Deathwatch: American Film, Technology, and the End of Life

IN A December 2002 British Medical Journal article a number of doctors, clinicians, and bioethicists were asked to nominate music they would play while facing their final moments. “People say that birth and death are lonely events,” write the editors. But “music can be a birth or death companion piece.”¹ Choices in the study range from the predictably classical (Rachmaninoff and Handel) to the unorthodox (Velvet Underground and Nico’s “Heroin”). Imagining a terminal playlist is an exercise in fantasy that suggests an intimate connection between melos and dying. Funerals, of course, have long been musical events and are frequently personalized through song. In a scene from High Fidelity (2000) a record-store owner, Rob (John Cusack), lists three songs he’d play at his funeral, adding this to a series of similar “Top-Five” lists of different ranked categories (breakup songs; side one, track ones; etc.). But music can be projected into the before and after. Those questioned in the medical journal were asked to score their future deaths as observers, underscoring Freud’s point that we cannot imagine our own deaths, that we are mere spectators—or auditors—at best.² The soundtrack’s musical theme has certainly become a standard presence in the Hollywood deathwatch. Figures perish to nondiegetic music, and it is a custom to certify death has happened with the soundtrack. Soft accompaniment of strings, for instance, tells us about the fallen and immobile soldier; a thunderous drone triggers the shocked discovery of a horror corpse. What is cued in these scenes, in fact, is what I have been calling registration: the onscreen beholder begins a process of comprehension as if following the soundtrack’s lead. In well-attended dyings, music envelops the body at its threshold, too, so that death appears a little less “lonely.” The watcher (on the screen, in the audience) is brought closer to the terminal sign—closer, not to its visible evidence, but to a kind of aural symbol.

What appears conventional to us now—the nondiegetic soundtrack as registrant—was not the norm established early on by the talking film. It would seem the microphone would present filmmakers with another on/off device with which to organize the instantaneous impression of life/death. In theory the sound camera can capture the last audible breath or spoken utterance, and it can produce the medial equivalent of such a stop. The body has an acoustic “off” switch, and the continuity of sound can be suddenly muted. But, once again, the theory of an instant does not translate well to screen practice: final words are reinforced by the slide to bodily inertia, and sudden silence leaves an echo. This is certainly not to suggest that music did not accompany silent film deaths. But it is important to listen closely to the source of a sound that is often understood by the films themselves as coming from the world onscreen. The audible intensity of certain deaths in the early period of Hollywood’s transition to sound (1927–31) suggests that filmmakers were trying to figure out where to place the microphone within the scene of dying. Rather than synchronize the body with its terminal sounds, early films experimented with a range of techniques of disembodiment, placing the origins of sound outside the body. Many early sound films define death as the lack of synchronization between sound and image, and this lack unfolds over time, much like the body’s stillness in earlier silent films.

This chapter listens to disembodied death sounds in the transitional period. Though Béla Balázs would famously bemoan the sound film as “speaking photographed theatre” interested only in synchronized speech, early sound films often explored gaps between sound and image.³ Several filmmakers resisted sound’s encroachment on the art of silent cinema, and some worked through the microphone’s potential as a new sensory instrument for the death sign. This is perhaps not surprising. Because cameras and early microphones were bulky to use, many films appear slow or static, characteristics I have ascribed to the deathwatch.⁴ And film sound has often been described as less concrete, less locatable than the image and more conniving in its impact. Mary Ann Doane suggests “the ineffable, intangible quality of sound … requires that it be placed on the side of the emotional or the intuitive,” rather than on the side of knowledge, culturally and cinematically linked to vision.⁵ True to form, the soundtrack extends the emotional range of the death scene beyond the duties formerly performed by the silent film registrant. It evokes what cannot be seen at a moment when vision breaks down, comes up short, or otherwise lacks precision. It draws our attention to the world outside the image, and there it supplants the registrant’s tactile and bodily response. Persistent sound takes on the former role of parallel editing to articulate relations between otherwise noncontiguous spaces. Supplementing vision, the soundtrack binds spaces, emerging as if between them, changing the spectator’s reliance on outer forms as the sole media for death. As a companion piece, the acoustic context alters the mood of the scene by participating in the erosion of that slash between Life/Death.

Let us continue, then, to uncover the cinematic image from its connotations of immortality and resurrection. Now we do so by listening more carefully to recorded voices and sounds. Invariably in this we invoke Edison’s early wax cylinder recordings of human speech that inculcated the impression of the phonograph as itself a “speaker.”⁶ The disembodied voice has an important role to play in ending life onscreen. I turn now to consider the microphone as a death technology. Not only seeing dying, in this chapter we will be hearing dying, but what we hear does not emanate from the screen body but rather from the external space surrounding it. In a move that introduces a new temporality, the sonic aftermath (that is, the musical or vocal registration) is synchronized with the death moment itself. Offscreen sounds of singing, talking, and crying registrants pave the way for Hollywood convention; nondiegetic music will later be placed where diegetic sound was at first, and once there, it will cue registration. What later goes unnoticed is hereby brought to the foreground. The soundtrack continues the course of silent film’s production of death through disembodiment, repetition, and posthumous motion, but it also changes the fundamental flow of the deathwatch.

Always troubling, “firsts” are illuminating even if what they introduce is a projection back from the future. We can see, and hear, something new in the long dying scene found in Alan Crosland’s The Jazz Singer, the first feature-length, part-talking film and the first to use diegetic sound to terminate life. Its story concerns Jakie Rabinowitz’s dilemma between pursuing his own career as a Broadway performer and following his Jewish father’s footsteps as a cantor—a dilemma intensified by his father, who demands that he choose. Moments before the opening of his stage debut, Jakie is confronted at home by a family friend, Moisha Yudelson, who begs him to sing as a cantor, in place of his dying father, in the synagogue adjacent to the cantor’s sickbed. A window in the bedroom bridges the two spaces. Singing will spread “sunshine,” Moisha insists; it might restore his father’s health. Jakie’s girlfriend, Mary Dale, a coperformer responsible for his big break, is present. Standing by also are Yudelson; Jakie’s mother (the cantor’s wife), Sara; and the show’s producer. This gallery of witnesses later attends to the dying cantor when Jakie is absent from the room.

All present know that Jakie performs his stage act in blackface; only his father in the next room presumably does not. Much has been written in the last couple of decades on The Jazz Singer’s extraordinary racial and ethnic dimensions, but my interest here lies in another matter.⁷ The film’s use of offscreen music at the dying moment marks a number of conversions in the film that have not yet been placed in relation to the death scene, one of which is Jakie’s return to the religious music of his upbringing. When Yudelson confronts him, Jakie is torn, looking desperately from Mary Dale and the producer (who in intertitle says, “You’ll queer yourself on Broadway, Jack; you’ll never get another job”) on his left to his pleading mother (again in intertitle: “If you sing and God is not in your voice, your father will know”) on his right. Back and forth, the film cuts accordingly. This dilemma seems more like a litmus test of Jakie’s godliness than of his fidelity to his earthly father. But before this moment of indecision the break between father and son has developed in a pair of scenes in which the offscreen music of the synagogue is heard within the apartment. When offscreen music returns in the death scene, it echoes these previous scenes of separation.

Earlier, when Jakie is a kid singing with a club pianist, Yudelson spots him and reports this injunction to the cantor, who arrives to snatch Jakie away from his performance. At home he spanks him offscreen after the child insists that if he is whipped, he will run away for good. Leave he does. At this moment the light inside the synagogue is turned on, as seen through the apartment window, illuminating the window and its Star of David. At the Yom Kippur service across the way, the cantor leads the Kol Nidre, the song of prayer for forgiveness performed on the Day of Atonement. While mother and father worship, Jakie reenters the house. His movements are accompanied by the sound of the prayer song; his father’s voice emanates from the synagogue offscreen. As he moves through the darkness lit by the adjacent window, his gaze rests on a photograph of his mother that he picks up. The film cuts again to his father singing, and then to his mother crying “underneath” the song, fearing she will permanently lose her son. The use of offscreen, or asynchronous, sound reinforces Jakie’s separation from his father and his mother’s private grief. Despite the cantor’s recent interdiction on modern music, the song binds the father’s offscreen voice to the son’s onscreen figure through a technical arrangement. We hear the union between two characters in separate places when we cannot spot it in the image.⁸

In a famous scene occurring later, the up-and-coming “Jack Robin” returns home to New York for a chance at a Broadway career. Sitting his mother down, he sings “Blue Skies,” accompanying himself on piano, first a midtempo, then a faster and more syncopated version. Arriving home, his father enters and interrupts the performance, silencing the voice and shutting off the sound with the shouted command “Stop!” Live sound disappears, and the film resumes silently. The father’s power to kill Jakie’s live voice has been plentifully observed.⁹ In his important work on the film’s use of blackface, Michael Rogin reads this moment ruefully in the Oedipal context of the son’s eventual replacement of the silent film father: “Jack’s father may have the power to stop speech in this film, but it will cost him and silent movies their lives.”¹⁰ The two men have an argument they could not have had earlier, when Jakie was a boy. Jack insists jazz is the voice of God but fails to convince his father, leading to a second and more severe reprimand: this time, he must leave home and never come back.

These scenes illustrate the cantor’s association with both musical types (secular, religious) and prepare us for his sonic departure from the film. The cantor is also the first to sing when he is not onscreen, audibly reminding the audience (and Jakie) of the son’s musical roots. Whether Jakie registers this influence is not clear, but the cutting between apartment and synagogue certifies that the voice is not enacting a sound bridge, joining two impossibly contiguous spaces. It is the actual singing voice next door that pervades the apartment.

The power of the cantor’s voice to fill the adjacent room with his presence evokes, and complicates, Michel Chion’s notion of the “acousmatic” voice, or the “acousmêtre”—the voice, that is, that speaks (or sings, in this case) but whose source is momentarily offscreen and not seen by the spectator in the moment of projection. This mismatch creates a moment of perception “in which we don’t see the person we hear, as his voice comes from the center of the image, the same source of all the film’s other sounds.” The acousmatic voice builds to particular effects in film; in fact, Chion claims it is “the cinema’s invention.”¹¹ Chion usefully distinguishes between “complete” and “partial” acousmêtres. The former is the more hard-core version, the voice whose origin is held in suspense from the viewer for the majority of the film and produces a long buildup to potential revelation, as in the screen wizard’s booming voice in The Wizard of Oz. The partial acousmêtre, soft-core by contrast, has only momentarily lost its bodily source and is closer to the more common term “voice-off.” He explains: “The already visualized acousmêtre, the one temporarily absent from the picture, is more familiar and reassuring.”¹² In keeping with Chion’s distinction, the cantor’s voice is clearly more of a partial acousmêtre: its source is locatable in the room across the alley, but it is nonetheless a voice that fills the image, as if emerging from its “center.”¹³ It is in two places, and thus not fully in either place, at once. The momentarily offscreen voice “can acquire by contagion” the complete acousmêtre’s godlike powers of ubiquity, panopticism, omniscience, and omnipotence. When the source of the voice is revealed (for example, when Toto snags the curtain aside to reveal the mortal man at the wizard’s mouthpiece), these powers dissolve. The effect of the singing voice “that wander[s] the surface of the screen” to guide our attention to the world unseen inspires Cantor Rabinowitz’s passing, all the more compelling given it comes so soon in the transitional period.¹⁴ Another acousmatic power, dimly outlined at this moment, has something to do with the funerary dirge.

In fact, the overlap of one man’s voice with another’s body is repeated, roles reversed, during the cantor’s dying. Jakie manages to make it back home in time to share intimate shot/reverse-shots with his father. The older man says (in intertitle), “My son, I love you,” and touches his face with a trembling hand, the first and last such gesture. Jakie decides to sing the Kol Nidre, but we don’t see him make the decision: a cut intervenes first, taking us to the stage where his premiere has been canceled. Cutting back to the cantor’s bedroom, we find out why. The first shot contains the four figures arranged in a tableau: father lying horizontal on the right, mother standing behind the bed and over him, Miss Dale standing by the window, and Yudelson sitting far left at the end of the bed. No one moves at first in the frame. This stasis feels suddenly strange because, emerging from the “center” of the image, Jakie’s voice begins the song of atonement. The effect of this tableau, then, is a pause to listen. If we count from here, the death scene consists of twenty-three shots, but breaking down the scene into parts fails to capture the soundtrack’s continuous flow, which is crucial for understanding the visible impressions of death.

Jakie’s out-of-sight voice leads the Kol Nidre that floods the tableau. After Miss Dale walks to the window, a couple of two-shots of the parents follow. In the second the father’s face shines with recognition of Jakie’s voice and what he takes to be its signal of religious and familial fidelity. A few images later, we see a long shot from inside the synagogue of Jakie singing in a position occupied previously by his father; two images later, a cut-in to a medium shot demonstrates that his lips are synchronized with the song we hear. Lips and sound are previously not synchronized in this same sequence, which is why we notice this synchronization. (Alan Crosland, the film’s director, lacked access to postsynchronization, a device we will consider later.) As if cued by the audiovisual match, the film returns to the father in bed, who stares up and beyond the top of the frame as Sara looks on. Three more shots of window or synagogue follow, and then the father (in intertitle): “Mama, we have our son again” (fig. 3.1). Crosland cuts back for his body throes and the reaction of the mother. His slide to inertia occurs in two steps: he calmly closes his eyes as if entering sleep, and he tilts his head back. Sara at first holds him. A return to the establishing tableau witnesses Sara’s histrionic hands-to-heaven gesture and her collapse onto the dead body—movements familiar to us from Griffith’s melodramas. Invoking the language of melodrama’s temporality, Linda Williams writes that Jakie sings “just ‘in time’ to make his father die happy but ‘too late’ to save him from death.”¹⁵ I want to emphasize the music associated with the “happy” dying man.

At first glance, this appears a silent film death with music simply played over it.¹⁶ The registrant’s repetition of the cantor’s apparent death does not end the scene, any more than it did in our earlier silent scenes that employed posthumous reframings or pans to lament a loss through spatial dispersal. But here the song supplies that residual effect, bringing the camera to it. Promptly we cut back to Jakie singing in the church and then to male members of the congregation. The song goes on. Miss Dale makes the point of these shots clearer: “A jazz singer, singing to his God.” The image of Jakie’s Jewish congregation recalls the earlier comparison he made between his father’s and his own audiences. He really gets into things during the Kol Nidre’s finale, improvising gestures that have been read as an incorporation of his “jazz” side with his Jewish sound. Williams in fact argues that the music in the film synthesizes two ethnic musical expressions.¹⁷ Improvising passionately, Jakie neither knows nor seems prescient of his father’s death having just occurred, a point that deserves more consideration (fig. 3.2).

The fact is that an important registrant—perhaps the most important—never appears in the death shot. Jakie is absent from the tableau, but his voice is present, breaking from convention. The silent film often organized the speed of dying through crosscuts between the body event and the (would-be) registrant. In The Jazz Singer, activity inside the scene—that of singing and listening—becomes the focal point. In light of the son’s paternally condemned syncopations within the plot, we should note that rhythm seems less important to the scene than duration. The father’s death moment, after all, is placed within the context of images that run together along a single thread of time—strung together by “live” singing—rather than a synthetic time construed through edits. Griffith placed the death moment within a higher order of shots, allowing the spectator to arrive at certainty when a survivor found the body. Here, a religious song provides that “higher order.” It is Jakie’s singing that orchestrates our visual attention, stitching together synchronized and asynchronous sound images, bridging bedroom and synagogue, father and son.

This doubling of the voice makes quite a difference, and two of Chion’s insights help us grasp its effects. The first has to do with the soundtrack’s usual privileging of the voice over all other sounds. In the hierarchy of film sound, says Chion, the voice comes out on top. Chion’s acousmatic voice is almost exclusively a talking voice. That we are talking about a singing voice, though, only buttresses his claim: in this novel part-talkie the voice functions as a musical sign of synchronization within the diegesis and its use opens up a perceptual ambiguity in the shot. The second evocative point is Chion’s designation of the acousmatic voice’s origins in partial places—“neither entirely inside nor clearly outside” the image.¹⁸ “Inside” the image of the cantor’s tableau, voices make no sound: the figures register and pose as in a silent film. But “outside” this same image, Jakie is singing in the synagogue in a reprisal of his father’s teachings. Any attempt to find a formal break in vitality is swallowed up by a second cinematic signifier—the song. While the father is onscreen passing from life, his son’s voice is passing from one building to the next. The Kol Nidre makes present Jack’s religious feeling at the moment the cantor slides to inertia, reassuring the older man that it is time to let go. The scene synchronizes the visible death sign with the ear—both our ear and the cantor’s.

One consequence of the song’s flow between rooms is that Jakie’s voice outlasts the onscreen registration in the bedroom, where concerned parties are gathered. It does not turn “off.” Exceeding the image, the voice, instead of the registrant’s touch, confirms the final impression. Jakie’s melody replaces the body and the registrant. Of course, singing can be heard and filmed outside the body, so it exceeds the body’s limits. Thus, the Kol Nidre begins “on time” for the cantor to hear it and proceeds into the domain of posthumous motion. The song encourages the camera to mark finality within diegetic space and outside the body.

But, at the same time, it feels odd to call Jack a registrant: he does not “see” or react to what happens in the bedroom. In fact, he seems to conjure the corpse’s replacement—a spirit version of the cantor appears behind him to mutely acknowledge the son’s deliverance. Where is the center, and when is the peak, of the cantor’s screen death? Jakie registers by overwhelming the scene, absorbing the impact of what happens in the bedroom across the alley. Because it can travel through space, says Doane, “the voice has a greater command over space than the look—one can hear around corners, through walls.”¹⁹ Singing, in fact, replaces here the melodramatic specialty of lamenting over the dead body, an act that would seem to encapsulate melodrama’s temporality of “too late.” This song reaches the dying man’s sensorium, however. The film acknowledges the age-old practice of singing dirges, hymns, or lamentations at or after death moments. As another fold in its synchronization project, The Jazz Singer reconfigures the funereal temporality that basks in “too late.” Here the dirge is “on time.”²⁰

Considering too that dying previously piqued the imagination as something the camera could not finish without people’s reactions, it should not surprise us that the soundtrack was seized as further useful for the task. Here we see a “good death” accompanied by the “live” voice enveloping it, as if a well-attended dying requires lively participation on the part of the cinematic signifier. Onscreen singing affords a vital presence that was by definition remote to silent cinema’s inherent muteness.²¹ And onscreen heard sound also brings our attention to the acoustic environment inside the scene. Balázs, for one, championed the sound film’s potential discovery of our acoustic environment: “The meaning of a floor-board creaking in a deserted room, a bullet whistling past our ear, the death-watch beetle ticking in old furniture and the forest spring tinkling over the stones. Sensitive lyrical poets always could hear these significant sounds of life and describe them in words. It is for the sound film to let them speak to us more directly from the screen.”²²

The conversation between film and spectator (“speak to us”) opens up a secret world of unheard presence. Balázs’s examples hover around an animate source of life that lies dormant because isolated from the perceiver, unheard because solitary with respect to the human observer. It stands to reason that Balázs sensed that the important death sounds to be discovered would emerge from around and not in the body when he wrote of the “death-watch beetle” in the “furniture.” Gathering environmental traces, the sound film can stage death without placing its sole focus on the sickbed tableau. Indeed, it can place death within more than one location. (This multiple placement was done in the silent film, but not simultaneously.) The Jazz Singer’s offscreen singing creates the impression that he who is absent (Jakie) “sees” death. The added presence of what cannot be seen expands on the medium’s temporal dilation of dying. Borrowing again from Williams’s dialectic of melodrama, I would say the Kol Nidre’s “live” spatial dynamic in turn changed “too late” into “on time.” Crosland completes the death scene outside the body, not with Sara’s registration but with Jakie’s conclusion of the Kol Nidre—all the more striking considering that Sampson Raphaelson’s 1925 stage version of The Jazz Singer had Jakie sing the Kol Nidre offstage for his “already-dead father.”²³

But we must return to the racial dimension of Jakie’s live performances in the film, and his appearance in blackface in the very next scene, to understand the use of this evocative song. Singing absolves his own guilt over causing (perhaps even wanting) his father’s death. Clothed in a cantor’s skullcap and robe, he becomes the center of attention, and his performance ends the scene. It is Jakie whom we see last before the elliptical fade-out, and it is his blackface performance (the only one in the film performed onstage) of “My Mammy” for his own seated mother that follows quickly and resumes narrative focus. The film substitutes Pappy with Mammy. Given its deus ex machina for Jakie’s dilemma and its catalyst for his ethnic reprise, the Kol Nidre can be read as both lethal and conversional. It cannot function in the film without associating itself with death. Doane’s “ideological truth of the sound track” strikes a new note.²⁴ Sound finishes a sentence begun by the body, but it remains invisible and thus not wholly locatable. The concept of sound production seizes the erosion of death as an instant, allowing us to perceive that the instant dims, fades, reverberates. In The Jazz Singer synchronized sound proves to be made for the cinematic dispersal of the death moment (as Jakie’s erasure of his father makes narratively clear). In its own terms the film suggests that screen sound and screen death may share a common terrain, a domain that is “not exactly unseeable, but unknowable.”²⁵ The touching appearance of the ghost-cantor merely manifests what is already heard. Jakie replaces his father in the one way that the older man can allow—by taking his place and duplicating his role.

Rogin has pointed out how the film breaks from its silent film ancestry by staging the death of the father, a figure here deeply invested in protecting tradition. The Kol Nidre’s enveloping of the still-dying father blends modern sound technology and tradition. It is diegetic sound that functions to give the cantor a good death, but it is the Kol Nidre, not jazz or the syncopated music of the American melting pot, that offers peace. One could say it “kills” the father. This is where my account differs from that offered by Rogin, who makes much of the fact that the film concerns the conversion to sound and the Jew’s conversion to popular mainstream entertainer. Rogin claims that Jakie can only overcome his father through the assimilation afforded him by blackface: “Blocked from overthrowing the father directly, he regresses to blackface’s imaginary realm of music, image, and the specular, histrionic self.”²⁶ Yet this is not exactly true. It is not blackface, but the Kol Nidre, that successfully gets rid of the father without guilt or remorse. For the moment, it seems Jakie is trying to make good on the perceived lack of his Jewishness. His father’s death gives him a chance to be center stage while embodying a musical style that to him is outdated.

As I mentioned earlier, Linda Williams suggests that the film’s “jazz” and the Kol Nidre are infused, so that when Jakie holds our attention as a cantor and extemporizes arm gestures of rapture, we are to find somewhere the incorporation of black melos in his performance. But, first of all, there is no purely black melos in the film, and one cannot put quotation marks around the Kol Nidre in the same way nearly every critic has done around “jazz” or “black” music. In other words Williams does not account for the undeniable ethnic specificity of this particular Hebraic song, and to argue that Jakie as cantor sings jazzily resorts to reading the scene merely for its visual stereotypes—that is, to not be listening to a film that wanted so much to be listened to, as Williams herself points out. Passing for white throughout, Jakie allows his father to pass as Jew. The film, the assimilation afforded by blackface, and the conversion to sound all depend on it.

This will not be the last time the voice overlaps so prosaically with the dying. Chion tells us that “particularly in the cinema, the voice enjoys a certain proximity to the soul, the shadow, the double—these immaterial, detachable representations of the body, which survive its death and sometimes even leave it during its life.”²⁷ For practitioners at the time, sound was already a sort of parallel dimension that usurped (Chion says “supplant”) the role of superimposition in silent cinema, which could visualize what someone heard or thought by laying it on top of an image of the person hearing or thinking. Superimposition was also the tool to produce images of the spectral world, employed at least as early as Porter’s Uncle Tom’s Cabin (1903) to represent death beyond the registrant’s touch.²⁸ But the idea of flow connotes dispersal and projection rather than superimposition’s palimpsest. If we were to sketch death based on the template of sound, it might resemble concentric receding waves that move away from a nodal point.²⁹ It would take place by moving through space.

The cantor’s death moment is broken apart into several moments, and more than one “passing” is noted: first there is Mary Dale’s comment “A jazz singer, singing to his God”; then the cantor’s “Mama, we have our son back”; then Sara’s hands-to-heaven gesture and collapse over the bed; then Jakie’s improvisational Kol Nidre climax; then the appearance of the father’s ghost behind him; then the completion of the song. There is quite a cascade of registration in this scene—many people get to do it, many times it gets done, and it is hard to say if Jakie’s faith or the cantor’s death is being registered. This relay of beheld transformations makes palpable the peak moment; the film seems to offer up Jakie’s impassioned climax as that moment, though the point is that there are dramatic options. Finally, it is difficult to mark a boundary between external and internal experience.³⁰ Who can say, when viewing the senior Rabinowitz’s pallor, if death is that terminal? The actor’s physiognomy qualifies both the music subsuming him and his emotional state. Around the cantor, inside appears to meet outside. The external appearance of death is sheltered by the melos enveloping it. Cinema continues to insist that our deaths take place elsewhere and not just in our bodies.

Theorists have turned to metaphor to determine the surplus value of the soundtrack. It is often imagined as an additional stream of information that brings vitality and duration to the image. As Doane explains, “Sound is the bearer of a meaning which is communicable and valid but unanalyzable.”³¹ In fact, the soundtrack was thought of as a continuous vital flow by early technicians who employed life-and-death metaphors to express its potential to reverberate through space rather than get killed, or absorbed, by stage materials.³² If vitality is enlivened by sound, we might say that “being dead” is enlivened as well, a condition supported by the ubiquity of undead creatures roaming around the early transitional period.

The concept of flow takes us away from the synthetic edits of the silent film; something persists on the visual level that cannot be broken down into fragments (shots). For this reason, says Christian Metz, we must say that sound, unlike the image, does not turn “off” as clearly, for it is not enframed like the visual elements we do or don’t see. Whereas the cinematic image frequently refers to a visible world just outside the frame, a sound whose source is for the moment not seen is only just that—momentarily not seen, but not unheard.³³ And unlike the screen that can fade to black to render darkness, film sound cannot as easily “die”—if by that we mean erase itself, for even silence emanates a sonic presence and the reverberation of echo.

Yet another conception of sound is as a third dimension beyond the width and height of the flat, two-dimensional image. This added element “presupposes a certain depth” and “is in contradiction with the flatness of the two-dimensional image.”³⁴ Literal sounds produced from within the diegesis can make palpable the space before the camera.³⁵ Early American film theorist Rudolf Arnheim put it this way: “Sound arouses an illusion of actual space, while a picture has practically no depth.”³⁶ The more we ponder sound’s capacity to make present the unseen, the more “it” (focused sound, the devoted voice or instrument) appears a good “companion piece” indeed for our final moments.

Listening may be crucial for those present in the dying chamber in The Jazz Singer, but in other films a concerted sound is not easy to find. High Fidelity’s Rob can score his funeral with Aretha Franklin and Gladys Knight, but cacophony is a disconcerting possibility. The connective sonic glue between bodies can come unglued at this last lived moment, which is, after all, a moment of separation. Rouben Mamoulian’s Applause (1929), a film praised for its systematic use of asynchronous sound, offers that inverse scenario. If The Jazz Singer presents death “on time” with the bonus of the added voice that lulls the central figure into crossing over, Applause presents just noise. The former, an Oedipal father-son melodrama, contrasts pointedly with the latter, a melodrama of maternal sacrifice.³⁷ Content is, as always, matched with form.³⁸ The films’ respective protagonists—Jakie Rabinowitz and Kitty Darling—represent clear psychic roles that for Peter Brooks epitomize melodrama’s polarized play of signs.³⁹ Their relationship to music extends these roles. It is no coincidence that their death scenes are so audible and so rousing.

Rather than entertain or feature vocal talent, the musical numbers sung by Applause’s female leads elicit our pathos for their entrapment in a patriarchal theater system, from the stage numbers filmed from the animalistic points of view of male spectators in the theater to Kitty’s “What Wouldn’t I Do for That Man,” hummed to herself more than sung aloud. Rick Altman has suggested that the backstage musical proved a convenience for filmmakers interested in showing off “the full range of the sound technology.” The tone cast by many early backstage musicals was one of sorrow, and Altman suggests that its decline in popularity by 1931 or 1932 may have been due to this: “It is tempting to conjecture that it declined not because the public wanted more variety …, but because the public tired of the equation of music with sadness.”⁴⁰ But before wearing out the audience, that equation was made with exuberance.

Applause tells the story of an “over-the-hill burlesque queen,” Kitty Darling (Helen Morgan), who struggles to maintain her own rank while making sure her daughter, April, does not follow in her own footsteps (the opposite trajectory parentally desired in The Jazz Singer). She sends the girl away to a convent. Years later, an adult April returns to live with her mother, but she begins to fight with Kitty’s abusive lover, Hitch Nelson.⁴¹ April falls for a sailor named Tony, temporarily docked in New York. The two make plans to marry, but when April sees her mother will be abandoned by the industry, she decides to turn Tony down and stay to help Kitty.

Melodramatic too-lateness is none too kind. Before April can tell her mother that she is ditching Tony, Kitty spots a bottle of pills in her medicine cabinet and ponders them in a suicidal frenzy. Meanwhile April is dumping Tony: she finishes a glass of water as Tony remarks, “You’ll be late,” and a dissolve shows Kitty downing the poison. She sits at first in her armchair staring out the window, juxtaposed with an urban ambience of car horns on the soundtrack. In a later shot Kitty appears to be driven by the offscreen sound of a siren to pick up her daughter’s picture from the table. Besieged by regret, she cries April’s name, grabs her housecoat, and leaves to find her. The digestion of pills sets up another possible combination of too late / on time.

April and Kitty are united backstage; April lays her down in her dressing room.⁴² She tells her mother about backing out on Tony, and a horror-struck Kitty screams, “Oh my God what have you done, what have I done?” Kitty poignantly “cries out” her recognition that she is too late to save her daughter from the world of burlesque, translating silent film’s gesture into an audible cry. We realize it is doubly “too late”—for Kitty alone, and for April, whose attempt to take care of her mother (the whole point of her return) is thwarted. Mamoulian cuts to a high-angle shot that reveals the positions of (another instance of) four people in the room: Kitty lies diagonally across the bottom of the frame, with April near her feet; the stage manager and Hitch have entered and stand on either side. The daughter gestures at them to go away, but the stage manager persists: “Kitty Darling is finished.” Another cut takes us back again to a medium long shot at eye level. The room’s acousmatized voices intensify the image of Kitty’s claustrophobia: offscreen, the managers and April argue over Kitty’s doomed career as Kitty whimpers softly.

Rather than as sonic glue that coordinates passing to an elegiac tune, Mamoulian records sound as atmospheric weight. More generally in the film, overlapping and poorly heard sounds reflect a lifelike sense of our hectic everyday acoustic worlds. “Unlike other directors of the period,” Lucy Fischer tells us, “he recognizes the inherent spatial capacities of sound and, furthermore, understands the means by which they can lend an aspect of depth to the image.”⁴³ Accordingly, Mamoulian builds layers of sound during Kitty’s dying.⁴⁴

First, we hear Hitch convincing the stage manager that April can save the specialty show. He says to April, “Come on baby wake up,” which resonates ironically with what we see: a medium close-up of the mother with eyes closed, convulsing. Following an offscreen drum roll—and in unison with April’s proclamation: “Oh, I’ll be good all right. I’ll go out and give them what they want. I’ll show them. I’ll show them.”—the camera blurs Kitty’s image entirely out of focus (fig. 3.3). This, it would seem, is the visual representation of her death, one that we have not yet seen, in fact. Cut to the room: the manager says, “Now you’re talkin’!” (recalling The Jazz Singer’s “You ain’t heard nothin’ yet!”). April leaves the room. Then, the camera does something rather intense: it tracks in to a close-up of Kitty’s motionless, upside-down head as we hear April begin singing from the distant stage. If we look close enough, we can see a sudden stilling of the wobbly camera; perhaps a freeze-frame was added to connote stillness. Could this be farther from The Jazz Singer? Rather than binding them, the soundtrack separates. Diegetic music plays throughout the scene but shrouds Kitty’s condition. The camera, not a living character, returns to the bed to register Kitty’s death.⁴⁵

The voice-off is distinct from the ways Mamoulian’s camera otherwise works with visible space through tracks-in and -out, rack focus, and a cluttered mise-en-scène. Mamoulian broke from then-current conventions of recording directly with a single mike in a static shot. Arthur Knight explains that Mamoulian “felt that the camera could and should move,” “fought against the prevalent notion that the source for every sound must be seen,” and used the camera as “more than a passive observer looking on while actors recited their lines” because he thought it should “help the audience find what was dramatically significant in a scene, picking out what was important with his camera.”⁴⁶ Mamoulian was an early explorer of film’s acoustic environment. Given the mobility of his camera and its relative autonomy from the soundtrack, it is perhaps not surprising that he would employ the camera itself as a registrant.

As with The Jazz Singer, Applause’s death scene fits into the film’s larger aural logic. During Kitty’s convulsive last movements the moving camera blocks out all present. Since it is produced by the lens, the image of Kitty’s death cannot be perceived by the fictional characters and is not registered by their offscreen voices. Even if they were looking, they wouldn’t see it. Those present continue their dispute over who will fill the star’s role. But the spectator witnesses these convulsions and the camera effect. We are forced to perform as a registrant, denied the usual gestures of grief. With synchronized diegetic sound, it becomes possible for figures in the scene to talk and not to pay attention to what the camera picks out as important.

But is the camera “more than a passive observer” of death? The ironic friction between sound and image creates a half-registered moment that slides into posthumous dance. Specifically, April decides to take her mother’s place in the number under way, taunting that she will “go out and show them,” though she is blind to what has just happened. While we are being signaled visually toward the mother’s death, the sounds that emanate from beyond the frame’s limits do not signify human registration but its failure. The camera moves in and does not take the talking bodies with it, marking an unobserved death. The spectator’s registration is even more interesting given Fischer’s account of Mamoulian’s spatial world as one “that seems visually and auditorially perceived.”⁴⁷ Mamoulian’s death scene—constructed through what Rick Altman terms “sound space”—provides merely a partial view.⁴⁸ To represent death without registration, Mamoulian blurs then refocuses the still body. What the world needs to absorb Kitty’s death is more than the camera, or the spectator, can offer. The scene hovers on the verge of diegetic absorption.⁴⁹

There appears no external agent to confirm that death has really happened—but Kitty is not quite alone, for her body is enveloped in a chatter that forces us to project her ending. The unregistered death continues to haunt the film. Instead of leading to reunion or arrival, the reaction shot and the body event never join—April never returns to the dressing room. Instead, we see first a brief shot of April singing and dancing superimposed over her still mother, the effect of which is rather base with Hitch looking on, licking his chops (fig. 3.4). After performing the routine she earlier watched Kitty perform and from which she recoiled in horror, April rushes to the wings, where she finds that Tony has returned to her; the two embrace in front of a large poster of Kitty from her “Parisian Flirts” show. Tony and she decide to move to Minnesota and take Kitty with them, where “we’ll always be together, all three of us, all three of us.” As their heads (and lips) drop out of the frame, the camera pans up to rest on a medium shot of the poster’s face (fig. 3.5). The movie ends by fading to black.

Certainly this is anything but satisfying closure. Mamoulian finds in the soundtrack a metonym for the insatiable, male-driven desire running the burlesque show. In fact, he seems to suggest that sound technology is forever “on” and has no “off” switch, that it, like show business, is incommensurate with death. Too-lateness emerges in Applause as an effect not so much of editing as of the spatial compositions in the frame and the narrating agency of the mobile camera. This does not eliminate repetition and disembodiment but merely alters their appearance. The final image of the talked-over poster mirrors the death scene, communicating the mother’s absence by showing a body that can no longer dance or sing (corpse, poster). The poster is opaquely commemorative. In the film’s final shot Kitty is pictured in the poster but muted; April is heard but not seen. The film leaves us longing for the union of body with voice.

Doane tells us that the voice-off “deepens the diegesis” to an extent that “exceeds that of the image, and thus supports the claim that there is a space in the fictional world which the camera does not register. In its own way, it accounts for lost space.”⁵⁰ In service of the construction of space more than the image, the voice-off articulates a sensory presence that the image cannot. According to Doane and Pascal Bonitzer, the depth of the unanchored voice can be a frightful thing, potentially disrupting the spectator’s seated mastery over the onscreen synchronization of voices with bodies. For Bonitzer, the film noir voice-off, that “dark prophet of the end of the world,” thematizes this disengagement.⁵¹ Kitty’s seems a “bad” death compared to Cantor Rabinowitz’s partly because the voice that flows momentarily away from its bodily source does not get absorbed by the dying body, thus leaving the spectator’s own bodily experience incoherent—“a body” that is usually “positioned as unified and nonfragmented.”⁵² This fragmentation represents the other side of synchronization and expresses an entirely different sensibility.

That early sound films make use of acoustic space at the dying moment is not a break from sound style. As Lucy Fischer has pointed out, Balázs saw sound in general, and especially silence, its absence, as a spatial experience. Since “there is never an absolute auditory void,” silence “can only be rendered when sounds are present.”⁵³ Silent films produced no true void of sound. Whereas silent films circumvent stillness, sound films approach silence. The camera’s link to related sound encourages it to navigate around the void—death emerges not at the presumed borderline between motion and stillness but in a series of spoken passages between rooms and speakers. The very edge between sound and silence is hard to imagine, perhaps impossible to produce. Sound always leaves an echo, gets absorbed by other substances, before it “dies.”⁵⁴ Voices fill the rooms in which the cantor and Kitty leave the diegetic world: though they still die silently, their rooms do not.⁵⁵ In the language of the first sound technicians, the dead body absorbs and keeps alive, thus does not kill, its surroundings.⁵⁶

Mamoulian’s Dr. Jekyll and Mr. Hyde (1931) also uses offscreen sound at dying. Having once more turned into Hyde (a transition he cannot himself control), Jekyll is caught and cornered by the police in his own laboratory. Monstrous Hyde is fatally shot and transforms back into Jekyll through a stop-motion sequence. It is the only such transformation in the film (from monster to human): death ensures the final erasure, and not merely the suppression, of Hyde. While we look at Jekyll’s head laterally framed by vials on his desk, we hear “Doctor” spoken by Poole on the soundtrack; then, “Jekyll” matches a cut out to a long shot in which Poole mourns standing among the police and Jekyll’s once-friend Lanyon.⁵⁷ The camera reveals a larger space in a beautiful iteration of posthumous reframing. The death occurs silently and is not registered until the soundtrack cues this shot of public recognition.

Kitty Darling’s persistently unregistered body is not a unique instance when the early sound film withholds synchronization to explore the borderline between vitality and its lack. Robert Spadoni argues that the quiet but “persistent corporeality” of Bela Lugosi in the title role of Tod Browning’s 1931 Dracula makes use of audience memory of earlier slow-moving sound films to enhance the screen effect of a walking dead figure. This “re-estranging” of synchronized sound, Spadoni writes, builds over the course of the film as Dracula’s voice becomes more removed from his screen body.⁵⁸ Of course, the silent film corpse figured a “persistent corporeality” that did not speak its status but was spoken for by the registrant. Dracula gradually disappears and stops speaking, as if the film loses its ability to track him, drawing our attention to others in the film to account for his body, vitality, and power of influence. Losing synchronicity makes it more difficult to account for the body (the opposite of the acousmatic voice that builds up to synchronization). It’s also difficult to kill someone—for example, Count Dracula—who is defined not by one body but by many (and not just forms of the same body, like bats and wolves, but other people under his spell, like Mina and Renfield). At the end of the film Dracula carries a catatonic Mina to his coffin, with Van Helsing and Harker on their heels. Dracula’s death is represented by Mina’s offscreen scream, signaling her return to normalcy, and Van Helsing’s proclamation that “Dracula is dead forever,” spoken while he is holding the stake he used offscreen on the vampire. In neither case do we see the vampire’s body.

The corpse lacks a voice, and the offscreen or acousmatic voice lacks a body. Horror plays with the formidable border of synchronization between corpse and sound. In The Mummy (1932) Imhotep is destroyed by a spoken curse, and a makeup transition turns him back into a mummy, invoking Bazin’s “change mummified” without the change. Since deadness is not certain in the image, the unseen voice possesses a panoptic power not mentioned by Chion—the power to see the body’s interior state. Synchronization falters at death: what one body loses (its voice), another body projects. The fact that the soundtrack plays such an important role in attending the deathwatch—whether live or postsynched, as in my next example—underscores the audiovisual image as “undead.”

Registrants don’t just sing or quibble; they also cry. Wails of grief remind us of Peter Brooks’s sense of melodrama’s “most transparent, unmodified, infantile form” of self-expression. As he remarks, “Desire cries aloud its language of identification with full states of being.”⁵⁹ Sometimes this emotional language becomes centerpiece to the scene of registration, to the extent that it confirms death beyond referential contact with the corpse. King Vidor’s Hallelujah! (1929) provides an example.⁶⁰ The Jazz Singer made specific use of Jakie’s singing the Kol Nidre—that piece of music so charged in the film as the sound of Jakie’s own ethnos, the origin he has left but now revisits, that of his now-dead father. Only after singing the Kol Nidre is Jakie able to occupy the position of the mammy singer, fully integrating himself into that role in front of his own mother. In Vidor’s film, one of the first screen musicals with an all African American cast, Mammy possesses preternatural knowledge, registering her son’s body before we even see it. Vidor had recourse to postsynchronization, allowing him to freely stitch images to a soundtrack. Visually, Mammy’s pathos becomes a rapidly disembodied response, as she herself drops out of the picture. Jessica H. Howard has argued that characters generally sing the songs in the film to catalyze their own transformations. Guided by the competing forces of religious zeal and sexual desire, Zeke and others build—over the course of singing—toward more and more ecstatic states. Since transformation through song creates an audiovisual “present” for both the singers and the spectators, the latter are in turn able to identify with the former and in some cases join in. Building from Altman’s point that audiences of the film would in fact have been familiar with many songs and encouraged to experience it as a sing-along, Howard comments that “one knows that it is precisely one’s act of joining in the musical present that renders the emotion inclusive and unmediated, perhaps—it seems—even by the medium of film.”⁶¹ Spectators were prompted to participate “live” in the musical dilemmas onscreen. “Live” death, too, is very much at issue in the early period, as the spectator is drawn closer to the lack of synchronization as a terminal sign that plays out in his or her own time.

Vidor’s film was praised in the trade press on its initial release for its realistic representation of black life in the rural South. And a Variety review mentions black noise—“moaning,” “crooning,” “carefree, syncopated singing”—as exemplary of that realism. The asynchronous sound of “the harmonious voices of the Dixie Jubilee Singers” (returning to the screen four months after their Show Boat debut on the stage) was “especially necessary … in the spirituals.”⁶² Ruth Morris, who ironically condemned the film as a successful matinee picture for women spectators and praised it for its intelligence and “composition,” suggested that “the dialog in itself is a musical accompaniment.”⁶³ Musical accompaniment seemed the attraction, hanging on to images and surrounding them with an intended aural authenticity.⁶⁴ That point was certainly made by the all-black musical Hearts in Dixie (1929), praised for its infusion of sound technology with (black) “personalities … ideal for this medium.” As Alice Maurice explains, the early sound film fetishized the “black voice,” not because such a voice was indeed naturally suitable for legible recording but because it brought “expectations regarding authenticity, the alignment of internal and external characteristics, and the evidence of the senses.”⁶⁵ Of course, the expressions of “black” grief in Hallelujah! are influenced by stereotyped codes already present; an embodied and expressive pathos is reaffirmed by sound technology. Mammy’s pathetic wails read as an older presumption about the black body as an overabundant physical presence. But if we listen and watch closely, they also point to an affect that the moving image can explore in interesting ways.

Hallelujah! tells the story of Zeke Johnson’s call to ministry in the wake of accidentally killing his own brother, Spunk, in a juke joint brawl. Spunk’s death scene constitutes another interesting step forward in the transitional period. Once again, determining the components of the “scene” moves us, along with the persistence and duration of sound, to the outside of the body. At a juke joint, where the seductive dancer Chick eggs him on, Zeke fights with Hot Shot over money owed him from a rigged craps game. Zeke pulls a knife, Hot Shot a gun. The two men wrestle, the gun goes off, the crowd flees. Zeke ends up with the gun and shoots haphazardly around the room, twice in a synchronized close-up (a shot whose omission from the film was unsuccessfully urged by Jason Joy of the PCA), then three more times. After the crowd disperses, Zeke hears his brother whimpering offscreen and bends down close to attend to and hold him. On the left of a medium close-up, Zeke crouches in shadow as he holds his brother’s lateral body, cast in light. The cut-in of the brothers marks the scene with an audible desperation, as Zeke pleads for the younger man to talk (“Speak to me, Spunk”) and the other agonizes (“Hold me, Zeke; I’m gettin’ cold”). After stepping outside to shout for offscreen help (“Someone help me! My brother’s dying.… O won’t somebody help me?”), Zeke returns to Spunk in the medium close-up, this time his face fully eclipsing his brother’s: “I’ll take you home to Mammy. She’ll help me.” He carries Spunk, still alive and mumbling, outside the juke joint.

As in Mamoulian’s film, talking here is synchronized (the conversation between the two men) but the death moment is not (no gurgle or rattle). Vidor’s postproduction, postsynch soundtrack, as Arthur Knight tells us, “explored the possibilities of the sound track to evoke mood and atmosphere” in a film shot for the most part silently. Vidor builds a tonal atmosphere through such things as “the rhythmic swell of Negro spirituals, a woman’s scream or a barking dog heard in the distance, the sounds of the swamp and the river.”⁶⁶ In both Applause and Hallelujah! the microphone listens for confirmation from the world outside the body. Because we do not see Zeke’s brother die, the film must eventually address his condition (or come to an end, as in Applause). It does so with Mammy Johnson’s piercing wails. Mammy’s crying all but thematizes sound’s capacity to extend grief beyond the usual scenic limits of the framed registrant.

First, the film damningly cuts to Hot Shot and Chick arguing over how to divide the money. Then, in a close-up we see Mammy wake up and look right: the reverse shot reveals an empty bed. Vidor cuts to a medium shot of Mammy—we might call this the first “reaction” shot, yet she is reacting not to the fact of death but to her son’s conspicuous absence. In the role of Mammy, Fannie Belle DeKnight intones, “O Lord, have mercy on my children,” over a reverse shot of the full line of children’s beds with the empty one at the end (a rather poignant composition). In a medium shot the family gathers around crying Mammy, soothing her and bowing their heads, pleading with her to stop crying, in a tableau of grief with no corpse. These sounds and gestures recall the earlier scene in which Mammy gathers the younger boys around her to pray at bedtime, singing each one to sleep cradled in her lap. Wailing reaffirms the “cliché whereby blacks sing out of misery and religious sentiment.”⁶⁷ Here, the static camera reinforces the tableau while human voices explode from its tightness. Whereas silent cinema cornered the conflict between movement and stillness, here we are aware of the conflict between voices and silence.

It is Mammy who is surrounded, not her soon-to-be-dead son: “O, my children ain’t come home.…” As the death wagon approaches, the family emerges into the yard. Mammy sits down in a folding chair where the voices continue, with one child’s “Please, Mammy, don’t cry” and Mammy’s “Have mercy on my soul, O Lord.” This last “Lord” breaks from the single-note stream of Mammy’s singing and enters the stratosphere. Once Pappy steps in to pray, we hear several voices, one over another, producing a cacophony of emotions ranging from pain to humility. Zeke, the wicked messenger, arrives: we see that he hears the sounds of his family’s grief, his face contorting slightly in recognition. Of course, reaction shots to offscreen sound throughout the film visualize emotional effects the music has on its listeners: earlier the younger sons smile at the beauty of Mammy’s singing, or Zeke stares lustfully with outstretched arms at Missy Rose while she plays an organ wedding march.

The cut-in to Zeke’s face shows his reaction not so much to death but more to the pain it causes in familiar voices. Zeke’s head and shoulders are framed from a low angle against a blank sky, lending a naked expressivity to his face. Stepping off the horse, Zeke approaches as we hear Mammy continue: “Look what I hear.” It is like an instruction to the spectator, for we are exactly looking at what we hear. Vidor cuts to the family’s tableau; Zeke enters finally from offscreen left while Mammy releases “O … O … O …” I do not quite know how to write the sound of Mammy’s grief. (It is striking that rewriting the sentence of cinema death from instant to process must accommodate emotional expressions that are grounded but inarticulate.) Zeke kneels down, hat off, and bows his head on his mother’s lap. Still no registration. With the last “O” Mammy collapses over Zeke’s body. Ned asks Zeke, “Where is our boy? Where is Spunk?” to which Zeke responds by pointing offscreen left. Unlike the rest of the family, Zeke does not open his mouth once. He never delivers the message. Mammy supplies the mute Zeke with a vocal release and emotional expression. Her voice is not always localizable and travels freely across shots, uncovering emotional depths in images of the wagon’s approach and of Zeke’s physiognomy. The rising intensity of Mammy’s “O”s highlights the power of sound to chart increasing intensity.⁶⁸ This whole scene represents the crescendo of the voice. Mammy transmits the certitude of finality without translating death into a registered fact.

We cut to Ned, who is approaching from the rear of the wagon and looking. His point-of-view shot begins from Spunk’s feet and extends along an imaginary z-axis into the frame, accompanied by the only modulated nondiegetic sound in the entire film—a soft drum roll announces the final revelation (without words) of the corpse. The quietness of the body’s unveiling contrasts poignantly with Mammy’s wails. Again Balázs: “After a sequence of a hot dance music the stillness of a sick-room affects us differently than if the preceding picture, too, had been a quiet one. Sounds which, as the saying goes, ‘still resound in our ears,’ may deepen and interpret the silence that follows them.”⁶⁹ In the momentary quiet that follows Mammy’s “O”s we hear the echo of grief catch up with its source. Mammy’s chant-singing “O”s repetitively, Jessica Howard tells us, “allows the ‘true’ bodily expression of Mammy to emerge (i.e., piety, grief). For example, repeating the words ‘Lord, have mercy …’ over and over again does not enhance their meaning or propel the plot forward.” Indeed, Spunk’s departure from the film has not been confirmed before—and, one could argue, without—her chants. The words she says in this peak moment do not convey a narrative message, Howard argues, but rather a sensory presence: “Thus, its meaning is unfixed and ultimately transformable. The chant, then, prioritizes the presence inherent in aural expression.”⁷⁰ The “other world,” part of Mammy’s vision, seals her authenticity.⁷¹

Throughout this sequence, those heard but not seen “impress us,” Balázs might say, with the “inevitability,” “approaching catastrophe,” and helplessness of a situation, guiding our dreaded movement in closer to the corpse. Because it is already outside the image, sound extends the act of registration beyond sight of the body. And because Mammy is made present through repetitive, nonnarrative expressions, she acts as an endless mortuary device. Voices usurp the spectacle from the body and the tableau. Because sound is not synchronized around Kitty’s death in Mamoulian’s film (as with Vidor’s), it remains an incomplete event still hovering between attraction and narrative, between the spectator’s act of observation and that performed by the usual onscreen beholders.

With the above reference to the cinema of attractions, I have in mind Stephen Best’s observation that the earliest film representations of African Americans denied narrative access to the black figures, framing them as performing subjects on display. Best points out that this technique of display—though clearly not unique to black subjects in early cinema—contained any kernel of black narrative mobility within an endlessly repeated action going nowhere, a “sea of representations of a temporally arrested blackness.”⁷² Not surprisingly, Mammy’s reaction to death temporally forestalls narrative registration for reasons having as much to do with her racialized expressivity as with Spunk’s narrative departure.

It is certain, however, that aural grief—of whites as well as blacks—is a lasting Hollywood habit. Like the singing caveats throughout this musical, Mammy’s wails in Hallelujah! do not supply narrative mastery; they are spectacular on their own. They lengthen the spectacle of death and stall narrative progression, continuing the static tableaux of grief to prolong the interval between the body event and the narrative’s resumption. Sound commands incompletion: it is not completely seen, its source not always fully defined. The early losses in the sound film have no clear coordinates because they do not focus on the body. To compensate for the death scene’s omission, Mammy’s desire literally “cries out.”

What we hear on the soundtrack outlasts the familiar bodily gestures around the body, and Mammy’s cries are central. The images of the processional and of the family’s approach to the wagon almost luxuriate in the late encounter cried out on the soundtrack. Applause and Hallelujah! withhold the union of the living and the dead, and they lack a tableau or pause to bring the observer near the body. Their dramatic peaks are located elsewhere—in the longing for synchronicity of sound and image. Mammy’s voice travels, but it has nowhere to go, no place to stop and rest. It envelops the body that has lost its voice for good. The Mammy figure has long been coded as a fantasy of coherence and origins; here, she offers a durable vocal expression at the end of her son’s dying, one that banks on the audiovisual image’s funerary qualities.⁷³ Is it a coincidence that two of the three singing acts on Rob’s High Fidelity funerary list are African American women?

Quiet bed deaths hardly exhaust the early terrain—note in passing the machine-gun killings that flooded the gangster genre or the offscreen screams of horror films. Spadoni looks to offscreen sound to rethink the impact of synchronization on the horror film, identifying “unseen bumps and offscreen screams” as constitutive of the genre rather than merely characteristic of it.⁷⁴ Offscreen screams point to the spatial dynamics of diegetic sound. When in The Bride of Frankenstein (1935), Henry’s assistant, Hans, is pushed by the monster off the top of the castle, his death is represented almost exclusively on the soundtrack as his scream pierces then fades while he falls. In other words the soundtrack conveys a sense of space that allows dying to unfold—albeit quickly in this film—as a progression into silence. Death here is the aftermath of a silenced continuous voice.

And earlier, in Frankenstein (1931), a film that literalizes posthumous motion in the figure of the monster, the death moment and process of registration are also sonorous. Both Fritz and Dr. Waldman die offscreen—the former after two piercing screams heard throughout the castle, and the latter after a cut to a conversation between Henry and Elizabeth. And after little Maria drowns, her father carries her corpse through the town’s musical wedding festivities for Henry and Elizabeth. The music contrasts poignantly with the silent stupor of his face and stride, as do the clanging wedding bells, which in this case double as a death knell. As Maria’s father continues his procession, the townsfolk gather in behind him (fig. 3.6). In one shot the camera tracks by three windows in which appear faces of townspeople. Maria is delivered to the Burgomaster’s front step. The town transforms from a party crowd to an angry mob as they become witnesses to the public registration of Maria’s death.

So convincing is the soundtrack as an unseen registrant that it is used to “kill” when sudden effects are cued to the impression of ballistic attack. Perhaps the close-up of Zeke’s hand, pulling the trigger of a gun, synchronized to a big bang—a type of close-up used by the early sound gangster film and many thrillers besides—was objected to precisely for its auditory fullness. The gun close-ups, as well as the sound of the scream from somewhere in the room, may have identified Zeke too strongly with the position of the physical assaulter because they embodied the acts of situational violence. At any rate the gun produced a new violent trick that worked with sudden sound. The death scenes in Little Caesar (1931) and Scarface (1932) all feature the newly available in-film sound of gunfire, but not surprisingly, not a single body is visibly punctured. Missing are “visual details” like “disfigured, dismembered, bloodstained and mutilated bodies, close-up views of dying men,” as Olga Martin, former MPPDA employee, observed in 1937.⁷⁵ If we see them at all, the actual performances of dying are quite standard (grasping of chest, falling over). Sounds associated with early violent death are acute; they do not continue to be heard throughout registration. They are sharp and instantaneous. In Little Caesar it is Rico’s “rocket” and its automatic eruptions that get the most attention. We do not see his first victim, Crime Commissioner Alvin McClure, die. He merely drops to the floor in unison with a gunshot, and we are told in a subsequent scene at the gangsters’ lair that he was “knocked off.”

In Little Caesar we do see, very rapidly, Otero and Rico drive by and shoot Tony on Father McLean’s church steps: the car drives past, and three puffs of smoke accompany three audible gun shots as the body falls down the stairs. But the composition of Tony in depth marginalizes this moment with the car’s foreground clearing of the frame and the sounds of the gun. Bodies die immediately. In short, the gun immediately announces a violent act from outside the body but does not indicate death or dying.⁷⁶ It is the liveness of the image that shows up in the wording of the PCA eliminations: “gangster shooting man,” “man being thrown out of car,” “man falling in doorway,” “men falling.”⁷⁷

The violent acts in Scarface are spectacularly choreographed: appreciate the bowling alley scene where in a deceptive single shot (there are two shots dissolved into one, continuing the tradition of The Execution of Mary, Queen of Scots) we hear the offscreen sound of a round of bullets as Gaffney releases the ball; the camera follows the ball away from the body until it knocks down nine pins, leaving the last one hanging. Then a second round of offscreen gunfire makes the point quite clear that the pin—which successfully drops on synchronized cue—continues the body event fully relegated to offscreen space. He first rolls a spare that then barely becomes a strike; the pin wobbles before toppling over. We may infer spasms from the pin: its falling, not the body’s, completes dying.

The “strike” completed outside—and as an extension of—the body event contains the X that appears throughout the film as a kind of signature of completed death. This Hawksian “grotesquerie” of Xs formed around or with besieged bodies emerges not through the means of violence but through the commentary the camera provides after the body event.⁷⁸ Hawks maintains for himself some of the flourish of gangster killing through his posthumous remarks—a whimpering offscreen dog after the Valentine’s Day Massacre; an offscreen screaming crescendo like a teapot after a booze truck is busted; a corpse forms an X with the shadow of a street sign that itself reads “Undertakers.” The camera takes away from the embodied performances of dying (which are short and sweet) to register the final configuration and “meaning” of the deaths. It is where the deaths occur that supplies commentary, not merely what we see the bodies do. Hawks’s camera connects an urban architecture of apathy to the visual and sonic immediacy of murder. As Richard Maltby reminds us: “The inventiveness may be the auteur’s; the discretion, however, belonged to the institutions regulating Hollywood’s regime of representation.”⁷⁹ Registrants still appear and are instinctively needed, whether provided by beholders or by the camera itself, but they do far less work than in nonviolent deaths, for the dyings to which they attend are significantly less protracted and ambiguous. The sound effect’s impression of instantaneous ballistics can be faked, too. In an early scene from Ernst Lubitsch’s The Love Parade (1929) a quarreling couple of adulterers is interrupted by the woman’s husband. Upon his entrance she holds a handgun to just below her breast and pulls the trigger, and a bang on the soundtrack causes her to fall over. When the husband takes the gun to his wife’s suitor and shoots, the second man stands still. Together, they eject the full cartridge from the gun, and the second man laughs. Looking back at the corpse, they see her with eyes open, glaring. Lubitsch is poking fun at the soundtrack’s verisimilar ballistic effect, all the more ridiculous given the earlier cutaway to an approaching street-side crowd when the gun goes off.

The early films explored in this chapter are not indicative of Hollywood standards of sound editing. Produced in an experimental era, each features its own uncommon novelty, from Crosland’s part-talkie (itself a rarity), to Mamoulian’s use of two microphones and a mobile camera, to Vidor’s access to postsynchronization. Rick Altman explains that at first sound technicians disputed where to place the microphone in the shot. One idea was to link sound to the image track each time there was a cut in the latter, for it was believed that if sound remained constant throughout different images the result would be unintelligible and a deviation from the way we experience sounds in real life. The first impression of sound-image match was, as Altman says, of “the constitution of a monstrous spectator.”⁸⁰ Monstrosity became the norm, however, with the standardization of the continuous-level soundtrack in 1930s Hollywood. As James Lastra argues, the model for sound recording in classical Hollywood “tends almost uniformly from the early thirties on toward the telephonic,” foregrounding the intelligibility of dialogue.⁸¹ The dominant two-shot of dialogue over the deathbed, and the ubiquitous soaring nondiegetic score throughout such scenes, would eventually create their own expectations and habits of viewing.

Asynchronous deaths continue in Hollywood, used to particularly good effect in Lewis Milestone’s All Quiet on the Western Front (1930), in which Paul dies in a close-up showing his hand as it reaches for a butterfly. Having returned home and remembered catching butterflies for his sister when they were children, Paul returns to the trenches to find all his friends dead or gone. In a momentary lull in battle, he spots through a hole in the barracks a butterfly and reaches for it. As an unseen enemy shoots Paul, the hand that has almost touched the insect recoils, and the camera quickly shifts right to separate (forever) the hand from the little creature. This all occurs with no musical accompaniment, and at the moment of impact the diegetic sound cuts off as well and does not return. What we see next is the memorably posthumous epilogue of retrofit replay in which in line the young friends walk away from the camera, one by one looking back at it, superimposed over a battlefield of marked bodies. In Old Chicago (1937) begins with Patrick O’Leary’s death in front of his Chicago-bound family while the youngest son sobs offscreen throughout the final speech. After his last sentence (“Let Chicago come to me as I couldn’t come to it”), Patrick sighs and closes his eyes, while on the soundtrack a wind begins to blow audibly. There are calls without response in the last two images: “Patrick!” and then, “Ma!”