Prologue
Taiwan, ca. 1920: Hidden behind the small wooden theater, waiting, Li T’ien-lu holds his beloved puppets. Only a few meters away, the fireworks are going off, informing the populace that the puppet theater is in town and that the artists are ready to start. The puppets too are waiting, and in a matter of seconds they will come to life thanks to Li’s unsurpassed artistry and dexterity. Before the deafening sound of the fireworks reaches the most distant corners of the humid valley, the musicians begin with their bells, cymbals, and percussion. Li’s own voice rises fending the air. The show begins. An entire scene unfolds before our eyes in an unedited, stationary shot. As the scene comes to a conclusion, music marks the ending and the curtain is drawn. Over the same music, the image fades to black and there appear on the screen three red, large Chinese ideograms—literally, “Play, Dream, Life.” Their presence carries a strong ritual significance. The film begins.
Thus ends the prologue to The Puppet Master (1993), by Taiwanese filmmaker Hou Hsiao-Hsien. Rolling the title over the diegetic sound of Li T’ien-lu’s musicians, this exquisite prologue makes explicit a suggestion that is the starting point of this essay: just as in the puppet theater, the music we hear during the credits of a film is a culturally coded, socially recognized sign, playing a specific ritual function. To Li’s audience within the film, the music indicates that the game of ma
ke-believe is over, sanctioning a shift back to a normal mode of perception. To the spectator of Hou’s film, the same music is instead an invitation to pick up where Li’s audience has just left off. By cleansing our current auditory field and creating a new sound environment, the presence of the music sanctions our own readiness to turn from mere bystanders into genuine appreciators of a representation, to cross the threshold that leads into a world of appearances. Music, we surmise, is an invitation to imagine, to transform the ensuing sounds and images into paths accessing imaginary places, people, stories.
Beginning Credits
It is now commonplace to claim that music was a necessary component of many ancient rituals, sacred and secular, public and private. But there can be little doubt that in many cultures music was present when initiating an individual to different kinds of rituals. Music—or something we would now define as such—must have helped the participants to revive long forgotten memories, for instance, to strengthen their sense of belonging to a community, to feel the potency of the Gods, or to let themselves loose into a world of fantasy (whether freely and privately, or in a socially coordinated fashion, as in modern spectacles).
Many modern-day rituals still begin with the sound of music: parades, religious celebrations, and performances of all kinds. Differences among these rituals notwithstanding, the continuation of this practice in so many different social occasions betrays a link between them, and it is a hint that at one time the spheres of the magical, the sacred, the civic, and the fantastic were not as far apart from one another as they seem today. Many of us have lost awareness of the secret relations that evolution, culture, and history have woven among our rituals. Many may not even suspect that they are rituals at all, or that they fulfill ancient, deep-seated needs. Only rarely do we experience deference and trepidation for that moment in which we prepare ourselves to participate in them.1
Our perception of something as seemingly prosaic as “credits” music at the opening of a film is directly affected by this kind of situation. In the very distant past, no longer accessible to our memory, music must have been understood as possessing a direct illocutionary force: invitation, persuasion, permission, or even command (in the form of sound) to engage in imagining.2 Given that performances have long become a common, habitual, social phenomenon, such illocutionary force has naturally lost much of its impact. True, under special circumstances music will silence a chatty audience or direct its attention to the fact that a spectacle is about to begin. Perhaps children experience the fullness of that address when they first become acquainted with cinema. But films soon become a habit—all the more so in our consumer society—and thus in our everyday experience the power, urgency, and freshness of that initial address has inevitably been lost.3
Sometimes a director styles the credits sequence in such a way as to revive part of the secret excitement that should be inherent in the passage into the realm of imagination. I have already referred to the prologue of Hou’s The Puppet Master. The beginning of Woody Allen’s Another Woman, a work from a very different cinematic tradition, also comes to mind. Like Hou, Allen begins the film rather abruptly with a pre-title sequence. We hear the voice of Gena Rowland’s character over a shot of her downtown apartment in New York. She says who she is and what she does in a plain, inexpressive tone. Then, just as abruptly as it had started, the pre-title sequence ends and the credits start rolling in the format familiar to Allen fans: white credits in a small-sized, elegant font over a deep-black screen. The soundtrack plays Erik Satie’s Troisième Gymnopédie, as orchestrated by Debussy, giving full expression to Allen’s yearning for “prelapsarian” movie magic.[4 info]
Allen’s musical choice may not seem very original—not only are the Gymnopédies Satie’s best known pieces, but they also have been exploited as a paradigm of cheap, easy sentimentality by the recording industry (an ironic fulfilment of Satie’s own nightmarish vision of the commodification of high art). In a more conventional credits sequence, the choice would have seemed predictable, the effect trite. However, Allen manages to turn a potentially weak beginning into something highly noticeable and powerful by simply moving the credits after the pre-title sequence.4 The delay caused by the pre-title sequence heightens our expectation for the credits proper. When the credits do begin, they naturally come as a release of previously accumulated tension. Moreover, the plainness of the décor and lighting, the abruptness of the editing, and the austere tone of the voice over in the pre-title sequence sharply contrast with the lyrical, euphonious nature of Satie’s music played in the credits. The slow-paced, rocking accompaniment figure, played twice alone (as in the original), is the initial lure we sense. When the notorious melody begins unfolding, the spectator is literally plunged into the poetic world of the film. Satie’s beautiful melody is rejuvenated and the role of music as a powerful mediator between spectators and the story world is greatly enhanced.
“Humanly Organized Sounds”
Strictly speaking, music is not needed to ease the spectators’ immersion into the world of the story. A large number of music-less opening credits testifies to this. In fact, there are cases in which the absence of music is necessary to the success and stylistic coherence of a film. In Woody Allen’s Zelig (1983), for instance, the 
silence “accompanying” the brief credits sequence conveniently underscores the fact that the film is a mock documentary. Music is also conspicuously absent from the beginning credits in all of Luis Buñuel’s later films—from Belle de Jour (1966) to his last work, That Obscure Object of Desire (1977). Far from being merely a capricious rejection of a movie convention, the lack of music in the credits foreshadows the character of the soundtrack as a whole. It is the correlative of the de-dramatized plot and subdued light, colors, and sounds of a world in which the ordinary and the surreal miraculously meet. In a gesture more radical still, Federico Fellini disposes of the credits altogether—except for the brief display of the title—at the beginning of his metacinematic film 8 1/2(1963). In this case, the choice is also highly symptomatic of the film as a whole. In the moment normally allotted for the credits, Fellini reflects the absolute coincidence between the film the protagonist Guido is planning to make and the film we’re watching.
These examples notwithstanding—and there are many others—the use of music over the beginning credits of a film is an overwhelmingly widespread phenomenon. Indeed, it has become a norm that very few films violate, and when this happens, as I have shown, that in itself is perceived as a marked, highly significant authorial gesture fulfilling a highly specialized function. Why? Though it may not facilitate our immersion in the fiction as directly and vividly as in the Another Woman example examined above, music invariably prepares, enables, and signifies the perceptual, cognitive, affective shift that marks the spectator’s psychic life the moment a film begins. It is like a passport into the world of the film, and on the other end it becomes a signpost pointing to our exit. The conventionalization of the practice has made the presence of the music necessary even though the effects that motivated its employment to begin with may be overlooked (or, better, “overheard”). At least, hearing music reassures us that a convention is being honored: it plays because one expects to hear it, and is thus necessary to the extent that it is part of the “decorum” of the spectacle. We might call this the “zero degree” of film music’s effects on the spectator.5
If music signifies a nascent relationship between the film and the spectator, then it must be perceived and employed as “humanly organized sound” (to use John Blacking’s expression). This is to say that upon acting on the listener, film music must mobilize the same mental resources, behavioral patterns, social and cultural conditioning involved in the reception of other kinds of cultural artifacts. Unquestionably one of the main components of the beginning of the spectacle, music does not directly provoke a state of mind appropriate to the appreciation of a film, like a mere physical agent acting on a passive physiological apparatus. Rather, music socializes the emotional and cognitive shift the spectator undergoes in preparing herself for a cinematic narrative. It is a socially constructed mediator, not a material cause in the strict sense.6
Enabler, conventional sign, and invitation to imagine: the significance of this cluster of overlapping functions cannot be underestimated. Not only are these functions the surviving traces of ancient rituals, they also constitute the necessary stepping-stone, both historically and psychologically, for the successful deployment of a whole array of syntactic, expressive, narrative, and symbolic functions within the ensuing narrative. The music played over the beginning credits, in other words, underscores the shift toward the imaginative understanding of not only images and sounds, but also of music itself.
The Ontology of Diegetic Music: Hearing as “Double Intentionality”
The music we hear during a movie comes from the loudspeakers in the theater and is the result of a number of stages: recording, editing, mixing, and playback. One hardly pays any attention to this, as the soundtrack is normally perceived in terms of a source internal to the scene depicted (be it a radio, a voice, or a musical instrument). This anchorage to a source is sometimes called “diegeticization,” and scholars often call this music “diegetic music” (from the word “diegesis,” which in film studies parlance has come to mean “story world” or “fictional world”).7
Despite the fact that during a film we never quite abandon the belief that the music we’re hearing has been recorded and subsequently manipulated, it is undeniable that anchoring music to a diegetic source lowers our moment-to-moment awareness of the production process as well as our awareness of the actual source of emission (the loudspeakers). In this respect, diegetic music functions in the same way as the objects, places, people who populate the movie world. They too “monopolize” our attention, diverting it from the fact that we are seeing a photographic image.8
The spectator is co-responsible for this process. When we hear music in a film as if it were produced by a band playing on-screen, we are abstracting from any considerations about the actual genesis of the sound and its subsequent manipulation during the filmmaking process. We are concentrating instead of what the sound is a sound of, in the context of the story world. This process of perceptual abstraction may have become a habit, but it isn’t imposed upon a passive, defenseless, perceiving subject. It is an act of selection, memory, and ultimately construction, and it is central to the appreciation of many forms of representation, such as the spoken theatre, ballet, or painting. Upon pondering the “mysterious way in which shapes and marks can be made to signify and suggest other things beyond themselves,” art historian E. H. Gombrich has famously named this process “illusion.” Illusion is a charged word, however, fraught with implications of error, deception, and even hallucination.9 Let us refer to this process instead as “imagining.”
It may seem strange to invoke the concept of imagination to explain the perception of diegetic music. What, one may wonder, is there to imagine? Isn’t the music “out there,” audible to anyone in the movie theatre? To be sure, hearing diegetic music at the cinema need not consist in the inner, mental recreation of the sonic manifestation of the music as such, for music is played aloud and is rendered with a high degree of resolution by the recording technology. When I use the word imagining, however, I am not referring to the notion of imagination as a kind of daydreaming, the conjuring up of mental imagery about things absent, or the mental, inwardly completion of things barely sketched as in the everyday usage of the word. Rather, I am using the word imagination as the ability to consider things that are not taken to be real (whether absent or not).
In terms of this definition of imagination, to take Laurence Olivier acting Hamlet on stage, for instance, is to imagine that he is Haml
et. Note that in imagining that he is Hamlet we are not forming a mental image of Hamlet; rather, we are taking Olivier himself, the actor in flesh and blood whom we see and hear on stage, to be Hamlet. Similarly, upon looking at a trompe l’oeil painting of a fruit basket, we are imagining that we are seeing a fruit basket. No inwardly completion of the visual object is necessary, since the painting is replete with information, spelling everything out to the smallest detail and in the most realistic fashion possible. Still, looking at such a painting is to engage in imagining to the extent that looking at what it depicts is simultaneous with the awareness that we are looking at a painting. In this respect, there is continuity between a trompe l’oeil and the most barely sketched outline calling for much projection on the part of the beholder. They are both props in a game of make-believe.10
So it is with a cinematic scene, and the music we hear in it (if any). Hearing the music coming from an on-screen radio in a film is an act of imagination to the extent that we are hearing the music as if it itself were radio music within the movie world, exploring it perceptually as if it were such, and responding to it accordingly. Of course, this also implies that we imaginatively attribute the music’s source to the radio we see or know to be in the scene, despite knowing full well that the music actually comes from a recorded track and reaches our ears via a set of loudspeakers. Imagining consists precisely of the capacity to regard the recorded sound as the result of a chain of virtual relations internal to the story world. We need not think of this as an inner, occult mental activity. It can be described, more simply, as a form of behavior.11
Imagining is to take an interest in appearances, while maintaining intact one’s awareness of the status of the physical substrata that make those very appearances possible—be they actors, pictures, printed words, moving images, or recorded sounds. The simultaneity of these seemingly contradictory stances is a puzzling phenomenon, difficult to express in words (see the epigraph by Ruskin above). Philosopher Roger Scruton has characterized this phenomenon as an instance of double intentionality. Upon looking at a picture of a face, says Scruton, “I am presented with two simultaneous objects of perception: the real picture, and the imaginary face” (87).12 Approaching the question as an experimental psychologist, J.J. Gibson too has formulated the problem in terms of a duality: “The picture is both a scene and a surface, and the scene is paradoxically behind the surface. This duality of information is the reason the observer is never quite sure how to answer the question, ‘What do you see?’” (280).
The contrastive pair of nouns “scene/surface” is not of much help in describing the perception of a diegetic soundtrack, not even metaphorically. But the notion of a dual experience is. Like a face in a picture, diegetic music is the object of twofold perception. The spectator is confronted with a real sound—the recorded sound of the music via the loudspeakers—as well as a virtual one—the music as the imaginary product of agents and causes internal to the movie world. An immense, unbridgeable gulf separates them, both ontologically and psychologically. It is because they are so incommensurable that the real and the virtual (or imaginary) sound do not exclude each other in our perception. They act upon two different areas of the psyche. They occupy adjacent yet distinct spaces. That is why one actually hears both.13
Dual perception is paralleled by dual belief. Note that there is no contradiction between believing the music to be a (virtual) element in the diegesis and believing it to be a (real) sound in a (real) space. Indeed, there cannot be any contradiction between these two beliefs, since they reflect two entirely distinct ways of approaching musical sound. They are mutually exclusive only in the sense that one’s attention can focus on only one at a time.
No matter how much we may be “caught up” in a story, our belief that the music consists of pre-recorded sound emanating via the loudspeakers is never called into question. The belief in diegetic music’s actual absence and impossibility at the moment of its apparent presence or reality—to paraphrase Ruskin—permeates the whole listening experience. True, such belief may not be entertained or asserted on a moment-by-moment basis. Even then, however, it will be present, alive, even nurtured. This is because of the peculiar nature of belief as a mental state. A belief is not so much an occurrent mental act as a disposition or readiness to act in a certain manner. As such, a belief remains “alive” across time independent of whether one consciously focuses on its content or provides external, behavioral evidence that one believes. It follows that to hold certain beliefs about the actual physical status of music need not entail entertaining them on a moment-to-moment basis, let alone articulating them in language.14
Just as our knowledge of the actual nature and provenance of film music is never quite discarded, so is “disbelief” about its fictionality never “suspended,” to quote Coleridge’s notorious expression. During a film, in short, beliefs concerning the actual status of the music heard are neither discarded nor suspended. They are, more simply, nonoccurrent. The perception of diegetic music differs from an illusion proper because it does not entail that we discard our beliefs about the actual status of the sound in the physical world we inhabit.15
That said, these beliefs make themselves felt. But how? As a readiness to differentiate a failure of the power system from a silence in the make-believe world of the story, for example; or as a kind of safety blanket enclosing the film-goer’s listening experience. The firm conviction that diegetic music exists in a purely fictional realm “frees” the filmgoer, as it were, enhancing her sensitivity to its dramaturgical and poetic roles. In this respect, the reception of diegetic music merely mirrors the reception of representation in general. The firmness of the spectator’s belief in the fictional status of the narrative, makes her more vulnerable to a range of responses that are not as readily available when we observe human dramas being played out in our real world.16 There is a sense, then, in which the intensity, richness, and vivacity of our absorption into a world of make-believe is evidence not of a deceptive, illusory state of mind but rather of the opposite—an all too encompassing understanding of the nature of representation and the subject’s relation to it.
Frame, Object, and Vector
The concept of imagination is central to our understanding of both diegetic music and the music we hear during the credits. Reference to imagination also allows us to acknowledge important differences between them. Ambiguity is possible and highly interesting, as when diegetic music is employed as a substitute for the title theme proper. However, the functions they fulfill are typically quite distinct: while credits music prepares, coordinates or signals our engaging in imagining, diegetic music is instead an object of that imagining. Metaphorically, credits music is like a frame around a painting, marking it off from the adjacent space, encouraging a certain imaginative appreciation of it, and highlighting its status as representation. Diegetic music, on the other hand, is more like an element inside that very painting, just like the nose of a face in a portrait, a piece of furniture in an interior, or a bush in a landscape. The music played by an on-screen instrumentalist is part of the soundscape of a scene in something like the way the trees outside are part of its landscape.
External, or “nondiegetic” music—what most people intuitively think of as film music—participates in both roles. It can be both frame and object, sometimes simultaneously. Like credits music, and like the muzak we incessantly hear in stores, bars, shopping malls, it reminds us where we are and what we are there for, especially when it plays profusely as in certain Hollywood films of the 1930s and 1940s, for instance, or most contemporary Hindi films. When this happens, nondiegetic music provides an acoustical frame through (and not just before and after) the film, thus bridging the gap between the beginning and end credits.
At the same time, nondiegetic music may also point to elements inside the story world.17 As many will recall, the shark’s approach during the first attack in Steven Spielberg’s Jaws (1975) is conveyed through the combined effect of a camera movement below sea level and nondiegetic music in the soundtrack. This example is quite typical, as nondiegetic music—unlike diegetic music—never points to itself as music but rather stands for something else within the story world. The Jaws example is also typical of what nondiegetic music most commonly stands for: processes or events, both mental and physical (like the approaching of an unseen character, the foreshadowing of a future deed, a shift in a character’s mood, the flashing of an idea, etc.). Finally, the example is typical in that the effect is achieved in collaboration with other components of the film (in this case, a camera movement) and its success is contingent on at least a minimal knowledge of the story’s events and circumstances.18
This does not exhaust our overview of the manifold roles played by nondiegetic music. We still need to consider what we all sometimes superficially refer to as “setting the tone” or “the mood” of a scene. In these cases music acts like a vector: it directs our attention toward a certain element or a particularly meaningful aspect of a scene, guiding us to a certain understanding and a certain emotional response to it.19 Nondiegetic music passes through our mind, as it were, before we process the content of the images and the words of the characters, filtering everything we see and hear, think and feel. We, the spectators, adapt to the “slant” thereby produced just as we naturally adapt to the point of view suggested by the camera. This parallels the shaping of our attention and emotional responses achieved through camera position and angle, framing and composition, and, of course, acting.20 Nondiegetic music also contributes to indicating how the scene relates to that which precedes and that which follows—spatially, temporally, and thematically—integrating what is done through editing (sometimes reinforcing it, sometimes working against it). We may refer to this as its syntactic function.
The “Inaudible” Orchestra
As is the case with diegetic music, the actual provenance of nondiegetic music (the recording sessions, and the manipulation it undergoes during a film’s production process) goes unnoticed during a narrative film. The fact that this mass of potentially distracting information rarely comes to our attention might seem a deliberate attempt to efface cinema’s sound technology. The actual source of the music is “effaced” not only to the eyes but also to the ears of the spectators. During a film the orchestra may well be invisible or ghost-like (as in operatic practice since Wagner). However, we might just as well say that it is “inaudible” as an orchestra. Think of how film music is orchestrated and recorded. Normally, the acoustics of the music are homogeneous across an entire film. Volume is constant. Reverberation is usually absent. Moreover, in many cases the orchestration is such that no soloists or instrumental groups can be easily discerned. “My second rule [concerning the use of music],” says director François Truffaut, “is to avoid instruments that are too easy to identify or visualize, like the piano or the harp”.21 We may or may not disagree with Truffaut on this point, but it is hard to deny that any combination of these features is instrumental in diverting our attention from the performing and recording aspects of the music heard.
A number of editing techniques also work to “efface” the presence of the music. This is particularly true with respect to the in- and out-points of the music cues (which are all but noticeable in most conventional narrative films). I am thinking of what is called sneaking in, the cueing in of mu
sic at low volume underneath dialogue; or of the overlap, i.e. the cutting of nondiegetic music during an overlap with diegetic sound. These scoring practices are common not only within Hollywood, but they also respond to needs and preferences spread across many different filmmaking traditions. Bertolucci’s film, The Conformist(1970), for instance, features a classic case of sound overlap right after the end of the credits. As soon as the title theme ends, a new music cue abruptly takes over. Its obsessive repetitiveness and extreme timbre point effectively to the state of psychological breakdown suffered by the protagonist. But the music also posed a problem: how and when to make the cut, without calling attention to it? The solution found was a most traditional overlap: the music stops as a train of the Parisian métro enters the frame. The train produces a loud noise and thus “covers” the cut in the soundtrack.
Hollywood composers have openly referred to the use of devices such as the overlap in terms of effacement.22 Efficacy of recording, compositional and editing practices notwithstanding, the role of the spectator in minimizing the obtrusiveness of the sound apparatus can hardly be underestimated. Representation demands that attention to the technical apparatus and the network of efficient causes giving rise to a sound be strongly attenuated. Interest in representation, as a broadly anthropological fact, is an interest in appearances to begin with, not in the physical or causal processes giving rise to the stimuli that form the ground of those appearances. It is in the spectator’s own interest that the music heard during a film be translated as effectively and continuously as possible into an incitement to imagine, or an attribute of an imaginary world, or a filter better to understand and fully respond to what happens in that world. This is a state of affairs reminiscent of the economy of language comprehension. During a conversation, language conveys semantic, affective, or pragmatic content taking up the whole of our attention. As a result, attention to speech as the product of a physiological apparatus or as a string of mere phonemes is considerably reduced, as if continuously deferred, despite the fact that the “apparatus,” i.e. the speaker, may be sitting in front of her interlocutor.
To the extent that the most common composing and editing conventions enhance continuous and single-minded participation in the narrative, then, they reflect concerns as old as representation itself. However, given the seemingly inexhaustible interest in representation across cultures and the propensity for imaginary absorption into a narrative, the question arises whether Hollywood filmmakers overestimated the disruptive power that the visible or audible traces of their work might have on the spectator.23 The degree of discretion and strategic placement of the music cues is so much in excess of the minimal conditions for following a narrative that it must be interpreted as a function of style (and not as an inevitable development dictated by the nature of human perception or cinematic representation). But why this style and not others?
Claudia Gorbman has come close to an answer through a reference to the alleged “displeasure” produced by one’s awareness of cinema’s technological mediation. After discussing how music wards off the “displeasure of uncertain signification” by anchoring a potentially ambiguous image in meaning, Gorbman goes on to say:
A second kind of displeasure that music helps to ward off is the spectator’s potential recognition of the technological basis of filmic articulation. Gaps, cuts, the frame itself, silences in the soundtrack—any reminders of cinema’s materiality which jeopardise the formation of subjectivity—the process whereby the viewer identifies as subject of filmic discourse—are smoothed over, or “spirited away” (recall Eisler and Adorno’s view of music as magical antidote to the picture”) by the carefully regulated operations of film music. (58)
A summary of “subject positioning” that Gorbman refers to and the debate that has developed around it lies beyond the scope of this essay.24 Suffice it to say that with their emphasis on cuts, gaps, or the frame as “disruptive” of the diegetic illusion “subject positioning,” theorists give—in my opinion—too much weight to the film as an already constituted object at the expense of the film as reconstituted by the spectator’s mind. I hope this crucial difference will emerge more clearly in the remainder of my discussion.
Gorbman’s passage would seem to contain an implicit answer to the question: “Why this style?” For if one of the main functions of film music is to “smooth over” cuts in the image track, gaps, silences, the effect of the frame, and so on, then by the same token music must also “efface itself,” as it were, “smoothing over” its own cuts, gaps, and silences. Hence the particular style of scoring we normally associate with mainstream narrative cinema. It may well be true that producers self-consciously effaced all traces of the recording, mixing, and playback process for the sake of unity and cohesion. However, the question remains whether their choice was dictated by what they deemed necessary or, rather, by stylistic preferences. If we were to conclude that the effect of widespread soundtrack practices is the impression of unity, cohesion and absence of any technological mediation, we would be implying that such practices are indeed necessary to create this impression to begin with.
A moment’s reflection tells us that this is simply not the case. The spectator can “smooth over” her viewing experience by restructuring it mentally as more continuous and homogeneous than it appears externally (either during the film or retrospectively, or both). She can appreciate a representation as unitary and coherent even when the apparatus is flaunted or even defective. After all, isn’t the frame of a painting fully visible without disturbing the one-ness and coherence of what one is contemplating? The unity, coherence, and “smoothness” of the contemplated object exist only in the beholder’s imagination—if at all—and are as much the result of her skill, interest, and level of attention as the external facets of the work. Correspondingly, a spectator can form the impression of harmony and integration between music and narrative under a variety of circumstances. When traces of the recording process do survive in the finished product, she can marginalize them. She can learn not to pay attention to them through habituation, just as when playing an old record one increasingly learns to ignore the sound of the scratches.
By the same token, ensuring that no attention be given to the performing apparatus per se need not involve the apparatus being “invisible” or “absent”. The existence of many theatrical traditions in which little is done to efface the musicians or their instruments proves that, pace Wagner, music can continuously and effectively serve the need of representation even when a musical ensemble is fully visible and the production process flaunted. Take opera performance practices before Wagner. Or take Hou Hsiao-Hsien’s The Puppet Master again. True, the musicians are normally behind the small wooden theatre, but they care little if someone sees them out of the corner of her eye.25
The use of music in such forms of representation as theater, marionettes, ballet, or opera is contingent on the idea that the spectator will form the impression of at least some degree of cohesion and mutual implication between music and the other components of the work.26 This is also true of many films whose scores differ markedly from the scores of the so-called classical period of Hollywood cinema. What is so peculiar to the latter is not the impression of cohesion and absence of mediation they produce but, rather, the fact that cohesion and absence of mediation are inscribed onto the soundtrack itself before the spectator can produce them.
What does this entail? There is something greatly satisfying about the subtle, carefully timed and highly functional scores of such films as Otto Preminger’s Laura (1941), Billy Wilder’s Sunset Boulevard (1950), or Alfred Hitchcock’s Vertigo (1958) (music by David Raksin, Franz Waxman, and Bernard Hermann, respectively). Even less celebrated films display a complex interplay of music and narrative and a rich array of effects. This is because Hollywood scoring practices have resulted in a tradition of matching music to images. Such devices as the overlap, the sneaking in, or the use of a title theme marking the genre of the film—to make three examples—are the expression of this tradition and contribute to a “language” of their own as rich as any other.
On the other hand, they also represent a style whose goal seems to reduce to a minimum the cooperation of the audience, as if the spectator had to be “spoon-fed,” so to speak, in order to weld music to narrative. To be sure, the obsessive need to efface the in- and out-points of the music cues eventually became (and still is) a sign of compliance with a crystallized style. However, it must have originated from the fear that the spectator might be incapable of transcending his awareness of the material processes at work in making the representation possible. In less fastidiously constructed forms of spectacle, the spectator’s relative lack of attention to the production process is also the result of the intensity of his imaginary involvement with what is being represented, not only the facets of the work or the circumstances of reception themselves. Focus on the represented content as well as its foil, momentary disregard for the apparatus, emerge out of negotiations and adjustments taking place between the spectator and the work (or performance). Habituation sometimes makes these adjustments unnecessary.
Not so in Hollywood mainstream cinema: instead of relying solely on the audience’s capacity to concentrate on the narrative, Hollywood filmmakers have deemed it necessary to cleanse their soundtracks systematically of any signs of intervention. By the same token, experimentation is discouraged, lest the spectator be “distracted.” This implies the simplistic idea that the material facets of the text alone are responsible for the cohesion and unity between music and narrative experienced by the spectator, irrespective of his capacity to construct that cohesion through imagining. Inscribed in many a Hollywood score is the derogatory view of an unimaginative, inattentive, listening subject, easily distracted by either the apparatus or any deviation from the norm, or both. Anxiety over unity and cohesion, then, is not a function of representation tout court. Rather, it is a function of a certain view of the listening experience at the cinema and, indirectly, the spectator’s mind.
Performance versus Reproduction
Until the advent of the sound film, musical accompaniment has always required an “apparatus” of sorts, be it an orchestra, a small ensemble, or a single instrumentalist. In this respect, there is continuity not only between the sound film and the silent film but also between cinema and other forms of representation. However, the painstaking, fastidious way in which the beginning and stopping points of the music were treated indicate that mechanical reproduction must have been perceived as a genuine break in the relationship between performing apparatus and visually depicted narrative. I take it that Hollywood composers, filmmakers, and sound engineers were following their listening habits when they set out systematically to efface signs of their intervention on the recorded soundtrack. In particular, I am referring to their understanding of the different aesthetic status of live and recorded music, respectively.
Those who have attended a film show with live musical accompaniment will have noticed how easily one’s attention shifts from the performing group to the narrative. This applies also to those cases in which the orchestra is fully visible (proof, if any were needed, that the invisibility of the musical ensemble is not a necessary condition for music/image relationships to obtain). Whenever there occurs a shift in attention away from the narrative and a refocusing on musical accompaniment per se, one is still witnessing a performance. This makes it relatively acceptable. On the other hand, attention to mechanical reproduction, as opposed to performance, always seems undesirable. Why? Though it would be an exaggeration to claim that live musical accompaniment possesses the “aura” Benjamin famously talked about, it is undeniable that spectators regard live accompaniment in terms of an age-old aesthetics of “the event.” This makes it a relatively acceptable, and sometimes even desirable, focus of attention.27 It goes without saying that an aesthetics of the event—as I have called it—was simply not available for the processes of recording and reproduction, which replaced live performance with the advent of sound.
This interpretation is meant to reopen, or at least historicize, the question of whether there is something inherently displeasing or disruptive about mechanical reproduction per se.28 It also implies that if an aesthetics of the recording or reproduction process were to emerge, the preoccupation with the traces of such processes would perhaps vanish. Admittedly, the prospects for such a development look bleak, except perhaps in the realms of the avant-garde film or video art. The history of sound recording practices outside cinema confirms that the dominant approach to the reproductive technology in the 20th century has privileged function over contemplation, effacement over exposure.
For a functionalist, recording and reproduction are not events worthy of attention in themselves. When they survive at all, he appreciates audible traces of the processes of recording and reproduction only insofar as they point to a musical performance, the time or place in which it took place, or the agents involved in them. He treats these traces as “evidence,” not as objects of appreciation. Not surprisingly, in most studio recordings, they are nowhere to be heard. In the recordings of as radical a performer as Glenn Gould, one hears his own singing with the music, his legendary squeaking, or the creaking of his broken chair. Everything else is cut out as unnecessary. This is in keeping with Gould’s intention to foreground his own presence—not the technology he is using or the manner he is using it—as the focus of the listener’s attention.
“Passing Unnoticed”
Mainstream narrative films often feature moments of loud music and abrupt editing which pass just as “unnoticed” as the faintest string underscoring. How is this possible? The answer lies in what we mean by “unnoticed.” What goes “unnoticed” is not the music tout court but some of the information it carries with it—some of its formal features, when and where it was recorded, what kind of ensemble it used, who composed it, how it was edited, and so on. This information could potentially be the focus of our attention, but it is instead pushed to the margins of our field of attention even when the music is loud and the editing abrupt. However loud the music or abrupt the cut, the loudness of the volume and the abruptness of the edit are perceived as integral to our experience of the narrative. They are translated immediately into components of what psychologists of reading call our “semantic representation” of a fictional world. It is only in this sense that the music “passes unnoticed.”
When this happens the spectator simply does not process the music as “loud music” or the cut as an “abrupt cut,” but rather as paths of access to, or attributes of, the represented content of the scene. When Norman Bates attacks Marion in the famous shower sequence in Hitchcock’s Psycho, for instance, the music literally assaults the viewer. But the viewer has neither time nor interest in dwelling on the music’s abnormal volume or the abruptness of the edit per se, as these features are immediately turned into signifiers of the murderous folly of the protagonist himself (especially upon viewing the film for the first time).29
The Psycho example indicates that it is impossible to understand the perception of film music by reference to the external facets of the stimulus alone. “Effacement,” as it has been called, obtains even when the music is all but soft and discrete, and the editing abrupt and unsubtle. To claim that Bernard Herrmann’s loud, obtrusive music for the shower scene is self-effacing may seem paradoxical. However, this paradox need not detain us long, as we can easily renounce the notion of effacement, along with the negative connotations of “inaudibility” or “unnoticeability” it carries, and substitute it with that of absorption in an imaginary world. Once we acknowledge that the goal pursued by filmmakers is absorption, it will seem both sensible and desirable to view loud, violent gestures on the same continuum with subtle editing and delicate underscoring. Whether there is absorption depends on the relationship between film and spectator, not on the features of the soundtrack alone.
Hitchcock’s famed collaborator had a special gift for spotting in a finished film places in which loud music or abrupt cuts would be both technically possible and highly effective. The shower scene in Psycho is a case in point. Herrmann’s perspicacity and confidence in using music in that scene becomes all the more apparent when one learns that Hitchcock originally did not want any music in it. In Vertigo, also scored by Herrmann, a sudden, loud statement of the habañera motif in the brass marks Scottie’s realisation that the necklace Judy is wearing is the same as Magdalene’s.
In the same film, the music underscoring Scottie’s first exploration of Carlotta’s old house is suddenly cut by the voice of the day porter. Far from calling attention to the cut itself, the halting of the music renders quite wonderfully the reawakening of Scottie’s consciousness to the external world after speculating and daydreaming about the alleged ancestor of Elster’s wife. Near the end of the film, the music stops even more abruptly as Judy, after confessing her deeds to Scottie on top of the Mission Tower, sees the nun coming up the tower stairs. The cut wipes out the music, conveying unequivocally Judy’s sudden shift of attention, coupled with surprise and fear, due to the appearance of the nun (only the faint sound of an Hammond organ is audible). The ensuing silence, one of the most tension-ridden silences in all of film, is broken by a few words whispered by the nun first, and then even more violently by Judy’s scream as she throws herself off the tower.
The final event of Vertigo is conveyed solely through aural means. From the moment Judy starts screaming until her death the camera alternates between close-ups of Scottie and medium shots of the nun. Thus Judy’s fall and her death remain unseen. As her body crashes on the roof, Judy’s scream also comes to a sudden halt. In its place, we hear a loud B played by tuba and trombones, accompanied by a single, loud blow of the timpani. The loud entrance of the music is unprepared; nothing is done to “cover” it. Herrmann must have been confident that the music would naturally fill in the psychological void left by the impossibility of actually seeing the impact of the body on the roof. So driven is the horizon of expectation at that moment, so certain is the outcome of Judy’s deadly fall, that when the low B strikes, it fulfills our need for some sensorial evidence of the unseen impact of the body on the roof. Moreover, the entry of the music is timed in such a way as to give the impression that the sound originates from the crash itself.
The dark sound of the brass also takes on a number of expressive overtones: it may connote Judy’s suicide as being sudden, incomprehensible, cruel; it may point to the irreversibility of her destiny, or to Scottie’s shock at witnessing her death, again. The semantic indeterminacy of the unison allows for a degree of interpretive openness, of course. The effect of the sound on each spectator depends on one’s interpretation of the film up to that moment and the role of the nun’s sudden appearance in Judy’s decision to kill herself. The orchestral sound also produces an effect of stylization, abstracting the incident, as it were, and tempering considerably the suicide’s most gruesome aspects.
Though it may at first appear like an isolated occurrence marking the crash of the body, the low B is in fact the first note of a two-note chromatic motif (B-C). The same music is heard over the shots of the dead body of Scottie’s colleague at the beginning, and of Madeleine’s body at the end of the first half of the film. It is a highly recognizable musical gesture that neatly marks the film at beginning-, mid-, and end-point. The motif is played repeatedly as the camera moves from a brief shot of Scottie to another medium shot of the nun. The music decreases in volume to underscore the nun’s reaction to the suicide. As the nun starts ringing the large tower bell, the volume of the music rises again so that the orchestra does not drown in the loud, metallic sound of the bell. It would be unfair to characterize the combination merely as a perceivable overlap of diegetic and nondiegetic sound. The two sounds—the brass and the bell—amalgamate to form a sonic complex in which the traditional border between diegetic and nondiegetic becomes fuzzy. On the one hand, the music seems like a component of the diegetic soundscape; on the other, the sound of the bells comments upon the scene nondiegetically as well as being a realistic element of it.
Outer versus Inner Sounds
Metaphorically speaking, our reaction to Judy’s suicide in Vertigo inwardly produces the equivalent of a loud, purely mental “noise” resounding all through our minds. Thus, when we first hear the low B in the brass and the blow in the timpani, there is a kind of overlap between the sound of the orchestra and the emotional turmoil provoked by Judy’s suicide. I believe that the metaphor captures a psychological truth, as it is precisely our own emotional state of shock, and not another sound, that “covers” the loudness of the soundtrack. The metaphor also shows that our vulnerability to potentially distracting facets of the music does not depend solely on such physical invariance as loudness, timbre, or timing of the cue. While loudness is a physical feature of the music, “obtrusiveness” is a psychological category. What is processed as “loud” physiologically may not be interpreted as obtrusive. This is because the extent to which film music infiltrates the imaginative experience of the diegesis, thus “passing unnoticed,” depends at each moment on one’s attentional focus and emotional situation.
There are many examples in mainstream narrative cinema of such “overlaps” between outer and inner sounds. Hitchcock’s equally well-known classic, North by Northwest, produced only a year after Vertigo and scored again by Herrmann, features a spectacular and exemplary moment in which a loud orchestral music cue is edited without preparation. I am referring to the oft-cited sequence in which the protagonist, Roger Thornhill (played by Cary Grant) is attacked by nothing less than a crop-duster plane while waiting for an agent in the middle of a huge, flat, deserted field in Indiana.
As those familiar with the film will recall, the agent Thornill is waiting for never turns up. However,the attack on Thornhill fails as the plane crashes into an oncoming fuel truck. At this point Herrmann’s music, scored for the whole orc
hestra, explodes at maximum volume, presenting a variant of the title-theme. After the crash, we see the truck drivers precipitously leave the vehicle and then the explosion of the fuel tank.
The example is particularly revealing because music occurs after several minutes of almost total silence, and yet nonetheless “passes unnoticed.” The recent memory of the eight minutes of almost total silence does nothing to heighten our sensitivity to the suddenness and bombastic quality of the music cue. The in-point of the music is “covered” by the sound of the crash itself, of course, but also by our involvement with the unexpected, spectacular conclusion of the episode. So fast is the escalation of events, such is our preoccupation with the crash and its consequences, that there is no psychological room to perceive the music as anything but strictly integral to our imaginative absorption in the action.
As it is cued in just after the explosion, the music seems to result from the deflagration itself. In other words, the timing of the cue establishes a virtual, imaginary cause-effect relationship between the explosion and the appearance of the music. Moreover, the music participates in the deflagration, becoming part of the diegetic soundscape of the scene, and magnifying the sensorial impact of the explosion on the spectator. Finally, as in the Vertigo example described above, Herrmann’s music actualizes, prolongs the spectator’s own emotional state of shock and excitement.30 As the scene comes to a close, the image track takes us far away both in time and space from the site of the accident. Accordingly, Herrmann’s music attenuates its role through a swift thinning out of texture and the deintensification of the thematic work in the strings and brasses. It is as if the music were voicing the spectators’ own releasing of emotional energy following the spectacular climax. A simple, two-note descending motif played by the low clarinet underscores the dissolve to the shot of the truck stolen by Thornhill parked in downtown Chicago.
A Cut into Silence…
Filmmakers, then, have also operated on the assumption that the spectator’s imaginary experience of a film can accommodate the most abrupt and seemingly “obtrusive” musical gestures.31 In Stanley Kubrick’s 2001: A Space Odyssey (1968), there is a sequence that not only confirms this assumption but also provides a glaring instance of how a filmmaker endows a simple procedure with new meaning. I am referring to the well-known sequence in which the black monolith appears for first time before an audience of astonished “primitive” creatures, half-monkeys, half-men. The music cue begins just before the first appearance of the monolith: a faint, dark sound in the background, rather like a low hiss, perhaps emanating from the object itself, over a medium shot of the waking creatures at dawn.32 As the creatures notice the object and gather around it, the initial sound rises in volume and becomes recognizable as music. It is Gyorgy Ligeti’s “Kyrie” from his 1966 Requiem: a sour, non-tonal piece for chorus and orchestra, moving slowly in massive, interlocking shards of sound in which no single voice or instrument is discernible. This is music that borders on the archetypical, perhaps the closest Kubrick could conceive to an utterly unfamiliar, transcendental, non-referential sound.33 
The sequence concludes with a shot of the monolith seen from below at dawn. The way the shot is composed conveys the mysterious, possibly supernatural, power of the object more strongly than any of the preceding shots. We see the upper edge of the object barely shielding us from the blinding light of the rising sun. The edge of the monolith, the sun, and the moon are all aligned toward the receding point of the composition. Explicitly centered, the figure of the monolith is made more overwhelming and foreboding by the distortion of the visual field due to the use of a grand angle. Finally, the music greatly enhances the effect of the image, since it is played at this point at maximum volume, literally saturating the spectator’s auditory field.
What follows is an abrupt, unanticipated straight cut both in the image and the sound track to a long shot of a large, desert plateau, clothed by a mantle of almost absolute silence. There isn’t a single reference to this extraordinary moment in the immense literature on this film. The sensorial shift is so sudden and violent that the deafening sound of the previous shot may seem to persist even after the cessation of the sound stimulus itself (a kind of auditory equivalent of an after-image). In narrative terms, the cut conveys with supreme economy that the new shot depicts the earth at an indefinitely later time—thousands, perhaps millions of years after the appearance of the monolith as seen in the previous scene.
I would argue that the abruptness of the cut does not lessen the intensity of the spectator’s absorption, since her attention is wholly captured by its meaning. The desolate landscape is conspicuously lacking in any signs of its position in time, and yet the indication that a large amount of time has intervened between the two shots is very clear. Kubrick decided not to employ the editing figures traditionally used to convey the passage of time, like the fade out, the dissolve, or the montage sequence. Instead, he devised a new, bold solution, one that he must have found commensurate with the scale and the scope of this crucial transition. The moment is crucial because it “trains” the spectator to understand that the temporal course of the story is vertiginously driven toward the future, familiarizing her with the pace at which man’s history on the earth will be narrated in the remainder of the film—by means of jumps of thousands of years at a time, until eternity.
A World of Appearances
As I have observed, the transition to recorded sound entailed the shift not only from a visible, live ensemble to an invisible, ghostly one, but also from performed to reproduced music. This did not prevent films from being washed in music in the manner of the silent film for quite some time. Many scores from the early 1940s are not immune to it.34 The soundtrack of Kubrick’s 2001: A Space Odyssey may be viewed as a partial revival of the practice of matching long stretches of pre-existing music to silent, dialogue-less visual sequences (most notably in the famous “spaceship-waltz” sequence, accompanied by Strauss’s Blue Danube). In The Gospel According to St. Matthew(1964), Pier Paolo Pasolini used excerpts from Bach’s St. Matthew Passion through entire episodes, creating the impression of isomorphism between musical and narrative development within each episode. Luchino Visconti used lengthy excerpts from Mahler’s 5th Symphony and Wagner’s Tristan und Isolde in both Death in Venice (1971) and Ludwig (1973), respectively, exploiting to great effect the ambiguity between music as a projection or memory of the protagonist and music as independent narrative voice.
Leaving individual differences aside, this style of matching music to narrative is reminiscent of the silent film aesthetics in that musical effects are diffused and the soundtrack, if not a performance, can be grasped and appreciated as a relatively independent component of the representation. It must be said, however, that since the 1930s, the dominant trend has privileged pointed, highly localized effects. Mechanicization allowed filmmakers to begin experimenting with increasing precision and control in their attempts to coordinate music and the image track. Though continuous and hyperexplicit in a way that would later become old-fashioned, scores from as early as the mid-1930s show that filmmakers were able to tailor music to narrative action with a precision impossible to achieve with any other medium before.
Scores eventually became more fragmented. Painstakingly timed, intermittent scores are still with us today. The fact that audiences hardly experience a film score as intermittent—let alone fragmented—would seem to justify Chion’s claim that “each audio element enters into simultaneous vertical relationships with narrative elements contained in the image (characters, actions) and visual elements of texture and setting.”35 In this essay I have been at pains to suggest that such “vertical relationships” arise out of the encounter between a set of visual and auditory stimuli and a perceiving subject, who recomposes them in an imaginary world of appearances.
Giorgio Biancorosso
Princeton University
Works Cited
Carolyn Abbate, “Outside Ravel’s Tomb.” Journal of the American Musicological Society 52:3 (Spring 1999): 465-530.
Altman, R., McGraw Jones, and Sonia Tatroe. “Inventing the Cinema Soundtrack: Hollywood’s Multiplane Sound System.” Music and Cinema. Eds. J. Buhler, C. Flynn, D. Neumeyer. Hanover: Wesleyan University Press, 2000. 339-359.
Anderson, Tim. “Reforming ‘Jackass Music’: The Problematic Aesthetics of Early American Film Music Accompaniment.” Cinema Journal 37:1 (1997): 3-22.
Benjamin, Walter. “The Work of Art in the Age of Mechanical Reproduction.” Illuminations. Trans. Harry Zohn. New York: Schocken Books, 1969.
Blacking, John. How Musical is Man? Seattle and London: University of Washington Press, 1973.
Bordwell, David. Narration in the Fiction Film. Madison: University of Wisconsin Press, 1985.
—. “Contemporary Film Studies and the Vicissitudes of Grand Theory.” Post-Theory: Reconstructing Film Theory. Eds David Bordwell and Noël Carroll, Madison: University of Wisconsin Press, 1996. 3-36.
—. “Exceptionally Exact Perceptions: On Staging in Depth.” On the History of Film Style. Cambridge: Harvard University Press, 1997.
Brandi, Cesare. Teoria generale della critica. Turin: Einaudi, 1974.
Brecht, Bertolt. “On Film Music.” Bertolt Brecht on Film & Radio. Trans. and ed. M. Silberman. London: Methuen, 2000.
Carroll, Noël. Mystifying Movies: Fads and Fallacies in Contemporary Film Theory. New York: Columbia University Press, 1988.
—. The Philosophy of Horror, or, Paradoxes of the Heart. New York: Routledge, 1990.
Chion, Michel. Audio-Vision. Trans. Claudia Gorbman. New York: Columbia University Press, 1994.
—. La musique au cinéma. Paris: Fayard, 1995.
Genette, Gérard. Narrative Discourse Revisited. Trans. Jane E. Lewin. Ithaca: Cornell University Press, 1988.
Gibson, James J. The Ecological Approach to Visual Perception. Hillsdale, NJ, and London: Lawrence Erlbaum Associates, 1986.
Gombrich, E. H. Art and Illusion. Princeton: Princeton University Press, 1960.
Gorbman, Claudia. Unheard Melodies. Bloomington: Indiana University Press, 1987.
Hagen, Earl. Scoring for Films. Los Angeles: Alfred Publishing, 1991.
Heath, Stephen. Questions of Cinema. Bloomington: Indiana University Press, 1981.
Marks, Martin. Music and the Silent Film. New York and Oxford: Oxford University Press, 1997.
Metz, C. Le signifiante imaginaire. Paris: UGE, 1977.
Nietzsche, Friedrich. The Birth of Tragedy and Other Writings. Eds. Raymond Geuss and Ronald Speirs. Trans. Ronald Speirs. Cambridge: Cambridge University Press, 1999.
Odin, Roger. “L’entree du spectateur dans la fiction.” Théorie du film. Eds. J. Aumont and J.L. Leutrat. Paris: Editions Albatros, 1980.
Paulin, S. “Richard Wagner and the Fantasy of Cinematic Unity.” Music and Cinema. Eds J. Buhler, C. Flynn, and D. Neumeyer. Hanover and London: Wesleyan University Press, 2000.
Price, H. H. Belief. London: George Allen and Unwin, 1969.
Rouget, Gilbert. Music and Trance: A Theory of the Relations between Music and Possession. Trans. Brunhilde Biebuck. Chicago and London: Chicago University Press, 1985.
Scruton, R. The Aesthetics of Music. Oxford: Oxford University Press, 1997.
Smith, Murray. “Film Spectatorship and the Institution of Fiction.” Journal of Aesthetics and Art Criticism. 53:2 (Spring 1995).
Souriau, E. ed. L’univers filmique. Paris: Flammarion, 1953.
Turvey, M. “Seeing Theory.” Film Theory and Philosophy. Eds. R. Allen and M. Smith. Oxford: Oxford University Press, 1997.
Walton, Kendall. Mimesis as Make-Believe. Cambridge: Harvard University Press, 1990.
Williams, Alan. “Godard’s use of Sound.” Camera Obscura nos. 8-10 (1982): 193-208.