diegesis 3

Diegetic/Nondiegetic: A Theoretical ModelAuthor(s): David NeumeyerSource: Music and the Moving Image, Vol. 2, No. 1 (Spring 2009), pp. 26-39Published by: University of Illinois PressStable URL: http://www.jstor.org/stable/10.5406/musimoviimag.2.1.0026 .Accessed: 03/07/2014 00:42

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

University of Illinois Press is collaborating with JSTOR to digitize, preserve and extend access to Music andthe Moving Image.

http://www.jstor.org

This content downloaded from 132.216.86.166 on Thu, 3 Jul 2014 00:42:15 AMAll use subject to JSTOR Terms and Conditions

From Music and the Moving Image Vol. 2, Issue 1.

Diegetic/Nondiegetic: A Theoretical Model

David Neumeyer

Two pairs of terms, diegetic/nondiegetic and synchronization/counterpoint, have remained central to discussions of filmmusic. The first of these in particular has strong connections to film studies and literary studies (specifically, narratology).A model that positions diegesis in relation to (between) spatial anchoring and narration can help sort out the shift fromopposition to function and make the diegetic/nondiegetic pair more reliable for interpretation of film music.

Introduction

In film music studies, two pairs of terms have proved to be remarkably hardy, no matter the systematic orideological environment in which they find themselves: these are diegetic/nondiegetic and synchronization/counterpoint. Attheir most basic level, the initial terms in these binaries refer, respectively, to spatial and temporal relations betweenimage and sound: the anchoring of sound in the physical world depicted in the film, its diegesis,1 on the one hand, andthe appropriate or apparently natural coordination of sound with a moving image, on the other. The opposing terms ineach pair, again at the most basic level, are negative: sound fails to be positioned securely in relation to spatialcoordinates (and is therefore extradiegetic or nondiegetic), or sound fails to align properly with the image (and is thereforein counterpoint with or plays against the image). Both of these contraries can, of course, be reinterpreted positively by thefamiliar process of favoring the marked term, or flipping the binary: we might assert, for example, that the shifting planescaused by the application of nondiegetic or nonsynchronized sound add depth by creating additional narrative registers(background music, voice-over, or, more broadly, a self-reflexive sense of constructedness).

The complexities and ambiguities that arise from trying to interpret specific instances in films using these binarypairs have led some observers to argue that they restrict analysis unduly, holding attention on a narrow set of formalfunctional relations and inhibiting a focus on other aspects, such as representation or subjectivity. Anahid Kassabian takesthis line with respect to the diegetic/nondiegetic: their opposition "suggests that film music can be categorized within adichotomous schema--grossly reduced as either 'in' (diegetic) or 'out' (nondiegetic) of the narrative world of the film. Thisdichotomy is insufficient; it cannot comfortably describe music that seems to fall 'in between' these categories, much lessaccount for its different character."2 The objection, it should be noted, is not to the terms themselves but to their pairing inan opposition. Kassabian prefers to adopt three industry terms used by Earle Hagen in his Scoring for Films: sourcemusic, source scoring, and pure or dramatic scoring.3 She then argues in favor of this triad because the "inside/outside,diegetic/nondiegetic dichotomy has difficulty with [many] musical events; they belong to a third 'inbetween' category.Moreover, it cannot analyze the degree to which music takes cues from other filmic events. Most importantly, it cannotaccount for the relationship between these two qualities." Although the move is difficult to reconcile with this triad ofterms, Kassabian proposes to replace the opposition diegetic/nondiegetic with a continuum, "a kind of inverseproportionality in which the more identifiably within the narrative the music is produced, the less liable it is to take its cuesfrom the events of the narrative."4

Industry professionals, following a similar line of thought, sometimes propose discarding diegetic/nondiegeticaltogether on the grounds that they are not used in practice: sound designer Randy Thom, for example, writes, "In thethirty years of conversations I've had with coworkers on feature films in the USA and Britain, nobody has ever used theword diegetic except to deride it as an academic term of little practical use."5 Thom acknowledges that, in postproduction,"the music mixer may ask whether the sax should be treated as 'source' or 'score' in order to know if it should be muffledand treated with artificial reverb or if it should be played cleanly and crisply" (2). On the other hand, "most filmmakers,whether they are directors, composers, or sound designers, are minimally analytical about their own work" (1), and, partlyas the result of the ambiguities that inevitably arise in films because of that characteristic, "the question of whether asound in a given scene is diegetic or not is often irrelevant to the effect the story has on its audience." He cites ahypothetical case of a character who on waking in the morning hears a saxophone playing, music that "will tend to beinterpreted by the audience as the acoustics of [the character's] soul, making the question of the diegesis moot" (2).This would be one of the "inbetween" instances in Kassabian's model--or what is nowadays referred to variously as point-of-view, subjective, metadiegetic, or internal diegetic sound/music.6

And sometimes the words themselves rankle: as early as 1959, film phenomenologist Jean Mitry, punning on theFrench terms, labeled diegesis and diegetic "indigestible" and complained that they "would seem to be tautologous withthe words drama and dramatic, which, etymologically speaking, are quite equal to the task indeed, do so in a way that

26


is clearer, simpler, and more easily comprehensible."7

In what follows, I will argue that the diegetic/nondiegetic pair is now well embedded in film music discourse,agreeing with Henry M. Taylor's assertion that "by now this terminology has been so well established that it would be futilenot to use it in its accustomed sense." 8 I will argue further that the terminology is useful: despite its several difficulties,the distinction represented in diegetic/nondiegetic is still fundamental to material relations of image and sound and tonarrative functions of music in the sound film. I will propose a model that defines and positions diegesis in relation tospatial anchoring and narration with a goal of sorting out the functions to make them more reliable for the interpretation offilm music. In this model, the diegetic lies in the middle of a process of cognition and interpretation, with the stagesanchoring diegesis narration. The first of these concerns the viewer's relation to the loudspeakers, the second therelation of film sound to the world of the film, and the third the relation of sound to narrative process (or narrativeunfolding). In this model, synchronization/counterpoint is analyzed in parallel with diegetic/nondiegetic--indeed, the formeris a subcategory of the latter, its "temporal arm," so to speak. In the process, the viewer-auditor acquires both spatial andtemporal information needed for the interpretation of sound in narrative. Anchoring is concerned with the spatial, diegesislikewise but with fundamental inputs into the temporal (synchronization/counterpoint) and agential (when sound isanchored in a character), and narration requires and coordinates the spatial, temporal, and agential.

Diegesis: From Narrative to the World of the Narrative

The term diegesis has a venerable history dating back to Plato and Aristotle, and it would be most satisfactory ifone could link their definitions and usages directly to contemporary ones, but alas that cannot be done. The modernhistory of diegesis and the diegetic is little more than fifty years old--and the modern treatment of the term is a significantdeparture from its ancient meanings.9 For Plato, the focus was on the narrator: the distinction between diegesis andmimesis was that the former is a direct telling of a story and the latter is an indirect telling through characters. ForAristotle, on the other hand, "mimesis is the umbrella term designating all poetic representation,"10 and Plato'scategories become subtypes, (confusingly) retaining the same names, diegesis and mimesis.

In modern usage, diegesis and diegetic are associated especially with the widely known structuralist narratologicalsystem of Grard Genette. According to Genette, he derived his use of diegesis not from Plato or Aristotle but from a bookon film by tienne Souriau, who included "digtique" in a list of eight adjectives that he describes as "indispensable" forclarity and precision in the analysis of film.11 In defining "digtique," Souriau detached the place (or physical space) of anarrative from its story; thus, the noun digse means "a universe rather than a train of events (a story [histoire]); thedigse is not the story but the universe in which the story takes place."12 Thus, as Genette points out, "[the Greek]diegesis has nothing to do with digse; or, if one prefers, digse (and I had no hand in this) is by no means theFrench translation of the Greek diegesis. Things can get complicated at the level of adjectives. For my part, I (likeSouriau ) always derive digtique from digse, never from diegesis."13 The distinction is lost in English translation,but to little harm if it is remembered that the contemporary meaning does not relate to or depend on the original Greekmeanings in Plato and Aristotle. Souriau's definition was particularly amenable to film, as it shifted attention from anarrator to the mise-en-scne and so greatly enhanced the status of visual representation in analysis of narrative.14

Regardless of whether or not digse had any roots in ancient writings, it enabled (perhaps even forced) theseparation of two levels of narration in which the diegetic/nondiegetic pair are easily situated: the narration proper andextradiegetic narration. (Extradiegetic is Genette's term for what is called the nondiegetic in film theory; intradiegetic =diegetic.)15 The result, then, is that "diegesis has moved out of the level of discourse to the story plane of a narrative, out of which discourse evolves."16 In Genette's conception, there is a parallel to Plato's distinctions that invites confusionabout the terms' origins: the visual representation (as if mimesis) against the extradiegetic narration (as if diegesis).17 Theproblem with that, of course, is that diegesis means the opposite in its contemporary usage: mimesis the diegetic(showing the world of the narrative), diegesis the nondiegetic (the narrator intrudes). The shift from narrative agency tonarrative space, however, was very clear, and, by the late 1960s, film semiologist Christian Metz could matter-of-factlysummarize diegesis in this way:18

[Diegesis] designates the film's represented instance that is to say, the sum of a film's denotation: thenarration itself, but also the fictional space and time dimensions implied in and by the narrative, andconsequently the characters, the landscapes, the events, and other narrative elements, in so far as they areconsidered in their denoted aspect.

Thus, film would seem to be strongly implicated in the rise of modern diegetic theories of narration, a direction thathad been encouraged, enabled, and extended by the Russian Formalists but was particularly developed by Frenchstructuralists and semioticians.19 Although the Formalists did study cinema in addition to their primary focus on literatureand language, the semioticians were the ones who systematically expanded the idea of language-like systems of

27


communication far beyond literature:20

[Roland] Barthes sought to apply Saussure's theory of signification to nonlinguistic systems, such as fashionsand advertisements. There followed, and still follow, semiotic studies of all manner of cultural phenomena.Media which had seemed analyzable only through mimetic assumptions came to be treated as analogous toverbal language.

It was also the semiologists (beginning with Christian Metz, Raymond Bellour, and Jean-Louis Baudry) who focusedespecially on cinema. As David Bordwell puts it, "Not until French structuralism and semiology emerge do we find criticssystematically employing linguistic theory to analyze film."21 Finally, then, both Genette and Souriau are credited byClaudia Gorbman in connection with her adaptation of the diegetic to film music in her semiotics-based system.22 Shesays that "Genette's and Souriau's definitions would agree that the diegesis means the space-time universe and itsinhabitants referred to by the principal filmic narration. At this point, then, we may summarize and define 'diegesis' asbeing the narratively implied spatiotemporal world of the actions and characters."23 Souriau's definition, as cited byGorbman, is slightly narrower (at least, before its illustrations complicate matters): "all that belongs to the narratedstory, to the world supposed or proposed by the film's fiction."24 I prefer to keep this narrower view for the present inorder to isolate the spatial (which is prior) from the temporal and agential ("actions and characters"). Thus, to paraphraseGenette's description, diegetic sound means "sound in the universe in which the story takes place" and nondiegetic(extradiegetic) sound means "sound not in that universe and presumed to belong to the level of the narrator."

As we have seen, then, diegesis is by no means a neologism, but its customary meaning in film studies iscertainly of modern origin. Of course, one might still object to Souriau's use of the term, if one wishes, but the definitionsof diegesis as world of the story and of extradiegetic as the level of the narrator (or of the narration) are now thoroughlyembedded in the scholarly study of narrative and of film (and also in their pedagogies),25 and thus, as Taylor put it in thequote cited earlier, there seems little to be gained--and I would say definitely something to be lost--by substituting otherterms for diegetic and nondiegetic in film sound and music studies. Other terms do not remove the complexities andambiguities that are an inevitable result of narrative process. Source/background, for example, does nothing to removeproblems such as the audio dissolve (music passing from source status to background but without erasing the source) orthe uncertain status of music that would be stylistically appropriate as source music in a particular scene but whose levelis noticeably too high for the physical space and whose source is not shown on screen. Note that Kassabian, in adaptingterminology from professional practice, did not discard the categories to which the terms are attached (source music diegetic; dramatic scoring nondiegetic) but simply added an intermediate term (source scoring) and then questioned therigidity of the opposition, which is another issue altogether.

In the next section of this article, I shall argue that the oppositive aspect is fundamental to the diegetic in its basicspatial aspect, but not necessarily in relation to narration. As Robynn Stilwell puts it, "because the border betweendiegetic and nondiegetic is crossed so often does not invalidate the separation. If anything, it calls attention to the act ofcrossing and therefore reinforces difference." 26 Fred Karlin and Rayburn Wright, likewise, do not share Kassabian'sdiscomfort with the dichotomy. In their textbook for professional training, they use "source music" and "underscoring" andoffer the common definitions: "Source music is music that the people on the screen can hear, while underscoring is themusic that they can't hear but the audience can." 27 But Karlin and Wright also say that "source music is most often usedas independent music without any underscoring to accompany or overlap it, yet it can also change its function at any timeand continue as score, acknowledging the emotions and events of the drama," and they offer a set of five commonoptions: "source music can function as score," "source music can change into score," "source music can crossfade toscore," "source music and score can be used simultaneously," and "source music can play the underscoring theme." 28

Karlin and Wright's treatment of the terms and their relations is very close to Gorbman's and in my view still themost productive way to work with them. Gorbman recognizes the basic spatial category of the diegesis (Souriau'sdigse) and the basic relation of space to narrative (intradiegetic and extradiegetic), but also the complex loose play offunctions between them (equivalent to Karlin and Wright's five options). These apply broadly to sound, with music as aspecial case: 29

It is not difficult to realize that the sound track takes many more liberties with the diegesis than does theimage track. Voice-over commentaries and verbally narrated flashbacks, both nondiegetic, punctuate manyfilm narratives. Sound effects, however, tend to remain diegetic (unless they accompany also nondiegeticimages). One reason for this lies in the ambiguity of many sounds when presented out of the context of theirsound source. Significantly, the only element of filmic discourse that appears extensively in nondiegetic aswell as diegetic contexts, and often freely crosses the boundary line in between, is music. Once weunderstand the flexibility that music enjoys with respect to the film's diegesis, we begin to recognize howmany different kinds of functions it can have: temporal, spatial, dramatic, structural, denotative, connotative--both in the diachronic flow of a film and at various interpretive levels simultaneously.

28


The anchoring diegesis narration model to be discussed in the following sections coordinates well with Gorbman'ssystem. Anchoring focuses on the spatial bias in the basic cognition of sound; diegesis then takes the real-world/film-world distinction and maps it onto the second-level cognition specific to film viewing; and narration is the collection offunctions that arise from the treatment of the opposition in the narrating instance of a particular film and that shift thebalance toward interpretation.

Diegesis and Spatial Anchoring

By comparison with the third stage of the anchoring diegesis narration model, the first and second stagesmight seem simple and direct, but interesting complexities do arise there, as well. Even in the basic cognition of spatialanchoring, two levels have to be managed simultaneously (see Figure 1). In the dreamworld of cinema, where sound-in-that-world is mediated by recording, mixing, and reproduction, the fundamental cognitive action for sound is neverthelessto attempt linkage to a physical object or, at least, to some (relative) coordinate in the relevant physical space. WhenChristian Metz took up the question of film sound, he argued that any specific or particular sound must be regarded as an"autonomous aural object." Yet listing the properties of a sound is insufficient to account for it as a perception in the filmiccontext because sounds cannot be identified independently of their source.30 Metz decides, therefore, that these "auralobjects" are complex and problematic because the visual is always privileged, because to "name" a sound requires boththe sound and its source yet the source remains primary, and because "spatial anchoring of [sound] is much more vagueand uncertain than that of visual events," a trait only increased by the physical separation of the film screen andloudspeakers.31

Figure 1. Stages in the anchoring diegesis narrationmodel.

The strong bias in human hearing toward connecting a sound to a physical source, as is well known, is a simpleand ancient matter of self-preservation, of using information to distinguish between the threatening and thenonthreatening. Even in the local multiplex, whatever else goes on and no matter how much we may be enthralled by thefilm, we must still be able to distinguish between someone yelling "Fire!" on screen and someone doing so in the theater.(Mark Kerins refers to the problem of contemporary surround-sound systems in connection with this cognitive bias: "Somefilmmakers have been afraid of the so-called 'exit door effect,' where spot sounds in the surround channels distract theaudience's attention away from the screen. This can be particularly problematic if the sounds are [mis]interpreted by theaudience, such as when the sound of a door slamming in the surrounds is mistaken for the theater door.")32 Thus,there are two levels of anchoring: in the "real" world (environmental acoustics) and in the screen world (loudspeakeracoustics); and two modalities: threatening and nonthreatening. Shifting the modalities from the real world to the screenworld invokes a sense of abstraction or distance that is one of the guarantors of narrative understanding and that issimilar in some respects to listening to music. For example, in a nontheatrical situation such as a city street, "Hey, youthere!" is potentially threatening, but music anchored only in a street musician is (generally) nonthreatening, and a specialcase, too, in that it makes an abstract play of sounds, more or less pleasing depending on the volume and on the abilityof the musician. Two distinct modes of listening are operating simultaneously, and in this separation of the concrete andthe abstract lies the seed of the diegetic/nondiegetic distinction.

Listening in the cinema, thus, involves at least four sound registers, experienced in this order: (1) (potentially)threatening diegetic sounds in the theater, such as "Fire!"; (2) nonthreatening diegetic sounds in the theater, such as

29


noises from the loudspeakers or your companion eating popcorn; (3) (apparently) diegetic sounds in the physical world ofthe film, such as characters speaking or singing; and (4) other (nondiegetic) sounds in the film's sound track. For soundregister 3, the sorting process for sounds continues at one remove (the real-world status of a film character speaking isnonthreatening, but there is narrative significance in a threatening/nonthreatening discrimination within the screen world--as, case in point, repeat the city street scenario as a scene in a movie you are watching), but the process is complicatedby sound register 4, the ever-present possibility of a narrator's intrusion in form of a voice-over or nondiegetic music. Theinitial reaction to sound--sorting of diegetic sounds into categories of attention, or something akin toforeground/background--sets the stage for, and leads to, the diegetic/nondiegetic distinction, which is simply a differentmode of sorting, in this instance into levels of narrative and narrator. In other words, anchoring leads to diegesis (oranchoring diegesis), and diegesis enables the narrative levels (or diegesis narration).

Finally, then, the diegetic/nondiegetic pair is conceived as an opposition at the stage of spatial anchoring but as amode of sorting at the stage of diegesis. We do not normally understand narration and narrator as being at odds witheach other, however, and thus paradoxically perhaps the same distinction that enables us to separate the narrative fromthe narrator, the intradiegetic from the extradiegetic, is reduced to a function once that distinction is made. In Figure 1, thefirst item under "narration--opposition" expresses this status change. A binary, of course, may be treated as a fixedopposition or understood dialectically--therefore, the "subclass: dialectical" in the list--or the opposition may be "dissolved"into a continuum with a potentially infinite number of intermediate stages, each one a distinct function (Kassabian'ssolution and no. 2 in the list under "narration" in Figure 1). That is to say, the opposition itself may serve as a narrativefunction, as for example with stinger chords, audio dissolves, abrupt shifts from diegetic to nondiegetic music, or anythingelse that draws attention to the difference between diegetic and nondiegetic.

Finally, however, I prefer Robynn Stilwell's solution, which is based on an idea she and James Buhler presentedduring an online interview: to maintain the opposition but acknowledge a distinct and complex space inbetween, the"fantastical gap" (no. 3 in the list). As she puts it, "Diegetic or nondiegetic may be a simple distinction, but it need not bea simplistic one. The fact that the boundary is crossed so often should not invalidate the integrity of the distinction; indeed,the manner in which the meaning in the distinction multiplies and magnifies in the crossing is indicative of its power."33This is entirely in line with Gorbman's comment cited earlier: "Once we understand the flexibility that music enjoys withrespect to the film's diegesis, we begin to recognize how many different kinds of functions it can have: temporal, spatial,dramatic, structural, denotative, connotative--both in the diachronic flow of a film and at various interpretive levelssimultaneously." Gorbman's statement is in effect a description of the fantastical gap, or "the border region [that] is atransformative space, a superposition, a transition between stable states. [M]ovement through the gap between diegeticand nondiegetic takes on great narrative and experiential import. These moments do not take place randomly; they areimportant moments of revelation, of symbolism, and of emotional engagement within the film and without."34 It isparticularly relevant to the model I have constructed here that Stilwell emphasizes this idea of movement or passage intoor through.

As an example, consider two short scenes from Catch Me If You Can (2002), which tells the story of FrankAbagnale Jr. (Leonardo DiCaprio), who runs away from home rather than face the trauma of his parents' divorce and hisfather's slow financial ruin. Frank is a risk taker who discovers that he has innate talent for imitating professionals (airlinepilots, doctors, and lawyers), and he develops great skill in forging checks. The story, told mostly in flashback, is aboutFBI agent Carl Hanratty's (Tom Hanks) campaign to capture Frank, who spends several years in prison but eventually,thanks to Hanratty's advocacy, works for the FBI cracking check-forgery cases. The two scenes occur midway through,just after Frank has humiliated Carl (who has considerable desk experience but very little in the field) by escaping fromunder his nose. In the first of these scenes, Carl gets his own back to a limited extent during a phone conversation withFrank; in the subsequent scene, Carl receives a crucial bit of information that puts the evidence chain in motion andeventually leads to Frank's arrest.

Carl is working alone on Christmas Eve (at 01:03:38 of the film); a radio sits on his desk and resonates weakly inthe large room, empty except for desks and metal fixtures. Bing Crosby and the Andrews Sisters (1950) sing "MeleKalikimaka" as Carl examines evidence. The phone rings; Frank is calling from his hotel room; in the traditional manner,the film cuts back and forth between the two men; very near the end of the conversation, as Carl avers that Frank will becaught and observes that Frank called him because he had no one else to call, underscoring enters with a melancholytheme that has already been associated with Frank's feelings about his father. Music follows its strict roles in the scene:diegetic music (with a visible source) and, later, nondiegetic music that gives a narrative reference (in the leitmotif) andsupplies additional information about Frank's mood (his body language suggests agitation or anger rather thanmelancholy). A simple cut brings a scene change to a diner: effects noise is prominent (sounds of the grill, dishes, rainoutdoors), music plays from an invisible but plausibly diegetic source ("He's So Fine" [Chiffons, 1963]), but limited dialogueis backgrounded. Carl is sitting in a booth studying printouts of criminal activity associated with "Barry Allen," the nameFrank gave him during their initial meeting several weeks earlier (when Frank impersonated a Secret Service officer andescaped). Carl's conversation with a young waiter brings dialogue into the foreground again, and Carl learns that BarryAllen is the name of a comic-book character; the light dawns, and nondiegetic music blocks out nondialogue sounds. Themusic is Frank's theme (also the film's title theme), and it continues and intensifies as Carl talks to a fellow officer on the

30


phone (in a public booth presumed to be outside the restaurant). In a parallel with the first scene, underscoring follows ondiegetic music, and the underscoring bears a narrative reference while also suggesting something about mood: a morerapid, furtive-sounding music plays as we can see that Carl is excited about the new information he has acquired.

In both of these scenes, the music's spatial aspect is clearly defined and carefully maintained (with one exceptionto be noted in a moment). The narrative function of the nondiegetic music with respect to emotion and narrative referenceis unambiguous. In this case, then, the spatial dichotomy of intradiegetic/extradiegetic coincides with narrative function,where the opposition (or juxtaposition) of diegetic and nondiegetic is part of the design for each scene, indeed, is brieflyforegrounded at the end of the first scene, with a cut to Carl in a crane shot (the only one in the scene) just after Frankslams down the receiver. The nondiegetic music continues, but Carl turns up the radio (whose music had disappearedcompletely) and for a moment the two musics vie for attention in the sound track. One could point to some smalldiscrepancies--the fantastical gap is all about discrepancies, ambiguities--but nothing that rises to the level of disturbingthe dichotomy that frames the sound track in both scenes. "Mele Kalikimaka," for example, might be interpreted to moveoutside its space- and venue-confirming role to the level of narrative: we will eventually realize that Christmas songs arean aural motif, an invitation to assign significance in retrospect to this first instance; and the song itself providesinformation about Carl through his musical tastes--but the latter is tautological at this point and the former is important toits final instance in the film, not here. One might also point to a possible extrafilmic reference: "He's So Fine" was thesubject of a successful plagiarism suit, which could be a reminder of Frank's check forgery. But, even if correct, thisreference is little more than an insider's joke, both obscure and unnecessary. The exception that I have already mentionedcomes just before the nondiegetic music enters: during previous cuts to Frank, we could hear the music from Carl's radiobehind his speech, but this time when Carl speaks there is no music, only a couple seconds of silence until Frankannounces that he needs to hang up and the nondiegetic cue enters. This is the fantastical gap in a passage fromdiegetic to nondiegetic, and its irrational character spills over into the opening of the crane shot, until Carl turns up (or is it"on"?) the radio and the diegetic/nondiegetic opposition (re)asserts itself radically. On the other hand, the passage isquick, and one might well argue that the impact is minimal.

Diegesis and Subjectivity

As the scenes from Catch Me If You Can already suggest, ambiguities arise readily in connection with subjectivity.Crudely put, for the basic purpose of diegesis a character is an object in which sound can potentially be anchored. Fornarrative understanding, however, this spatial placement is complicated by the fact that this object is also an agent. Evenmore problematic, this narrative agent always has the potential to take over narration (Genette: "Insofar as the narratorcan at any instant intervene as such in the narrative, every narrating is, by definition, to all intents and purposes presentedin the first person");35 this narrative agent may in fact (choose to) occupy the diegesis itself. As one of many possibleexamples, in Million Dollar Baby (2004) we learn only in the final seconds that Morgan Freeman's character, presentthroughout the film, was actually writing a letter about the events to his sister--the film was his visualization of thoseevents while he wrote the letter.

It might well be possible to catalog the several modes of this interplay between diegesis and agency,36 but I thinkit is more productive for interpretation to assign all ambiguous instances to the fantastical gap. In the anchoring diegesis narration model, anchoring is about the physical "I" of the viewer in relation to the "physical" I of an on-screencharacter (or, in by far the most common instance, speech in the theater as opposed to speech understood to come froman actor on screen). Diegesis, on the other hand, is about character-agents and the narrator, and here, of course,representation is the central issue.37 Diegetic sound is at its clearest in an on-screen character's speech, an object'snaturalistic sound, or a character's singing or playing a musical instrument on screen; and nondiegetic sound is clearest invoice-over narration by an absent narrator or in orchestral underscoring using an ensemble without any obviousconnection to the story. For everything else, there is a (variable) level of ambiguity. At the stage of narration, those levelsare interpreted (and positioned against the opposing poles of the diegetic/nondiegetic). In the passage from anchoring todiegesis, then, the diegetic has the advantage of simplicity, whereas the nondiegetic is always more complex (because itsnarrative level has to be constructed--that is to say, the sound has to be interpreted). In the passage from diegesis tonarrative, the roles are reversed: the nondiegetic has the advantage of (relative) simplicity (since it is already narrative infunction), whereas the diegetic is always more complex (because it provokes the question, Is this environmental sound ordoes it have [is it meant to have] narrative significance?). See Figure 2 for a summary. The first task of diegesis is toconnect to and sort out anchoring at the (secondary) level of the film (the primary level being the individual seated in atheater); the second task of diegesis is to enable interpretation.

31


Figure 2. Interpretation of diegetic/nondiegetic in thestages of the anchoring diegesis narration model.

In the first scene from Catch Me If You Can, the status of "Mele Kalikimaka" is unproblematic (apart from thesmall, unexplained shifts in volume already discussed): the radio is on Carl's desk, inches away from him, and twice wesee him grasp and turn the volume knob. The interpretation of the song's narrative function, however, introduces anambiguity: we are invited to guess what this music, heard in every instance while Carl is on screen, might tell us abouthim, and--still more vaguely--whether that information is of any narrative significance. It is only in a larger context that wefind a way to resolve the question: musical selections heard here and elsewhere confirm Carl's decidedly middlebrowtastes, which fit his character as a relatively low-level bureaucrat, the antipode of the flashy, supremely confident JamesBond character that Frank imitates in the immediately preceding scene. The nondiegetic music near the end of the scenewould initially seem to be as direct as the music from the radio, but that is so only because the conventions ofinterpretation involved are so well engrained that they have become naturalized. First, the music's diegetic status must berejected (once again, the persistent convention of traditional orchestral timbre helps with this). Then, since nondiegeticmusic occupies the register of the narrator, explanations for its "message" must be fashioned. In this instance, conventionis again helpful: a melancholy music enters as we see a close-up of Frank at the moment he reacts to Carl's assurancethat he will be apprehended (see Figure 3). By that point, the narrative functions of the nondiegetic music are secure andinterpretation can proceed uninhibited. (As I suggested earlier, I would favor reading this scene in terms of the oppositionbetween the characters, FBI agent and criminal, with an associated opposition of diegetic and nondiegetic musics.)

Figure 3. In Catch Me If You Can, Frank reacts to Carl'scomment.

32


In the diner scene that immediately follows, the diegetic music is similarly easy to position, but the narrativeassociation is much harder to make. For one thing, "He's So Fine" appears immediately, so that our first choice is tointerpret it as environmental, as an aural component of the mise-en-scne. The diner itself is not important to thenarrative (even if restaurant scenes occur often enough to constitute a potential motif), nor are any of the persons we seeinitially (Who is "He"? Surely not the short-order cook). Indeed, the sound track might be said to cue the listener to anegative result: the environmental sound of rain significantly interferes with the music, and, later of course, the volumelevel retreats in favor of Hanratty's dialogue with the waiter and the entrance of the nondiegetic music, whose status isquickly certified by the orchestral timbre, as in the first scene, and then as quickly interpreted as point of view--Hanrattyreacts to the new information (see Figure 4) and, thanks to the narrative reference in Frank's theme, we realize that agentand criminal are still essentially on a par.

Figure 4. In Catch Me If You Can, Carl reacts to informationreceived.

Spatial and Temporal

The anchoring diegesis narration model asserts that diegetic and nondiegetic are concerned first with spaceand thereafter with time (specifically, the time of narrative unfolding). This accords well with Stilwell's assertion that "theborder crossing [between diegetic and nondiegetic] is not so much an event as a process, not simply a crossing, or evenpassing through distinct intermediary states, but a trajectory, a vector, a gesture. It unfolds through time, like film, likemusic."38 In the paired scenes from Catch Me If You Can, the border is crossed abruptly, without attention to anyintermediary space. In Stilwell's first extended example, the arrival at Skull Island in King Kong (1933), on the other hand,the process moves at a deliberate pace and with considerable attention to the fantastical gap: Max Steiner's mimeticunderscoring that begins with the appearance of the ship in thick fog later enfolds drumming that the characters take asdiegetic (and is eventually proven to be so).

The opening of the recent film Atonement (2007) goes further, to foreground the fantastical gap and play it againstthe diegetic and nondiegetic in service of developing an aural portrait of Briony (Saoirse Ronan), the young protagonist(but also the film's accidental villain). The first sounds are effects--birds chirping, heard against the last two of four studiocredits and quickly joined by a sharp mechanical sound (something like "zzzzt"), which older viewers will recognize as thecarriage return of a manual typewriter and will connect to the atypically simple monospaced font of the third and fourthstudio credits. The letter-by-letter presentation of the main title in the same font (see Figure 5) against the sounds oftyping is no surprise, then, nor is the typing out of "England" during the charmingly self-reflexive "establishing shot" of alarge English country house that follows (We see a scale model of the actual house in which the action takes place).From this point, the camera pans slowly across what is clearly a child's room until it reaches Briony; all the whileobviously diegetic typing has continued off screen, and now we see Briony in action. A cut in to close-up is followed by

33


another to an extreme close-up showing little more than Briony's eyes (see Figure 6)--in the middle of this disconcertinglyintense shot, and with the typewriter sounds continuing undiminished, music starts in the form of a single repeated pianonote not rhythmically synchronized with the typing. The typing stops, Briony pulls out the sheet, gathers her papers, andwalks out of the room in a determined, marchlike gait. As soon as the typing stops, the nondiegetic music takes over andwe hear Briony's theme (also the film's signature theme) over her procession through the house, first to the kitchen andthen to her mother's drawing room (after a detour to talk to a gardener, Robbie [James McAvoy], on whom she has acrush and whose life she will affect tragically before long). All the while, typing sounds continue, now synchronized withthe music (indeed, part of the music); all stop abruptly with two loud chords, both stingers tightly synchronized, the firstwith a door closing and the second with a jump cut to Briony and her mother seated in the room. The mother's twoexclamations of the word "stupendous" immediately following are obviously meant to be parallel to the two stingers: thelong series of accents that began with the first carriage return abruptly shifts away from Briony to her mother and suggestsin the narration, too, the attitude of an adult patronizing a child.

Figure 5. In Atonement, the main title is typed onto thescreen.

34


Figure 6. In Atonement, an extreme close-up of Briony.

The use of a diegetic sound advance against studio logos and credits has become a well-established conventionover the past generation. The bird sounds are thus easily anchored in the diegetic and encourage us to hear the typingnoise also as diegetic because they continue behind it.39 Foregrounded typing noise and background bird soundscontinue until Briony stops and the nondiegetic music takes over the aural foreground. Since we see the typewriter andnote that the room's windows are open, neither is problematic. A process of slippage, though, moves the diegetic into andthrough the fantastical gap. Typewriter noises are linked to the nondiegetic graphics, especially in the rhythmicallysynchronized unfolding of the film's title, but also in the establishing shot, where "1935" appears, improbably, at the soundof a single keystroke--this last, though subtle, is the one moment when we can most readily understand the typewritersounds as nondiegetic, as tied to the narrative register. Insofar as the initial typewriter noises lean toward the nondiegeticand are foregrounded, we can take them as indicative of a probable typewriter motif or perhaps even a theme connectedto writing. The fantastical gap fades back into the diegetic when we see Briony's back and the typewriter over hershoulder, but ironically this plainly diegetic moment also confirms the probable narrative significance of the typewriter andof writing, a realization reinforced by the single piano note, which must necessarily be understood as point-of-view soundbecause the cue enters over the extreme close-up. From this point till Briony gets up and walks out of the room, thediegetic and nondiegetic are presented in parallel, separated but obviously linked by rhythmic similarity. Thereafter,nondiegetic music prevails, and most effects (especially Briony's footfalls but also doors opening and closing) are closelysynchronized with the orchestral cue. The fantastical gap is in full play here, though this kind of muddying the boundariesis not uncommon; we might say that the narrative register of sound takes its cue from the diegetic, as in Mickey Mousing,and returns to it an interpretation of priorities (here the strong rhythms of Briony's walking and the suggestion of a self-focused abruptness in her character). Only when it is clear that the erstwhile diegetic typewriter sounds have become partof the nondiegetic music is the fantastical gap once again erased--almost. The situation at the end is exactly the reverseof the beginning: plainly diegetic sound only slightly confused by some nondiegetic details has been replaced by plainlynondiegetic music only slightly confused by the residues of earlier diegetic sound.

In terms of the process outlined in Figure 2, then, uncomplicated but offscreen diegetic sounds (birds, typewriter)provoke questions about narrative significance because of their nondiegetic elements; these questions are laid aside whenthe sound sources come on screen (the typewriter directly, the bird sounds indirectly through the open windows) but ariseagain when the diegetic sound is folded into the nondiegetic music that dominates the latter half of the scene. The onesupports the other: at the stage of anchoring diegesis, the diegetic sounds at the beginning support the nondiegetic (byminimizing their intrusion and thus minimizing questions of status), and then later on, at the stage of diegesis narration,the nondiegetic similarly supports the diegetic by assigning a clear and obvious narrative role to the typewriter noise as anelement of the march or "hurry" music for Briony's rapid progress through the house.

All in all, the two categories are well defined but just off, and the wavering between reality (diegetic) and fantasyor story (nondiegetic) not only supports a definition of Briony's character but is, in retrospect, an almost alarmingpremonition of the flaws that have tragic consequences not only for Robbie and Briony's older sister Cecilia (KeiraKnightley) but also for others in the household, including Briony herself.

Conclusion35


Whether tienne Souriau misunderstood Plato and Aristotle's meanings of "diegesis" or deliberately recast thosemeanings is difficult to know at this point, but in any case the shift from narrator to the world of the narration was aninspired choice for the cinema, where the depiction of a physical space--the principle of mise-en-scne--is a fundamentaltool. Indeed, mise-en-scne (or diegesis in Souriau's sense) might be considered a third term for narrative agency,between Plato's mimesis and diegesis. If diegesis (in Plato's sense) is a story told directly by a narrator (and thus heavilymediated), then in cinema that would mean a film dominated by voice-over narration and/or first-person address of a maincharacter (as is common in Woody Allen films, for example). Mimesis, as a story indirectly told through characters'actions, would correspond to a film that minimizes voice-over narration, nondiegetic music, and editing. The middle term,mise-en-scne (or Souriau's diegesis), suggests a strong role for the "narration of the visual," or continuity editing, towhich sound contributes in its characteristic roles: anchoring sounds in the physical world, distinguishing between narrativelevels, and encouraging interpretation. Regardless of whether one adopts this three-term structure, the anchoring diegesis narration model can be used to separate out the spatial and temporal (in anchoring and the move to diegesis,and in diegesis and the move to narration, respectively) and to combine them in a process that leads from initialperceptions to interpretation (or criticism).

Endnotes

1 The term spatial anchoring comes from Metz (1985, 158).

2 Kassabian, 4243. Flinn goes further: she explicitly rejects diegetic/nondiegetic as a tool for interpretation (1112).

3 Kassabian, 43; her citation is to Hagen (90). Buhler, Flinn, and I discuss some of the historical circumstances ofindustry practices in the introduction to Music and Cinema (1819).

4 Kassabian, 4849.

5 Thom, 1. The page reference is to the article's PDF file. Five articles are gathered under the title Forum 2:Discourses on Diegesis in Offscreen.com 11.

6 "Metadiegetic" is Genette's term for secondary narration (see note 22); Gorbman discusses its adaptation to filmmusic (2123). "Internal diegetic" is used by Bordwell and Thompson to refer to internal speech (1985, 19394), but anapplication to music is inherently problematic because it is often impossible for the viewer to distinguish between musicthat is being rehearsed mentally by a character and (nondiegetic) music that is meant to represent the character'semotions.

7 Mitry, 72. Hagen, cited by Kassabian (44), uses "dramatic scoring" to refer to nondiegetic music.

8 Taylor, 1. The page reference is to the article's PDF file. Five articles are gathered under the title Forum 2:Discourses on Diegesis in Offscreen.com 11.

9 Bordwell sketches the intermediate history with respect to both mimesis and diegesis in Narration in the FictionFilm (39 and 1617). See also Genette's comparison of Plato's and Aristotle's definitions of these two categories (1982,12833), his commentary on terminological confusion (1988, 1718), and Taylor's account in "Success Story."

10 Taylor, 2.

11 Souriau, 6. The eight adjectives are afilmique (belonging to the real or everyday world of the viewer), cratorial(belonging to the auteur), digtique, cranique (belonging to the screen), filmographique (what is on the film strip itself),filmophanique (belonging to projection), profilmique (the part of the real world set before the camera), and spectatoriel(belonging to the viewer-auditor's response) (79). L'Univers filmique is actually an edited volume under Souriau'ssupervision: its twelve chapters were written by a total of eight authors (Souriau himself among them). Ironically for thepresent purpose, neither Franois Guillot de Rode's chapter on sound nor Jean Germain's chapter on music uses theword digtique.

12 Genette, 1988, 17. According to Taylor, there is a dispute over priority for this application of digse to film (23).

13 Genette, 1988, 18. Despite Genette's explanation, the connection between Plato or Aristotle and the diegetic in36


film studies and narratology continues to be rehearsed in the literature; for example, we read in a recent introductory text:"[Christian] Metz draws on the notion, borrowed from the classical Greek tradition of literary commentary, of the 'diegesis.'In the Poetics, Aristotle uses diegesis to refer to a mode of representation that involves "'telling' rather than 'showing.' In1953, Etienne Souriau revived the term to designate the 'recounted story' of a film, after which it was elaborated byGerard Genette before Metz imported it into film theory" (Stam et al., 38).

14 Metz, 1974, 98. According to Metz, film folds showing (mimesis) and telling (diegesis) into one.

15 Genette, 1980, 228. Gorbman adopted the terms in her early film music and film sound publications butchanged "extradiegetic" to "nondiegetic" in Unheard Melodies (22).

16 Hauch, 534.

17 As Hauch (ibid.) puts it, "For Genette, diegesis and mimesis are not two distinct narrative modes, but rathervarying degrees of mimetic representation which differ according to the distance between informer and information;diegesis is defined by a maximum of the presence of the informer and a minimum of the quantity of information, andmimesis by the opposite."

18 Metz, 1974, 98.

19 Bordwell, 1617.

20 Ibid., 17. The reference is to Barthes's Fashion System.

21 Bordwell, 1718.

22 Gorbman, 1987. The citation is to Souriau (7). Metz also remarks that the term diegesis "was introduced intothe framework of the cinema by Etienne Souriau" (1974, 98). Genette credits Souriau in Narrative Discourse: An Essay(27 n.) and works out the terminology in detail in section 5 of the same volume (21262). The three basic terms fornarrative level are extradiegetic, diegetic (or intradiegetic), and metadiegetic. Pseudodiegesis is a category of themetadiegetic (236). Genette also introduces terms that relate the person of the narrator to the narrative--heterodiegeticnarrative, homodiegetic narrative, and autodiegetic narrative--and combines them with narrative level (24448). Anexample would be extradiegetic-homodiegetic, which in film would be a voiceover narrator who tells his own story.

23 Gorbman, 21; emphasis in the original.

24 Ibid.

25 It is worth pointing out that Bordwell and Thompson used diegetic/nondiegetic in the first edition of theirtextbook Film Art, well before Gorbman introduced the terms to film music studies. The sound chapter from that firstedition is reproduced in Weis and Belton (18199); a chart of spatiotemporal relations is on page 197. Now in its eighthedition (2006), Film Art has been and remains by far the most successful introductory film textbook.

26 Stilwell, 184.

27 Karlin and Wright, 186; emphasis in the original. In his introductory textbook, Karlin puts it more starkly still:"The score, also called 'underscoring' or the 'background score,' is all music on the sound track except that which iscoming from a source on- or off-screen" (68).

28 Karlin and Wright, 187. These are the headings of sections: I have rendered them here in lower case form, nottitle case.

29 Gorbman, 22.

30 Metz 1985, 15455. While he does not necessarily contradict this point, Chion, adopting a post-1970s viewgrounded in the aesthetic of sound design, argues for the musicality of the sound track (in Audio-Vision).

31 Metz 1985, 156, 158. This discussion of Metz is drawn from my commentary in "Melodrama" (86).

37


32 Kerins, 3. The page reference is to the article's PDF file. Five articles are gathered under the title Forum 2:Discourses on Diegesis in Offscreen.com 11.

33 Stilwell, 200.

34 Ibid.

35 Genette, 1980, 244; emphasis in the original.

36 Genette in fact does this: his basic categories for the narrator's status in relation to characters areheterodiegetic (narrator absent from the story) and homodiegetic (narrator present in the story) (ibid., 24447).

37 Under the circumstances it is not surprising that Stilwell's examples focus on subjectivity: it would clearly seemto be a rich field for the ambiguities and complexities that result from movement into and through the "fantastical gap," andmusic is uniquely positioned to take advantage of the situation, given its common role as a commentative replacement forthe voiceover narrator. In the film music literature, readings that touch on spatial ambiguities of music are very common,but Brown focuses specifically and explicitly on the issue in one chapter of his book Overtones and Undertones (6791).

38 Stilwell, 18485.

39 Joe Wright's previous film, Pride & Prejudice (2005), also uses a sound advance consisting of birds chirpingthat starts up immediately after the Working Title logo; these sounds, however, are more prominent and are diegeticallyanchored shortly afterward by an establishing landscape shot.

Bibliography

Barthes, Roland. The Fashion System. Trans. Matthew Ward and Richard Howard. Berkeley: University of CaliforniaPress, 1990.

Bordwell, David. Narration in the Fiction Film. Madison: University of Wisconsin Press, 1985.

Bordwell, David, and Kristin Thompson. Film Art: An Introduction. 8th edition. New York: McGraw-Hill, 2006.

------. "Fundamental Aesthetics of Sound in the Cinema." In Film Sound: Theory and Practice. Ed. Elisabeth Weis andJohn Belton, 18199. New York: Columbia University Press, 1985.

Brown, Royal S. Overtones and Undertones. Berkeley: University of California Press, 1994.

Buhler, James, Caryl Flinn, and David Neumeyer, eds. Music and Cinema. Hanover, NH: University Press of New Englandfor Wesleyan University Press, 2000.

Chion, Michel. Audio-Vision: Sound on Screen. Trans. Claudia Gorbman. New York: Columbia University Press, 1994.

Flinn, Caryl. Strains of Utopia: Gender, Nostalgia, and Hollywood Film Music. Princeton, NJ: Princeton University Press,1992.

Genette, Grard. Narrative Discourse: An Essay in Method. Trans. Jane E. Lewin. Ithaca, NY: Cornell University Press,1980.

------. Figures of Literary Discourse. Trans. Alan Sheridan. New York: Columbia University Press, 1982.

------. Narrative Discourse Revisited. Trans. Jane E. Lewin. Ithaca, NY: Cornell University Press, 1988.

Gorbman, Claudia. Unheard Melodies: Narrative Film Music. Bloomington: Indiana University Press, 1987.

Hagen, Earle. Scoring for Films: A Complete Text. New York: Criterion, 1971.

Hauch, Linda. "Diegesis." In Encyclopedia of Contemporary Literary Theory. Ed. Irena Makaryk, 53334. Toronto:University of Toronto Press, 1993.

38


Karlin, Fred. Listening to Movies: The Film Lover's Guide to Film Music. New York: Schirmer, 2001.

Karlin, Fred, and Rayburn Wright. On the Track: A Guide to Contemporary Film Scoring. 2nd edition. London: Routledge,2004.

Kassabian, Anahid. Hearing Film: Tracking Identifications in Contemporary Hollywood Film Music. London: Routledge,2001.

Kerins, Mark. "Constructing the Diegesis in a Multi-channel World." In Forum 2: Discourses on Diegesis--On theRelevance of Terminology. Offscreen 11, nos. 89 (AugustSeptember 2007).http://www.offscreen.com/biblio/pages/essays/soundforum_2/ (accessed 19 May 2008).

Metz, Christian. Film Language: A Semiotics of the Cinema. Trans. Michael Taylor. New York: Oxford University Press,1974.

------. "Aural Objects." In Film Sound: Theory and Practice. Ed. Elisabeth Weis and John Belton, 15461. New York:Columbia University Press, 1985.

Mitry, Jean. The Aesthetics and Psychology of the Cinema. Trans. Christopher King. Bloomington: Indiana UniversityPress, 2000.

Neumeyer, David. "Melodrama as a Compositional Resource in Early Hollywood Sound Cinema." Current Musicology 57(1995): 6194.

Souriau, tienne. L'Univers filmique. Paris: Flammarion, 1953.

Stam, Robert, Robert Burgoyne, and Sandy Flitterman Lewis. New Vocabularies in Film Semiotics: Structuralism, Post-Structuralism, and Beyond. London: Routledge, 1992.

Stilwell, Robynn. "The Fantastical Gap between Diegetic and Nondiegetic." In Beyond the Soundtrack: Representing Musicin Cinema. Ed. Daniel Goldmark, Lawrence Kramer, and Richard Leppert, 187202. Berkeley: University of CaliforniaPress, 2007.

Taylor, Henry M. "The Success Story of a Misnomer." In Forum 2: Discourses on Diegesis--On the Relevance ofTerminology. Offscreen 11, nos. 89 (AugustSeptember 2007).http://www.offscreen.com/biblio/pages/essays/soundforum_2/ (accessed 19 May 2008).

Thom, Randy. "Acoustics of the Soul." In Forum 2: Discourses on Diegesis--On the Relevance of Terminology. Offscreen11, nos. 89 (AugustSeptember 2007). http://www.offscreen.com/biblio/pages/essays/soundforum_2/ (accessed 19 May2008).

39


diegesis 3

Documents

jstor terms

moving image

film music studies

music andthe

anchoring of sound

terms conditions of

interpretation of film

conditionsfrom music