6

The Cinematic Present:
From Intolerance to Interstellar

I

In 1915, at the height of his fame, D. W. Griffith proclaimed the triumph of film over all other forms of art: “It is the ever-present, realistic, actual now that ‘gets’ the great American public, and nothing ever devised by the mind of man can show it like moving pictures.”1 Movies, he implies, have the advantage of immediacy. They show their subjects directly, without the intervention of signs or symbols, and without any lapse of time, unlike written narrative, which seems to admit its essential belatedness by conventionally telling stories in the past tense. Apart from its boosterish exaggeration, this may seem an unexceptional claim, especially in the context of 1915, when Griffith was trying to gain respect for a medium still disdained by many as a toy. But it may also sound a little strange now, when Griffith is considered the first director to have pulled together the basic elements necessary to film narrative. After Griffith did his work, according to the standard account, films no longer depended on a series of attractions, one pure present at a time, but coordinated their moments in a series, so that complicated stories could be told. The “actual now,” in other words, was no longer “ever-present” in his work but was split up, delayed, displaced, and doubled or tripled in the service of suspense, pathos, social commentary, or other narrative demands.

A film as complex as Intolerance (1916) involves four different storylines, ranging over 2,500 years of history, with each storyline further divided in various ways, between characters and places, so that the conflict and suspense within each time period is counterpointed to that in two or three of the others. In what sort of “now” can all this be taking place? As the scenes shift from characters in modern dress to those in outlandish Babylonian costumes, how is the audience supposed to avoid the impression that the Babylonians are further away in time? And yet it was surely Griffith’s notion that all of this, being present to the camera in the same way, was also equally present in time. The gap between the purity of his claim, that movies occupy a single, solitary now, and the formal and temporal ungainliness of his most ambitious movie raises a basic question about film itself. What kind of present does a film establish for its audience, and how does that present relate to the prevalent use of film to tell stories?

This turns out to be a controversial issue among film theorists, though it is not often raised as such. One thing that Griffith meant when he claimed that movie audiences can “see everything—positively everything” was that seeing everything has the effect of seeing everything at once.2 This claim was frequently echoed by early commentators on film. Hugo Münsterberg, writing in 1916, describes the temporal effects of film as if shots were superimposed on one another and not arranged in sequence: “With the full freedom of our fancy, with the whole mobility of our association of ideas, pictures of the past flit through the scenes of the present. Time is left behind.”³ Ten years later, Terry Ramsaye said much the same thing: “The photoplay of today moves backward and forward through Time with facile miracle from the Present into the Past and Future by the cut-back, flashback and vision scenes.” In this way, film satisfies “the human wish to live in the Past, Present and Future all at once.”⁴ Each film, then, is a modern Christmas Carol, turning the ordinary present into the eternal present, balancing all of time on the head of a pin.

Later theorists have offered a much more mundane version of this idea, based on the commonsensical notion that whatever a film shows at any moment must be shown as if it were happening in the present. So Seymour Chatman says: “It is commonplace to say that the cinema can only occur in the present time. Unlike the verbal medium, film in its pure, unedited state is absolutely tied to real time.”⁵ Stanley Cavell quotes Alain Robbe-Grillet to the same effect: “The cinema knows only one grammatical mode: the present tense of the indicative.”⁶ However obvious this may be, it has been disputed by many. For it is just as commonsensical, of course, to assert that, as a photographic medium, film only represents things that are no longer present. For this and many other reasons, Gilles Deleuze insists that “the postulate of ‘the image in the present’ is one of the most destructive for any understanding of cinema.”⁷

Cavell and Deleuze are as influential as they are in part because of the complex negotiations they effect between these two commonplaces: that film transpires in real time and that it preserves what is now past. In The World Viewed, Cavell describes at length “that specific simultaneity of presence and absence which only the cinema will satisfy.”8 What he has in mind is something like this: the temporal and spatial absence of the filmed subject is overcome, turned into a conviction of its presence, by our confidence in the automatism of the camera, a confidence that depends, in its turn, on our absence from the negotiations between lens and subject. As he puts it, “photography maintains the presentness of the world by accepting our absence from it. The reality of a photograph is present to me while I am not present to it: and a world I know, and see, but to which I am nevertheless not present (through no fault of my subjectivity), is a world past.”⁹ So a film shows what is past but makes it present, not just because the film reels, the projector, and the screen are present with us, but also because the filmed subject itself exists in a present tense now coterminous with and indistinguishable from our own.

Deleuze objects to this very common and apparently sensible idea that film exists primarily in the present, but only because he also objects to common assumptions about the nature of the present itself. For Deleuze, following Bergson, the present must always be internally differentiated, since to be registered as a present it must also already be a past. In these two modes, the present is successive to itself, “still present and already past, at once and at the same time.”¹⁰ The image is that which allows us to see this internal successiveness in its full amplitude, extending out as it must to encompass all of time. As Andrey Tarkovsky says in his treatise Sculpting in Time: “The image becomes authentically cinematic when (amongst other things) not only does it live within time, but time also lives within it, even within each separate frame.”¹¹ Even the smallest increment of film time, the single frame, thus contains some time within it, as does the smallest increment of actual time, the present.

This is not perhaps as radical as it sounds, nor is it necessarily very far from what Griffith had in mind. Münsterberg, at any rate, maintained that the separate scenes of a well-crafted narrative film seem to be “proceeding in the same instant.” Three different times and places may be presented on-screen, but “it can hardly be said that we think of them as successive. It is as if we were really at all three places at once.”¹² Münsterberg does not take this down to the level of the individual photogram, but Deleuze, following Tarkovsky, does: “What is specific to the image, as soon as it is creative, is to make perceptible, to make visible, relationships of time which cannot be seen in the represented object and do not allow themselves to be reduced to the present.”¹³ To see time transpire, even in the instant of a single image, is the apparently impossible feat accomplished by the great filmmakers that Deleuze lionizes in The Time-Image.

This inevitably means, though, that all the past and future are available in one present. Thus there is a tendency in some of the filmmakers that Deleuze prefers toward what he calls a “‘perpetual present’ cut off from its temporality.”14 Films that aspire to this temporal state do not necessarily need to have any of the internal organization usual in most films. As Tarkovsky admitted in relation to Stalker, “I wanted it to be as if the whole film had been made in a single shot.”¹⁵ The omnipresence of the “ever-present now” means that narrative devices are really unnecessary. Why should there be any elision, condensation, or rearrangement when the past and future are immediately available in the moment itself? Thus Deleuze has little interest, in The Time-Image at least, in narrative issues or methods, concentrating as he does on the description and analysis of moments in which the film has come to a virtual standstill. He warns that “the simultaneity of a present of past, a present of present and a present of future” does not mean the suppression of all narrative. But it does nonetheless abstract narrative “from all successive action,” so that even narrative is rolled up inside the moment, along with the rest of time.¹⁶

In contrast to this kind of concentration, the splits, displacements, and rearrangements that characterize conventional film narrative must seem like diversions at best. Thus Raymond Bellour points out how little Deleuze has to say about “the interruption of movement. The often unique, fugitive, yet perhaps decisive instant when cinema seems to be fighting against its very principle.”¹⁷ For Bellour, the most basic fact about a film is that it is made up of still images and thus of interruptions. Though these interruptions disappear in the process of projection, they tend to reappear at successively higher levels of organization. It is even possible for some filmmakers to “pull or lure the film, cinema, toward the point they designate.”¹⁸ That point is the instant, a constitutive and yet alien stillness at the heart of film. In this analysis, then, film is essentially associated with the present, not because every filmed event takes place in the present, but rather because the present lurks as an inevitable presence within filmed movement. Where Deleuze sees the present as always immediately linked to past and future, Bellour understands it as a kind of invisible barrier between them, a stubborn independence in the instant that prevents it from being assimilated. If the basic cinematic tendency as Deleuze understands it is toward the film of a single shot, the films that Bellour favors are broken by interruptions. The question for Bellour, then, is “what kinds of instants does the interruption of movement imply?”¹⁹

II

Questions like this show how basic explanations of the structure of film inevitably involve assumptions about the nature of the present. Nothing demonstrates this more vividly than the stubborn persistence of belief in the persistence of vision. Though serious theorists from Erwin Panofsky on have started with motion in their attempts to define the specific powers of film, it is still not entirely clear how the illusion of motion is derived from a series of still images, and the explanation most commonly offered is quite wrong.

For serious investigators from Plateau to Marey, the answer to this mystery was fairly simple. The familiar presence of afterimages left on the retina by a particularly bright light or a vivid color suggested that every image left its impress on the eye. Thus a series of still images could fuse into the illusion of continuous motion because each frame left an afterimage on the retina to be fused somehow with the one following.20 Marey successfully established this as the standard explanation of cinematic motion despite the fact that it violates common sense in several ways. There doesn’t seem to be any way for an afterimage, burned into the retina, to fuse with another, incoming image, so that an average of the two will be registered, a process that would seem to require the stimulation of parts of the retina by light that is not actually present. Nor does it seem logical that this process should continue for some steps, producing clear motion instead of a blurry mess. Nonetheless, persistence of vision became the conventional explanation, trotted out in the opening paragraphs of film histories from Ramsaye to the present day.

Even as Marey was establishing retinal persistence in its position of authority, though, the optical and psychological science of the day had displaced it. Münsterberg, professionally aware of the latest research as of 1916, rehearses the familiar stories about stroboscopic discs and afterimages only to debunk them. He cites Sigmund Exner’s experiments, which had also impressed William James, to the effect that perception of motion is a specialized process, separate from perception of shape or color.²¹ Exner’s experiments with still and moving lights seemed to make motion perception an automatic process, even perhaps a physiological one, but Münsterberg also cites the later research of Max Wertheimer, which offered a cognitive explanation. As Münsterberg puts it, “the continuity of the motion results from a complex mental process by which the various pictures are held together in the unity of a higher act.”²² Wertheimer is still often cited in contemporary studies, though “complex mental process” has generally been replaced by complex neurological processes.²³ But some basic conditions for understanding motion vision had been laid down by the time Münsterberg wrote: the perceptual process behind the movies is the same as that behind the perception of ordinary motion; both rely on sampling and summation that occur well after light hits the retina; motion is perceived in itself and not just as a feature of shape or color.24

These more complicated explanations have made little headway outside certain specialized sciences. Even Münsterberg, who went to great lengths in his book to give a scientifically respectable account, succumbed in the end to the theory he had debunked. When he comes to the “cutback,” or what we would now call the flashback, he concludes: “The whole technique of the rapid changes of scenes, which we have recognized as so characteristic of photoplay, involves at every end point elements of suggestion which, to a certain degree, link the separate scenes as the afterimages link the separate pictures.”²⁵ Afterimages persist in his account as they do in much film criticism of the present day, long after their very existence has been disproved. They retain their prestige, not just in undergraduate film surveys, but also in the most sophisticated studies of the visual arts. Rosalind Krauss, who has written a closely considered account of the optical unconscious, assured her readers in 2011 that “cinematic motion is based on the physiological fact of the ‘persistence of vision’ by means of which any visual stimulus induces a ghostly copy of itself (called afterimage) which remains on the retina, as though suspended before our eyes, masking the slippage from that stimulus to the next.”²⁶ Münsterberg’s slip suggests an explanation for such otherwise inexplicable ignorance: persistence of vision provides a vivid metaphor, visual itself, not just for the way in which film narrative links together its separate scenes, but also for the way in which ordinary temporal experience knits its moments into a continuity.²⁷

The notion of afterimages must have appealed in the first place because it so strongly resembles one of the most ancient metaphors for memory: the imprint. Plato established a tradition by comparing memories to impressions on a wax tablet.²⁸ A popular medieval metaphor for memorial vestigia was the footprint.²⁹ This habit of imagining memory as an inscription in some medium was considerably reinforced by the advent of recording technologies, some of which, especially the phonograph, seemed to work just as Plato’s wax tablet was supposed to work.³⁰ Thus it became increasingly easy to think of the appearance of successive recording media as a biological and psychological revelation, as the mind progressively externalized its methods of retention, revealing that it had been a sort of photographic plate or phonograph record all along.

If long-term memory worked in this way, by inscribing itself into some medium, it made sense to imagine that short-term memories would work in the same way, that everyday experience was given its continuity by the gradual and successive imposition of one sense impression on another. Thus the influence of Fechner’s Nachklange on psychologists such as James. For Fechner, every sense impression was a mini-memory, with an echo trailing after that connected it to later impressions. The tendency of impressions to subsist for an instant thus made ordinary experience possible. This much longer and deeper tradition subtends and reinforces the purely optical theory of afterimages, supporting it even as science obliterates the basis for both. Thus Giorgio Agamben argues, in a masterful essay on Aby Warburg, that “alongside the physiological Nachleben (the persistence of retinal images), there also exists a historical Nachleben of images.”31 The purely optical afterimage would have been fairly easy to displace if it had not been an instance of a much more complex idea about temporal experience in general.

This is one of the things that film theorists from Münsterberg to the present mean when they say things like “the photoplay obeys the laws of the mind rather than those of the outer world.”³² Münsterberg, of course, would not have included in his equation of film and mind any of those ideas like condensation, representation, and displacement derived from Freud and used by later theorists, nor could he have included any of the Lacanian concepts that go into theories of the gaze. But, underneath these, there is fairly common agreement with a principle enunciated by Annette Michelson in 1971: “The illusionism of the new, temporal art reflects and occasions reflection upon, the conditions of knowledge; it facilitates a critical focus upon the immediacy of experience in the flow of time.”³³ What this means is that the temporal relations that make up a film resemble those that make up ordinary experience to such an extent that they make that experience available in a way that is otherwise structurally impossible. At the heart of that experience is the relation between “immediacy,” or the present, and “the flow of time,” or the continuity of one present with the others around it.

Probably the most uncompromising version of this claim comes from Deleuze, who disdains the notion that film indirectly represents the flow of time. What he calls “direct cinema” achieves “a before and an after as they coexist with the image, as they are inseparable from the image. This is what direct cinema must mean, to the point where it is a component of all cinema: to achieve the direct presentation of time.”³⁴ Though the philosophical background of this idea in Bergson is far from phenomenology, Deleuze is very nearly paraphrasing Husserl here, especially in his insistence that the flow of time is presented directly to us and not through representations. For Deleuze, film accomplishes something that Husserl felt was fundamental to ordinary temporal experience: it constructs a present in which the past and the future are immediate and immediately available. The now, in other words, is “ever-present,” just as Griffith claimed. What happens, then, when Griffith’s narrative methods divide and multiply the now, interrupting one now with another, stalling or speeding up the flow of the immediately present? Is narrative, in fact, a basic violation of the fundamental ontology of film? Or does it display for us a different present, one that does not flow but persists in its singularity?

III

For Münsterberg, the structure of the photoplay depends on a kind of persistence of vision, working at progressively higher levels of organization. Separate scenes, as he puts it, are linked together “as the afterimages link the separate pictures.”35 Two decades later, Rudolf Arnheim made the same point in reverse when he described film as “a montage of single frames.” The famous montage effects associated with Eisenstein, he says, are achieved by carrying over the principle of persistence of vision to the “macroscopic” level.³⁶ In this analysis, the most fundamental unit of film criticism is the single frame, and the relationship of that frame to the frames around it is an analogy for all the other effects of film editing, not just montage as such. This apparently natural progression from frame to film motion to film continuity in general resembles the similar progression from present to temporal experience to history in general that is proposed by philosophers such as David Carr.³⁷ The similarity is probably not coincidental, given the general tendency to explain film on the analogy of human temporal experience. Barthes recognizes this when he describes the common belief that “the filmic world” and “the real world . . . continue to flow by in the same constitutive style.”³⁸

For Barthes himself, however, the photograph breaks this flow, since it is a present turned entirely to the past, without any visible relation to the future. The individual frame, then, remains stubbornly separate, and its stuttering relation to the frames around it might remind us of an essential discontinuity in ordinary temporal experience. Or, more famously, it might remind us of what Barthes calls “flat Death.”³⁹ This apparently commonsensical idea, that the photograph, because it is still and distinct within the flow of time, inevitably represents death, is common to many distinguished theorists of film, including André Bazin and Christian Metz.40 The most vivid pieces of evidence for this view are portraits like the famous photo of Lewis Payne that Barthes includes in Camera Lucida, and the association of photography with death in general would seem to depend on a reduction of the full range of photographic representation to the particular case of the portrait. Metz, for one, admits that “there are other kinds of photographs: landscapes, artistic compositions, etc.,” but this does not stop him from making the portrait exemplary.⁴¹ Still, it is not entirely clear why a photograph of a landscape or an eggbeater should necessarily remind us of death, except perhaps in the way that everything can potentially remind us of it. Perhaps the stillness of the photograph, lurking within the flashing frames of a film, represents a different problem, present within a life and not just coming at the end of it.

From at least the time of Bazin’s famous essay “The Ontology of the Photographic Image,” film has been understood as repairing some essential damage done to human experience by the stillness imposed through photography. No longer is the object “enshrouded in an instant,” as Bazin puts it, since film motion delivers it from its “convulsive catalepsy.”⁴² Behind this idea is a long history, especially prominent in France, hailing film as a transfiguration of the stillness of photography, a history that also extends forward to Metz, who considered film motion to be “a destruction of the photograph.”⁴³ The most thorough critique of this tradition is to be found in the work of Garrett Stewart, who does not dispute the distribution of stillness and motion across the relationship of photography to film, just the idea that film can ever entirely subsume the still.⁴⁴

Restoring motion to stillness as it does, film also rescues a real present from the past. Bazin offers a metaphor persuasive enough to have become quite popular: the photograph immures its subject “as the bodies of insects are preserved intact, out of the distant past, in amber.”⁴⁵ To make these insects move is to animate them, once again, in the present, so that Metz, among many others, thinks of film as always happening in the present tense.⁴⁶ Some fairly large part of the general notion that films take place in the present must rest on this prior association of the still photo with the past. Popular anecdotes about the Lumière brothers astounding their first audiences by starting a film with a still image reinforce the idea that film kicks the photograph out of its immobile past and starts it living in a real present.

Lately, however, this canonical separation of photographic images into the still and the mobile, the past and the present, has been questioned. Under the influence of video and digital imagery, critics have rediscovered a whole range of forms that are intermediate in some way: printed sequences of photographs; film stills and other uses of the photogram; films published as illustrated books.47 As Stewart and Bellour have shown how commonly the still photograph appears within films, partly as a reminder of the photogram that otherwise seems invisible, critics of photography have begun to insist that there can be a fair amount of movement in a “still” image. An early, influential essay in this respect is Peter Wollen’s “Fire and Ice,” which insists: “The fact that images may themselves appear as punctual, virtually without duration, does not mean that the situations that they represent lack any quality of duration or other qualities related to time.”⁴⁸ A still image, in other words, can represent what it cannot depict, a more modest version of Deleuze’s notion that still images within a film can directly depict times beyond their frame. In any case, it is much more common now to see photography and film as points along a continuum, rather than polar opposites, and to look, under the influence of Bellour, into the space between film and photogram, for degrees of motion or stillness.⁴⁹

Taken far enough, this rethinking can result in a complete reversal of the hallowed association of the photograph with the past, film with the present. As Olivier Lugon points out, it was common in Germany between the wars to advertise photomontage as a film in stills, and to think therefore of the individual photo as the concentration of movement rather than its antithesis.⁵⁰ Reanimated in this way, the photograph could also be reinserted into a mobile present, pulled out of an immobile past. This is part, perhaps, of Barthes’s idea, expressed in “The Third Meaning,” that a film still “throws off the constraint of filmic time.”⁵¹ What Barthes means is that a projected film enforces a certain time scheme onto its viewers, while the still frees them into a present that, in a sense, is wider in scope than the time of the film. Meditating on the still, the viewer abides in the present, while the film piles up in the past. In “Observations on the Long Take,” Pier Paolo Pasolini takes the same idea to the next higher level of film organization. According to Pasolini, film takes place in the present only insofar as it is limited to single takes: “The long take, the schematic and primordial element of cinema, is thus in the present tense.” Once these are multiplied, subordinated to one another in an organizational scheme, “the present becomes past.”⁵² An immersion in time, of the sort also admired by Barthes, becomes a perspective on it, and the pure present of the single image is sacrificed to an organization that brings back the whole notion of before and after and thus plunges the film into the past.

On one hand, then, the individual photogram is an inescapable reminder of the past lurking within the flowing present of the film. On the other hand, though, the single frame, lifted out of context, has the full amplitude of an expansive mental present, while the film enforces a death march through time on a helpless consciousness. How far this conversation has come from Griffith’s blithe assurance that film provides us with a limitless now. Perhaps the source of this disagreement and division is not to be found in the history of film criticism itself but rather in the terms it borrows from ordinary experience, as if we understood that. Appeals to the present or the past imply that these are known quantities, to which the structure of film can be matched and thus explained. Deleuze is one of the most powerful film theorists because his theory of the time image rests on a well-developed concept of the present and its relation to the time around it. But the Bergsonian account of time on which Deleuze depends is just one of several contending theories that have attempted to make sense of the present. Thus the contentious variety of ideas about the relation of still to film motion may derive from a more fundamental indecision about the experiential relation of one moment to the next.53

The present thus persists within film as a nagging problem, a constant reminder of our inability to understand what seems such a natural and fundamental aspect of experience, the progression from one instant to the next. In this sense, the individual film frame may represent death, as so many have insisted, but only insofar as the instant of death is a very drastic instance of a hiatus that is always possible within the apparent continuity of experience. With the invention of the photograph, Tom Gunning has argued, “the instant, previously only an abstract philosophical conception of an indivisible atom of time, now took on visible form.”⁵⁴ The still photograph thus presented in a form that became ubiquitous and inescapable the problem of the present. Sergey Tretyakov once called the photograph “an infinitely fine scale that has been scratched from the surface of reality with the tip of the finger.”⁵⁵ Bellour calls the photograph “a thin film of time in a pure state.”⁵⁶ Both are groping toward a description of a magnitude just this side of zero, so close to it, in fact, as to be identical, while still somehow remaining a magnitude. This is a conundrum that is only exacerbated in the structure of a film, where the individual frame may very well seem to mark, not the instant, but the difference between instants, as if time itself were to be found between the frames.⁵⁷ But this then raises the other nagging issue, that of continuity between instants, since moments whose own duration vanishes toward zero can hardly establish an ideal basis for continuity. Inability to explain film movement is just an instance of a larger inability to explain the perception of movement in general, which itself rests on a profound lack of understanding of the continuity of time. The stillness found inside the moving image is, in this sense, a stubborn persistence in the present, a persistence of it as an unexplained fact at the heart of human temporality.

IV

Even as it organizes still frames into larger and more complicated units, film preserves all sorts of reminders of a recalcitrant present. In its earliest and most undeveloped form, according to the influential formulation of Tom Gunning, film is nothing more than a sequence of present-tense instants. The film of pure visual display, what Eisenstein called the attraction, is limited, in Gunning’s words, “to the pure present tense of its appearance.”58 At this stage, a film is nothing more than a pile of instants, all at the same level of temporality. Anything like narrative development is discouraged in favor of a simple alternation: now you see it, now you don’t. Such has been Gunning’s influence that it is now a commonplace that the tension between attraction and narrative is fundamental to the history of film. Like most commonplaces, this has also attracted its share of critics, and even Gunning has not represented his dichotomy as an utterly exclusive situation.⁵⁹ So it is now generally agreed that the attraction represents one pole of a tension in all films between the present tense of pure display and the more complex temporality required by narrative.⁶⁰

Contemporary narrative films therefore often contain residual instances of the pure present tense of the attraction. One of the more obvious of these is the freeze-frame, which Bellour and Stewart treat as a reminder of the photographic stillness at the heart of film motion. But the freeze-frame also demonstrates something uncanny in the photograph itself, the way it holds the present, allowing us to contemplate it in a way that is impossible for ordinary consciousness. Freeze-frames force the viewer to study the photograph for a fixed period of time, so that the instant comes to have an amplitude it never has as actual time passes. What is seen on-screen then is the paradox of an instant that takes some time, as if every freeze-frame were an essay on the perennial question, how long is the present?

There are many such moments even in the most conventional narrative film, in which the camera comes to dwell on something purely for its visual interest. Roberto Rossellini referred to the effect of such moments as “denarrativization,” and he claimed that, for him, the narrative was just a means of getting from one moment of suspended contemplation to the next.⁶¹ Probably the oldest and most common of such denarrativizing devices is the close-up. In his famous essay on Griffith, Eisenstein focuses first on the close-up, and he marvels at its capacity to pack a whole set of ideas and reflections into a small space, to make the audience or reader see what otherwise would have to be explained. But this power of ocular concentration is precisely what makes the close-up a nonnarrative element, what made it such an unwelcome innovation, at least according to the creation myth as Griffith told it. He notes ironically that it was considered “anarchistic” because it broke up the obvious flow of action across a legible physical space.62 Later, the close-up came to be celebrated, almost fetishized, for precisely this reason, because it broke up the integrity of the scene and cut into the flow of the narrative.⁶³

This apparent opposition between the momentary stillness of the close-up and the flow of narrative rests on the assumption that the latter is achieved primarily by continuity. But it is one of the basic paradoxes of narrative organization that continuity is achieved by means of carefully modulated interruptions. In classic montage, the jump cut makes this interruption obvious. In doing so, it makes the viewer aware of an interval, a space between the images, which is technically atemporal and nonchronological.⁶⁴ In cases of maximum discontinuity, as in films like Dziga Vertov’s Man with a Movie Camera (1929), the continuous time in which the film viewer is ostensibly situated is completely replaced by this blank non-time, which is the matrix within which the individual shots are situated. Instead of flowing, then, the film stands still within a static time frame constituted by the atemporal jumps from shot to shot.

The commonplace cuts in an ordinary narrative film apparently work in a very different way, constructing a coherent temporal space that supports the shots, floating them in a medium that flows onward even when the camera eye is closed. But this coherent space is not an automatic effect of a sequence of cuts. Cutting, as Kristin Thompson explains in the standard account of classic American cinema, was both an opportunity and a problem: “Unless the filmmaker finds cues for conveying the spatio-temporal relationship between shots, the effect of the cut is a perceptible break between bits of subject matter.” Early filmmakers developed a set of rules and procedures the purpose of which was “taming and unifying” the otherwise anarchic effect of the cuts.⁶⁵

Perhaps the most venerable of these rules is the one laid down very simply by Rudolf Arnheim: “If two sequences of the same action are to be understood as occurring at the same time they may simply be shown one after the other, in which case, however, it must be obvious from the content that simultaneity is intended.”⁶⁶ Arnheim is careful to note that succession does not automatically imply simultaneity, but as the convention took hold, it became common enough that it came to seem automatic. Thus, according to a basic formula first proposed by Christian Metz, “alternating of images equals simultaneity of occurrences.”67 It is worth stopping for a moment, though, to note how odd this convention is, and what a kink it puts it the apparently automatic flow of film narration. To make the formula work, filmmakers must evoke in the minds of the viewers a discontinuous time zone, quite apart from the one visible on-screen, in which successive moments come to be simultaneous.

This odd, nearly atemporal present appears overtly in some of the earliest films in which Griffith established the norms of continuity editing. Griffith’s first extended sequence of parallel editing, according to Gunning, comes in The Fatal Hour (1908), a film whose title announces its dependence on the clock. In it, a detective is captured by the criminals she has been pursuing, and she is tied up in front of a pistol, which is linked to a large clock and set to fire at 12:00. Apprehended at about 11:40, the criminals graciously reveal their plan to the police, who set out, lickety-split, to stop the clock before it can strike the hour. Though the clock is in view for much of the film, it does not, in fact, mark real time, and the amount of time advanced on the clock from shot to shot does not correspond to the actual time taken up by the action in the parallel shots. It is fairly obvious, in any case, that to be at all useful on the set, where the filming would have taken several hours, the clock could not have been operational. So the notion promoted in Gunning’s account, that in this film “each shot finds its place in an irreversible temporal logic,” must already be qualified just a bit.⁶⁸ More fundamentally, the process of crosscutting between two simultaneous actions, that of the clock and the detective and that of the police racing to rescue her, sets up a complex temporal order. As Gunning puts it, “the order of shots no longer indicates a simple succession in time, but the staggered process of simultaneity.”⁶⁹ Though the clock on the wall tells the viewer that time marches on ineluctably, the cuts to the racing police in their carriage indicate a slip backward in time, to show something happening simultaneously elsewhere. This is an early and crude example of a technique that would later, Gunning says, be perfected, but it does show in a simple form what would later become a temporal convention in classic Hollywood cinema. As Metz’s formula puts it, two scenes succeeding one another would be understood to be happening simultaneously.

This is probably the most basic rule by which the disorderly set of instants delivered by the cinema of attractions is turned into a coherent time-scape. But the order it imposes is incomplete, for the nonsuccessive present of the attraction is not really hammered into orderly succession, one thing logically following and replacing another. Instead, simultaneity and succession are shown to coexist within a very uneasy portrayal of the present. In the time-line represented by the clock, the present is fleeting, each tick of it replaced by another in linear progression. In the simultaneity of montage, however, it is flexible in size and not limited to a purely linear progress. With the switch from one line of action to another, the same present, in fact, succeeds itself, and the testimony of the clock on the wall makes this twist in time all the more shocking. It is not so much that the pure present of the attraction remains as a residual presence within the narrative as that the incomplete imposition of narrative reveals some of the intrinsic peculiarities of the actual present of ordinary experience.

These are all the more obvious in the early Griffith films in which parallel editing has a logical or ideological purpose rather than a temporally dramatic one.70 In most of these cases, the ne plus ultra of which is of course Intolerance itself, shots of high and low life are alternated to create what Gunning calls a “moral dualism.”⁷¹ One of the best examples is A Corner in Wheat (1909), based on material from Frank Norris. In one series of shots, farmers struggle against the elements, and the urban poor line up for handouts, while in another a financial wizard exults and celebrates as he corners the wheat market. There are no dramatic connections between the two sequences, and yet it is clear that they are supposed to be causally related and in some large sense simultaneous. But what sort of simultaneity links these sequences, and in what sense can it correspond to what Griffith announced as the triumphant “now” of film? The farmers seem to live and work in a static time frame, while the financier’s machinations seem to take only seconds. In one way, it might seem as if these divergent time-lines are unified by reference to the present of the viewer, but that viewer is elevated, by the freedom granted through film editing, to a time and space not apprehensible by any actual human being. If there is a present here, it is the eternal present traditionally available only to God.

At one point, in fact, time stops still as a breadline of haggard beggars freezes into a kind of tableau vivant. The implication is that those in the breadline are immobilized by their poverty but also that their situation is timeless and eternal.⁷² But the frozen tableau, which looks like but is not a freeze-frame, is also a reminder of the photogram that is the static base of even the most dynamic film. In another way, though, the strange immobility of the breadline seems to be an inadvertent revelation of the odd syncopation of screen time that often marks parallel editing, especially the situation of the off-screen character, the Lonedale operator, for example, who seems to wait in exactly the same pose for some seconds, as the train approaching to save her eats up the miles. In this case, it looks as if the camera filming A Corner in Wheat has switched back to the second storyline too soon and caught the protagonists, waiting patiently as the primary storyline develops. Of course, there is no reason for these characters to wait, for the relation between them and the financier is not a temporal one in the first place. So what appears on-screen is actually the static present within which logical comparisons are made, the odd, invisible simultaneity at the heart of parallel editing, the necessary moment in which two successive things are made simultaneous so that we can tell they are successive.73 This is the present, not by any means an ordinary now, in which Intolerance takes place.

V

For many years, Intolerance was considered to be not just Griffith’s masterwork but also the pinnacle of American filmmaking. It has also been compared with some frequency to those hugely ambitious, historically encyclopedic works in literature: The Waste Land, The Cantos, Ulysses.⁷⁴ Unfortunately, many such comparisons tend to find these works similar in their failure, especially their failure to cohere. The most extensive of these is Miriam Hansen’s, which begins with a sharp judgment of the film as “a gigantic ruin of modernity.”⁷⁵ She thinks of Intolerance in the way literary scholars frequently think of an unfinishable project like The Cantos, as a work ruined by its commitment to modernity. To some extent, though, the commitment to modernity is also a commitment to the present in formal terms, to local, immediate, intensive effects rather than articulated narrative or logical design. The relation of the present to the past thus becomes a structural problem.

At its base, Intolerance is actually a good deal simpler than The Cantos or The Waste Land. There are four distinct storylines, taken to represent ancient, sacred, medieval, and modern times. These focus on four events: the fall of Babylon in 539 BCE; the crucifixion of Christ; the massacre of the Huguenots on Saint Bartholomew’s Day, 1572; and a modern story, datable by car license plates as occurring not later than 1914.⁷⁶ In very general terms, the four stories are meant to be the same story: the eternal conflict Griffith stages between love and intolerance, a concept that is very largely and loosely construed throughout the film. This general ideological matrix is supposed to hold the film together as Griffith intercuts the four time-lines with one another: the betrayal of Babylon to the Persians by jealous, sectarian priests; the crucifixion of the Man of Peace, which Griffith had wanted to ascribe to some narrow-minded Pharisees until he was forcibly reminded of the crucial role of the Romans; the religious animosity between Catholic France and the Huguenots; and a rather tortured and strained sequence of events in the modern period, in which an industrial magnate spends so much money on the charities of his spinster sister that he must cut wages at his factories, throwing people out of work and into poverty, where one of them falls in with bad companions, is falsely blamed for the death of their chief, and is almost hanged. Just how the last few stages in this drama are supposed to be caused by the sexual frustration of the spinster sister is one of the greater logical mysteries of the film.77

The stories, then, are meant to be typical examples, moments chosen from a continuing conflict between elemental forces, and thus there are a number of parallels between them. But there are also differences. In one way, the overall story seems to be one of decline, since the Babylon of Belshazzar is portrayed as a model of tolerant authoritarianism, in which the people are left to their innocent pleasures, while the modern period is ruled by puritanical capitalists, who begrudge their underlings even a moment’s enjoyment. On the other hand, though, only the modern story ends happily, since the innocent worker is not hanged, and this seems to suggest that a modern justice system is worth something after all. As Hansen suggests, parallelism as a technique tends to slide between comparison and contrast, and even comparisons can be tricky when they are made across time.⁷⁸ Are the similarities between Leopold Bloom and Odysseus meant to show how far humanity has fallen since classical times or that classical times weren’t necessarily as grand as they have been said to be?

In any case, parallelism between the four storylines has a tendency not to reinforce but rather to undermine narrative coherence. At the most basic level, as Hansen puts it, “Intolerance, like few other American films, exemplifies the truism that linking requires cutting: for every connection between periods, one narrative is necessarily disrupted at the expense of another.”⁷⁹ Even when actual cutting does not occur, the implicit presence of three other time-lines has the tendency to call into question the solidity of the one currently on-screen. As Hansen explains it, this “transhistorical, temporal omnipresence” has a tendency to weaken the “spatial coherence and closure” on which the fiction effect depends.⁸⁰ But this might also be put in purely temporal terms. Having four different historical presents alternate on-screen vastly magnifies the disorienting effect of ordinary parallel editing, for there is no time to which all four might be referred, within which they might be composed. This has an especially attenuating effect on the modern story, which should seem to be happening more or less now, in a time the ultimate conclusion of which has still to be determined, but which ends up in the temporal distance with the Persians and the Huguenots, whose fates were settled long ago.

Even within the individual stories, Griffith’s rather casual attitude toward the norms of continuity editing and his devotion to décor and spectacle mean that narrative continuity is often overwhelmed by detail. Many scenes, especially in the Babylonian sequences, are so crowded it is hard to know where to look. Close-ups have a distressing tendency to seem like insertions into rather than selections from the larger canvas. Joyce Jeseniowski picks out the camel that ambles through the opening shot of the Judean sequence, whose qualifications for particular emphasis are not immediately clear.81 To take another example, the toothless old man vignetted into the marriage market scene in the Babylonian story has little spatial relationship to the shots he interrupts, and thus he seems the tool of a timeless comment about male appreciation of female beauty and not a dramatic elaboration. In short, Griffith’s devotion to immediate visual impact and to ideological comment tend to set up a static undertow that impedes the development of a coherent narrative line.

The thrilling conclusion, the final chase scene that should tie all these loose narrative threads into a single satisfying bundle, actually leaves them even more obviously frayed. The chase, at this point in Griffith’s career, is obligatory, and this one must have seemed an opportunity to raise the ante, to have four chases in one. But there are problems with some of the older narratives that make this unfeasible. Surely it would be indecorous, even sacrilegious, to figure Christ’s passion as a race to the cross, so the Judean story essentially disappears from the last two reels of the film, except for one shot of the Crucifixion that punctuates the tension of the modern execution scene. The Huguenot story becomes a variation on the pattern established with The Lonely Villa, but without vehicles, as Prosper works his way across Paris in the vain hope of saving Brown Eyes from the slaughter. So the pure chase scene is concentrated on the Babylonian narrative, in which the plucky Mountain Girl speeds her stolen chariot toward Babylon to warn Belshazzar of the invasion, while a small group in the modern story tries to catch the Governor’s train, to secure a stay of execution for the falsely convicted Boy.

The special oddity of this chase scene is that in some weird way the two stories seem to be racing one another. The message of the editing that created such scenes in the past is that two parallel, successive shots are, in fact, happening simultaneously. This was always a seriously strained fiction, but when the two time-lines are 2,500 years apart, the simultaneity acquires a fantastic aspect. Of course, the viewer knows quite well that the stories are separate, but the speeding chariot of the Mountain Girl and the speeding motor car can easily seem to be speeding to the same place, in the same time, as they are apparently speeding to the same purpose. This misimpression may be underscored for the few viewers who notice the tire tracks left by the camera car in the path of the Mountain Girl. The Persian army is also racing toward Babylon on a back road heavily traveled by motorcars. For these viewers, the sense that the two stories are being superimposed on one another may become actual and not just notional. But the comparison may well remind the viewers that, in a real sense, the modern story is in the past as well, and that whatever passes for the present on the screen is, in fact, something done and finished some time ago.

There is a very real tension, then, between the linear narrative of the chase as such and the implications implanted by the parallel editing. On one hand, there is a continuously unrolling present, kept open to the future by the devices of suspense, while on the other hand, there is the vague, amorphous, and curiously static present implied by the parallels. Even the narrative by itself evokes a decidedly nonnarrative pleasure, the pure thrill of the racing elements, which pound away, scene after scene, exciting even when they make no discernible progress. But the grand narrative of Intolerance encloses this relatively simple nonnarrative experience, the pleasure of the image as such, within another, more complex experience, located in a different kind of present. This present, external to the four narratives and independent even of the time of viewing, is the focus of a number of specific elements within Intolerance.

The inter-titles, for example, though they were composed by the very funny Anita Loos, adopt an “omniscient scriptural voice” that speaks from a position above and beyond all the narratives.82 The narrator editorializes throughout, making it clear that 16^th-century France is a “hotbed of intolerance” and that Monsieur La France is “effeminate.”⁸³ The narrator also quotes, sometimes with and sometimes without actual quotation marks. And there seems to be another level even beyond the narrator, a level that is the source of numerous notes appended to the inter-titles themselves. Some of these are little self-advertisements, like the oddly distancing news that the Babylonian set is a “replica of Babylon’s encircling walls, 300 feet in height and broad enough for the passing of chariots.”⁸⁴ Some are protectively informational, like the news that the harridans grouped around the edges of the Babylonian marriage market are “women corresponding to our street outcasts, for life, the wards of Church and State.”⁸⁵ And some are frankly tendentious, such as the note claiming that the code of Hammurabi protects the weak from the strong.86 In any case, these impersonal addresses from a position beyond even the narration of the inter-titles establish a time frame that is as external to the baffled viewers as it is to the characters in the film.

A more visual element with the same effect is the transitional device that Griffith adapted from Whitman. “Out of the cradle endlessly rocking” is a sort of motto for the film as a whole, a common inter-title, and a visual vignette marking transitions from one time to another. The vignette is composed of a long-haired girl, played by Lillian Gish, seated at a huge cradle, with three ghostly women in the background, representing the Fates. It puts birth in the foreground but also matches it with the death dealt out by the Fates, so that there is a full circle of life in one self-contained picture. The endless rocking of the cradle, as well as the repeated appearance of the vignette without any apparent variation, suggest a circular, repetitious time scheme, and there is nothing in the scene itself to place it in a particular time, though the whole thing has a vaguely archaic look. On its first appearance, the cradle scene is accompanied by an additional inter-title that reads: “Today as yesterday, endlessly rocking, ever bringing the same joys and sorrows.”⁸⁷

It is fairly obvious, then, that the cradle scene is Griffith’s attempt to provide a temporal matrix for the four storylines of his film, to link together what otherwise would look too various. But, as Hansen says in her long, detailed analysis of the cradle motif, what is supposed to link has always had the tendency to divide, so that the cradle has always stood out as just the sort of old-fashioned, nonnarrative element that prevents the film from cohering.⁸⁸ This also means, though, that the cradle best represents the peculiar sort of time in which Intolerance takes place. William Drew’s notion that it represents “the eternal present” makes the paradox clear, for the cradle scene never advances or changes, and thus it occupies the same moment of time for all time.⁸⁹ If the contiguity of birth and death is meant to suggest that both take place in an instant, this is still an instant that takes forever, or at least the length of the film. For the eternal present of the cradle scene is also the most appropriate time frame for Intolerance as a whole, which should take place in a present distended to the size of all history. The ultimate frame in which the elements of Intolerance are supposed to be contained is the instantaneous one of logical comparison, a present that is inherently nonnarrative, though it pretends to contain all of time.

Griffith’s gamble in Intolerance is that film can transform the ordinary, immediate present of human experience, with its problematic relationship to past and future, into the eternal present traditionally afforded only to God. This is what Ramsaye claims it accomplishes when it satisfies “the human wish to live in the Past, Present and Future all at once.”90 The phrase “all at once” makes the paradox clear, for there is finally no difference between expanding time to the size of eternity and collapsing it into the present moment. In either case, everything happens all at once. This is, more or less, the conclusion of Deleuze’s analysis of Intolerance: “Time as interval is the accelerated variable present, and time as a whole is the spiral open at both ends, the immensity of past and future. Infinitely dilated, the present would become the whole itself; infinitely contracted, the whole would happen in the interval.”⁹¹ What Deleuze calls “the continually diminishing interval between two movements or two actions,” the infinitesimal present of the cut itself, also tends inevitably to encompass the full immensity of time.

In somewhat more practical terms, this devotion to the eternal present means that there is no very satisfactory way to end the film. The chase requires a conclusion, and the Boy is finally saved from hanging, but the cradle goes on rocking nonetheless. If the back-and-forth tug of war between love and intolerance is, in fact, timeless, “the same today as yesterday,” as the sign proclaims when the workers in the modern sequence are shot down by the police, then the salvation of the Boy is just an episode. The eternal present established above and beyond the chase scenes can be concluded only by apocalypse or, as it happens, by double exposure. Intolerance switches for its last few moments to an apocalyptic battle scene, intercut with shots of crowds behind prison walls. These are dissolved by a double exposure, so the prisoners can run through walls suddenly become transparent, and then angels are superimposed over the battle scene. Finally, a cross comes to be superimposed over the entire scene. As a technique, the superimpositions suggest a resolution of the back-and-forth cradle rocking of the parallel editing, which can finally come to an end when two things can actually be seen at once. Only with this end is it possible literally to see what is implied throughout the film, that everything in it should ideally be visible in one instant of revelation.

For Griffith, then, there is a present within film, retarding its attempts to mount a narrative, and another present outside it, containing all the narrative lines and allowing them to be compared. The trick is somehow to seize the local present so fiercely that it reveals itself as that larger present, as if the eternal time outside the film had penetrated into it as a kind of epiphany. This is apparently the sort of thing that Deleuze sees in Intolerance, but which is achieved much more successfully in the later films featured in The Time-Image. If Griffith’s own time images, his random shots of camels and grinning beggars, were more compelling, then presumably he would have achieved the kind of thing Tarkovsky aims at, when he claims that each single frame contains the whole time of the film.

VI

Almost a hundred years after Intolerance, another Hollywood blockbuster combined four different plotlines in a simultaneous race against time. For David Bordwell, the similarities between this film and its distant precursor are strong enough for him to call Christopher Nolan’s Inception “something like the Intolerance of the twenty-first century.”92 Inception’s plotlines are situated in a descending set of dreamscapes, not in widely separated historical ages, but it depends just as heavily as Intolerance, according to Bordwell, on old-fashioned commonplaces of crosscutting. The principle that may still have seemed a little new in 1916, that parallel lines of action crosscut with one another should be understood to be progressing simultaneously, structures the final act of Inception as it does that of Intolerance. However, both films also expose the artificiality of this apparent simultaneity, Griffith’s by stretching it across 2,500 years of history, Nolan’s by inserting each subjective dream time within another. The same question is raised by both films: in relation to what time can the four different plotlines be considered simultaneous? For Bordwell, the answer is the same in both cases. Inception, he claims, brings us back “once more, to that sort of God’s eye view of an eternal present created in Intolerance.”⁹³ What made some sense for a conservative believer in 1916, though, can hardly have the same relevance for an apparently secular filmmaker in the 21^st century. Perhaps that century has its own ways of moving from the limited present always visible on-screen to the eternal present situated somewhere beyond it.

The published screenplay of Inception constantly illustrates the complexity of these issues as it tries to guide the reader up and down the layers of time included in the film. At one point, in the course of a long retrospective monologue by Dominick Cobb (Leonardo DiCaprio), the script helpfully notes that we are “back in the present.”⁹⁴ But this dramatically simplifies the actual situation, which involves a voice-over from one time, describing on-screen events from another time, which depend for their exposition on long, interpolated scenes from a more distant past, all of which is inserted into the lowest of four levels of nested dreams. Even in the simplest possible version of this arrangement, the “present” to which the script returns us is actually a dream in which time is moving several magnitudes more slowly than it does four levels up, where the ostensibly real present awaits our eventual return. In a more conventional film, with ordinary flashbacks, the present remains a constant from which the deviations can be seen to depart, but in a situation as complex as this one, where what counts as the “actual” present is off-screen for more than half the film and the deviations from it have deviations of their own with deviations of their own, the present loses its temporal authority and its capacity as a touchstone.

Nolan has made a particular specialty of negotiating through such complex temporal structures, from his first film, Following (1998), to his most recent, Interstellar (2014). But he is also considered a leading exponent of a new genre, the atemporal or anachronic film, which itself might be considered a subset of the more general category of puzzle films.95 Suffice it to say that Inception stood out less prominently from the general run of contemporary films than did Intolerance. This general turn to more complex temporal structures has been traced to a number of causes, the most obvious of which is the advent of recording technologies beginning with VHS tape.⁹⁶ Where a film critic such as Cavell had once to rely on a fairly faulty memory, now his successors could watch and re-watch at will, stopping the action whenever necessary. As Nolan puts it, tape allowed the viewer “to control the time-line.”⁹⁷ It is hardly a surprise, then, that films came to be constructed as if they had already been rewound, stopped, and replayed. Such films are hardly atemporal, since they contain more time, or at least more times, than their more conventional counterparts. What they lack is a secure present, an absolute moment in screen time against which the others might be calibrated. In films like Following or Memento, everything appears on-screen as if it had already been watched at least once, so that even the present tense of the viewer is made to seem belated and out of sync. In other words, the celebrated cinematic time machine has been inserted into another time machine, with different powers and different rules, and the result is a new take on the nature of the present.

In Nolan’s first two films, the formal disorder of the time-line is linked thematically to some pathology in the main character. This is most overt in Memento, where Leonard Shelby (Guy Pearce) has apparently lost his ability to form new memories, though he retains quite vivid recollections, true or not, of the rape and murder of his wife. Without the ability to remember what he has done, Leonard is prone to repetition. Appropriately enough, the action he repeats most is telling others that he cannot remember. More seriously, he also repeats the action that has become the goal of his life: wreaking revenge on the killer of his wife. For all we know, Leonard may have done this several times, with different suspects, by the time the film begins with the murder of Teddy (Joe Pantoliano), which is, in fact, its last scene chronologically. Since he cannot remember and the film is restricted to his point of view, we have no idea.98

In narrative terms, Leonard lives a purely anecdotal life.⁹⁹ But his temporal stasis has a peculiarly restless quality because he is driven by a goal, in the future, that he can never know himself to have attained. As Leonard puts it, he is stuck in the present: “I want time to pass, but it won’t. How can I heal if I can’t feel time?”¹⁰⁰ It is common within the film and in the criticism of it to consider this condition fatal to Leonard’s humanity, since the formation of basic emotions, such as fear or anger, seems to depend on temporal relations that he cannot construct.¹⁰¹ But Leonard does actually feel these emotions, particularly anger, and he is far more aggressively goal-oriented than anyone else in the film. The problem for him is that he can never actually discharge the future, and so his present remains perpetually open, as if with a wound that cannot heal.

Viewers and critics generally consider this to be an abnormal condition, but Nolan has identified his film so closely with its main character as to make it difficult to judge. Memento alternates between black-and-white scenes that tell the fairly limited story of Leonard getting ready to leave his motel room and color scenes, moving steadily backward in time, that tell the story, in reverse order, of how he came to murder Teddy. Presented in reverse order, the color scenes put the viewers in the same position as Leonard, as each scene erases the memory of the one before. Though the black-and-white scenes move forward in what would seem to be normal time, they all occur before any of the color scenes, and so they do not establish a norm from which the color scenes could be seen to deviate, which would be the usual effect of crosscutting. The usual format for crosscutting of this kind would be an alternation between present and past, such that the past sequences could explain those in the present and the present could serve as an anchor in time, so that the past can be seen as past. Nolan’s very different method, Bordwell maintains, sacrifices “the omniscience that usually comes with crosscutting.”¹⁰² In other words, there is no time-line in Memento within which the two different sets of scenes could be included and so compared. Leonard’s inconclusive present is all there is.

Nolan has also packed Memento with reminders that this sort of wounded present is basic to film itself. The most obvious of these are the Polaroids that Leonard tries to use to document his activities. The general failure of these bits of visual evidence may be indicated in the first scene of the film, the only one to be shot in reverse, which shows a Polaroid being sucked back into the camera. A mnemonic method that erases rather than preserves evidence is a good visual metaphor for the film to follow. The Polaroids are also generally considered to be a good metaphor for Leonard’s condition, a life lived as a series of disconnected snapshots.103 But what is a film, if not a series of disconnected snapshots? The Polaroids, like the still photos studied by Garrett Stewart, are reminders of the photogram, made invisible within the smooth course of filmed movement. But the photogram itself is just the smallest in a whole series of filmic discontinuities, including the cuts between scenes that Nolan makes more painfully obvious by presenting the scenes out of order.

This, finally, is the most interesting discontinuity in Memento itself, since Nolan has taken great pains to extract and exaggerate those aspects of film that most painfully represent his character’s pathology. What reads as mental illness at the level of content, however, appears at the level of form as aesthetic achievement. Nolan’s great distinction as a director is founded on his insistence that conventional narrative chronology is dispensable in film, as it has been for many decades in literature.¹⁰⁴ The same disconnected, anecdotal quality that makes Leonard’s life a hell on earth is taken as an index of freedom for the filmmaker. Of course, Leonard is a sort of filmmaker himself, with his storyboards on Polaroids, his casting calls and endless rehearsals, his campaign to get the script executed and recorded once and for all. In his obsession, he reaches a kind of stillness, an incessant interruption, at the very heart of film. Is this what Nolan is aiming at in structuring Memento as he does, or is there another reason for presenting his film as if it were a still photo disappearing before our eyes?

At the very least, Nolan turns out to be consistently interested in Leonard’s situation, which he reproduces with Cobb in Inception. He, too, is obsessed with a dead wife, trapped by guilt about the past in a present that cannot advance. In literal terms, Cobb cannot return to his former life in the United States because he is suspected of having caused his wife’s death. His situation in this respect is represented by a repeated scene of his children, locked in a tiny loop of action, with their faces turned away. This loop is, in a sense, what remains of the true movie of Cobb’s life, interrupted by his wife’s death and prevented from progressing by his unresolved legal situation. Cobb’s only hope is to use his professional skills as a kind of hacker, invented by Nolan for this film, whose specialty is breaking into other people’s dreams. Since this is exactly the kind of thing that implicated him in his wife’s suicide in the first place, Cobb is doubling down on the most dangerous bet of his life.

On one level, Inception is a version of the “one last job” variety of heist film. In return for a resolution of his legal situation, Cobb will use his professional skills to implant a potentially self-destructive idea deep within the dreams of the heir to a great industrial empire. The only way to do this, as it turns out in Nolan’s very intricate plot, is to nest this idea three levels deep in the subject’s dreaming mind, not in a dream, where it would be detected instantly, but rather in a dream within a dream within a dream. Nolan’s version of the dream world turns out to be quite rational in at least one respect, that time within the descending dream levels runs in precise ratios, slower and slower the deeper you go. Cobb’s plot thus requires very careful synchronization of three and, as it turns out, four time-lines running at different rates. Abstracted a bit from all this folderol about dream levels, Cobb’s task is fairly clear. In order to earn his release from the time loop he is caught in, to get out of its own subjective present, he must reestablish an objective time within which the four dream levels could be synchronized. If this synchronization fails, then Cobb himself and all his associates will be marooned within the dream, perhaps forever.

In general terms, then, the stakes could hardly be more obvious, for Cobb and for the film itself. The subjective time of dreams, as Nolan calculates it, is always slower than waking time, exactly twenty times slower, in fact.105 At the very bottom, lower than Cobb hopes to penetrate, lies Limbo, where time hardly advances at all. To save themselves from being pitched into this endless present, Cobb and his team must carry out a precisely timed plan on three dream levels simultaneously, though time runs at drastically different rates on the three different levels. On one hand, then, there is the chaotic, endless present of pure subjectivity, and on the other, the supremely objective present within which three subjective times can become one. Nolan’s gamble as a filmmaker is roughly parallel to this one, since he is asking his audience to construct and maintain a temporal point of view that hardly exists at all in the film itself. This is much trickier than ordinary crosscutting, where the different scenes can be assumed to progress at the same rate, though it is roughly similar to what Griffith does in Intolerance, where the plotlines are so far from one another in historical time.¹⁰⁶ In both cases, the same question arises, what sort of present could accommodate such drastically different temporalities? Bordwell suggests that there must be an omniscient level of narration in order to convey the dramatically different time-lines of the different levels.¹⁰⁷ On what is that omniscience to be founded? What is the opposite of Limbo?

These abstract difficulties appear as concrete effects in the film itself. The conceit, at this point, is that in order to extract themselves from the dream—to wake up, that is—the team must synchronize a set of “jumps” or shocks, on the model of the odd start that sometimes awakens a sleeper just on the edge of consciousness. And yet how are four different times, running at different rates, to be synchronized? What would synchronization mean in such a case? What is more or less punctual at level one would be twenty times longer at the next and so on. How could an alarm sounded at level one be synchronized with its counterparts twenty or four hundred times longer at the other levels? Nolan more or less advertises this problem by making the dream levels selectively permeable, especially to sound. Thus an Edith Piaf tune played at the top level sounds like some kind of deep sub-bass on the next level down and like distant wind on the next below that.108 If, as it seems, the first note of the song takes a good twenty minutes to register at the deepest level, how is the timing to work? Any short action performed at the last level will be of infinitesimal duration at the top, making the timing all the more crucial.

The characters’ necessary concentration on timing, along with their very frequent references to clock-time, make it all the more obvious that screen time does not equal elapsed time within any level. At one point, Cobb calculates that the team at his level has another sixty minutes to complete their task, which may be a bit of a shock to the audience. Even his recalculated time of twenty minutes turns out to be much more than the screen time left for this scene.¹⁰⁹ Cobb cannot be making this calculation using his watch because watches in the film seem to keep something like “real time,” so that they can be seen to move with unnatural slowness at the lower levels. Of course, if this were carried out with real consistency, watches at the lowest level would move so slowly that their hands could not be used to calculate at all, for they would tend to look completely still.¹¹⁰ Selective use of slow motion raises some of the same questions about screen time. It is fairly clear that action at the highest level must be slowed down somehow so that there will be enough time to show the necessary action at the lower levels. Nolan moves to slow motion at one point, virtually suspending the action at the top level, but he mostly uses selective cutting to eke out the second level, until close to the end when it too shifts to slow motion. This means that the first level must go to “extreme slow motion” to stay more or less in sync. But then he also uses slow motion at the lowest level, apparently just for dramatic emphasis.¹¹¹

In other words, there is no time, most certainly not screen time, within which the dream levels might be compared and thus measured against one another. Of course, Cobb does ostensibly complete his task, even after a series of fantastic complications, and he is reunited with his children in something more or less like real time. But the end of the film, which cuts away from the spinning top that is supposed to tell Cobb whether he is in a dream or not, puts even this in question. This cut to black is, according to Nolan, the final kick, imposed from “outside the film,” as if there were another temporal level yet.112 This seems to admit that everything within the film is finally relative, even what passes for experience outside the dream. And though this admission might occasion all sorts of speculations about the essential subjectivity of human experience, what it makes clear in relation to film is that the ordinary mechanisms of screen editing have already qualified out of existence the kind of absolute simultaneity on which Cobb’s plan depends. As was obvious from the first in simple examples like The Fatal Hour, the temporal matrix within which shots are composed is infinitely elastic. The clock may stop when the camera looks away, or it may race ahead. The present tense of film, the tense installed deep within it by the gaps between the frames, which appears again in the cuts and shifts of editing, is finally indistinguishable from Limbo.

The gravity of this problem is evinced by the cosmic methods necessary to resolve it in Interstellar. In several ways, this film represents a massive inflation of the basic situation of Nolan’s earlier works. In this case, the Earth itself suffers from the pathological condition that afflicted Leonard Shelby and Dominick Cobb. Infected with a rather vague blight, the Earth can no longer support the human population, which then languishes, without a future, in a reduced version of the recent past. Stylistically, the first act of the film also lives in the past, as Nolan consciously evokes the Dust Bowl and the films and photographs that immortalized it. The only way out of this terminal condition of temporal stasis is to leave the Earth itself, an apparent impossibility that has been made marginally possible by the convenient appearance of a wormhole out past Saturn. The rest of the film then becomes a conventional race against time, a structure old at the time of Intolerance, but in this case the race really is against time and not just against the clock. The distances from this planet to any other likely to support life are so vast that in order to make the trip feasible, time itself must be defeated, first by means of the wormhole and then by the agency of a massive black hole on the far side of it.

These cosmic structures also set up a narrative situation reminiscent of Nolan’s earlier films, especially Inception. In this case, the massive gravitational pull of the black hole asserts a drag against time, so that every hour spent exploring a planet in its vicinity might equal seven years back on Earth.¹¹³ Thus the astronauts, on their expedition beyond the wormhole, do not age perceptibly at all, while their relatives back on Earth grow up, grow old, and, in some cases, die. On-screen, of course, everything runs at the usual pace, and the storyline back on Earth has to be severely edited and truncated so that it can stay more or less in sync with the much slower time-line beyond the stars. Crosscutting in this situation is like hitting a moving target, the difficulties of which are advertised whenever the two time-lines communicate. In one such scene, Murphy Cooper (Jessica Chastain), now a grown woman, sends a message into the wormhole for her father, who has hardly aged at all, though he left Earth when she was a little girl. Her message continues in voice-over as the scene cuts from her lab to the spacecraft Endurance, now empty, as its crew drops to the planet below.114 A single sentence thus extends over three different time-lines, moving at three dramatically different rates. The implication is that Murph is speaking as the crew leaves the Endurance, and some of the poignancy of the scene depends on this dramatic convergence of events. But this is, of course, a mere effect, covering the practical reality that Murph’s entire message, in its original form on Earth, would only have taken the merest millisecond on board the spacecraft.

In what sort of space are we, the audience, situated, such that we can experience these convergences between subplots that are not even time-lines any longer but completely different times? Nolan’s ambition in this film is to answer this question, posed but not resolved in his earlier work. The answer lies on the other side of the black hole into which Cooper plunges when the expedition seems to have failed. What he finds there is the modern version of God’s eternal present, an extradimensional space within which all times are simultaneous and equally available. A certain amount of explanation has already grown up around this device, which is more or less an expansion of a tesseract, or four-dimensional cube.¹¹⁵ For better or worse, little of this is available in the film itself, where it appears that Cooper has fallen into a cross between the New York Public Library and a laundry chute. Suffice it to say that he has found himself in a kind of space that affords access to any time-line, including the ones we have already seen earlier in the film. In fact, we now re-watch these scenes from the other side of the screen, as it were, and realize, along with Murph herself, that the mysterious gravitational effects that made her childhood bedroom seem so uncanny were, in fact, messages from her father, communicating via gravity, which strings a line of communication between different times and spaces.

There is no God in this film, but there are godlike beings, who are apparently responsible for the wormhole and for vaulting Cooper across dimensions and back into his own past. But these gods turn out to be us, evolved beyond the confines of time and space.116 Cooper’s expedition through the wormhole is apparently the first step in this process, one result of which is the placing of the wormhole itself. This sort of logical nonsense has an accepted place in time-travel fiction of the 21^st century, where it is sometimes assumed that the relativity of time is such that it could loop back on itself so that an effect could become its own cause.¹¹⁷ Here, though, it may seem a little desperate, as if the awful recursiveness of human influence on our planet could somehow turn out to be not a trap at all but rather an escape. In any case, this cosmic loop back through time is a vast, massively redemptive version of the time loops suffered by Nolan’s earlier characters. The vicious circle of lives like Leonard’s or Cobb’s now becomes a virtuous one, and the recurrent present in which they were trapped reappears as a quasi-scientific version of eternity.

In short, relativity plays in Interstellar the role that God plays in Intolerance. The very fact that so worried Einstein, that in relativistic terms there is no Now and therefore no one standard by which to synchronize different times, turns out to be the escape hatch for a humanity trapped in a dead end of its own making. The positive turn that Nolan gives this fact is exemplified by the use of watches as communicators. Initially, Cooper gives his daughter a watch to match his own, as if this were a symbol of faithfulness, though she immediately realizes that the two watches mean exactly the opposite, since they cannot remain in sync once her father leaves Earth.¹¹⁸ This classic exposition of the facts of relativity is then reprised at the end of the film, when Cooper uses gravity to manipulate the second hand of Murph’s watch, sending her data from a distant future that will allow her to achieve that future.¹¹⁹ From the point of view offered by the tesseract, the watches are in sync, not because they keep the same time but rather because time has ceased to matter in this extradimensional space.

Nolan thus extracts a double happy ending, for his characters and for humanity at large, from what might otherwise seem the rather dire notion that time is not a constant. Though it might be said that the only way he escapes the unhappy ending that seems so certain at the beginning of the film is to avoid ending altogether, by turning the human story into a loop. Avoiding the apocalyptic, as it turns out, was one of Nolan’s basic ambitions for Interstellar. In an interview published with the screenplay of the film, he congratulates his brother for transcending “the negative aspects of the continuing human story.” Such is our pessimism right now about the human condition that there is always “this apocalyptic mode to storytelling.”¹²⁰ Though Nolan must mean this in fairly practical terms, his words also have some bearing on the formal shape of his film. To avoid the apocalyptic, in this sense, is to avoid ending at all, which can only be achieved if time can somehow be made to stop. For all his storytelling verve, then, Nolan is constitutionally biased against the demands of narrative, particularly its inevitable allegiance to the end. The cure for a pathological isolation in the present, for an individual or a planet, is not narrative but rather a larger, more perfect present, preferably an eternal one. God’s eternal present, the expedient that Griffith uses to wrap up his multi-story film, has this disadvantage, that it appears on Earth in the form of apocalypse. The eternal present that Nolan adapts from relativistic physics has this fundamental advantage, that it does not require the end of anything.