Peer Reviewed

Exclusive to M/m Print Plus

Story, Discourse, Dunkirk

It will surprise no one to see wartime treated as an especially narrative problem. Indeed, given the long and apparently necessary relation between war and narrative, a relation that goes back at least to the Iliad and the in medias rage of Achilles, it is probably harder to think of them apart, harder to resist the urge to see both old and new wars in the ready and comfortable terms of already available narrative models: war as an epic or a revenge plot or a rescue mission or a buddy film or an echo of a previous war. “War,” says Norman Bryson, “is perhaps the most ancient, and certainly one of the most powerful, of rhetorical topoi.”[1] That said, it might pay still to think about war not simply as a narrative but rather as a narratological problem. It might, in other words, pay to ask in technical terms: is the narrative representation of war different in kind from the narrative representation of other objects and events? Does the enormity, the violence, the inhumanity of war put particular pressure on and thus alter or upend what we can normally expect from narrative structure or narrative technique?

Something About War

Fredric Jameson has pointed towards two ways to think about this question in his recent discussion of “War and Representation” in The Antinomies of Realism. First, he distinguishes between, on the one hand, war texts that—as with Stendhal and Tolstoy—“defamiliarize” war, laying bare its real horror and, on the other, war texts that “are content to reconfirm the stereotype,” to fit war into preexisting normative or ideological frames.[2] Although Jameson focuses in his essay on several particular generic variants, what’s important here is the simple fact that there could be a choice between these and other discursive modes. The very existence of that choice underscores an issue I’d like to discuss: how the content of war affects the particular relation between story and discourse, the what and the how of a given narrative. That the same war stuff can be rendered as either critical or stereotypical—or as what Mikhail Bakhtin would call either official or unofficial—highlights the relative autonomy of story and discourse and perhaps even heightens that autonomy, leading us to ask whether the narrative representation of war might be different in kind from other kinds of narrative. Second, Jameson raises a larger, different, but ultimately related “suspicion” about war, “the suspicion that war is ultimately unrepresentable” (Antinomies, 233). Or, as Kate McLoughlin puts it in her Authoring War, “War . . . resists depiction, and does so in multifarious ways.”[3] As opposed to seeing war as more amenable to one or another type of narrative, this is to see war as something sublime, radical, or abject, as something essentially resistant to narrative representation. No longer a matter of one instead of another relation between story and discourse, war in these terms refuses the relation and perhaps reveals ideological or historical limits immanent to narrative as such.

As, both Jameson and McLoughlin suggest, however, that resistance, when it meets the imperative nonetheless to say something about war, might lead, despite itself, to new or innovative modes of narrative discourse:

Yet, even as it resists representation, conflict demands it. The reasons that make war’s representation imperative are as multitudinous as those which make it impossible: to impose discursive order on the chaos of conflict and so to render it more comprehensible; to keep the record for the self and others (those who were there and can no longer speak for themselves and those who were not there and need to be told); to give some meaning to mass death; to memorialise; to inform civilians of the nature of battle so as to facilitate the reintegration of veterans into peacetime society; to provide cathartic relief; to warn; and even, through the warning, to promote peace. (McLoughlin, Authoring, 7)

Just as narrative can defamiliarize war, showing how its violence exceeds or undermines ideological expectations, so, it seems, can war defamiliarize narrative, encouraging us to see its historical limits as well as the relative autonomy of the story/discourse relation. “The point,” writes Jan Mieszkowski, “is not that war is ‘unrepresentable’ but rather that war itself is partly a contest between different paradigms of representation that are not easily coordinated with one another” (Watching, 35).

In my recent book, War Pictures: Cinema, Violence, and Style in Britain, 1939–1945, I examine instances in which the cinematic effort to represent war revealed formal and conceptual limits immanent to both the concept of total war and the wartime feature film; I looked at moments when “impacted, difficult, or aberrant instances of cinematic style assert themselves against the thematic, aesthetic, or political unities that would seem otherwise to organize [a] film.”[4] I want here to look again at the pressure that war puts on representation and how that pressure might distort or enable different relations between story and discourse. I’ll revisit a couple of older examples before turning to Christopher Nolan’s Dunkirk (2017). What’s different about Dunkirk is not only that it was produced long after its represented war, and so could understand the historical event and its effects as an ostensibly finished thing, but also that it uses a highly wrought style—a kind of cinematic eccentricity—not to speak to the necessarily divided politics of war but rather to imagine war as essentially unnatural.

Like many films of the period, David Lean and Noel Coward’s In Which We Serve (1942) relies on flashbacks. After the HMS Torrin sinks at the film’s start, some surviving sailors, each representing a distinct socioeconomic class, clutch round a Carley float (fig. 1). A series of radial flashbacks then reach back into the sailors’ pasts, bringing us always back to the film’s putative present. This structure is essential to the film’s message: imagining sailors from different classes as points ranged around the same circle’s circumference, the film clings to a two-Englands ideology of class discrimination and imagines wartime as a pseudo-utopian environment in which those discriminations might be tactically suspended.

Noël Coward and David Lean, In Which We Serve
Fig. 1. Noël Coward and David Lean, In Which We Serve (British Lion Films, 1942).

That said, while this configuration both suspends and maintains socioeconomic difference, it creates problems. The film has some trouble remembering which of its memories belongs to the memory of which sailor, so stories that begin within the focalized context of one flashback sometimes end impossibly within another. Memories jump narrative tracks, disrupting parallel metadiegetic frames. The film’s structure thus produces a number of what Gérard Genette refers to as metalepses, moments of transgressive transit between existentially distinct diegetic levels.[5] Just as the raft suspends social differences along the diegetic spokes of its flashbacks, so does it undermine the narrative structures that otherwise support that ideological work. Narrative paradox allows Coward and Lean to have their war film and yet tacitly to imagine what might lie beyond its ideological horizons.

Michael Powell and Emeric Pressburger’s The Life and Death of Colonel Blimp (1943) also relies on flashback, as the film’s protagonist, Clive Candy (Roger Livesey), thinks back from the film’s present in 1942 to see how previous wars made him the man he would in time become. This structure lends the film a feeling of progressive inevitability, which is, however, belied by a melancholy counter-plot of compulsive repetition: even as we follow the progress of Candy’s movement through the Boer War, the First World War, and the Second World War, the morbid because identical reappearance of his romantic ideal as different women (played always and inexplicably by Deborah Kerr) in each war threatens to undermine that progress. In other words, while the film depends on an official argument about how the past develops into the present, its unofficial story of eternal recurrence suggests that none of that matters and that love and death will surely thwart even our best-laid plans. While the film imagines war’s violence as a narrative middle suspended between beginning and end, it evokes an antinarrative idea about history as repetition that undermines as it underlies the film’s other, more sanguine order.

In both cases, the ideologically riven content—the story-stuff—of the Second World War puts pressure on the contingent temporal relation between that stuff and a narrative discourse that cannot entirely contain it, resulting in moments of anti-narrative or unofficial figuration, moments of cinematic eccentricity or “historical irruption” or “cinematic excess.”[6] The films rely on what D. A. Miller has recently called “hidden pictures”: “These microstructures are too coherent as structures to be unplanned, but so deviously concealed that, once they are observed, the ostensible story is completely lost to view. Minimal, meager, without narrative interest or utility, the hidden pictures are exercises of pure style, of style devoid or uncaring of substance.”[7] Or, following Georg Lukàcs, we might think of these moments as unofficial and “internal discrepancies” paradoxically necessary to official forms: “Internal discrepancies in artistic form are manifestations of distortions in life patterns and result from unresolved (and therefore especially compelling) social contradictions.”[8] Like Hitchcock’s hidden pictures or Lukàcs’s internal discrepancies, moments of cinematic eccentricity in Lean, Powell and Pressburger, and others embody as film style a productive tension between the official and the unofficial work of narrative, representation, and wartime propaganda.

Another Fine Mess

I’ll turn now to my main and more recent example: Christopher Nolan’s Dunkirk (2017). Where the films I’ve already written about deal with the nervous relation between the contemporary experience of war and its tactical representation as cinema, Nolan’s film is a period piece, an effort to capture and to celebrate—as memory, event, parable, or heritage product—an episode often taken as emblematic of Britain’s war: the small-boat evacuation of the British Expeditionary Force from Dunkirk from the end of May to the beginning of June, 1939. Although Dunkirk was a success at the box office, early reviews registered a stark ambivalence, with critics falling into nearly opposed camps: either the film was a moving representation of war and its real effects on real people, or the film was a sterile and maybe boring exercise in technique for its own sake. Whether you think the film good or bad, this ambivalence makes a certain sense as Nolan relies on a conscious if obscure use of a similar ambivalence throughout his film.

For instance, unlike some other recent examples—Steven Spielberg’s Saving Private Ryan (1998), Terrence Malick’s The Thin Red Line (1998), or David Ayer’s Fury (2014)—Dunkirk is about violence but avoids graphic displays of its effects on the human body. Although soldiers and sailors are in fact shot, blown up, set aflame, and drowned in Dunkirk—and although the film is ostensibly about suffering and surviving on the field of battle—we get little sense of the material precarity of the human body, which other films have relied on to get at the historical or ethical real lying beyond or beneath the hygienic consolations of national memory. As opposed the dehumanizing violence of Fury, where shots are disordered, obscure, dirty, and chaotic—and where bodies and faces are brutally and routinely deformed—Dunkirk relies on the paradox of the “clean mess,” on highly wrought scenes that make violence palpable as the negative or implicit potential of obsessively organized images.[9]

Consider our first sight of the beach itself, seen from the wondering perspective of Tommy (Fionn Whitehead), the young soldier whom we follow throughout the film (fig. 2). The beach, the lines of soldiers awaiting evacuation, the four white posts sunk into the sand, the tidily piled rows of crates and bags and even bodies, the shadows cast at angles neatly perpendicular to all of the above: on the one hand, we might simply read all of this as evidence of the British military’s laudable if belated commitment to order and the rules of the game in the face of almost certain annihilation; on the other hand, the labored purity of the image, its presentation of war as something mathematically as opposed to dynamically sublime, seems to set form free from content and to imagine violence as a structural antithesis dialectically latent in aesthetic design. Seen in these terms, violence is reduced almost to the mix of anxiety and desire that we feel when we see something that is both especially beautiful and evidently breakable.[10] Even after the beach is attacked from the air, Nolan returns to an almost identical image of two of the four posts as they divide and frame the scene into a resiliently exquisite triptych (fig. 3).

Fig. 2. Christopher Nolan, Dunkirk (Burbank, CA: Warner Bros. Pictures, 2017).
Fig. 3. Christopher Nolan, Dunkirk (Burbank, CA: Warner Bros. Pictures, 2017).

This tension between the gorgeous cinematic image and the expectation of and, perhaps, the desire for its disfiguration is present in the film’s larger commitment to suspense as a primary affective mode: “I wanted to produce a film that was almost entirely based on the language of suspense, which I think is the most visual of cinematic languages, which is why I think Hitchcock has always been held up as possibly the greatest director of all time.”[11] Seeing the beach not as a battlefield but rather as a beautiful and brittle thing that implies its own necessary but deferred destruction supports this ethos of suspense. The beautiful or clean thing that must be upset in time is, in other words, an aesthetic equivalent to the ticking timebomb in Hitchcock’s familiar distinction between surprise and suspense:

Let us suppose that there is a bomb underneath this table between us. Nothing happens, and then all of a sudden, “Boom!” There is an explosion. The public is surprised . . . Now, let us take a suspense situation. The bomb is underneath the table and the public knows it . . . The public is aware that the bomb is going to explode at one o’clock and there is a clock in the décor. The public can see that it is a quarter to one. In these conditions this same innocuous conversation becomes fascinating because the public is participating in the scene.[12]

It is not simply the clock that makes this a scene of suspense; it is, rather, the play between the passage of time, the delicacy of “innocuous conversation,” and the spectacle of anticipated destruction that makes what was expedient “fascinating” instead. Early in Dunkirk, this sense of suspense as the unrealized but inevitable mess made out of a pristine thing is given bathetic expression in the unresolved micro-drama with which the film begins. When we first meet Tommy on the move from the war-torn town to the oddly curated beach, he is looking for a place to defecate. Each time he comes close, something prevents him from finishing: gun shots, a soldier burying another soldier, and so on. As far as we can see, Tommy never manages to make his mess, a fact that makes grossly visceral the resistance to a nonetheless inevitable closure on which suspense relies.

This tension between violence and aesthetic design is all the more pronounced at the level of the film’s intricate narrative structure. As many have recognized, the layered temporal form of Nolan’s film is weirdly and maybe unnecessarily complex. Instead of simply presenting its events in chronological order, Nolan organizes his film around three distinctly focalized and crosscut perspectives: the perspective of the soldiers on the beach, the perspective of a civilian steamer crew making its way to Dunkirk, and the perspective of a pair of RAF pilots in flight. Each section is introduced in the same style with an elegant and numbered sans-serif title superimposed on the image—“1. THE MOLE/one week,” “2. THE SEA/one day,” and “3. THE AIR/one hour”—emphasizing, first, the spatial and social differences between the three perspectives, second, their status as parallel as opposed to embedded diegetic schemes, and, third and most obscurely, the different temporal scales at which the three perspectives are presented (fig. 4).

Fig. 4. Christopher Nolan, Dunkirk (Burbank, CA: Warner Bros. Pictures, 2017).

It is that temporal difference that makes the film so complicated; for, although each of these layers takes up a clearly defined and different amount of time at the level of story—an hour, a day, and a week—they each occupy roughly the same amount of narrative discourse: roughly a third each of the whole film. In fact, Nolan’s orthographic presentation of the intertitles as fractions, with the place name appearing as a numerator above the line and the time span appearing as a denominator below the line, recalls Genette’s attempt to express the vagaries of narrative duration in the algorithmic terms of proportional “relationship between a temporal dimension and a spatial dimension,” a relationship that corresponds to that which stands between story and discourse (Narrative Discourse, 87).

As a result, while Dunkirk seems to advance from one point to another and to follow a consistent line from beginning to end, its different temporalities require the film’s focus to move jaggedly back and forth between past, present, and future. The same events are sometimes presented from two or three different perspectives, and it is not always clear or, rather, it is variously and incoherently clear what motivates the film’s choices when it comes to its presentation and distribution of events. Sometimes discrete shots are linked by visual pattern or color, sometimes by rhythm alone, and sometimes by an excess of emotional or psychological intensity. Although the film seems—thanks to its reliance on exact crosscutting and its pulsing soundtrack—to move insistently forward, its convoluted plot produces significant narrative disruptions as it negotiates the scalar relation between its three wartimes.

Nolan’s handling of these events and perspectives is nonetheless a remarkable organizational achievement; and, precisely because its underlying motivations lack coherence, it is something like that beautiful but fragile beach, a thing whose extravagant neatness somehow implies and perhaps provokes its own eventual and disarranging mess. For instance, the film is avowedly obsessed with measuring, marking, and embodying time. Hans Zimmer’s score is built both around the sampled sound of a ticking clock and an auditory illusion known as a Shepard Tone (a pair of melodic lines an octave apart are dynamically arranged to produce the effect of an insistent, interminable, and thus impossible tonal ascent).[13] Farrier, the RAF pilot (Tom Hardy), keeps careful track of his dwindling fuel with a watch and a white pencil (fig. 5).

Fig. 5. Christopher Nolan, Dunkirk (Burbank, CA: Warner Bros. Pictures, 2017).

Commander Bolton (Kenneth Branagh) frets about the delayed appearance of ships, air support, and material. A number of soldiers (not sailors!) discuss and mistake the timing of the tides, thinking that it comes and goes in three as opposed to six hours. On the one hand, this obsessive attention to the passage and orderly measurement of time is part of the film’s cultivation of suspense; all its extradiegetic ticking and anxious expositional time-telling help to make what is ultimately a lot of waiting around into something dramatic and inherently narratable. On the other hand, and unlike Hitchcock’s time bomb, the film’s handling of time is finally less inexorable than it is enigmatic. While the rhetoric of the film’s many references to clocks, watches, tides, the oscillation of day and night, and its own announced scheme might contribute to a feeling of time’s relentless passage, its underlying logic often works in an opposite direction and renders time not as relentless but rather as repetitive, dilated, or stalled.

On a first viewing, many of the film’s most studied temporal effects can look like mistakes or moments of uncanny or unnatural recurrence:

1. We see the same air attack on the same destroyer three times, but, given the uniform look of military equipment and the film’s relatively static representation of conflict, you might think they’re two or even three different boats. Or you might mistake two different boats for one and the same.

2. Because they come from “The Sea” and “The Land,” adjacent scenes that appear to move the plot forward at the level of discourse in fact move backwards at the level of story as they invert one character’s—Cillian Murphy’s—before and after. In a moment that is only apparently surreal, he appears wet and shell-shocked before his ship has sunk.

3. As the small ships finally approach the beach, Dunkirk cuts back and forth between unfocalized or ambiguously focalized shots of the various boats and focalizing shots of Mr. Dawson (Mark Rylance) as he seems to look towards the shore from his ship, the Moonstone. As, however, attentive viewers will know, the Moonstone never makes it to the beach but instead stops mid-Channel to rescue men forced to abandon a capsized destroyer; despite appearances, Dawson is looking not only towards a different place but also from a different time; it is a metaleptic look. (And, of course, his boat would be named after a novel that depends on the splintering of narrational and thus diegetic perspectives.)

4. As Dawson and his crew drag men aboard, the film cuts between the perspective of the Moonstone and the previously distinct perspective of Farrier flying above. Because these cuts represent not only a shift in perspective but also the appearance of a perspective formerly aligned with one diegetic system (“The Air”) within another diegetic system (“The Sea”), they are also a point at which either the film breaks its own rules, crossing its diegetic streams, or it sees its three temporalities meet finally in the “present.”

Indeed, the decisive collapse of three perspectives into one is given ponderously Christological expression minutes later as one of many rescued soldiers is revealed to be our Tommy (fig. 6 and fig. 7). As he is pulled from beneath the water’s surface into the Moonstone, his appearance aboard the boat and thus within what had been a distinct diegetic level brings the hour, the day, and the week into the same narrative current.

Fig. 6. Christopher Nolan, Dunkirk (Burbank, CA: Warner Bros. Pictures, 2017).
Fig. 7. Christopher Nolan, Dunkirk (Burbank, CA: Warner Bros. Pictures, 2017).

His apparent resurrection, which also brings together the soldier as species and the soldier as individual (both a Tommy and our Tommy), is offered as a moment of both narrative and symbolic resolution. This heavily freighted moment—water, birth, death, etc.—seems to offer an instance of what Jameson would call a “symbolic [resolution] of [a] real political . . . contradiction” insofar as it allows the film to solve an apparent problem in an especially spectacular fashion; because, however, the apparent problem is of the film’s own making (the perspectives were alienated from one another because Nolan wanted it that way), it is hard to say to what real problem it could really refer.[14] It is, perhaps, a symbolic resolution without its corresponding “real political contradiction.” I’ll return to this point in a moment.

A Special Type of Illusion

Now, I’ll admit that I’m being sort of thick about this, and that even if these problems are real, many viewers have seen and enjoyed Dunkirk without losing themselves in—perhaps without even noticing—the labored complexity of Nolan’s game. That said, things that sometimes seem easy sometimes aren’t; and, if there’s much to enjoy in Dunkirk, the film nonetheless leads us to at least the edge of paradox and threatens to undermine what we tend to consider a “natural” relation between story and discourse. I’m confident saying this because anyone familiar with Nolan’s other films—Memento (2000), Inception (2010), Interstellar (2014), and others—will know that temporal paradox is what a Christopher Nolan film is all about. Indeed, Nolan’s reputation as auteur depends on this type of cinematic legerdemain. That said, where the scalar relation between dreams and dreams-within-dreams or experiences at one or another distance from a black hole serve as his conceptual content and a basis for his narrative experiments in these other films, there is no similar supernatural or unnatural motive in Dunkirk. In Inception or Interstellar Nolan relies on science fiction to produce objects at the level of story-time that would authorize his temporal experiments at the level of discourse. In Dunkirk there are no black holes. There is only the war.

Think of the impossible staircase in Inception. As Arthur (Joseph Gordon-Levitt) explains, “In a dream, you can cheat architecture into impossible shapes. That lets you create closed loops, like the Penrose Steps. The infinite staircase” (fig. 8). The Penrose Steps (a thought experiment concocted in the late 1950s by the father-and-son team of Lionel and Roger Penrose) produces the Escher-type illusion of four connected flights of steps that take you up or down an apparently infinite and squared spiral; if that is the paradox, in Inception the dream and its rules are what motivate the paradox and, as it were, promise to resolve it into narrative form. And, while there are no infinite staircases in Dunkirk, there is, as I have already said, the Shepard Tone, an auditory illusion of infinite melodic ascent, with which Roger Shepard explicitly compared the Penrose Steps in his foundational 1964 essay; indeed, he refers explicitly to Lionel and Roger’s 1958 paper, “Impossible Objects: A Special Type of Visual Illusion.” However, where Inception motivates its impossible objects with the unnatural logic of the dream, there can be no similar motivation behind the impossible objects of a historical film like Dunkirk.

Fig. 8. Christopher Nolan, Inception (Burbank, CA: Warner Bros. Pictures, 2010).

Nolan’s other impossible or “spooky” story objects are all good examples of what Brian Richardson, Jan Alber, and others have taken as necessary to “unnatural narratives,” which, because they thwart the rules governing space, time, and causality at the level of story, put pressure on the “natural” conventions of the story/discourse relation. Jan Alber writes that “the term unnatural denotes physically, logically, and humanly impossible scenarios and events.”[15] Unnatural events of this kind are, of course, Nolan’s bread-and-butter, the extraordinary story-objects—black holes, dream machines, etc.—that underwrite his experiments with narrative discourse. What’s striking about Dunkirk is that it assumes that a real war can produce similarly paradoxical narrative effects. There is, in other words, an implicit claim about the narrative status of wartime in Dunkirk that we need to consider if not to embrace. Where, in Nolan’s other films, obviously, even stereotypically unnatural story objects authorize unnatural narrative forms, Dunkirk seems to assume that there’s something about the real story of war—trauma, violence, evil, the struggle for survival, original sin—that sets narrative discourse free to follow the path of unnatural temporal and auditory paradox.

What, though, does it mean to think of wartime—at least in this narrow narratological sense—as unnatural? In the case of wartime films like In Which We Serve and Colonel Blimp, cinematic eccentricity allows one to understand, to manage, or simply to express at the level of film style ideological contradictions that would be difficult otherwise to articulate. These films manage not only to express those real contradictions but also to reach towards (as Jameson might say) the “ideology of form” that shaped the lived experience of those contradictions as they indexed the historical and conceptual horizons of total war (Political Unconscious, 98–100). What, with Dunkirk, does it mean to treat war not as a real source or index of political or historical contradiction, but rather, as an impossible something that would seem somehow to exceed the laws of time and space? What does it mean to allow an autonomous aesthetic surface to refuse or to sidestep the real problems it ought rather to address? If Dunkirk fails, it is not because it is a sterile formal exercise. On the contrary, its management of time is wickedly intelligent, a bravura feat of logistical organization and auteurist filmmaking. No, if the film fails, it fails because it treats war as something unnatural, as an object or event that, like a black hole or a dream, can sufficiently motivate unnatural narrative experiments. If wars resist representation, it is not, alas, because they are unnatural or unreal; wars resist representation because they are violent expressions of real and yet “unresolved (and therefore especially compelling) social contradictions.” Insofar as it relies on undermotivated and unnatural discursive experiments to represent what’s resistant about war, Dunkirk fails—where my earlier examples succeeded—because it fails to acknowledge the reality of that resistance.


[1] Norman Bryson, Word and Image: French Painting of the Ancien Régime (Cambridge: Cambridge University Press, 1981), quoted in Jan Mieszkowski, Watching War (Stanford, CA: Stanford University Press, 2012), 33.

[2] Fredric Jameson, The Antinomies of Realism (London: Verso Books, 2013), 233.

[3] Kate McLoughlin, Authoring War: The Literary Representation of War from the Iliad to Iraq (Cambridge: Cambridge University Press, 2011), 6–7.

[4] Kent Puckett, War Pictures: Cinema, Violence, and Style in Britain, 1939–1945 (New York: Fordham University Press, 2017), 7.

[5] Gérard Genette, Narrative Discourse: An Essay in Method, trans. Jane E. Lewin (Ithaca, NY: Cornell University Press, 1980), 234–37.

[6] The term historical “irruption” comes from Antoine de Baecque in his Camera Historica: The Century in Cinema (New York: Columbia University Press, 2012) and “cinematic excess” is of course the subject of Kristin Thompson’s classic essay “The Concept of Cinematic Excess” in Narrative, Apparatus, Ideology: A Film Theory Reader, ed. Philip Rosen (New York: Columbia University Press, 1986), 130–42.

[7] D. A. Miller, “Hitchcock’s Understyle: A Too-Close View of Rope,” Representations 121, no. 1, (2013): 1–30, 12, emphasis in original.

[8] Georg Lukács, Writer and Critic and Other Essays, trans. Arthur D. Kahn (New York: Grosset & Dunlap, 1971), 9.

[9] I owe the phrase “clean mess” to D. A. Miller, who used it years ago in a conversation to describe the impossibly ordered disorder of a kitchen spill (pecans!) in Alfred Hitchcock’s Marnie (1964).

[10] I’ve written about a similar effect in Jean Renoir’s Rules of the Game (1939). See Puckett, Bad Form: Social Mistakes and the Nineteenth-Century Novel (Oxford: Oxford University Press, 2008), 155–56.

[11] Christopher Nolan, “‘Dunkirk’ Director Christopher Nolan: ‘We Really Try To Put You On That Beach,’ interviewed by David Greene, National Public Radio.

[12] Francois Truffaut with Helen G. Scott, Hitchcock (New York: Simon and Schuster, 1983), 73, emphasis in original.

[13] Roger Shepard accounted for this musical illusion in his 1964 paper “Circularity in Judgments of Relative Pitch,” The Journal of the Acoustical Society of America 36, no. 12 (1964): 2346–53.

[14] Fredric Jameson, The Political Unconscious: Narrative as a Socially Symbolic Act (Ithaca, NY: Cornell University Press, 1981), 80.

[15] Jan Alber, Unnatural Narrative: Impossible Worlds in Fiction and Drama (Lincoln: University of Nebraska Press, 2016), 25, emphasis in original.