts, and_^ about the past, are really the result of the characters' attempts 'to ne^o- ~ tiate a mutual reality'. This challengesjliejeferential theory of meaning^ as regards not only facts^bufalso^personality', which is notjndiv _h i _ (ojTndiyisible), but instead 'tsliTunction ola compromise negotiated In." 7 a particular relationship'.42 Although speeqh-act theory can be relevant tQ_the,di.sciission of .din. dialogue, its range of application is much more restricted._It does not ? adequately serve the various kinds of dialogue detailed later in this^ chapter, principally because dialogic exchanges in cinema tend to to much shorter than in theatre. It is revealing that Boon, lnlTcTiapfi The Dialogue Text 145 Heiigned to elucidate the quality of film dialogue, selects a scene from f%eti? a staSe Play transposed to the screen by the same Suiter; David Mamet. In the scene in question, one of the real estate jljllesuien, Moss, persuades his colleague, Aaronow, to participate in a Bobbery: ^ARONOW: I mean are you actually talking about this, or are we fust... *RI0SS: No, we're just ... AARONOW: We're just 'talking1 about it. fMOSS: We're just speaking about it.43 Ijach character is self-consciously aware of using language to create relationships, in the literal sense of being particularly interested in defining precisely what words like 'talking', 'speaking', and 'saying' mean. This ■jspart of a verbal negotiation of a contract by which .they attempt to tgjablish precisely the rules according to which the discussion is to be pbnducted. In the conversation above it appears that Moss has estab-shed a fine linguistic distinction, in which 'talking' is serious business jPiile 'speaking' is merely idle or hypothetical banter. But this turns Kilt not to be so at all: Moss almost immediately reassures Aaronow th ii 'We're just talking1, thereby setting up an opposition not between |foiking' and 'speaking' but between 'talking' and 'talking1. It soon transpires that even this remodelled distinction is of no use to Aaronow, Mho is startled to discover that 'we sat down to eat dinner, and here I'm \tnminaleven though T thought that we were only talking'.44 aronow has been duped not just by the rule-governed nature of jituli ^ue, but by what the Russian linguist Roman Jakobson called •th- ■■hatic^ or 'contact' functionjjfjrerbal communication. These are jmessages serymgjjrlmarily to establish, to prolongJ^orip_discontinue nrmmjcatiojri,jo_ check whefeerthe channel works', and which^may splayed by a profuse exchange ofjjtualized formulas, by entire glialogues with the mere purport of prolonging communication'.45 ir -now repeatedly checks with Moss to confirm that the channel is forking, and that the rulgj_arejclear. but Moss in fact has effectively |evered the channel and made up the rules to suit himself. I The possibility of theatrical _simuIation challenges the very idea |>f "successful" pexfojematives, which depenorloTlheir effect orT a I 4w u% distinction between the genuine and the counterfeit. In fact, Austin's 1 vation of 'infelicities' acknowledges the possibility of a mimetic, iincere replication of_a^peech,.act; and 'infelicity is an ill to which all 146 The Screenplay The Dialogue Text 147 acts are heir which have the general character of a ritualjorcerem .-n ,| all conventionsgets'.46 The danger is that the conventional procedure/^ which'coristitute'the successful performance of an illocutionarv ac by themselves eliminate the possibilityof esta^sjiing^the smcenwS of the person who performslTiem, and Kennedy is~c^rtamly righTto argue that '"Sincerity" can seldom be taken for grantgdjn^raniatii3 djalogue'.47 The same is true oUlm_dMsgiIes, but since they ^&^ordinarily briefer than those of stage plays, the opportunities for trdf.^ jj it ing the establishment,jmaffi ^^relationship are relatively liinUedT^ven'Maniet's screenplay adapting tion of Glengarry Glen Ross follows the familiar cinematic method ofS i breaking up some of the lengthier duologues by repeatedly. lnterculS ' ting, between two scenes, each of. whichjjs self-contained ln_the stafo play version. Polyfunctionality Of course, neither dramatic nor film dialogue is restricted to the estalt§ lishment of personal and spatial relationships. Pfister observes tha| dnmiatic_Jari£uji^^ and distinguishes six kindsfl lf(,'5v, 'referential' (as when a character gives a report of events that happen* 1 ^^'off^tageT^n the past); ^q^ressiye' (the character reveals informaticn_ about his or her thoughts or emotions, either to another character caT* in the case of soliloquies, to the self and the audience); 'appellative' ( when a character addresses another in an attempt to influence or pfiijjjS suade - essentially, this is a speech-act function); phatic,- 'metalinguj 1' (a variant of the phatic in which the code itself becomes the obje. of discussion: the dialogue about 'talking' and 'speaking' in Glengwy-* Glen Ross provides an excellent illustration); and 'poetic', which refer to an 'external communication system and not to the communin tion processes taking place between the various figures'. An examrjlj is Shakespearean blank verse, which must be addressing the external^ but not the internal communication system, since 'if the opposite we true, the figures would presumably express their astonishment 'it rhi«" "unnatural" manner of speaking'.48 & A given utterance may possess more than one of these tun kiWjand all of them may be used in film as well as in stage dialogu v difference lies in the proportion of speecxx-thal-belongs to each * i* fiction. These proporti jenre: a highly jealisti dramTsuch as Nil by Mouth will display a preponderance of appj and_r>haji£sp^ech as the characters struggle to maintain their pi i ptionships, while a Marx Brotherscojriedy will make great play with ^ Tnetalingual and phatic functions. As Kozloff observes, however, im dLali3gue_generally minimises the jgtotic function.49 Post-Pinter, |na for reasons discussed more fully below, the expressive function jg come to be treated with great suspicion by many writers for both and film. The expressive and referential functions are instead Spies of television soap operas, operating as a kind'of short-hand for jaracter development and action respectively, in a genre in which gnificant quantities of drama have to be written and filmed on a daily basis. ^ .Vhile all of these linguistic functions are equally available to stage id screenwriters, the greater visual flexibility of cinema means that fibgue ten^d^oj^jnore "compressed in the screenplay. The resources oLediting and cajnera enable the film director 'to selec^^mpjiasize, ii jdgrciit, distract, reveal, or deform the filmgoer's interpretation', while Jthe phenomenologicanDss^njD^^ space and tlityallows fhe~spectators' cathexis with the characters more free jjl aft^- this also impacts on We"pfopbrtional distribution of the func-ihs. Broadly speaking, film writing tends to take advantage of the greased opportunities for visual representation to minhnise certain ads ofjliajogue. Scenic representation substitutes for the referential tfaction; a gog^s^rjenplayJs_Jikely to be deeply suspicious jafreh/ing ^the expressive function to exhibit much^ruth-value in character fraction; and^cTiaract^feTationships may be developed by means Vt'j\sv/ Dt metonyjnic visual represen^tatidTraM""'fcenlc_ juxtaposition, with a onsequently less^r_proportion of appellative andph¥fl&3TaioguTthan sxommonly found in theatre. Gertain gangster movies, however, especially those of the post-Pinter 14 have made great play of the sense of threat that can be generated 'the phatic function, as in the unforgettable conversation about athburgers in Pulp Fiction: JULES: Looks like me and Vincent caught you boys at break- "ft, fast. Sorry 'bout that. What'cha eatin'? 8|RETT: Hamburgers. JULES: Hamburgers. The cornerstone of any nutritious break- s' fast. What kinda hamburgers? ■BRETT: Cheeseburgers. JULES: No, no, no, no, no. I mean where did you get 'em? L McDonald's, Wendy's, Jack-in-the-Box, where? BRETT: Big Kahuna Burger. "^'•■r^iiAs, 148 The Screenplay \Ji JULES: Big Kahuna Burger. That's that Hawaiian bui &W& I heard they got some tasty burgers.51 .."Jj The dialogue appears disproportionate in three different \ . thai prolonged examination of trivial topics (Tarantino announced hini$eg| to the world in the conversation about Madonna at the beginning o||$l Reserwir Dogs [1992]), the imbalance between this verbal fri\olit\ arigT" the dramatic situation in which the reader or spectator inters that mfi&M der is imminent, and the quantity of such dialogue in a medium thajja routinely presumed to emphasise the visual. '^WM The dialogue about hamburgers creates a sense of threat, not omm because we have already seen Jules and his partner Vincent | paiffl themselves for violence against Brett and his associates, bu :caiii||| of an effect comparable to that of the extended shot in cin i i. Trj|j theory of suture argues that the reverse shot in classical H I wcH editing exists partly to quell a potential unease. A single shot lniphcifw poses questions: who is looking at this, and why? The reverse I t reJJ9 suringly fills in the empty space that might be occupied by tins hy|||| thetical voyeur (and is in fact occupied by the camera), revealing th|j nothing is there that shouldn't be present in the diegetic woild of fflj film. This creates the illusion that there is no narration; the c its jujj exist, and they are not being shown to us by a mediating agent 4mt The above dialogue in Pulp Fiction creates an effect similur toit^H of an unanswered shot. After a while - the discussion about t fooBJj continues for two pages - the reader is likely either to wond *y_J| much time is being expended on thi^part^ular dialogue (in UiectTtfajj narratoT becolmeTpr^ of whom such question may^[ asHd)TW~wTiraeg^ "' m , attention in itself (fulfilling the ^poetic.', iuncjion). A conversation! i'^. «t wbincTordinarily be regarded as phatic - idle chit-chat as a mean '"c keeping the communicational channels open - is thereforeboth^o hc ajidperformative, since in context it constitutes a form of aj ssid§l Pinter can be credited with first developing the theatnea | si|J| tieToTsuch dialogic forms in what have been termed his ^conucUesj^ menace', but it is arguable that the screenplay routinely pla reatjg emphasis than theatre plays onJhe_£oe^Jtociion. Becau: he 'r|jj istic stage play relies on dialogue to develop character relati i ship||j certain suspension of disbelief is required on the part of th< lciieM Monologues and dialogues are liable to be lengthier and moi ynt^S cally articulate within the internal communication system th..Ln.:i|ffl be expected in 'real life', and constant references to the metalinguisraj The Dialogue Text 149 jq t kjv if Rd poetic functions would break this illusion by making the audience J|fare of its own status within the external communication system. , gec^use^tne screenplay can more ne^aMy_develop such relationships ^t^t ^meanToFvisual repfel;BTtaTicTri7alaTogue more frequently has the Igfect of addressing the externalaudience^as^weU-as, oiTgggrrjrtstead p*%~*f jptthe internal audience. Such cinema-specific verbal phenomena as JETpne-linCT~and the voice-over conseouently tend, to calLattention gfetHgrr^ written; and in this lies much "ig. textual specificity and pleasure .of..the screenplay. Hpfister's poetic function has much in common with the eclectic range Mi possibilities that Kozloff groups under the function of 'aesthetic r^cttOi-t'. As well_as carefully patterned dialogue, she includg77n_ this jgjtegory jokes, irony, and internal storytelling. As the example of Pulp fiction shoWs,"Tiowever, any element of screenplay dialogue can take fh a poetic function simply by virtue of being expressed within such a Bghtly controlled form. "Ijuologues St ■Jbraham Polonsky's script for Force of Evil provides another frequently Kited example of a script in which the dialogue attains a poetic quality, |argely because of its rhythmic cadence. Film noir in general, indeed, Ipnds to be marked by dialogue that^ draws attgjrtion to its own conviction. Partly this is derived from some of the sourcVrioveisTairicrtne jBct that 'hard-boiled' writers frequently gravitated towards Hollywood ffiSlmselves. More importantly, it is because the world view of these iBmsJs of a ruthless existential masculinity that affirms itself in what Jgrrimgway called 'grace under pressure'. This is frequently shown not ~— |jk>hysical action but in the abiUtvto respond to situations of extreme £-(^ lit. gbtiorial intensity with verba! toughness anri samxfmiri, whichJsjjo ^ jpu^d thatitseems not to issue Jfom_wj£hiixJJie situation itself, but j., , JptelcTto^e a comlrientjipon it~by a character possessing an almost j^P "^cric^thic^^ events. " ' ~ " Hor exarnpTeTiiTDow^e Indemnity (Billy Wilder, 1944), Walter Neff calls tiHhe home of a client, Dietrichson, to sell him a renewal on his car ftosurance. Finding that he is out, Neff immediately becomes captivated fiy Dietrichson's wife, Phyllis, and starts flirting with her: -vis U. i Hut IhYLLIS: PEFF: There's a speed limit in this state, Mr. Neff. Forty-five miles an hour. How fast was I going, officer? 150 The Screenplay . PHYLLIS: I'd say about ninety. NEFF: Suppose you get down off your motorcycle and give-; me a ticket. |j| PHYLLIS: Suppose I let you off with a warning this time NEFF: Suppose it doesn't take. j PHYLLIS: Suppose I have to whack you over the knuckle NEFF: Suppose I bust out crying and put my head iin vojj shoulder. Wk PHYLLIS: Suppose you try putting it on my husband's sh i I NEFF: That tears it." The impossibly smooth patterning of the dialogue takes place in PtMer's: 'external communication system', but internally shows the chanuM*^. playing a kind of verbal pojejjnjvhich they must keep raisinglfog? stakes on the same root phrase. This strategy, by whicherotic,„or_\jQM tensiojijsjw^o^^ ' W articulate language, has been anticipated in the preceding scene, to which we shall return later, in which Neff's voice-over hints not oM at the nature of his relationship with Phyllis, but also at the plot tS they will hatch against her husband. The above exchange between im and Phyllis is a flashback inside the frame of Neff's voice-over, and b|f his monologue and the characters' dialogue function as commerifa ies upon the scene, even though the internal communication systf unfolds in the here and now. This has the overall strategic effectjj constructing the characters as possessing a sufficiently extreme degflj of emotional detachment to make their almost whimsical deasior|P kill Dietrichson appear at least aesthetically credible. The one-liner In 2005 the American Film Institute published as part of its centena celebrations its list of the top 100 quotations in the history of Ameri j movies.53 Almost without exception, these were short, pithy one-lin| of the kind that frequently acquire a resonance beyond the filrff which they are first uttered. One-liners often mrnishjhejnost, pron r^nKsigmfier of a film or a^tarTteing^ecycled as a movig^s 'tag linejj ^,1 in the 'Eastwood/Stallone/Schwarzenegger^modeJLol^pJmtatiOT^r ^ cl^tio^ iconic status of_theirpersonae, buF^so^^the trailexwhose_chmjc| pointll^^ And many lines from cinerr have crossed the boundary into broader areas of cultural and politic The Dialogue Text 151 gjjfe, as when an anti-trespass law passed in several American states in h.985 became popularly known as the 'make my day law', after the Spree words uttered in Sudden Impact (Clint Eastwood, 1983) that most comprehensively define the character of Harry Callaghan (Eastwood) in Igie^Dirty Harry' movies. In^2003JheJVnto use of famous SJnes from the movies ma campaign to highliglit~the importance of glgeeTiwriters, In one sense, this gave writers their due as the providers ghjne'-ofTne most pleasurable qualities of films, as any number of ^anthologies of film quotations attests. Yet the Guild's own arbitration |proKdures_foi_screen^redit a much.„.uigte liuivto dialogue.55 The Writers Guild itself, then, tends to de^empTiasise *fhe one quality of screenwriting that might most clearly establish a toiler's individual style, and that most successfully translates from the wriiten text into the cinema. ; Qne reason for this is simply that the nature of much Hollywood yriting in fac^je^c^s_aiithoriaijd^ntn^_with the spoken dialogue feco£iing_an agglomeration of lines jromjTianv j1^ajr^g_snn^^ at another is a consequence of the reduced importance of deixis pjr^ech^tsjn_comp^ Theone-liner jquently offers a ^ardonic or ironic^comment upon a^cenerrather Jan contributing to its dramatic development: the use of such lines to Jpsea scene is ubiquitous in the James Bond movies, for example. As p>phy observes in 'Read My Lips: Notes on the Writing and Speaking |Film Dialogue', the one-liner always has the effect of being recited, |d thereforejm'rfen, rather than of simply being spoken. 'In all the portant scenes in a Bond movie, [Sean] Connery throws a heavily-^gtedjine of dialogue that is either the dry coda or wet cade^iceto gne absurd act of espionage violence. Timing is crucial not in the jjse of dramatic rhythm but in the structural placement of narrative fe.-... [I]n a Bond movie words speak louder than actions because |rds announce action.'56 Although such lines may be deployed for many different reasons in pilous kinds of film, Brophy's analysis indicates that they are predomi-ntly genre-specific. They belong within the category of what Kozloff Is 'toppers^ - they are, literally in the case of many horror films, jterjm&sjftatjermmate tliedialogug, the_scene, ancLoften the verbal jgonent. Bond is like the heroes of many detective, action-hero, and |ence-fiction movies, such as Dr. Who, or Sam Spade in Humphrey |gart's incarnation, who 'surrender themselves to the power of the Itfen by evaporating themselves on-stage and in place manifesting '■ "-J.M.I 152 The Screenplay The Dialogue Text 153 on-screen the presence of the script, of the structural organic ,,t the narrative, of the written word'." In Superman (Richard Doiiner" 1978), Christopher Reeve 'looked graphic while speaking literally, a^ though you could almost see the speech balloons emanating from hjj2^ mouth'.58 In The Terminator (James Cameron, 1984), the eponymous^" robotic antihero has to search his memory bank to find the. most^ appropriate verbal response to a given situation. The database iunc tions as a kind of searchable screenplay, moving the character.: uncf^ arguably Arnold Schwarzenegger) a step further from realism, sine the terminator 'doesn't quote dialogue - he quotes the act of deliver," ing dialogue'.59 The horror film furnishes further striking exampl ,f characters who are largely defined in terms of their mode of deliverir what is transparently scripted speech. In The Shining (Stanley Kubrick; 1980), Jack Torrance (Jack Nicholson) 'is literally possessed by literal-quotations', and 'appears to delight in ironic quotation', while in Nightmare on Elm Street (Wes Craven, 1984), 'Freddy is a blank page.' a cypher [sic] of scripted one-liners, almost to the extent that he uT; only killing innocent children so that he can crack a joke about theit"„ demise'.60 Monologue and internal storytelling >t(. ,){- Part of the appeal of the on£Jine_r is that it has the^effectj^puje^ i<^i^ style. Typically .dehyered^ieadj^^ of violent" emotion or action, it makes the speaker seem un^utterablv 'cool', ~w °f Conversely, 'ejcm^s^iyel^peech, in which the character seemingly provides a moment of verbaWeJi^evelation, is apt to sound ■ :l ^rtjvrvt and suspect. The problem with the expressive function of dialogue-TTv^. fs^ound up with both speech acts and the ontological status ct the % ~T^T~~ screen or stage event as something that always occurs in the dramatjc .J| ~ present. Any statement a speaker may make about himself or herself|| v will always be perceived as an attempt to secure something from IS addressee present within the scene. If the speech is not doing this,\|j -Kw^ it can only be addresj^_tg_the_exte^ in a chmas^yj.ct_ofc'* -V^s-*.. au1±L0rialjgyjosition. ' ■.-if ^ Thi^fecognition is perhaps most strongly associated in the theatre,; ''" ^again, with Pinter. As he famously remarked early in his career, '[t]hejl desire for verification ... is understandable but cannot always be satiny fied. ... A character on the stage who can present no convincing argu?j| ment or information as to his past experience, his present behavioural or his aspirations, nor give a comprehensive analysis of his motives feSST legitimate and as worthy of attention as one who, alarmingly, can 3|:'aJl these things'.61 Holding to this principle .demands eliminating ^osih^iijnch^mg^ny speeches that reveal in earlier experiences prmariyej^ncident that would provide a psychological explanation ;the character's-behaviour. For Mamet^ similarly, any such speech is [iply a technical flaw in the writing,Jbecause it needlessly interrupts gl^ctloiiTrrorder to display f^ling^or _emotion he memorably §misses as the 'death of my kitten' speech,62 or what one of his men-UIS, Sidney Lumet, ridicules as 'the "rubber-ducky" school of drama: ^Someone once took his rubber ducky away from him, and that's why 's a deranged killer'".63 inhere is one major exception to Mamet's otherwise rigorous adher-ice to this rule when Bobby Gold, the secular Jewish detective who is tie protagonist of Homicide (1991), reveals to Chava, a female member bf a Jewish resistance group, his own self-loathing: 'They said I was a ussy, because I was a Jew. Onna' cops, they'd say, send a Jew, mizewell id a broad on the job, send a broad through the door ... All my god-pined life, and 1 listened to it ... uh-huh ...? I was the donkey ... I lallthe "clown" ....'64 It is a noticeably unconvincing speech, however, hd peihaps deliberately so. Gold appears weak at this moment - it is p;-and not Mamet, who is trying to generate an affective response by Sorting to a rubber-ducky monologue - and he is about to discover fat Chava will betray him, leading to the climax in which he is brutally ^abused of the notion that any of his fellow Jews, let alone one who is |o a woman, will be moved to sympathy by his account of being made £feel like a 'pussy'. This is an excruciating 'death ofiny kitten' speech Var excellence, and Gold is duly punished for it. ES^E^X-^iSiSS^eneed not either describe character, place, or JptMlshj^^ ijatirmtfnpj^^ in fe^isjiis^ioxuol pjiejljners, whaTKozioff terms JmemaLsiGrytening' > fednoj_bejxp^essiye, but caninst^dj^ * oerrtary_onjthe^ the most striking examples are the stories '"' "^digressions delivered by Orson Welles in a number of different films. ie story about the sharks in The Lady pom Shanghai (Welles, 1947) is j>ne; another is the unforgettable parting speech of Harry Lime to Holly £artins in The Third Man (Carol Reed, 1949): When you make up your mind, send me a message - I'll meet you Pany place, any time, and when we do meet, old man, it's you I want Hfto see, not the police ... and don't be so gloomy ... After all, it's not 154 The Screenplay The Dialogue Text 155 that awful - you know what the fellow said ... In Italy for thirty vears under the Borgias they had warfare, terror, murder, bloodshed - they produced Michaelangelo, Leonardo da Vinci and the Renaiss , In Switzerland they had brotherly love, five hundred years 0f democracy and peace and what did that produce ...? The cuckoo-clock. So long, Holly.65 It is in the nature of screenplay texts that this, one of the mosf famous speeches in all of cinema, exists in published form onl\ , footnote. The text printed first by.Lorrimer and reprinted by Faber was derived from the shooting script; material deleted in the film is indicated in square brackets, and interpolations are recorded as i. ,r._ notes. Accordingly, the cuckoo clock speech, which was impro ■ j by Welles himself during the filming, quite properly appears onh asf' a note at the foot of the page, and is presumably transcribed fr m , viewing of the film rather than from any textual material supplied by, Welles. To think of the speech as a footnote is also peculiarly appropriate, since it is a marginal comment both within and about the film. It , not further the story; nor does it develop the relationship between Lime and Martins, but instead terminates it in the manner of a classif 'topper'. As such, it comments on the moral world of the story and;,o| Lime himself. Because it is spoken (and written) by Welles, however, 1 has the distinct feeling of being uttered by two different speakeis to two! different audiences: by Lime to Martin also by Welles to the audience outside it, for it is absolutely in keejnnf with^e't^autSMl^aafted anecdotes and stories associated with,-tgfc.| Welles persona both on- and off-screen. The speech has become so well-'*" known that, despite its tangential nature, it is now difficult to think c The Third Man - or, unfortunately, Switzerland - without bringing it t> mind. It is thereby representative of film dialogue in general: marginaj and therefore essential. Voice-over For very similar reasons, Welles is also among the most prominent examples of film-makers who obsessively return to the voice-over Kozloff notes that voice-over^narration^appears mo^^ommprilv^r films made by writer-directors; prominent among these are figure" like BillyJ^jder and"Woody*Allen, and it is significant that as well asji writing and directing, both Allen and Welles usually deliver their oy jff-screen narration. Welles had come to cinema not only from theatre l^t also from radio; 'two-thirds of Welles's finished feature films use v lee-over', and both in his own films and as a narrator in films directed jfe[n others, 'we see clearly the imprint thaj^rad^narration madeon ■Welles, and the influence that his delight in narration has had on the Eistory of American cinema'. As he remarked, 'I know that in theory hlu word is secondary in cinema but the secret of my work is that eve-thing is based on the word. I do not make silent films. I must begin cwith what the characters say'.66 jilm criticism has long regarded voicejOTer_withijuspicion, for rea-yiis (or 'prejudices') that Kozloff helpfully summarises. Many are again ^variations on the ^afjcity^hesis, but the assumption that under-^es most of them is that the presenfcrtira of .1^ pmaHipulative than the often overtly narrational function of language :?grjjo"undtrack. This distinction, as Kozloff demonstrates, is false. Just as Showing' is always just another way of 'telling', so 'ail [voice-over narration] does is superimpose another type of narration on top of a mode Jhat is already at least partly narrative'. Moreover, far from being merely |a clumsy expositional device, or simply redundant, 'all complementary Kairings of narration and images provide more information than would |ave been available from either alone', the result often being an ironic 3terplay_betwegn_the two.67 Rartly to rebut the common charge that voice-over narration is a literary' device that calls attention to writing, Kozloff compares jt ttStead to the onstage narrators sometimes fqund^in ttieatricad jworks, "the grounds that the narration is both spoken and intermittent. The Bfferences, however, are far more significant. In plays such as Tennessee IgUhams's The Glass Menagerie (1944) and Arthur Miller's A View from the fudge (1956), the njirrjrtorjs^ajvis^ The vcic^oyer-narrator ^definition is not, since such narration consists in 'oral statements, anveying any portion of a narrative, spoken by an unseen speaker situ-ped in a space and time other than that being simultaneously presented the images on the screen'.68 The physical presence of the stage nar-Jor is one reason why, from Plautus onwards, s/he is generally the |gcus of the audience's attention while speaking. By definition, s/he ^dresses the audience; whatever else is visible on the stage behind or Side her is of lesser importance during the narration. The cinematic |quivalent is the character who speaks to the camera directly, as in Alfie ewis Gilbert, 1966). In voice-oyer narration, by contrast, the speaker jgJbsent, and audience attention is. divided between "soundtrackand ifsual image. 156 The Screenplay The Dialogue Text 157 The only theatrical analogy that springs readily to mind is the 'Voice' heard in some of the shorter plays of Samuel Beckett. In Footfc, (1976) the sole figure on the stage, a woman called May ('M'), p& to and fro while the voice of another woman ('V'), apparently that. f her mother, speaks 'from dark upstage'; in Rockaby (1980), similarly , woman identified simply as 'W sits in a chair, rocking to the rhythmic accompaniment of her own 'recorded voice'.69 The combination . i precisely calculated movement and the amplified voice of an unst u speaker is highly cinematic, and the versions arranged specially tor videotaping, starring Billie Whitelaw, are extraordinarily powerful. Yet the Voices differ from voice-over narration because they are, however mysteriously, part of the diegetic world inhabited by M and W, whcu1 interact with them, responding to what they say and even command^ ing them to speak. Beckett is most unusual in experimenting with "* the severance of words from action on stage, but the separation of diegetic and non-diegetic worlds is never as absolute as in voice-over narration. The Beckettian Voice is an interlocutor, rather than a naira-tor, and its function illustrates the centrality to Beckett's vision of ihe 'narrator/narrated', with the protagonist's actions seemingly prompted by a voice that appears to issue simultaneously from within and outside the self. Dramatising this perception constitutes the entire action of Film, written in 1963 and the only one of Beckett's works intended directly for i eponymous medium. Rarely described as a screenplay - but that is wj i it is - the five-page 'outline' of the action is equalled in length by thesJj prefatory material and notes, which describe the proposed method ror"g realising cinematically the ontological drama, in which 'the protagonist is sundered into object (O) and eye (E), the former in flight, the lalter in pursuit'. Numerous diagrams show the precise spatial relations!) i between O and E, essential to a film that depends on the conceit tha < > will experience the 'anguish of perceivedness' if E, following behind ■ breaks the 'angle of immunity', which Beckett sets at 45°.70 This bre ■ the illusionistic frame of cinema: the camera becomes the gaze to be avoided. Film looks like no other screenplay before or since. Although a sci ipt for a film that is to be silent save for a single 'sssh!', it is typical ■ screenplays in its struggle with the inadequacy of the word to find ary appropriate textual form in which to represent a complex relationship between the object perceived and a perceiving or narrating ag«m Similarly, Rockaby is prefaced by a diagrammatic representation of I acing of the feet, and Footfalls by extensive directions orchestrating e lighting, which just as much as the Voice appears both to prompt "fhd to be prompted by W. In all three cases, the framing of the visible ietion by extensive textual matter represents an attempt to resolve the difficulties of approximating the duration of the action as the spectator |Tto perceive it, while also indicating for the reader the nature of the lelationshipjpetween visible action and the offstagg_or_off-screen voice. é^s_rel^ in the screenplay jft-compared to the film, and to coj^ař)^ g^^fflm_experiences the'soundtrack and the_image. simultaneously, including of course in the case of a scene accompanied by a voice-'ffver. Disconcertingly, however, voice-over_n^tli^^logue text of a screen play^annotjromfortab^^ do 'Jwe of twojthmgsj^lJher the description may precede the voice-over íf^F^icejversa^jjar the scene text must be presented in one column and ifce dialogue text in ;another. Meither čase..thfiIičenTtext ínsišfs""ťh"at ííieevemsjtecjjjbjd_ajre^mifoldjn^rijhe^present^tense, while simulta-"ueoušty - oi nearly simultaneously - the voice-over casts them into the pSt and installs its^own moment of narration as the present. We are in ^iili^líELí^-í^-y^^^^y' of déjá vu. jBThe difference between this and on-screen narration is well illustrated By the screenplay for Double Indemnity. In the source novel by James M. jgain, Neff writes his confession as a memoir. Wilder and Chandler -prompted, perhaps, by the recognition that voice-over creates a presumption of direct oral transmission71 - instead came up with the Brilliant idea that he would speak his confession into ff dictaphone, fis narration begins after he struggles, wounded, into his office at the Insurance company in dead of night: "He presses the button switch on the horn. The sound stops, the irrecord revolves on the cylinder. He begins to speak: fH NEFF: Office memorandum, Walter Neff to Barton Keyes, Claims Manager. Los Angeles, July 16th, 1938. Dear Keyes: I suppose you'll call this a confession when you hear it. I don't like the word confession. [...] ijp&Ehe confession continues for a whole page, interrupted only by two |pf descriptions of Neff looking at his wounded shoulder and taking a ij§g on a cigarette. The layout then changes: 158 The Screenplay The Dialogue Text 159 DIETRICHSON HOME - LOS FELIZ DISTRICT Palm trees line the street, middle-class houses, mostly in Spanish style. Some kids throwing a baseball back and forth across a couple of front lawns. An ice cream wagon dawdles along the block. Neff's coupe meets and passes the ice cream wagon and stops before one of the Spanish houses. Neff gets out. He carries a briefcase, his hat is a little on the back of his head. His movements are easy and full of ginger. He inspects the house, checks the number, goes up on the front porch and rings the bell. NEFF'S VOICE It was mid-afternoon, and it's funny, I can still remember the smell of honeysuckle all along that block. I felt like a million; There was no way in all this world I could have known thai murder sometimes can smell HI honeysuckle... The double-spacing in the second column indicates an^effortjtOiHi-chronise the delive^y^fjthe lines with the visuals in the first; elsewr in the screenplay both columns are single-spaced, again providing an.^ approximatejndication of timing.72 The screenplay reader, however, isj." presented with not an image and a soundtrack, but with two forms of writing. The necessity of doing this exposes voice-over as a particularly cinematic device, but it also creates a highly unusual, and in a cerlain sense impossible, textual form withinjhe screent^yJLtselfLJUs_as if the eyes wereTbeing" asked to scan_a^coiumn each and then report back. The difference in register between the two columns ..emphasises li differences" between prose^description arid oral narration. More :| ~ cifically, Neff's voice gives not a statement of the action but an inter-* pretation of it. Objections to voice-over tend to state or imply that it introduces a subjective, literary form of narration, whereas the can simply records what is put in front of it. If this were so, however, the ^ left-hand column would present us with an irresolvable contradiction On the one hand, it states without inflection the succession of im; that are to appear in the film. On the other hand, it is every bit as £j| bound up with literary, verbal narration as is the material in the see I column; it possesses the characteristically metonymic, paratactic si I of the scene text as discussed in Chapter 7. a, Style that doesn't draw attention to itself is the most manipulative Style of all. Neff's narration, being a confession, has a kind of honesty: le has already told us he is a murderer, and here he elaborates on fee emotions that will lead him to become one. The depersonalised Hescription in the left-hand column, by contrast, insidiously sketches |ut an ideology. The houses, the children, the baseball, and the ice Ifeam van metonymically represent a clean-cut, ail-American life, liealthily balanced between home and sports, with the nuclear family |f-its centre. These are the images to be presented on the screen, while |he voice-over leads us towards the homewrecker, the femme fatale, the pller. The opposition could not be clearer, but it is not an opposition jfetween visual truth and narrative fiction. It is between narration that pikes the form of a sequence of images chosen to create one effect -perhaps a reality effect, to borrow Barthes's term, but still an effect - and |n oral narration designed to draw us into complicity with the speaker. jpor Neff is a sympathetic character, while even the plastic, psychotic jgbject of his attraction, Phyllis Dietrichson, possesses a ghoulish fasci-pation. At least they are not boring, and perhaps that is why theorists Influenced by the specificity thesis have a problem with voice-over: if it g-well written it becomes intrinsically interesting, effectively challenging the hierarchical dominance of the visual. How dull the images in gie left-hand column are, how relatively drab the language that creates fhem, and how stiflingly conformist the world they represent. That, at pst,- is part of the meaning of Double Indemnity, just as it is part of the meaning of film noir in general. You can have the American Dream, it perns to be saying, but once inside that antiseptic domestic nirvana |pu'll want to commit bloody murder to get yourself out. §r action as speech ¥~ ' [he verbal sign-system of screenplays, combined with the convention tfaata page of text equals a minute of screen time, means that reading j^dialogue-intensive scene will be very different from seeing the same ;£ene in a film, irrespective of how the text attempts to visualise it, |tjecause the reader's attention will focus on the language rather than ae action. In The Usual Suspects (Bryan Singer, 1995; screenplay by Christopher McQuarrie), Verbal Kint watches as his lawyer and the pros-■icutor engage in pre-trial negotiations about which charges are to be ^Brought against him.73 The scene text reports simply that 'Verbal's eyes rf&How the voices back and forth', and the dialogue text scrupulously er 160 The Screenplay The Dialogue Text 161 repeats before each of the lawyers' speeches that it is delivered of screen. The scene will therefore be visualised as a close-up on Verbal face, or possibly just his eyes, in keeping with the convention, that the' expressive potential of eyes receives priority in classical cinema. Such;« shot will keep the spectator's attention focused on Verbal, and a director * has the option of mixing the lawyers' voices either high or low on tf • soundtrack to signal to the viewer whether or not the words themselvi are significant; it may be that we simply need to register that Verbals^ peculiarly attentive to what is happening to him, with the precise topK. under discussion being of little importance. In such cases a screenplay may simply report that a discussion is taking place, without providing r, the dialogue itself, in which case it will be improvised and mixed in tt film to indicate its low priority. For example, towards the end of the lawyers' exchanges, their voices 'mumble off-screen. Verbal fidgets in his chair', and the written dialogue resumes with the information that. Verbal is to be charged only with 'Misdemeanor one'. Up to this point in the scene, McQuarrie's screenplay records M of the lawyers' dialogue, without interruption from the scene text;;; It' occupies two full pages, topped by the one-line report about Verbal eyes, and tailed equally laconically with the report that 'Verbal lets on a long-held sigh of relief. When reading those pages, rather than seeing the film, the visualisation of the image is likely to be subordinate-to, the dialogue, in which Verbal's lawyer ruthlessly negotiates immunif from prosecution in return for his testimony. The dialogue lays bare some of the intrigues within, and jealous competition between, the political networks in New York and Los Angeles, a theme that emerg more prominently in the script than in the film. Regardless of themal considerations, however, reading and viewing the scene will be two markedly different experiences. On second reading, moreover, other interpretive possibilities become apparent. The lawyer scene is experientially different from many in 7 < Usual Suspects, a script in which Verbal's voice-over forms the principal mode of narration. At a crucial point, towards the end, Verbal tells a story of how Keyser Soze came to acquire his terrifying reputation -story which, he tells us, may or may not be true, though he himself believes it. Soze had returned one day to find his wife and children violated by a Hungarian mob, with whom he was engaged in a tnrj war. Rather than let his family live with the humiliation, SSze kills both them and the gang, aside from one that he allows to flee to begin circ lating the story of Soze's terrible vengeance. The events are described m&. Verbal's voice-over as well as in a series of images in the scene text: i & He kills their kids, he kills their wives, he kills their parents and their fetparents' friends. *, We see glimpses of Keyser Soze's rampage. Bodies upon bodies in homes , and in the streets. Then, the fires. | Stores and homes burn, engulfed in flames. >T_He burns down the houses they live in and the stores they work in, "p he kills people that owe them money. And like that, he was gone. Underground. No one has ever seen him again. He becomes a myth, :'_~a spook story that criminals tell their kids at night.74 Because the images that would arrest the attention of a spectator in the ■linema lack detail in the scene text, Verbal's proportionately more prominent narration accordingly receives greater attention in the screenplay. The stunning final revelation is that much of what we have seen in the jilm is just a tale that Verbal has been improvising serendipitously from scraps of texts pinned on a notice board behind his questioners. Scenes . layed out before our eyes must now be retrospectively reinterpreted as ids inventions. This is a little less shocking in the screenplay, because the textual sign system has concentrated attention on Verbal as a narrator. It "still surprises, however, because while the unreliable narrator is a familiar .onvention in prose fiction, it is almost forbidden in cinema. 1 There is a crucial difference between The Usual Suspects and a spate cif superficially similar films that followed in its wake, including Fight Club (David Fincher, 1999), The Sixth Sense (M. Night Shyamalan, 1999), and ,4 Beautiful Mind (Ron Howard, 2001). In each of these later e samples, the spectator finally recognises that many of the events pre- iously shown are to be interpreted as the projection of events in the mind of a central character possessed by an extreme subjectivity (an i'lea deliciously parodied in the story attributed to Donald Kaufman in Adaptation). The protagonist is mentally ill, or dead, and does not realise ■ hat the world in which he appears to move is, to a large extent, his own mental construction. The reassuring solidity of the cinematic world, i hich film audiences have come to accept as real in a 'suspension of disbelief, dissolves. In other words, these films are variations on a kind of cinematic expressionism with a long history, stretching back at least '5 far as The Cabinet ofDr Caligari (Robert Wiene, 1919), in which the ii leaning of the entire film was altered by the controversial addition of a Name story that casts all of the events as the delusion of a madman. 162 The Screenplay In The Usual Suspects, however, the principal narrator, Verbal, is „ , deluded. He invents the plot that the spectator sees to gain a tact (| advantage in the here-and-now; he is attempting to secure his escape Although by the end the ontological status of many of the events remains uncertain, some, if not all, are to be understood as the fabucav tions of what had until this point appeared a relatively minor charai i r There is a significant difference between this and a lie told by a characters on the stage, since the latter does not alter the perceptual space of the4" theatrical set. In the theatre, words will always be scrutinised for their; reliability, and all speech acts will change the relationships between thfe-*f characters on stage, but they will not physically alter the stage itself In'-*!? The Usual Suspects, however, the audience is finally forced to reinter,- ^ whole sequences as visual representations of a story Verbal is making up: the action in such sequences is a representation of Verbal's speech,V The challenge then lies in determining what degree of reliability to give to any of the scenes in the film. The exchange between the lawyers is clearly an incident that is not -invented by Verbal, since it is in response to the legal procedures that he'-I begins to fabricate the story. Equally but conversely, the episode of Sozgf? and the Hungarians is explicitly presented as a story told in voice-over by Verbal, who stresses that he is merely reporting a tale that may or may notiS be true. There is a certain complicity here between Verbal as an embedded narrator and the problematic 'image-maker' who is the impersonal nar--^-* rator, or implied author, of The Usual Suspects. Just as the former stresses that nobody knows what Soze looks like, so the latter reports that ■when" Soze enters the house, '[w]e are never allowed to see his face' (p. 90). i hi definitive statement has a different effect when followed in the iijrn,_ because film tends to imply that such decisions are a choice of the director rather than an instruction in the screenplay. Nevertheless, as the unreliability of this particular episode about Sdze has clearly been signalled to*^ both reader and spectator, there is no great difficulty at this point. The problem emerges at the end of seeing the film or reading the;j text. Now alert to the unreliability of everything Verbal says, there is a, '| compulsion to go back and examine retrospectively the verifiabihly 01 every scene in the film. In many cases this is far from straightforw r I Previously, Verbal has recounted under police interrogation the history of another mass killing that we know to have taken place from se> eral* pieces of independent corroboration, and which has left the authoi lties^ dumbfounded. Knowing that although the police investigation is 'teal' much of Verbal's account is a fabrication, a question now surrounds thfti transitions from one to the other, as in the following: The Dialogue Text 163 KUJAN sjvfow what happened after the lineup? Verbal sneers at Kujan, unable to change the subject. ,rEXT - POLICE STATION - NEW YORK - NIGHT SIX WEEKS PRIOR 'Jieaton stops at the top of the front steps of the police station and lights a ;cigarette. Edie comes out behind him, fuming mad. (p. 46) [though the syntax indicates that the scene outside the police station presents Verbal's response to Kujan's question, the absence of any ^textual indicator of subjectivity (such as a dissolve or a voice-over) intro-ni.es an ambiguity. Customarily, film permits the conflation of these two issibilities: a transitional device indicating subjective memory may segue ^perceptibly into an objective record of events, with the narrator's recol-eetion taking on the status of accepted fact. If there is reason to doubt the eliability of the witness, the spectator will usually be made aware of this, as i ippens even in such a problematic case as Rashomon (Akira Kurosawa, 1950), But it is extremely unusual for the spectator to be compelled to emterpret a scene as a lie. The notorious precedent is Stage Fright (Alfred Ltchcock, 1950), in which a character's narration introduces a flashback fat is only revealed at the end of the film to have been an untruth: the jaracter himself had committed a murder, and the scene presented Bntlie screen is the story he has told to cover his tracks. The film fails -cause the scene is not integrated into any larger structure that would ill the cinematic narration as a whole into question; one scene, and one me only, is a deception, and as the audience can have had no way of lowing this, the effect is of a cheap contrivance. We Usual Suspects is different, because it calls into question the reliability of its narrational strategies in general. It is not simply the historical mas;in.favour of cinematic realism that creates an illusion of truth, it is ^aa result of the unfolding of cinematic time in the present tense. Each element of the story that the discourse reveals to us takes place in the Dntinuous present, even when it is presented within the frame of a flash-b iek: The same is also the case in the screenplay text. If The Usual Suspects vere written as prose fiction, the narrator would cast the confrontation |tween Kujan and Verbal into the past tense ('Verbal sneered at Kujan'). ie scene with Keaton and Edie would not only be narrated in the past \ 164 The Screenplay The Dialogue Text 165 tense but would also be revealed, by the presence or absence of invert* i-commas, to be either an event recounted by the narrator or embedd^ narration spoken by Verbal. This would alert the reader to the need for caution in assessing what degree of credence to give the narration. Instead^the^creenglay_^reader JsTaced with_an ajnbigujty that i temporarily irresolvable other than by the, conventional raesumrtf^p in favour of the truth of the image. This is supported by the statement that the events happened 'six weeks prior', an assertion that is not attributed to Verbal. The director will have to decide whether and ho\ to indicate this time frame to the spectator, but (slightly unconventi6h-~ ally) McQuarrie has given no indication within the script of how trii is supposed to be done. Accordingly, the words are likely to strike the-reader as a small piece of omniscient narration, although in retrospei it appears that this is probably another of Verbal's fabrications. Faced with these doubts, attention shifts from the unreliability <| Verbal's narration to that of the screenplay itself. The problem arisi because some of the images and events are not to be interpreted either as mere fabrications by Verbal (since there is independent corroboration ~ in the police reports), or as unmediated representations of events that have really happened in the story world. Instead, they are a visual interpretation of his words, and (if we are to make diegetic sense of the filri not an interpretation supplied by the director, nor by his auditors (tl police, but also the spectators in the cinema), but by Verbal himself This becomes apparent in the final twists, after Verbal has left the poli< station. A fax machine receives a copy of an image of Keyser S6ze drawn by a survivor, following which Verbal is picked up by a man in a car:1 INT. DISPATCHER'S OFFICE \ Jasper Briggs pulls the sheet out of the fax machine and turns it ovt revealing the composite sketch of Keyser Soze. Though crude and distorted, one cannot help but notice how much it loot like Verbal Kint. EXT. STREET The car stops. The driver gets out. It is Kobayashi, or the man we have come to know as such. (p. 133) The conjunction of the two images plays havoc with the different between the semiotic systems of screenplay and film. That the sketc H Keyser Soze resembles Verbal immediately begins the process of Prospective analysis, as we will start to consider whether the actions attributed to the former could in fact have been carried out by the latter, ffie image of Kobayashi, however, creates mayhem. Throughout the ..pipt, we have known of Kobayashi, Keyser Soze's lawyer, only from ' ^rbal's account. A shot near the end shows us that Verbal has simply jbrrowed the name from that of the manufacturer of a coffee cup. iThe arrival of 'Kobayashi' in the getaway car at the end, however, ^complicates matters. Verbal has taken the signifier 'Kobayashi' and "attached it to an associate, who must 'really' be called something else, in. a textbook illustration of the arbitrary relationship between signifier jhd signified in Saussurean linguistics. The problem is that signifieds ire mental concepts, raising the bewildering question at the end of The isual Suspects: whose mental concept is the signified of 'Kobayashi'? On reading the screenplay, one will have formed a certain visual impression >f Kobayashi. Or not: The Usual Suspects is very perfunctory in describing tjje physical appearance of its characters. For reasons noted Chapter 7, sfereenplay texts in general rarely offer the concrete visualisation of Hiaracters routinely found in realist fiction. In any case, the reader legitimately assumes a certain interplay with the text in the creation if the character. In this concluding moment, however, the screenplay seems suddenly to have usurped that autonomy and told us what we Th"ave been visualising all along. The film is no less disorientating, but for the opposite reason. Now the aiswer to the question 'what does the mental concept "Kobayashi" look ^Jjke?' is 'he looks exactly like the English actor Pete Postlethwaite, heav-ly suntanned'. Suddenly, someone other than Kobayashi, but with an 'identical facial appearance, emerges at the very end of the film. Again, the obvious explanation is that this is because 'Kobayashi' is Verbal's fiental concept. The bafflingly unanswered questions that remain after ither reading or seeing The Usual Suspects are therefore the result of the unresolved interplay of three different ontological fields: that of realism (the police and the lawyers, searching for clues within the diegetic ©rid of the film); that of the reader or spectator, who when told a |tory naturally supplies the mental concepts for herself; and that of the Jarrator (whoever that is) of the screenplay or film, who has usurped tps autonomy of the reader by asking us to accept that certain scenes ire presented directly from the inside of Verbal's head. That this does Jot fully add up is partly due to the idea that Verbal has actually been aprovising a story from signifiers pinned to the notice board that do yJ|ot fully cohere within a consistently and coherently imagined world: 166 The Screenplay The Usual Suspects exploits with exceptional subtlety the resoun es of„ dialogue within the screenplay. The quasi-realistic scenes concerning the investigation initially follow a dramatic structure well known m detective fiction, whereby the authorities act as readers constructing a discourse which attempts to decode a story 'written' by the crimm d-Like many contemporary films, such as those analysed by Temenuga" Trifonova and discussed at the end of Chapter 1 of this book, this I- i [ to the disturbing possibility that there is also, or instead, a non-realistic pre-text for the story, which is nothing other than the screenplay itselff"5 McQuarrie's brilliant innovation is to introduce to this fascinating but relatively familiar idea the conceit that the film is an act of oral improvisation. Instead of referring back to a story, Verbal is to be regarded ds actually creating the discourse that we see on the screen. This intro" duces a new level of interaction between two aspects of the screei | , in which the voice-over of the dialogue text is seen to be responsible for the creation of the scene text (and of the dialogiie of the otrTeTcliaracters. within it). In so doing, however, McQuarrie's script creates a new kind of palimpsest, in which a story that would make complete sense a appears to be almost within view, but is at the same time being rubbe I out by the voice-over. Instead of the convention that a voice-over provides expressive revelation, Verbal's provides tactical concealment In creating these interlocking dramas betweenL the di^loguejtext