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

We could take this opportunity to define the deathly nature of the cinematic image as the offscreen space to Gorky’s frame, the place “beyond” the border where life flows. What can be seen can be unseen. But there is more to consider in this anecdote. In Gorky’s eyes cinema produces not only life’s phantom trace but also the potential for death. The fact is that cinema has been “dying” ever since Gorky’s encounter in at least three obvious ways: by representing evanescence in much the way Gorky describes; by subtending itself to the next new screen technology and thereby dating itself; and by staging scenes in which living figures (humans, animals, aliens) and even “living” ones (robots and computers) can be seen in the act of dying. In some specific cases dying has packed in the crowds at movie theaters. (It is ironic that the second form of passing away has bolstered the medium from possible extinction.) What moves and entertains audiences in these spectacular instances—say, Psycho’s protagonist’s being slain midway through the film, Saving Private Ryan’s smorgasbord of combat horrors, Kill Bill 2’s joyful staging of the Bride’s execution of the Crazy 88 members, for example—seizes on the cinema as an apparatus uniquely capable of unfolding the process of dying before our eyes.

Movie technology has remained linked to dying since at least the 1895 Edison Company reenactment of a death by beheading (The Execution of Mary, Queen of Scots). We watch protagonists, secondary characters, unnamed groups, and digitized hordes perish in movies and on television. David Thomson muses that “the American 18-year-old has seen 20,000 acts of killing in movies and on tv.” According to Thomson, actors achieve legendary status for the way they wilt—James Cagney’s footwork, Edward G. Robinson’s “cadenzas” and “whirling expirations,” and Lillian Gish’s self-induced sicknesses.² Some genres, we know, are especially concerned with death, so we pick them when we are looking for it: slow and pathetically attended in the weepie, sudden and with painful explicitness in horror. We find websites devoted to amassing “greatest” movie death scenes, such as Tim Dirks’s Filmsite, where a skull and crossbones icon references a scene’s ranking on other lists (Total Film Magazine’s “50 greatest”) and files it into generic categories (i.e., “melodramatic,” “horror,” etc.).³ Not every filmed death is so frivolous. Other websites specialize in the “caught on camera” and “faces of death” variety, storing and selling actual footage, recorded, in most cases, by film and video cameras, not to mention the countless “snuff” or “mondo” films and videos still available for purchase and rental.⁴ The archival potential of the web to categorize fatal footage is nowhere clearer than on YouTube, where one can play and replay a variety of human and animal clips, featuring accidents and executions by chair, rope, and gun. And what the camera captures is not always a rote shorthand: despite the tendency to think of special-effects violence as the norm, the dying we see on film is often not over in the blink of an eye but enjoys an endurable tenure. It is quite often nonviolent in nature. Slow or static “dyings” indicate a more watchful capability of the camera than do spectacular killings that show spilled blood and lost limbs. Cinema’s ability to stage, time, and dilate the process of dying underwrites more generally its propensity to do so suddenly and violently.

Films routinely visualize the transformation of human bodies from “alive” to “dead” in both fictional and real sequences. Not much competes regarding our access to mortality; movies have all but replaced actual experience. Thomson guesses “we have witnessed a lot of killings … I wouldn’t doubt 100,000” to conclude “I have seen just two dead bodies. From what I can gather, asking around, two is on the high side—enough to be thought a little morbid.”⁵ We know, for instance, that the percentage of Americans who die in hospitals or nursing homes steadily increased last century, from 20 percent in 1900 to 50 percent in 1949, 75 percent in 1980, to about 80 percent in 1997 (about a quarter of these inside nursing homes).⁶ If it is placed outside the home, Philippe Ariès suggests, death becomes “forbidden”; it is “effaced,” made “shameful” as an interruption of social life.⁷ The trend to move it out of sight coincides with efforts to render and monitor death with machines that promise to indicate precisely when it occurs, like the EKG and EEG monitor, or devices that promise to inflict it tractably and instantaneously, as in electrical forms of capital punishment. The cinema camera has played an influential but unacknowledged role in this history. The same technology that has been used for entertainment has also been used in medical practice. From biopsies and body imaging tests to posthumous autopsies, both analog and digital imagery have gained evidentiary and indexical status, ushering the analytical gaze back into the past or deeper beyond the barrier of skin. Medical doctor Robert Artwohl revisited the footage taken by Abraham Zapruder of President John F. Kennedy’s assassination, wanting to disprove Oliver Stone’s use of that footage in his 1991 film JFK as evidence for a multiple gunmen theory. Artwohl published in the Journal of the American Medical Association a kind of “autopsy” performed on the Zapruder movie, arguing through frame-by-frame analysis that Kennedy’s reaction to the bullets evinces continuity.⁸

Perhaps most instructively, the use of video and digital screens to monitor critical condition is customary in the hospital room, pointing to an interdigitation between “screen” and “real” death. The moving image, understood in an expanded sense (analog, video, digital), has supplied both amateurs and professionals with a reference point. Mr. Wilde may well have appreciated the fact that death has become imitative of art as well. It is difficult to tell where “real” ends and “movie” begins.⁹ The very fact of the “snuff” film—a genre of enacted deaths purported to be real—indicates that a perceptual ambiguity between the real and the staged thrives in screen culture.

The overlap between dying and moving images knows a formal dimension. Still image traditions of painting and photography can intimate moments before or just after, leaving something quite fundamental to the imagination. But with cinema comes the lure of capturing a fuller transformation. Indeed, the moving image itself transforms. Like the body, it moves.¹⁰ Even so, this elongated visual form does not illuminate the secrets of the death moment. The excess of motion makes difficult any attempt to ascertain such a moment. As Vivian Sobchack writes of the Zapruder film: “Played again and again, slowed down, stopped frame by frame, the momentum of death escapes each static moment of its representation and frustrates our vision and thus our insight.”¹¹ Looking for a “moment” belongs to an earlier conception of death as a single commutative interruption of life, a notion that still persists though it has been updated by science. It is also a notion that has found its visual illustration in the photograph. Writing of photography’s essential temporal qualities, Roland Barthes suggests that the materiality and stillness of the image reinforces its marking of the past as present in the form of a visible instant. This instant is mirrored by the click of the camera’s button: “Death, outside of religion, outside of ritual, a kind of abrupt dive into the literal Death. Life/Death: the paradigm is reduced to a simple click, the one separating the initial pose from the final print.”¹² Though motion pictures produce instantaneous effects, they do not produce instant death, undoing the switch-like break between two states of being, or what Barthes, invoking the click of the photographic camera’s button, writes as “Life/Death.” Film strives to maintain that slash but blends two separate states. The slash blurs out of focus. If we assess film’s ontological connection to death by way of photography, we realize a very different quality confronts us in the moving photograph. For Barthes, photography’s stillness, its rigidity as an object and temporal document, denotes the exsanguination of the corpse itself. Cinema, however, cannot achieve such a suspension of time (or, for that matter, a pure arrest of motion) even in its approximation of photographic stillness we call the “freeze-frame.” Cinematic death-as-stillness cannot be thought, or at least, it cannot be thought photographically. So where does death belong in the moving image if not in visible corpselike stasis or time’s frozen halt?

Cinema death takes place somewhere between a precise moment and a complex progression in time. The camera is connected to other machines—gadgets ranging from toys to medical instruments—that have been thought to bring death closer to the doctor, the scientist, or the viewer but that raise new concerns or cast new doubts. “Our technology has not illuminated death; it has only expanded the breadth of our ignorance,” writes Dick Teresi in his study of brain death.¹³ For centuries machines have been used to supplement touch and sight in various ways: to determine a person dead, to specify the death moment, or to give the beholder a sense of temporal mastery. A full history is not tenable here, but a sampling of some of the earlier moments in medical history and popular culture when experimentation and doubt converged around machines will help place cinema along a specific technological trajectory, one that allows us to understand its death sign as both precise and fluctuating.

Two different but related ambiguities stand out in this history: one, the overlap and confusion between machine and human mobility (and vitality); two, the understanding of electricity as both a vital and lethal force. Distinguishing between animate (or “living”) and inanimate (or “dead”) objects, as we will see, becomes part of the question of determining when living bodies transition from one state to another. An important place to begin is the seemingly self-moving automata of the eighteenth century, those human- or animal-like clockworks that imitated gestures ranging from writing to defecating, playfully complicating the line between human and machine. In 1737 French inventor Jacques de Vaucanson created a doll that played the flute more quickly than any human could, performing not lack but superiority. The public display of these toys was often accompanied by an opening of the cabinet or doll to reveal the machinery inside, proving that the machines were not running on souls and that movement could be understood mechanistically.¹⁴ There were other jovial dolls, including a tambourine player, a poem writer, and a chess enthusiast, but as Gaby Wood argues, these androids also questioned what makes a human body a human body. In a (possibly fictional) story told about Descartes, the philosopher carried with him on a sea trip to Sweden an animate doll named after his daughter Francine, who had died seven years earlier. The story—whether or not true—clinches the point that machines were perceived to imitate and outlive us. Wood argues that the automaton induces the uncanny because it mimics human movement but lacks an understanding or experience of time: “Men are mortals, clocks are not.”¹⁵ In this line of thinking it is the beholder that links machines to mortality: androids embody death because they present movement that seems everlasting, a privilege the viewer cannot enjoy. As Wood points out, the word replicant—the modern android found in Ridley Scott’s Blade Runner (1982)—combines into an uncanny mixture both the terms replica and revenant.¹⁶

Anxiety about mechanical embodiment—the revenant side of the equation—can be found harboring cinema’s invention, in both fact and fiction. Auguste Villiers de l’Isle-Adam’s fantastical account of Thomas Edison in L’Ève future (1886) expresses both a desire for and fear of artificial life. In this science fiction tale, the scientist invents for his friend Lord Ewald an android with the faculty of speech to replace Ewald’s attractive but docile and intellectually vapid fiancée. In E.T.A. Hoffmann’s earlier story “The Sandman” (1816) a young student, Nathanael, falls for a girl he sees in a window, only to find later that she is a doll cocreated by older men. In such stories males fall in love with female dolls and die before consummating the relationship. If, following Annette Michelson, we look to the android in L’Ève future as a proleptic glimpse at the representational structure of the cinema, we must acknowledge that it is an incomplete, even optimistic, one.¹⁷ Those who think about films as films, as material objects, are accustomed to regarding them as fragile and impermanent. Reels deteriorate and fade; they can be destroyed. Machines do die, even mechanical dolls. (In fact, they often “die” more than once, a point to which I will return.) The viewer can also be reminded of death by the machine’s breakdown, by a dying that it projects. Eighteenth-century mechanists such as John Fothergill believed that suspended animation was a death that might be cured by restarting—that is, resuscitating—the body, like rewinding a stopped clock.¹⁸ Scientists in the competing vitalist camp argued that death had not really occurred in suspended animation, that the vital principle was still present as a potential for life.

Eighteenth-century gadgets (human dolls, birds, and shitting ducks) blurred the line between the mobile body and machines, but the line was further eroded in the nineteenth century with electricity, a technology that would be used both to resuscitate and to kill. Along with mechanistic conceptions of life and death, new thoughts circulated about what vital forces animate us. Mary Shelley’s father, William Godwin, chronicled doll makers in Lives of the Necromancers, but Shelley’s 1818 horror story Frankenstein was equally inspired by the use of electricity to animate apparently lifeless bodies. Some uncertainty about determining death has always existed, warded off by an enduring belief that breath and pulse were surefire indicators of life.¹⁹ But for physiologists and physicians in the eighteenth century, these old criteria were suspect. In 1740, Danish anatomist Jacques Bénigne Winslow warned about the dangers of premature determination of death, insisting that putrefaction, not even rigor mortis, was the only reliable indicator. Such a claim was echoed throughout the following century by the likes of Bruhier (himself obsessed with verifying death) and Diderot.²⁰

A number of medical discoveries yielded this suspicion. William Harvey’s description in 1728 of the heart’s pumping and circulation of blood led some to question the absent pulse as a death sign. According to medical historian Martin Pernick, eighteenth-century medicine experimented with mechanical and technical discoveries on the apparently lifeless body to regain certainty in the face of resuscitation methods.²¹ Victims of drowning were successfully resuscitated; the first treatise for artificial respiration appeared in 1740. But electric shock (a technique more directly related to this study) soon dominated research on restoring heart and nerve functions. The first human case of resuscitation came in 1755. Giovanni Aldini—nephew of Luigi Galvani and one inspiration for Mary Shelley’s Doctor Frankenstein—publicly displayed electricity’s powers to cause twitching and apparent movement on an executed corpse in 1803.

Deathlike states were suspicious because of electricity’s demonstrated power of animating the inanimate and resuscitation’s efforts in restoring life when it looked like it had been lost. Physicians studied and speculated on suspended animation, catalepsy, and coma in order to detect vitality’s lack. Both technical and invasive, nineteenth-century death tests introduced such distressing items as a steel needle stuck deep into muscle tissue, a heated cautery deeply inserted into the throat, and a nipple pincher.²² The nineteenth-century fear of premature burial helped legitimize the medical profession in the public imagination. As Pernick argues, new and milder instruments of detection (the stethoscope was introduced in 1819 by René Laennec), as well as new methods of resuscitation (not just electroresuscitation but also open-chest massage), helped allay premature burial panic. This is not a story of Foucauldian shift, of one paradigm’s replacing entirely an older one. Previously favored methods were, and are, still valued—discoloration, rigor mortis, breathlessness. But technical expertise bolstered the profession when dealing with intangible and invisible signs. The medical gaze turned inward, too, with the later inventions of the electrocardiograph (1903) and the electroencephalograph (1924), both of which pictured external information about the body’s hidden vital activity. By the time the EEG was introduced, failure to respond to electrical stimulation was considered an infallible indicator, and by 1930 one physician defined life as the ability to generate electrical current.²³

There is, of course, a flip side to the use of electrical machines to supplement old methods of determination. In the 1880s electricity was also used to induce death, first experimentally by amateurs and scientists, later by American law as an accepted means of capital punishment. Whereas Aldini electrified corpses to promote his uncle’s galvanic fluid, Harold Brown electrocuted animals in Edison’s lab to promote the lethal effects of George Westinghouse’s alternating current. (We will meet Mr. Brown later.) This current was later used to execute William Kemmler on August 6, 1890, and despite eyewitness reports alleging a terribly bungled incident, electrocution quickly became the law of the land. Six states still have recourse to the electric chair. With the chair, a machine thought to deliver a spectacle of instantaneous vital cessation produced a horrifically imprecise and gradual process.

Cinema—born of electricity—comes to play in this medical history of staging animation, studying vitality, and watching it end with the aid of technical instruments. What distinguishes cinema from these earlier devices is visual and temporal—it projects an image removed from the spectator by the screen, and it produces a more sustained temporality. We should not look to movies, however, to fix by magic the problems of modern medicine, to do it “better.” Despite new technologies, public confidence in the medical institution has waned since the rise in the 1960s of brain death as a criterion for removing patients from life support and the concomitant increase in organ transplantation. Debates over whole-brain vs. higher-brain criteria continue, and, given the 2008 white paper report of the President’s Bioethics Committee arguing in effect that higher-brain death need not be established in order to meet the criteria for whole-brain death, the debates seem likely to continue.²⁴ Though technology has been used to measure vitality and monitor its absence, it has also contributed to an understanding of death as a process of multiple deaths, a staggered trajectory. To quote the narrator of Science Channel’s Back from the Dead, who describes a flatline—that most common moving medical image in the popular imagination—on a hospital monitor: “This is not the time of death, but it is the time of dying.”

Let us return, then, to cinema, where we watch the transformation from alive to dead unfold, or the “time of dying.” Looking closely, we see death happen here both cleanly and turbulently, precisely and momentously. A combination of instantaneity and flux situates film within this modern trajectory of monitoring and visualizing cessation more precisely with supplemental devices. The first and last chapters of this book bring to bear on film two such devices—the electric chair and the heart monitor. These two devices have their differences, clearly (the chair is lethal and inflicts its vital cessation), but both are thought to be capable of immediately visualizing death as occurring in time. The other chapters take up film-specific technologies—narrative editing, synchronized sound, postsynchronization—as related instances of a machinelike precision. The link between the movie and the medical screens is particularly enlightening because of their mutual projection of moving images. Just as films visualize a mobile death sign for spectators, so, too, the monitors in intensive-care units register the flowing waves of the heart’s vitality and their gradual weakening and slide to “flatline.” Both these moving images “confirm.” The EKG has become so user-friendly, so self-evident, that it competes with the dying body for our attention during a deathwatch. The screen has become an uninvited guest within the final tableau. At an interdisciplinary conference on death and dying at the University of California at Berkeley, Dr. Guy Micco described his need to turn off the still-bleeping machine after declaring dead his patient “Mr. Reggie”:

Micco’s anecdote is corroborated by another story from Dr. John Murray, a medical doctor in San Francisco, who recounts monitoring vitals in the final hours of his patient “Mr. Guzman.” Here we can see the difference between a lay and a professional encounter with the monitor screen. Though the head nurse turns off the “television screen” beside Mr. Guzman’s bed, she keeps it on in the nurse’s station because, as Murray explains, “We must know what is going on at all times.” He offers an explanation of this difference: “She turns the bedside monitor off because we have learned that relatives and friends tend to pay more attention to it than to their loved one. People become obsessed watching the electrocardiographic squiggles that announce each heartbeat marching across the screen. Although the physiological nuances are undoubtedly overlooked, the deeper meaning of the electrical signals of life is hard to miss when they slow in frequency, gradually widen in appearance, and then finally stop altogether.”²⁶

We will return to this narrative of slow, widen, and stop. Suffice it to say that the “deeper meaning” is indeed difficult to miss, for though dying is gradual rather than instantaneous, the monitor tracks its narrative with marked changes—from wavy to linear. That is, it translates, or rather reduces, a gradual and uneven unfolding into a sequence of clearly marked stages, like points on a line. But in both anecdotes the monitor must be turned off because it replaces the body and the drama that customarily surrounds it. Though it is delivered on a “television,” the electrical output of the heart is tracked cinematically, that is, as an indexical sign that moves over time, one that is connected to its referent through physical contiguity. These anecdotes complicate our faith in the monitor by reminding us that it can give mixed messages—an image of fatality that does not match the present knowledge of the body, or the familiar bleeping of vitality set against the visible inactivity of the deceased. Both hospital monitor and movie screen are thought to be able to visualize death precisely, and both come up short.

Those who have tried to locate an essential connection between cinema and death divide into two camps—one searching for a visible quality within the filmic image, the other questioning what death might “be” before assuming what it might look like. In the first group we find important works by Garrett Stewart and, more recently, Laura Mulvey, who read cinema death in relation to photography. For both scholars static images in film, like the irruptions of the freeze-frame, conjure up an immobility that is always deferred by the medium in its essential flow forward; if the movie “pauses,” it will unpause.²⁷ Though he sustains an interest in medium specificity, studying “the visual phenomenon of death as only cinema, never photography, can capture it,” Stewart does proceed with confidence about when death occurs, or rather, that it occurs as an event. He conceptualizes death photographically. It is “the moment of the character’s conversion from subject to object” (154), for instance. Whether or not it occurs as a moment is not a question he seriously probes because film is not expected to reflect somatic death, much less teach us how to think about it. When there are corpses onscreen, death is abundantly evident, like a concept depicted (i.e., written, painted). The visible corpse inserts into narrative cinema a before-and-after; this insertion naturally requires the corpse be immediately apprehended.

Mulvey, too, reads cinema death as narrative stillness: “Of all the means of achieving narrative stasis, death has a particular tautological appeal, a doubling of structure and content.”²⁸ Thus her reading of Marion Crane’s death in Psycho privileges the stillness of the corpse—simulated by a close-up of her still-open eye—as a marker of narrative paralysis. Mulvey’s language imparts to the camera an act of objective reportage: she writes of the “momentary pause to register [Marion’s] transition from living human being, in whose story the audience has been deeply involved, to corpse.”²⁹ For Kaja Silverman this shot marks a pivot away from our identification with the victim Marion Crane, which is “sustained up until the moment when Marion is definitively dead, an inanimate eye now closed to all visual exchanges.” At this point “we find ourselves in the equally appalling position of the gaze which has negotiated Marion’s murder.”³⁰ Begun in photography, the search for nonbeing ends in the symbolic confirmation of corpses. But looking at film’s ability to “depict” death as an object misses the extent to which those in front of and behind the camera have tried to figure out whether it has a moment, if there is a way to visualize it as one instant or many, and, if many, how to prioritize one over others, and if so, which one to privilege—in short, it misses the point that the act of dying is an object, too, with no certain culmination or boundary. Nor is there much room left to wonder if the onscreen corpse is “definitively dead,” an anxiety that movies tirelessly exploit.

In the other camp we find a diverse variety of commentators who share a mutual interest in dying as a unit, or expression, of time. Film commits images to a time-based form, making repetition possible. The act of replay would seem to thwart an analysis of death as a temporal event, indicating more deferral. (If death can be repeated, it is not death; it is something else.) But two essays that have recently inspired critical work on screen violence show us that death’s temporal relationship to film yields significant insights into the medium. Though not written at the same time, André Bazin’s “Death Every Afternoon” (1949) and Pier Paolo Pasolini’s “Observations on the Sequence Shot” (1967) were each inspired by the act of watching a recording of an actual death.³¹ Other scholars have discussed these essays but have not emphasized their working through film form to understand a transformative process we might call dying. For Bazin, watching the footage of a bull’s gorging in Pierre Braunberger’s The Bullfight (1951) brings photography to mind, but as a contrast rather than a correlate. Working from the idea that we experience our deaths as unrepeatable, Bazin remarks: “We do not die twice. In this respect, a photograph does not have the power of a film; it can only represent someone dying or a corpse, not the elusive passage from one state to another” (30). Film’s capacity to make a repeatable viewing experience out of that “elusive passage” amounts to an “obscenity” on par with the depiction of a sexual act. Like sex, death brings forth a subjective time, “the absolute negation of objective time” (30). It cannot be experienced publicly. Influenced by Henri Bergson’s notion of durée, Bazin sees objective time as unfolding irreversibly. Activity can be reversed and repeated in cinema; however, death is the ultimate irreversible experience, so it is understood as a supreme occasion for film to demonstrate its ability to reverse time. Death marks cinematic temporality.

Pasolini’s film inspiration is the Zapruder footage I mentioned earlier. Repetition is not the crux of his argument.³² But looking closely, we find correspondences between Bazin’s essay and Pasolini’s. What concerns Pasolini is editing—specifically, the ability to arrange different views of an unfolding event from individual shots after they are taken.³³ Edited sequences might give us a privileged glimpse of death because they are constructed in postproduction. Unmitigated real footage, like that of Kennedy’s death, affords the recorder a chance to make legible the actions that constituted the event. It all happens too fast to compute exactly what happened. Pasolini envisions multiple cameras at different angles recording the assassination, all the “other visual angles” that could capture unique slices of the event. These angles are understood as the embodied perspectives of other witnesses, so the subjective shot is personified as a partial gaze. In this hypothetical scenario “subjective sequence shots” could be assembled by a “clever analytical mind” into one final, newly edited, sequence (234, 235). By “choosing the truly meaningful moments of the various subjective sequence shots and consequently finding their real ordering,” the editor would create “objectivity” through the partial views, a “montage” (235). Life’s final gesture does not bequeath meaning—it is not completed—until the isolated perspective is joined with multiple others. The editor becomes a “narrator,” a storyteller that confers past tense to real actions otherwise stuck in the present. With help Zapruder could turn the film of Kennedy’s dying into a montage of Kennedy’s having died. If all human gestures are understood ultimately after the end of lived experience, it follows that editing performs a unique deathwork on uninterrupted footage. Kennedy’s final gesture—the act of dying—poses a metonymic problem that editing solves.

Though Pasolini discusses here an actual death recorded on film, his montage of reality remains hypothetical. It does not pertain to unedited footage that precludes hindsight or rearrangement; in fact, it seems to favor the fiction film, that hypostatic collection of present pieces assembled into a newly configured past. This is miles from Bazin’s recoil at repeated viewings of actual executions. Replay becomes obscene when it induces repeated fatal incidents, like the shooting of a gun: “In the spring of 1949, you may have seen a haunting documentary about the anti-Communist crackdown in Shanghai in which Red ‘spies’ were executed with a revolver on the public square. At each screening, at the flick of a switch, these men came to life again and then the jerk of the same bullet jolted their necks. The film did not even leave out the gesture of the policeman who had to make two attempts with his jammed revolver, an intolerable sight not so much for its objective horror as for its ontological obscenity” (30). The filmmaker’s missing intervention into the murder is thematized by the onscreen jamming of the revolver. The delay becomes a sadistic indication of deferral’s failure. This brings Stewart’s aphorisms to mind: cinema is “death in serial abeyance,” and elsewhere, “photography is death in replica, cinema is dying away in progress.”³⁴ But cinema in Bazin’s anecdote demonstrates there is really no such thing as an abeyant death. Reading Bazin and Pasolini together, we learn something else about both edited and unedited footage: the former can control the unassailable disorganized present tense in which we live, whereas the latter can change forever the present’s status as unrepeatable. (One might also assume that it can slow it down and speed it up, and that slow motion and fast motion would similarly challenge the inalienable temporal flux upon which durée rests.)

The larger point is that these two theorists who look at death to understand film’s temporal logic move quickly to replay. For Pasolini film can make the present past, not by whimsically blending subjective shots into one another so that they melt into durée but by breaking them up and putting them back together. The goal of the sequence shot is not to destroy the present but “to render the present past.” For Bazin it is a cinematic specialty to repeat durée. His essays make clear elsewhere the asymptotic relation of unedited long takes to lived time, but he takes a strange turn with the bull’s fall.³⁵ At essay’s end he suggests that screen death does not replace the real event offscreen. Both gain powerful significance exactly because the cinematic does not replace the real: “This is why the representation on screen of a bull being put to death (which presupposed that the man has risked death) is in principle as moving as the spectacle of the real instant that it reproduces. In a certain sense, it is even more moving because it magnifies the quality of the original moment through the contrast of its repetition. It confers on it an additional solemnity” (31). Film moves the spectator closer to death: it is more moving because it cannot actually touch lived time’s privileged interruption despite its unique ability to harness death’s temporal process. This is not death in perpetual abeyance but in gradual approach. In fact, Bazin speculates cinematic representation elegizes its failure to bring death to the viewer. Punning off the French word for remorse (remords) with his term re-morts, he hails the “eternal dead-again” of the cinema, reminding us that lived time, as opposed to cinematic time, ends. The essay’s title phrase plays on this difference as well. Cinema can never arrive at death just once (which is to say it can never arrive). This is the reason cinema becomes “death every afternoon”—it moves us toward, but does not reach, its goal. In the place of direct experience, it magnifies, retimes, and repeats the spectacle object. In a move that reinforces Dziga Vertov’s kino-eye, we might say simply that the movie camera “sees” death differently than we can unassisted. Behind his language of resurrection, Bazin suggests that film cannot capture death—not because it cannot show the loss of life but because it cannot do so only once.

Frequently in such discussions, and in my own close readings that follow in this book, there appears a friction between human experience and a technology-assisted perception that runs, very often, on electricity. That contrast is obvious in electrocution and hospital accounts, but it is equally common in movies. Death’s evasion of the unassisted gaze compels that gaze to look toward machines for precision and authority. Pasolini points out one mechanical surplus of the camera when he mentions hindsight. The process of single-take dying can only be finalized by editing—the editor/narrator finishes the process by looking back from beyond the position of any one embodied spectator. Only the camera can provide the illusion of being in two places at once. Replaying death for Bazin is likewise beyond embodied human perception: the result is an eternal dying (remorts, or we might say as remoriens), not resistance to fatality. Simply put, the spectator cannot do anything eternally; thus film creates a time beyond human occupancy. The technique of the photographic camera earlier accompanied Barthes’s impression of the image’s fatality, as if the shudder occurring in the spectator were triggered by the revelation of a mechanical nature beneath the human.³⁶ But the moving image invokes the mechanical differently, committing an unwieldy process to a separately observable strip that sets up the spectator to search for a real occurrence that may not have existed before.

While Pasolini and Bazin both refer to actual footage, we can discover the deathwork of editing and what I will call “posthumous motion” (replay, repetition) in fiction films as well. If the single shot is linked to dying, it wants to terminate, to come to a close and empty out the signifiers of life. Editing is Pasolini’s solution; repetition is Bazin’s. But we might also see life’s absence in other figures and images. What I see in movie death scenes—from the goriest to the weepiest—is a mixture of the techniques of repetition and retrospective arrangement. Sobchack claims the fiction film tames the perceptual excess of death by clarifying the moment through careful staging or special effects prepared for the camera. Made up of iconic and symbolic signs, “representations of death in fiction film tend to satisfy us—indeed, in some instances, to sate us or overwhelm us so that we cover our eyes rather than, as with the Zapruder film, strain to see.”³⁷ This seems true. But just because fictive death “is experienced as visible within representation” does not mean that it is anymore visible. (It is often experienced as visible only by other observers in the scene, in fact, who screen the body for us.) One needs others to die, at least cinematically. The final moment is not clearer in the fiction film. There, too, death occurs but stays hidden by the body. No external fold can show the inner lining. In those cases in which movies stage the end from “inside” the consciousness of a dying person, we often get an externalized object or image—a reverse shot, a fade-to-black—that indicates a something or somewhere “else.” The fiction film pressures the image to harness the death moment because such a moment is visibly present while not directly visible—it can be seen on other surfaces, and as a consequence of its projection there, it is repeated to compensate for a lagging precision. These new moments must inculcate the impression of an actual occurrence, though that occasion is spurious. The screen body dies several times and in several places.

To take an apotheosized example of a violent death by way of contrast, consider the various images of finality that conclude Bonnie and Clyde. Here the camera pauses to show Bonnie’s slow-motion hand buoyantly resisting stillness, juxtaposed to the car. Now the camera reiterates the ricocheted bodies by shooting them through a cracked windshield. A long shot takes us out of the compressed scene of violence to bystanders who slowly approach the bodies for a closer look. Now the camera focuses on the face of a single beholder—Ivan Moss, the man responsible for arranging the gangsters’ ambush. The camera shows little interest in confirming the fact that either Bonnie or Clyde is dead by privileging a moment or examining their bodies. Rather, the variegated process under way is one in which the camera searches for multiple and powerful ways to embed their deaths in the world, to produce meaningful recognitions, to follow the reverberations of the aftermath. The film slows down, speeds back up, magnifies, cuts up, rearranges, repeats, and moves away—all of the techniques flagged as cinematic specialties by our two earlier theorists. Whether real, fictional, or somewhere ambiguously in between, death in film is not conclusively over without further supplemental gestures performed on the outside of the human body.

When state prisoners were electrocuted in the late nineteenth century, their still-moving bodies exceeded the instantaneity on which electrocution was founded in the first place. Capturing the absolute moment on the hospital monitor does not equate magically to its occurrence. Death must still be pronounced by the doctor, sometimes in vain. It must be registered. The camera shares much with these other technological gadgets. Although premised on accuracy, it resorts to other symbolic forces (doctors, lawyers, witnesses) to confirm death’s having happened. Much like in the legal and medical contexts, death in film is posthumous, its resolution after the fact, its prioritized moment obscured. Philippe Ariès writes of the mechanical cessation of care in the modern hospital: “Death has been dissected, cut to bits by a series of little steps, which finally makes it impossible to know which step was the real death, the one in which consciousness was lost, or the one in which breathing stopped.”³⁸ Vivian Sobchack connected the “little steps” of technical support to those of filmic montage when she first wrote of documentary death in 1984: “the contemporary dissection of death into a series of ‘little steps’ … is paralleled by the initial recording of death by film moving through camera and projector in twenty-four ‘little steps’ per second and, finally, in always disappointing post hoc attempts to ‘find’ and ‘see’ the exact moment of death in nonfiction films through a close inspection of every frame recording the event.”³⁹

The registrants that step in to apprehend a dying has concluded offer the symbolic function of medical and legal witnesses that work for the state. Given the modernity of its context, it is important to note that the registrant functions as a witness whose power exceeds—sometimes defiantly—the boundaries of the state. Registration is embodied but dispersed, in excess of any one body, including the human body. And insofar as the spectator behaves as a registrant, the moment of apprehension is not isolated but recurs through multiple viewings, a repetition not unlike Adam Lowenstein’s model of horror spectatorship as dynamic in affect because multilayered in time.⁴⁰

Death confounds representation, or as Elisabeth Bronfen and Sarah Webster Goodwin tell us in their introduction to an anthology on the topic, “Every representation of death is a misrepresentation.”⁴¹ But film’s lack persists as a technical scourge. Pasolini implies the Zapruder footage unfolds too fast; Sobchack muses that recorded hospital deaths unfold too slowly.⁴² How does cinema get dying “right”? Even when the footage is clocked at real time and unedited, the ability to replay the footage changes the temporal frame of the original occurrence in the way it allows access. We look to fictional scenarios to organize and clarify the end of life, for they certainly have greater recourse to the profilmic (again, Sobchack claims the fictive “satiates” whereas the real “evades”). It is true the fiction film can draw out or eclipse the death moment, combine various views, and stage registration multiple times, and that is why I take up a serious study of entertainment films in this book. But the distinction between fictional (diegetic) and nonfictional (indexical) film is itself, of course, problematic. It is often difficult to tell that difference. With death footage, that ambiguity is often the point of making the film. What is thought to be documentary often turns out to be fictive, as with the “snuff” film, and there are instances when actual footage is shot and edited according to narrative codes previously established, such as Jan Oxenberg’s documentary of her grandmother’s death, Thank You and Goodnight.⁴³ Both fictional and actual footage “produce” rather than “represent” death as an object extrinsic to the beholder, and in this process the spectator becomes internal to the object through identification with the registrant. The multiple posthumous markings we find there, the compensations for photographic lack, are not simply representational crises but encounters between human comprehension and machine-based accuracy. Without these markings onscreen, we could not have death as we know it.

This book departs from the tradition of writing about film violence.⁴⁴ Of course, ubiquity is a good reason to focus on spectacular death.⁴⁵ Perhaps intimate watching lacks entertainment value, as bell hooks suggests: “Dying that makes audiences contemplative, sad, mindful of the transitory nature of human life has little appeal.”⁴⁶ In any event I need to move away from violence (that is, the infliction of mortal wounds in a causal chain of bodies) because I am in search of an intimate relation between the dying person and the registrant, and between both parties and the camera. This is not to say that registration does not attend murder, only that its appearance in slow or static “dying” dilates—allowing us time to see clearly—the posthumous and disembodied effect of screen death more generally. This is also not to say that the difference between a violent and nonviolent death is absolute. Simone de Beauvoir, for one, refuses to acknowledge such a thing as “natural death” in her account of the final stages of her mother’s fight with cancer.⁴⁷ But seeing dying provides an analytical suspension that captures the film medium’s grappling with time, terminal stages, and the moving body. Films that do sustain an engagement with the final few moments of life often resort to supplemental features of editing and mise-en-scène to seal the deal; these features, in turn, visualize change in the world outside the body, not on it or in it.

To emphasize the technological nature of the camera’s search, I move through a reading of films from historical periods when technical innovation is evident—the turn to narrative, for instance, or the transition to sound. This history reveals that technical developments, as much as actor talent or period style, contribute to the impression of when, where, and how life ends. Screen death, in other words, has a formal history. Formal possibilities emerge at each stage of innovation before settling into standardized technique. Lists such as Tim Dirks’s “greatest death scenes”—and static claims like Sobchack’s distinction between “indexical” and “fictive” or Stewart’s and Mulvey’s on medium ontology—show little regard to technical developments in the cinema. A full history is not possible here. But there is more than one “death,” and thinking of its “madeness” reorients the way we think about cinema and the dying body.

To think about that relationship, I have set up my own deathwatch—a sustained look at movie scenes in which the image seems to foreground the anticipation for death and the search for a perceptible break in vitality.⁴⁸ To answer the question of when death has really happened, it is imperative to look at scenes that seem to be invested in answering that question, whether this interest is attributed to the camera, to the embodied gaze of onscreen or offscreen beholders, or to narrative form more generally. American film structures the death moment around switches, instantaneous effects, or formal cutoffs—different barriers, or partitions, to divine Life/Death that, in turn, express a quality set against one privileged moment. This book is organized around medial shifts in that architecture. Starting with the earliest motion pictures, chapter 1 connects the electrocuted body to cinema’s first attempts to visualize death. Early films often wanted the moment the body stops moving. Trick reenactments of beheadings, hangings, and lynchings were common early on, such as Edison’s Execution of Mary, Queen of Scots (1895) and American Mutoscope’s An Execution by Hanging (1901). Clearly salacious, execution films staged the substitution of human with dummy body to inculcate the impression of an instant cleanly marked. When the Edison Company turned to electrocution as a film subject (Execution of Czolgosz, 1901; Electrocuting an Elephant, 1903), a new image emerged that expressed the duration of dying without substitution. Both types of execution films employed registrants in the frame to conclude the scene. The on/off electrical switch organized perception around a clean break, but in both real application and its translation to the screen, this switch produced an untenable process that had to be verified after and elsewhere.

The middle chapters take up still “dyings” vis-à-vis technical possibilities specific to the medium—narrative editing, synchronized sound, voice-over, and noniconic screens. Each of these technical accoutrements affords filmmakers a new “switch” to pull. Chapter 2 looks at early story films of D. W. Griffith (1908–13) to consider the changes that occur within the scene of registration once movies bridge together several shots to unfold a story in a fictional world, or “diegesis.” Narrative editing harnesses dying as a temporal process, but there is always the threat of incompletion. In Griffith’s films at Biograph Studios, the dying moment is structured around its onscreen perception, affording the registrant, and us, additional moments to gather around the body, or the space where the body was last seen. The techniques of repetition and disembodiment in the earlier moving images are still at work in the narrative features. But there is more. In films such as The Country Doctor (1909), The House with Closed Shutters (1910), and others, we see the first signs of a response to death emerge within the filmic image, something in the scene after registration. A number of nonhuman or nonfigural registrants appear as well, marking what I call “posthumous motion” as a transitional or repetitive trope.

With registration and posthumous motion in mind, chapter 3 moves to some important films from American cinema’s transitional period to sound. The soundtrack offers filmmakers another on/off device that sets up potential precision (the last word or audible breath, for instance). What develops is a strategy to use offscreen sound as a registrant. The early period is an experimental moment—one loaded with movies that make thematic connections between the soundtrack and the dying body or between the soundtrack and the registrant. These early films clinch the importance of recorded sound as an accoutrement, years before the nondiegetic musical score would become the norm. No longer merely seeing dying, in these films we are hearing dying. And what we hear does not emerge from the dying body but rather from the acoustic space surrounding the body, contrasting sound with image in various ways. The Jazz Singer (1927), for instance, covers the Cantor’s final gesture with his son’s offscreen singing. In Applause (1929) Kitty Darling dies while voices chatter and quibble offscreen, unknowingly. Sound envelops the dying body in an acoustic fold and employs the voice’s posthumous dimension to extend screen focus outside the body. Chapter 4 continues this thread by looking at voice-over narration, another manipulation of synchronized sound but one that poses a variation on earlier films. From John Ford’s How Green Was My Valley (1941) to film noir to more recent noir descendants (The Sixth Sense, 1999; Donnie Darko, 2001), American cinema shows a perdurable interest in the disembodied voice to “open” the experience of dying from the inside. Often the disembodied voice triggers a doubling of the death moment, a second chance to approach an ending that could not be anticipated the first time around. As a consequence of filmic repetition, the voice-over keeps the process of dying in permanent suspension, though it seeks exactly to provide closure.

I bracket this study with offscreen death technologies, returning to another electrical machine in chapter 5—the EKG or EEG monitor. Films that feature flatlining characters, from 2001 (1968) to Steel Magnolias (1989) to The Matrix (1999), stage the interface between a younger screen technology and its parent source. What the cinema screen offers that the hospital screen cannot is a representational form that exceeds indexicality. Movies do not need to read the lived body truthfully as do hospital machines, yet they place the flatline within visual and narrative sequences that stage in effect the verification of death via screen technology. The movie camera reads the monitor screen—positioned outside the body—to grasp an exact instant it has long sought. But, time and again, the lineated waves and steadied bleeping do not announce the terminal moment with certainty, no more on film or television than in reality. Passing as a faithful indicator, the flatline allows the camera to resume its play with figural forms, but these forms reveal anxieties about the attendance of machines at our deaths.

This cinema history of death and dying is based on close readings of American films and moments of cultural significance. However, it is not my claim that the history of formal innovation is unique to this cinema; it is only here that I draw the line. Nor am I searching for a uniquely “thematic” occupation with death in these films. If, like Leslie Fiedler in his canonical work on the American novel, we look for a characteristic national sensibility, I’m not certain we would find it by looking at causes of death or by recapping plots.⁴⁹ We would do better, it seems to me, to look at the connections between plot and form, between theme and the technical possibilities of the medium. In doing just this, I do hope to question the usual characterization of American cinema’s representation of death as violent and sensational. This description—often intended as a critique of Hollywood’s deflection of our attention away from mortality—does not hold up when we look at the historical record. Slow or static “dyings” appear at key transitional moments in American film history, and it is to them that I would like to draw my reader’s attention. Dying in my examples is less spasmodic and spectacular than one might expect. For example, it is true that chapters on execution, on film noir, and a section on the gangster film would seem to belie the rubric of sustained dying. However, though criminals and antiheroes might suffer rather violent gunshot wounds, they are seen dying for a protracted moment in time. Theirs is a violent death, to be sure, but it is also often a stoic and talkative dying. Again, I am watching scenes in which the camera—under someone’s operation—also seems interested in the final moments, even if the blow to the body would seem fatal and death foregone. It is therefore sometimes necessary to cross generic dividing lines, not to suggest genre is irrelevant but to underline an enduring interest throughout American film history in human and visual/spatial responses.

Throughout these readings I look at the relationship between embodied human perception and mechanical surplus. In the context of increasing technology, the registrant figure in American cinema has something to tell us about the need to perceive death as a human, even communal, event rather than as an integer or commutative announcement. To conclude this study of the technical conditions for the aesthetics of a mobile death, an epilogue considers recent recordings (film and video) made in hospital or at home that structure real death “cinematically.” These moving images appear as a kind of afterthought or redress to Hollywood. It seems that the screen body might replace the mortal body as locus for registration, with the camera as registrant. Acts of replay and disembodiment in real recordings would seem to suggest an end point for cinema’s excursion into dying as an experiment in form. Death has not only lost its bite; it has lost its fundamental terminality. Movies have played a part in the erosion of that slash—real or perceived—that we imagine separates death from life.