In this paper, I examine the ending of Ian McEwan’s novel Atonement and Joe Wright’s film adaptation, considering the ways in which the shift in medium necessarily entails different rhetorical strategies which, in turn, entail different ethical judgments on the narrative’s central figure, Briony Tallis. Using the tools of rhetorical narrative theory, I argue that McEwan’s novel presents a particular challenge for adapters because its plot, its ontological play, and its ethics are all inextricably entwined with the affordances of verbal prose narrative, in the form of a novel-within-a-novel and a first-person diary coda. In order to transpose McEwan’s story to the screen, then, Wright must not only transpose the plot, but must do so by relying on the affordances of cinema to reproduce and reinterpret the novel’s rhetorical effects. But this move from prose to filmic discourse also necessarily shifts the ethical focus of the narrative from Briony’s own responsibilities as a character and an author to the audience’s investment in the fictional worlds which she creates, thereby also shifting McEwan’s indictment of Briony onto the film’s viewers. In order to see both the logic of Wright’s transmedial adaptation and its ultimate ethical effects, I focus on two questions in particular: first, how do McEwan and Wright each prepare their different audiences for the radical reconfiguration of the narrative’s twist ending, each relying on a different set of medial affordances; and second, how are these different audiences affected by the ethics of these disparate endings once the twist is revealed? Further, addressing these questions leads me to reflect on the troubled concept of fidelity, considering in what ways it can be recuperated as a theoretical tool.

1. The problem with fidelity

The call for papers of this issue of DIEGESIS focuses on whether “categories which are non-specific with regard to media [are] useful for disciplines which often define themselves by the peculiarity of their media-related subject.” In focusing on this topic, the editors touch on what has been the defining inquiry in the field of adaptation studies: how can we best understand the relationship between a source material and its adaptation, especially when this relationship transcends media boundaries? And this question, historically, has been a question of the fidelity of an adaptation to its source material.

The meaning of fidelity as a theoretical concept has a troubled history that parallels the development of adaptation studies as an academic field. From the emergence of adaptation studies—beginning with the publication of George Bluestone’s seminal Novels into Film in 1957—fidelity was conceptualized as an assessment of an adaptation’s formal features (such as perspective, narration, tense, and plot) as measured in terms of the formal features of its source material. (For this reason, I refer here to this approach as “formalist fidelity,” by which I mean the analysis and evaluation of an adaptation in terms of the formal features of its source material.) Not surprisingly, then, analyses of adaptations—almost exclusively film adaptations of prose literature—tended to focus on how closely the adaptation hewed to the original, implicitly asserting that the highest aesthetic achievement a film adaptation could accomplish was to ape what had already been successfully done in prose, measuring both cinematic aesthetic value and the success of the adaptation in terms of the formal affordances of verbal narrative.

My unfavorable description of formalist fidelity here echoes the predominant discourse in adaptation studies as it exists today, which is characterized by a rejection of fidelity in favor of a poststructuralist emphasis on intertextuality.¹ The poststructuralist reproach of formalist models of fidelity argues that fidelity is a theoretically unintelligible concept because it relies on a comparison between semantically distinct entities—comparing cinematic apples to prose oranges. As such, an approach to adaptation that centers around fidelity is essentially and unfairly biased toward the affordances of the source material, usually prose narrative. Instead of focusing on fidelity as an organizing concept, the poststructuralists argue, the field of adaptation studies should center instead around intertextual reference, recognizing that all texts—sources and adaptations alike—are infinitely connected in “an endless process of recycling, transformation, and transmutation, with no clear point of origin” (Stam 2000, 66). This argument so clearly articulated the problems with formalist fidelity that it remains the critical orthodoxy in adaptation studies to this day, and to discuss fidelity at all is considered naïve, accepting Barthes’s “myth of filiation” and foolishly trying to apply prose tools to cinematic texts (1977, 160).

I want to argue that the poststructuralist critique was certainly right, so far as it went, correctly identifying the theoretical problems with formalist fidelity. But the issue with this critique—and with the current state of adaptation studies as a field—is that instead of working to recuperate fidelity as a critical concept, the poststructuralists simply discarded it. And in doing so—in reorienting the field around the deceptively simple fact that all texts are intertextual—the poststructuralists eliminated the one thing that distinguishes adaptation studies from textual study in general: the special relationship between an adaptation and its source material.² In other words, even if we recognize that all texts are intertextual—one of the most valuable insights of poststructuralism—we still need to develop a way of talking about the particular kind of transmedial intertextuality we call adaptation. And further, we need to develop a way of talking about this particular kind of intermediality which will allow us to discuss the similarities and differences between adaptation and source: in other words, fidelity.

The proposal I’d like to develop here is that the tools of rhetorical narrative theory offer a way of recuperating the concept of fidelity and discussing narrative across media without falling into the false formalist comparisons that ensnared earlier adaptation theorists. Such an approach would focus on the rhetorical relationship between implied author, implied audience, and text—and in particular, on the affective, aesthetic, and ethical experiences that the implied author’s design of the text guides audiences to have. Considering these rhetorical elements of both source and adaptation will allow us to analyze the relationship between the two texts without assuming that a failure to recreate the formal details of the source text—or indeed, the choice not to recreate them—is in and of itself evidence of a failed adaptation. Indeed, by analyzing authorial purpose and audience response, we can recognize that different formal means can produce similar rhetorical ends, similar means can produce different ends, and that a shift in authorial purpose from source to adaptation does not necessarily imply a negative critical judgment.

Moreover, if a rhetorical approach to adaptation can productively recuperate fidelity as a meaningful concept for adaptation studies, then this rhetorical approach also stands as evidence that theoretical approaches that are not media-specific can indeed be productive tools, even for disciplines which are themselves defined by a particular medium. In fact, as I hope my analysis will show, rhetorical narrative theory offers a particularly nuanced set of tools for dealing with intermediality, in that it highlights the ways in which media-specific affordances can be used to for authorial purposes and to guide audience experiences in ways that are not media-specific and can, in fact, be transposed across medial boundaries.

To demonstrate what sort of reading this rhetorical approach to fidelity might produce, I want to focus here on the ending of Ian McEwan’s Atonement, which has been both praised and criticized for radically reconfiguring the ontological status of the entire storyworld generated in the first three sections of the novel. I argue that this shift in ontological status has ethical consequences not only within the storyworld, but also for the implied audience’s relationship with Briony Tallis, the novel’s central character, and the implied McEwan. As such, the ethical complexity of this twist ending—especially its complete dependence on Briony as a homodiegetic narrator—creates a particular challenge for adapting this narrative to film, a challenge that we can only address by discussing fidelity in rhetorical terms.³

The difficulty of Atonement’s ending is evidenced in the reception of Joe Wright’s 2007 adaptation of the novel, where reviewers identified the conclusion as a particularly troubling spot. Even laudatory reviews referred to the ending as “forced” (M. Scott 2007, par. 13), while some more scathing reviews called it “one of the most patently manipulative conclusions in movie history” (Beresford-Howe 2007, par. 2).⁴ The reversal of the ending prompts one reviewer to criticize not only the conclusion, but the entirety of the film as “about as substantial and authentic as the diffused, air freshener-ad light that keeps bathing the characters in rays of synthetic sunshine”—an essential artificiality we’ll return to below (Pevere 2007, par. 10). In this article, I argue that the discomfort caused by the conclusion of Wright’s film is directly traceable to the challenges of adapting McEwan’s twist ending. Because the film version of Atonement cannot engage in the kind of character-narrated diary construction that informs the ambiguous ethics of the novel’s end, Wright relies instead on the formal affordances of cinema to produce a similar rhetorical effect. But this move from prose to filmic discourse also shifts the ethical focus from Briony’s own responsibilities as a character and an author to the audience’s investment in the fictional world she creates, thereby transposing McEwan’s indictment of Briony onto the film’s viewers.

In order to see both the logic of Wright’s adaptation and its ultimate ethical effects—and to demonstrate the value of a rhetorical approach to fidelity—I will focus on two questions: first, how do McEwan and Wright prepare their different audiences for the radical reconfiguration of the twist endings; and second, how are McEwan’s and Wright’s audiences differently affected by the ethics of these endings once the surprise has been revealed? By considering the ways that Wright solves the problems presented by McEwan’s novel, we will come to better understand both the novel and the film; further, this analysis will demonstrate the value of rhetorical theory in discussing narrative across media, particularly as it relates to the concept of adaptation fidelity.

2. The design of McEwan’s Atonement

McEwan constructs his novel in three major parts, placing the twist ending in a short coda in the last twenty pages of the book. In Part 1, thirteen-year-old Briony Tallis witnesses the beginning of a romantic relationship between her older sister Cecelia and Robbie Turner, a childhood friend and son of the Tallis family’s charwoman. From her window on her family’s estate on a hot summer’s day in 1935, Briony watches and misinterprets a confrontation between Cecelia and Robbie, in which Cecelia strips down to her underwear and dives into a fountain to retrieve fragments of a vase that the two have struggled over and broken. Briony’s misunderstanding is further compounded as the day continues—first, when she reads Robbie’s sexually explicit note to Cecilia (mistakenly sent instead of an apology for the broken vase), and second, when she later misidentifies Robbie as a “sex maniac,” telling police that he is the shadowy figure she saw sexually assaulting her cousin Lola. As a result of Briony’s lie, Robbie is imprisoned, separated from Cecelia just as they are finally able to acknowledge the extent of their feelings for one another.

Part 2 of Atonement shifts from Briony’s narrative perspective in 1935 to Robbie’s in 1940, when he has been released from prison in exchange for his enlistment in the British Expeditionary Force at the beginning of World War II. This section of the novel relates Robbie’s desperate efforts to survive the BEF’s retreat to Dunkirk in the face of the approaching German army and return to England—and to Cecelia, who refused to believe Briony’s story and wrote to Robbie in prison, promising him she would be there when he was released. Accompanied by two other soldiers, the injured Robbie struggles across the French countryside toward the coast, in constant danger from German bombs and bullets. Given hope by memories of Cecelia’s letters and the promise of their future together, Robbie finally makes it to Dunkirk; Part 2 ends with him falling asleep in the basement of a ruined building, assured by his companion that they will be evacuated at first light.

Part 3 shifts perspective again, back to Briony’s experience. Driven by guilt for her lie and its devastating consequences for both Robbie and her sister, Briony estranges herself from her family and refuses her preordained future in Cambridge, opting instead to become a trainee nurse in London. Working under difficult conditions caring for the maimed and dying soldiers returning from the front lines, she attempts to do penance for her crime. In her spare time, Briony also attempts to atone by writing transparently autobiographical fiction, including a novella entitled Two Figures at a Fountain which is read enthusiastically but ultimately rejected by no less than Cyril Connolly. When she hears that her cousin Lola is marrying Paul Marshall—friend of the Tallis family and, Briony now realizes, Lola’s actual rapist from that night in 1935—Briony first attends their wedding, and then locates Cecelia, who is also working as a nurse. When she arrives at Cecelia’s flat, Briony finds Robbie there, psychologically scarred by the prison cell and the battlefield, but nonetheless intact. The section ends with Briony attempting to atone for her crime, promising the pair that she will confess, both to her family and the police, that her original testimony was a lie, in spite of the fact that Lola’s marriage to Marshall ensures that no legal action regarding the rape could possibly be taken.

Oddly, the last page of Part 3 contains an authorial signature: “BT / London 1999” (McEwan 2001, 349). Readers turn the page to find not “Part 4,” but a coda echoing this signature, entitled “London, 1999.” Here they find a diary entry from the now elderly Briony, a celebrated author who has most recently finished writing an autobiographical novel—the manuscript, made up of the first three sections, that the audience only now discovers it has been reading. Diagnosed with vascular dementia and afraid of losing both her memory and her capacity to write, Briony has finished her novel knowing that it cannot be published while the Marshalls are still alive, due to the inevitable libel suit that will result from the manuscript’s allegations. However, the reader also learns that while most of Briony’s novel is factual, she has changed the ending; in an attempt to avoid what she calls “the bleakest realism,” Briony has suppressed the fact that Robbie never escaped from Dunkirk, dying of sepsis from his shrapnel wound on the eve of the evacuation, and that her sister died months later in the collapse of a bomb shelter in London (2001, 371). In fact, the two were never reunited, and Briony neither confessed her crime nor asked their forgiveness. The novel ends with Briony at a production of her simplistic childhood play The Trials of Arabella, written for that summer’s night in 1935 and performed only now on her seventy-seventh birthday; she fantasizes about Robbie and Cecelia sitting together watching the play with her, claiming that such a reunion is “not impossible” to imagine (2001, 372). Thus the twist is revealed, and, as David Lodge puts it, “[w]hat seemed to be a conventional realistic novel turns out after all to be a postmodernist metafiction” (2002, 85).

In adapting McEwan’s narrative for the screen, Wright remains scrupulously faithful to the novel on the level of plot, hitting all of the expected beats and altering nothing of significance about the characters, events, or progression. When Wright reaches the coda, however, the diary format of “London, 1999” offers him an opportunity—indeed, almost requires him—to foreground the affordances of the cinematic medium rather than attempting to replicate the first-person immediacy of Briony’s journal. Before turning to Wright’s ending, though, I want to focus on how each text sets up its surprise; this will help us not only to understand the logic of McEwan’s narrative, but also how Wright’s reliance on cinematic affordances sets his audience up for the different ethical consequences of his ending.

3. Atonement’s allusions

McEwan employs two different strategies to ensure that, upon finishing the novel, his readers will view the twist ending not as a cheap trick, but as a surprise that was cleverly prepared for over the course of the narrative. First, as many scholars have noted, McEwan composes Briony’s novel in such a way as to invite readers to see a range of literary allusion in the text; while these allusions are generally attributed to McEwan and not to Briony, this technique generates a pattern of authorial control and awareness that encourages readers to see the coda as the work of a masterful implied author, as opposed to the act of destructive lunacy referred to by Boerner. And second, he cunningly includes hints of Briony’s own revision of her novel, inviting attentive readers to trace the shadowy palimpsests of earlier versions behind the final draft. Having reconfigured the text as a novel-within-a-novel, McEwan guides initiated readers to see these traces as subtle preparation for the “London, 1999” revelation.

Because much of the critical work on McEwan’s novel is dominated by and focuses explicitly on its allusiveness, I will focus the majority of my attention here on his strategy of embedding Briony’s revisions into both her own novel and the whole of Atonement. That said, it is worth focusing, at least briefly, on how Atonement’s allusions act as preparation for the metafictional turn in the coda. As has been observed by a variety of scholars, McEwan’s novel is rife with literary allusions, beginning with an explicit mention of another text in an epigraph from Northanger Abbey.⁵ In addition, the characters themselves often refer to other works in both their thoughts and conversations; these include references to Fielding and Richardson (a conversation between Robbie and Cecelia), Auden’s elegy for Yeats (carried by Robbie during the Dunkirk retreat), and allusions to a variety of modernist literary figures, most especially Virginia Woolf (in the rejection letter Briony receives from Horizon, discussed below). In addition, more oblique references have been highlighted by critics. Kathleen D’Angelo reads Briony’s play The Trials of Arabella as a covert reference to Charlotte Lennox’s The Female Quixote, with which McEwan “reminds readers to be critically engaged with the text” (2009, 92). Responding to Robert McFarlane’s observation that “the question of how the past is represented in language has become the central obsession of British fiction over the past three decades” (2001, 23), Pilar Hidalgo argues that McEwan’s novel is thematically engaged with the whole of British literary history, including Mansfield Park, Howards End, and Brideshead Revisited (2005, 83). Further, Anna Grmelová goes so far as to argue—through reference to implicit allusions to Auden’s Musée des Beaux Arts and Breughel’s Icarus—that “there is hardly a dichotomy between the first, metafictional part of the novel and the second, ‘realist’ one as it is sometimes claimed; both parts of the novel are discursive” (2007, 157).

This last point is significant. While there is, of course, some degree of difference between various critical claims, implicit in all of these analyses is the argument that the allusions McEwan imbeds into the text serve not only as a guide for savvy audience members as to how to read the text, but also as an anticipation of and an explanation for McEwan’s surprise ending. Grmelová’s claim quoted above is predicated on the fact that Auden’s poem—first published in 1940, the year of Robbie’s death—would have been unavailable to him; therefore, his knowledge of it is necessarily metafictional and is an anticipation for attentive and knowledgeable readers of the metafictional status of the first three parts of the novel. Likewise, given that “McEwan positions Atonement against earlier narrative models […] concerned with the author-reader relationship, specifically the 18th-century novel and the modernist novel,” D’Angelo ultimately argues that the coda sets up an “implicit argument about the value of critical reading” (2009, 89, 103).

For my analysis, the significance of these claims is that, whatever their conclusions, they all read these allusions—both explicit and implicit—as preparing the readers for the emotional, textual, and ethical consequences of the coda’s surprise. Of course, these allusions can only lay this groundwork for those readers who recognize it, and while it is plausible that McEwan’s implied audience will pick up explicit references to major authors like Fielding and Woolf, it seems to me less likely—though not impossible—that the echoes of Rosamond Lehmann’s Dusty Answer will be a touchstone for most readers, as D’Angelo claims. However, McEwan’s second strategy—including traces of Briony’s revisions at strategic points throughout the novel—ensures that even readers who miss his pattern of literary allusion will nonetheless encounter a design that retrospectively points to the radical reconfiguration of the coda.

4. Traces of Briony’s revisions

McEwan points readers toward Briony’s revisions at two major points in the novel: first, in the rejection letter Briony receives from Horizon, and second, in Briony’s own reflection in “London, 1999” on her writing process. The former anticipates the revelation that everything but the coda is the product of Briony’s pen, while the latter, occurring as it does after the coda’s revelation, underscores how deftly the trick has been pulled off. The rejection letter, penned by Cyril Connolly, is reproduced in its entirety in Part 3, immediately after Briony sits by the bedside of a dying French soldier and plays a role in his deathbed hallucination by pretending she is his fiancée. Though Connolly rejects Briony’s novella, he is also effusive with his praise, remarking on the fact that both he and Elizabeth Bowen, among others, read the entire work with great interest. He commends her deft images—the phrase “the long grass stalked by the leonine yellow of high summer” is singled out for particular praise (2001, 312)—and compares her interest in “the crystalline present” to Virginia Woolf, though he speculates that Two Figures owes too much to the Modernist author.⁶ Connolly includes minor corrections throughout—he suggests that the story works better if the young girl isn’t aware that the two figures have broken the vase, wonders if a Ming vase is “rather too priceless to take outdoors” (2001, 313), and corrects Briony’s reference to Bernini’s Triton as being in the Piazza Barberini, not the Piazza Navona—but his major critique is that nothing much happens in Briony’s story:

“[W]riting can become precious when there is no sense of forward movement,” Connolly advises, and he encourages Briony to include instead “an underlying pull of simple narrative” (2001, 312). While he hopes she “keep[s] some of the vivid writing about light and stone and water” (2001, 313 – 14), he presses Briony to include some kind of progression and conflict that move the story forward:

Though this is a major critique requiring significant revision, Connolly encourages Briony to “take our remarks—which are given with sincere enthusiasm—as a basis for another draft” (2001, 314). He even extends an invitation to discuss her draft further over a glass of wine, and notes—on the third page of his letter—that his rejections are rarely longer than three sentences. Finally, Connolly ends his letter with a few comments on an artist’s obligation to ignore the war and an oblique reference to Cecelia; apparently someone at the Horizon offices has recognized the surname Tallis.

Connolly’s letter stands out as a cerebral exception to the grim and bloody business of Part 3, focusing as it does on Briony’s immersion into the bleak hospital world of triage, amputation, and death. But it also draws particular attention from McEwan’s implied audience because it confirms for readers that Briony is interested in reproducing the events of her childhood in fiction. From the beginning of the novel, readers know that she has an interest in writing based on her composition of The Trials of Arabella, a juvenile play that featured a young woman who runs away from home, falls ill, and is healed by a disguised prince whom she subsequently marries. Likewise, passages in Part 1 describe the young Briony’s dawning realization of “the unbearable idea of other minds and the superiority of stories over plays” and her subsequent rejection of the immature Arabella aesthetic (2001, 41). In addition, her Horizon submission is mentioned tangentially earlier in Part 3, though its title and contents are not revealed. With Connolly’s letter, Briony’s desire to revisit and reshape her past becomes most explicit for the implied audience, though only the most astute among these readers would suspect that everything up to this point—including the appearance of Connolly’s letter—is part of this project.

This is sufficient for most readers on their first time through the book, but consider this same passage through the eyes of a reader who has read the coda and reconfigured this passage and the text that precedes it as Briony’s novel. If readers are willing to do the verbal legwork, they will find that McEwan has actually given the whole of the surprise away in Connolly’s letter, which offers not just traces of Briony’s earlier draft, Two Figures by a Fountain, but also points to the fact that the text we’ve been reading is a further revision of the Horizon submission. First, Connolly’s suggestion that Briony shift her focus from “the quality of light and shade, and [...] random impressions” to the “underlying pull of simple narrative” has clearly been followed. In fact, some of Connolly’s proposals for the kinds of conflict that could ensue likely struck Briony a little close to home, as his suggestions that the young girl act as a messenger for the couple, that her misunderstanding might affect the couple’s future, that she might expose them to her family, and that her intervention might “come between them in some disastrous fashion”—all of these, in addition to providing dramatic tension and narrative propulsion in Briony’s text, are in fact true.⁷

If these broad strokes aren’t enough, the industrious reader will find that even Connolly’s smaller suggestions have been taken into account in Briony’s manuscript. She has corrected her mistake about the fountain, correctly referencing “the half-scale reproduction of Bernini’s Triton in the Piazza Barberini in Rome,” as Connolly specified (2001, 18). Likewise, Briony has taken his suggestion about the vase’s provenance into account, though she ignores his suggestion that “Sèvres or Nymphenburg [might] suit your purpose,” opting instead to describe it as “Meissen porcelain” (2001, 313, 23). Even the line Connolly admires—“the long grass stalked by the leonine yellow of high summer”—has been retained almost verbatim; this line naturally appears on the same page that the young Briony witnesses Cecelia dive into the fountain—without, as Connolly suggests, noticing the broken vase (2001, 38). Of course, readers will likely miss these minute details on the first read of Atonement; to catch them would require the ability to recall obscure textual details—the name of an Italian piazza, the design of a vase, a particular line of prose—three hundred pages after they were encountered, a daunting task for even the most attentive readers. Moreover, McEwan wouldn’t actually want first-time readers to catch these details, even if they could; doing so would completely undermine the shock coming a few pages later in “London, 1999.” Rather, the only readers who are equipped to make these connections are those who are returning to the passage after reading the coda, thus understanding the way in which McEwan has built Briony’s revisions into the text itself.

In addition to functioning as a textual clue, Connolly’s letter also operates on a thematic level, anticipating the radical generic and ontological reconfiguration of Atonement’s coda. The historical Connolly was indeed editor of Horizon beginning in 1940, the year Briony submits Two Figures; his appearance, as well as that of Virginia Woolf and Elizabeth Bowen, in McEwan’s fiction simultaneously offers metafictional possibilities while also being easily read within the conventions of historical fiction, wherein real people and places comingle with fictional characters. Of course, first-time readers are likely to interpret Connolly’s presence as precisely the sort of allusion to be expected from historical fiction—which, after all, is what they believe themselves to be reading. Herein lies the particular cleverness of this clue: it confirms for both audiences that they are reading exactly the kind of generic text they suspect, whether they are uninitiated readers who have not yet read the coda or readers who have retroactively reconfigured McEwan’s novel as a postmodern historiographic metafictional narrative interested in disrupting the boundaries between characters’, implied audiences’, and flesh-and-blood readers’ worlds. In this way, Connolly’s presence in the fictional storyworld confirms both the stable historical fiction uninitiated readers think they are encountering and also foreshadows the ontological rupture of “London, 1999”—but only if readers are already aware of this rupture. During the first reading of Atonement, the appearance of historical figures not only fails to give away McEwan’s game, but further confirms readers’ misinterpretation as they naturalize Connolly’s appearance as a convention of historical fiction.

Lastly, Connolly’s letter—and the revision of Two Figures that it prompts—is effective in one final way, by providing a psychological motivation that helps explain Briony’s resistance to “bleak realism” that inspires her to write around Robbie and Cecelia’s historical deaths. McEwan—and also Briony—place her reading of Connolly’s letter immediately after her traumatic experience at the bedside of Luc Cornet, a French soldier dying of a horrific head injury. As Luc dies, his wound causes him to hallucinate that Briony is in fact his fiancée; reluctantly at first, and then more willingly, Briony participates in the delusion, confirming Luc’s false memories of the first time she came into his family’s bakery and reassuring him that she returns his love. After Luc’s death, a shaken Briony returns to her room in the hospital:

The reproduction of Connolly’s letter immediately follows, thus linking her fantasy of an impossible future with Luc and her eventual final draft of Two Fountains that includes both an accurate account of her crime and a fictionalized version of Robbie and Cecelia’s future that elides their deaths. In fact, Briony’s fictional future with Luc directly parallels the image that ends the coda: Briony’s fantasy of Robbie and Cecelia at her side, alive, together, and attending the first staging of The Trials of Arabella. In this way, the combination of the trauma of Luc’s death and the questions posed by Connolly’s letter both foreshadows (for uninitiated readers) the revelation that Briony has decided to suppress the couple’s tragic fate in favor of a happy ending and, at the same time, provides psychological motivation (for reconfiguring readers) for this decision, shaken as Briony is by Luc’s death in the moments before she begins considering how to revise her novella. Although Briony writes in the coda that all drafts prior to the final March 1999 manuscript “disguise[d] nothing—the names, the places, the exact circumstances—I put it all there as a matter of historical record,” reconfiguring readers can trace her final commitment to a happier fiction to this traumatic experience, in which her fictional game with Luc failed to cover over the bleak realism of his fatal injury.

Underlining the importance of Connolly’s letter as both an anticipation of and an explanation for the surprising revelation in “London, 1999,” McEwan actually repeats this strategy in the coda, after his metafictional trap has already been sprung. While the true fates of Robbie and Cecelia aren’t revealed until the closing pages of the novel, the first surprise—the fact that Briony is the author of the entire book up until this point—is exposed in the first part of the coda, which features Briony traveling to the Imperial War Museum to check some final sources for her manuscript. In particular, the coda details a series of notes and corrections to her manuscript made by a veteran of World War II. These notes largely address minor historical and idiomatic details, correcting the colloquial term for a “twenty-five-pounder gun” and the American “on the double” to the British “at the double,” highlighting errors in the details of RAF headgear, and irascibly chiding Briony for a typo that suggests that a Stuka—a German dive bomber—could carry a thousand-ton bomb: “Are you aware that a navy frigate hardly weights that much? I suggest you look into the matter further” (2001, 359 – 60). These corrections function in much the same way as Connolly’s earlier letter; now that even first-time readers are aware that the first three sections of the book are Briony’s work, the entire audience is capable of recognizing that these, too, are traces of an earlier draft of the manuscript. In fact, these changes have been incorporated into the manuscript the readers have just finished, and while again the details are too minor to be easily recalled, the industrious reader can indeed go back and find the changes in Part 2.⁸

Moreover, if Connolly’s letter inserts a recognizably historical figure into McEwan’s fictional storyworld, Briony’s reliance not only on the corporal’s corrections but the entire Imperial War Museum’s archive expands this ontological fluidity. For while Connolly’s letter gestures inward at the text—confirming Atonement’s generic status first as historical fiction, then in retrospect as postmodern metafiction—the presence of the Imperial War Museum among Briony’s sources points outward toward its paratexts and its author. Readers who carry on past the coda of Briony’s diary entry will find McEwan’s acknowledgements—signed “IM” in an echo of Briony’s “BT” at the end of Part 3—which begin thusly: “I am indebted to the staff of the Department of Documents in the Imperial War Museum for allowing me to see unpublished letters, journals and reminiscences of soldiers and nurses serving in 1940.” Thus, what readers believed was an actual archive fictionally employed by Briony, herself a fictional author, to compose a narrative that readers have just discovered, first, that they are reading, and second, that is partly a fiction within the storyworld, turns out to in fact be an actual archive employed by the flesh-and-blood author, with this research then fictionalized into the text of a fictional author.⁹ In this way, McEwan insinuates even the paratextual elements of his text into the service of the radical reconfiguration of the coda, positioning Briony as both a parallel figure of the author and underlining her fictionality, the creator of one text and the object of another.

5. The intermedial challenge of adapting Atonement

While McEwan prepares his readers for the radical reconfiguration of the coda by composing for Briony a novel that is deeply allusive—foregrounding the relationships between authors, readers, and texts—and that contains the evidence of both her textual revisions and the ethical issues that prompt these revisions, these are strategies that are categorically unavailable to Wright as he adapts the novel to film. In this sense, McEwan has it easy; after all, the novel that Briony writes is materially indistinguishable from McEwan’s Atonement, and thus he can easily disguise his use of a fictional intervening agent between this text and his own authorship. (After all, a page of prose by a flesh-and-blood author looks no different than a page of prose by a fictional one.)

As the author of a cinematic text rather than a prose one, however, Wright faces a very peculiar choice. He could remain faithful to the ontological structure of McEwan’s novel and alter Briony as a character, turning her into a filmmaker who has directed a mostly autobiographical film about her crime as a child and its subsequent effects. This choice retains the architecture of McEwan’s narrative, in which Briony generates a text that is embedded within a framing text of the same type. This adaptation strategy would potentially enable Wright to retain McEwan’s approach to setting up the coda’s surprise, first, by including cinematic allusions in the same medium as Briony’s text, and second, by incorporating cinematic critiques of earlier cuts, or drafts, of the film that would enable the audience to detect the revisions of Briony the filmmaker. This choice would maintain the diegetic hierarchy of McEwan’s novel, but it would do so by radically changing major plot elements, including the core identity of the novel’s central character. Alternately, Wright could remain faithful to the plot details of McEwan’s novel by retaining Briony’s role as a novelist—a choice that would meet a formal standard of fidelity—but this strategy necessarily generates a cinematic text that is not the product of Briony’s direct authorship, thus radically shifting the agency behind Atonement’s deception of its audience.¹⁰

To put this difficult problem another way, what we have here is a question of content and form, of story and discourse: Wright can recreate one or the other, but not both. In the storyworld of McEwan’s novel, Briony is directly responsible for both the story elements of parts one through three—details like events, characters, and settings—and for the discursive elements of the same text—factors like temporal order, focalization, diction, and syntax. If Wright wishes to retain Briony’s responsibility for both the content and the form of her portion of the film, she must be transformed from author to filmmaker; in this way she can still be the authorial agent behind both the film’s story and its discourse—elements like lighting, cinematography, casting, editing, and sound design. This would maintain fidelity to Briony’s function as the metafictional center of McEwan’s novel, but at the cost of abandoning fidelity to a crucial element of Briony’s character: her work as a novelist.

It’s not surprising, then, that Wright rejects this strategy outright, as it would lead to a fundamental restructuring not only of Briony’s character, but of much of the rest of the storyworld. (Would The Trials of Arabella become a home movie? Who would replace Connolly in rejecting Briony’s first filmic draft?) This would also require a significant departure from the historical referentiality of McEwan’s novel, as Briony’s choice to tell her story on celluloid in the 1940s, as well as the likelihood of the film industry allowing such an opportunity to a female director, are implausible in terms of cinematic history. Instead, Wright opts for a text that remains faithful to Briony’s role as a novelist, but in doing so, strips her of her authorship of the discursive elements of the film. In Wright’s Atonement, Briony is still the author of an autobiographical novel—here explicitly named Atonement—which relates the story of her lie many years before and the terrible consequences that flow from it, and this novel still obscures the tragic deaths of Robbie and Cecelia.

In choosing to adapt Briony’s novel into a film without concern for naturalizing the presence of the film as an adaptation within the storyworld, Wright replaces Briony as the implied author of the filmic discourse and must therefore find a cinematic equivalent of the diary entry that will provoke the narrative’s diegetic reconfiguration. The strategies of allusion and revision that McEwan uses to prepare his audience for the coda’s surprise are no longer available to Wright, because they no longer refer to Briony as an authorial agent.¹¹ McEwan’s use of literary allusion and embedded revisions operate as groundwork for the “London, 1999” radical reconfiguration precisely because they point, implicitly and explicitly, to Briony’s process as the author of the text the audience is reading. Without a filmmaking Briony, adapting McEwan’s strategy to include allusions to cinematic history or traces of early edits of the film might produce interesting aesthetic results, but it would be unintelligible as an attempt to prepare the audience for the revelation of Briony’s status as authorial agent, even for an audience retrospectively examining the film for clues.

6. Wright’s unpredictable ontologies

Rather than adopting McEwan’s strategies of allusion and revision, then, Wright must find other formal means to the same rhetorical end of preparing the audience for the surprise that is coming—at least, if Wright is interested in recreating the rhetorical effects of McEwan’s ending. I argue that he accomplishes this by foregrounding the exact issue that causes him problems: the presence of a teller, an authorial agent who is constructing the cinematic discourse. He foregrounds this discursive agency in two different ways. First, Wright engages in a kind of diegetic instability throughout the film, using elements of cinematic discourse (like cinematography or the presence of a soundtrack) to suggest an ontological permeability that asks viewers to rapidly shift between diegetic levels. And second, he foregrounds the particular synthetic affordances of film as a narrative medium in order to emphasize the cinematic nature of his text, including extra-narrative flourishes that do little to contribute to the narrative progression but continually remind viewers that they are watching a constructed text.¹² By foregrounding the cinematic and continually gesturing at the porous membrane between diegetic and nondiegetic elements, Wright invites the audience to contemplate how the narrative is being manipulated by an extradiegetic authorial agent, even if this agent is no longer Briony, but the implied Wright himself.

Wright’s diegetic play begins from the opening moments of the film, in a scene that foregrounds quite explicitly the ontological shifts that foreshadow the coda’s revelation. The film begins with the usual pre-title credits, presented in a typewriter font. Curiously, the sounds of carriage returns play on the soundtrack, though there is no obvious diegetic source for this sound. The soundtrack and the credits then come together as the film’s title is spelled out one capital letter at a time, each letter accompanied by the satisfying clack of a typewriter key. After the title, the first image of the film appears: a tight shot of a child’s dollhouse, over which is superimposed—again, one keystroke at a time—a subtitle indentifying the place as England and the time as 1935. The camera then pulls back and turns, following a meticulous procession of toy animals and people that lead away from the dollhouse’s door, coming to rest finally behind a young girl seated at a desk. She pecks away at an old Corona, typing “THE END” on a sheet of paper, then ripping it out of the typewriter and sliding it, with the rest of her manuscript, into a folder labeled “THE TRIALS OF ARABELLA / by / Briony Tallis.” Attended by an ominous and percussive piano theme, Briony marches out of her room and through the house, searching for her mother. Along the way, the camera alternates between focalizing shots over her shoulder and anticipatory set-ups that predict her movements and wait for her around corners and through doorways. Along the way she encounters busy servants, including Robbie’s mother Grace (though she is at this point only identified as a member of the kitchen staff), and Robbie himself, who already knows about Briony’s play (through “jungle drums”) and wants to read it, though he expresses hesitation about actually attending the performance that evening. Finally, Briony locates her mother in the drawing room and closes the door, shutting the trailing camera out of the room and bringing the martial soundtrack to an abrupt stop.

The following scene leaps forward in time to Mrs. Tallis finishing the play and heaping praise on Briony, who worries that her older brother Leon won’t like it. But before moving on, there’s quite a bit of diegetic play to unpack in this opening scene—especially for audience members who know, in retrospect, that this sequence is a cinematic analogue to Briony’s fictional novel. Importantly, this diegetic play occurs exclusively on the level of discourse; in other words, while the ontological flux doesn’t point to a filmmaking Briony, audiences are continually reminded of the overt manipulations of some authorial agent who is constructing this narrative. This is first apparent on the visual track, which depicts Briony’s dollhouse as a kind of set, in front of which she has artfully and painstakingly arranged all of her toys in careful order. Contrast this to the careless heap we might expect to find in a child’s room: this image first suggests the deliberate work of an agent who has arranged these figures into a careful procession, then reveals that agent to be the young Briony, at precisely the moment that she finishes another composition in the form of Arabella. But while this reveals Briony as someone interested in created ordered tableaus, it doesn’t yet suggest any ontological instability. The march of toys operates instead as a thematic anticipation of the coda’s revelation of Briony’s authorship, suggesting not only her habit of rearranging representations of people to suit her liking, but also placing her in the company of another all-powerful author, orchestrating the orderly procession of animals two-by-two.¹³

Like much of McEwan’s Atonement, this image suggests an ontological shift only in retrospect, but in this case audiences don’t have to wait until the coda to reinterpret this as an image of diegetic play. In the scene immediately following Briony’s conversation with her mother, we again leap ahead in time to a point later in the afternoon, where Briony and her sister lie on the lawn talking.¹⁴ This scene begins with a crane shot that pans down from the Tallis house and across the lawn, coming to rest almost directly above Cecelia and Briony. Particularly striking is the appearance of the house itself, which is virtually identical to Briony’s dollhouse, the first image of the film which appeared on the screen less than three minutes earlier. (See fig. 1.) This short gap between the two images ensures that the audience will recognize the similarity. Of course, this is easy enough to naturalize; clearly someone—Robbie, perhaps?—has constructed for Briony a scale model of the house she lives in. But the sudden appearance of this graphic match of Briony’s dollhouse is likely to produce at least a moment of ontological disorientation in the audience, a disorientation that leads to further thematic parallels. Like the dollhouse, is the Tallis house primed to become a stage on which Briony tells her story? Briony and Cecelia occupy the same position in front of the manor as Briony’s toys did to her dollhouse; do they also occupy the same functional position as figures manipulated into position by a designer? For initiated viewers, the answers to both questions is an emphatic yes, and for a first-time audience unaware that Briony is controlling the content of this narrative, too, the diegetic drift from dollhouse to manor house prepares the ground for this revelation.

In addition to the parallel images on the visual track, this sequence also toys with the film’s ontology in the soundtrack through the use of the typewriter. As I described above, the sounds of a typewriter’s carriage return and keys appear first over the credits, with no clearly diegetic source. This is not particularly unusual, however; films regularly introduce diegetic sound while the opening credits roll, with the source only becoming apparent when the visual track appears. Atonement’s audience is likely to naturalize the typewriter sounds in this way, though this reading is quickly undercut by the fact that this supposedly diegetic sound is apparently capable of having an effect on extradiegetic elements of the discourse: specifically, the film’s title and the subtitles superimposed on Briony’s dollhouse. This suggests that the typewriter is an extradiegetic sound, and further, that we are witnessing the composition process of the authorial agent. For reconfiguring audiences, this is clearly a reference to Briony’s authorship of the text; ironically, for uninitiated audiences, the fact that we discover this sound emanating from the diegesis in the form of Briony’s typewriter undermines this hypothesis, rather than confirming it.

The appearance of Briony’s typewriter on the visual track should provide firm ontological footing for the viewers, but Wright quickly pulls the rug out from under their feet. Briony types, and the minimalist piano theme mentioned above begins on the soundtrack. Then, as Briony assembles her manuscript and begins hunting for her mother, the solo piano is joined by a martial beat. It may take audiences a moment to notice that this percussion is produced by the carriage returns and keystrokes of a typewriter, and that this beat is introduced to the composition at almost the exact moment that Briony stops typing; the music—both the piano and the typewriter—then increase in intensity until Briony locates her mother and shuts the pursuing camera out of the room. Thus the typewriter, originally ambiguous, then firmly interpolated into the diegesis, becomes unambiguously extradiegetic, a part of the soundtrack that is the sole province of the implied author of the filmic discourse. Reconfiguring readers may think of this as a cinematic analogue for Briony’s writing process, though again, it’s important to remember that in Wright’s adaptation, there is no way to attribute discursive elements like extradiegetic music to Briony herself. Rather, this diegetic play on the soundtrack foregrounds and complicates the notion of an ontologically distinct authorship—just who is in charge of this story?—without actually holding out Briony as an answer. Additionally, viewers are invited to entertain the hypothesis that someone is writing in prose the story that we are now seeing visually, a medial crossover that models Wright’s own visual adaptation of McEwan’s, and Briony’s, verbal text.

This is therefore unlike McEwan’s embedding Briony’s earlier drafts in his novel, which implicate Briony herself in the production of the narrative, although it is similar to his use of allusions in that both techniques call attention to the presence of a sophisticated authoring agent behind the discourse without specifying anyone in particular. But in this case, Wright is relying on a specifically cinematic affordance—the dual track narration of both visual and audio tracks—in order to generate this ontological instability. In fact, this example, which I have unpacked in depth, is only the first of a pattern of Wright manipulating the soundtrack to create an ontological dance in and out of the diegesis.

A few minutes later, for instance, the typewriter pattern—in which diegetic sounds migrate onto the extradiegetic soundtrack—occurs in reverse. Cecelia has gathered flowers in anticipation of her brother’s arrival and places them in the heirloom vase that will soon be broken in her struggle with Robbie. The vase rests on top of a piano, and Cecelia contemplates it for a moment. On the soundtrack, a piano melody similar to Briony’s theme recurs, this time without the typewriter beat. Building to a frenzied crescendo, the melody repeats a single note until stopping suddenly when Cecelia leans over and plucks a piano string, producing the same note that has been repeating on the soundtrack. In this case, what was purely extradiegetic music suddenly intrudes into the storyworld and is completed by a character who is not ontologically capable of being aware of it in the first place. Though the pattern here is reversed, the effect is the same: this diegetic shift points to an extradiegetic agent who both is responsible for the music itself and orchestrates Cecelia’s actions such that she completes the melody, even though this extradiegetic agent cannot be read as Briony, even in retrospect.

One final example of this phenomenon will be of use because it combines both Wright’s diegetic play and his interest in foregrounding the cinematic. Leaping ahead in the film to when Robbie arrives with his two companions, Mace and Nettle, at the beach at Dunkirk, he finds a Boschian pandemonium of men, animals, and military equipment. Robbie’s arrival on the beach initiates a long take which lasts five and a half minutes and accompanies Robbie and his companions through the chaos of the retreat. Beginning with a tracking shot that follows the three men as they try to get information about the retreat, the camera escorts them past a grotesque sight: cavalry officers lining up their horses to be shot to prevent them from falling into the hands of the advancing German force. At this point, Mace stops to watch in horror while Nettle continues on with Robbie, past documents being burned and a beached ship, crawling with soldiers, one of whom swings from the mast shouting “I’m coming home.” At this point, a mournful string arrangement can be heard on the soundtrack. Robbie and Nettle continue moving through the bedlam, passing a soldier performing a gymnastic routine on a pommel horse, another sunbathing, another two fighting. The camera moves up the beach to the boardwalk, framing a turning Ferris wheel in the background against a gazebo in the foreground. Focusing on a group of men in the center, the camera moves close enough to hear them; they are singing, and as the camera circles them before moving back out of the gazebo, audiences recognize that the melody the soldiers sing is the same as the string arrangement on the soundtrack. Pulling away from the gazebo, the camera returns to Robbie, rejoined by the two other soldiers, as he passes a carrousel crowded with men, followed by a line of trucks whose radiators are being destroyed, a gesture parallel to the executed horses at the beginning of the take. The three men arrive finally at a bar overlooking the beach and go inside, leaving the camera to turn back and survey the space they have moved through. This final image takes in the entire panorama, as again, from a great distance, the audience hears from the beached ship the shouted refrain: “I’m coming home.” (See fig. 2.)

We see here the same diegetic blending on the soundtrack as the typewriter and the piano examples; the orchestral arrangement on the extradiegetic soundtrack morphs into the hymn sung by the choir in the gazebo—itself a surreal facet of the mise-en-scène, if not an ontologically disruptive one—then migrates back to the soundtrack. The ontological disruption here is further compounded by the song itself; unlike the typewriter clacking or piano note from earlier examples, this hymn, Dear Lord and Father of Mankind, is recognizable to the audience as part of their own history.¹⁵ For uninitiated viewers, the hymn first appears to be an allusion on the part of the implied author to a melody that audience members could be expected to know. As the take progresses, though, the melody reveals itself to be an extradiegetic anticipation of an event within the storyworld—the choir’s singing—which is itself a fictional representation of the historical events of World War II, suggesting an ontological continuity with the viewers’ world. Again, the effect is the same: the overtness of this diegetic play points to the presence of an implied author explicitly orchestrating the soundtrack and the storyworld in concert, even though this implied author cannot be identified as Briony.

7. Foregrounding the cinematic

Beyond being another iteration of Wright’s diegetic play, the Dunkirk long take provides our first example of Wright’s other strategy for setting up the coda’s surprise: foregrounding the synthetic affordances of cinema in an especially overt way. This sequence is mentioned in a variety of reviews, though it receives particular criticism from A. O. Scott, who writes that “[t]he impression left by a long, complicated battlefield tracking shot is pretty much ‘Wow, that’s quite a tracking shot,’ when it should be ‘My God, what a horrible experience that must have been’” (2007, par. 7).¹⁶ Scott is certainly correct that the long take draws attention to itself, but while he reads this as an aesthetic flaw to the extent that it disrupts the audience’s mimetic experience, I argue that this is an example of Wright’s efforts to highlight precisely those discursive elements of film that are largely invisible under standard Hollywood conventions. In doing so, this take distances viewers from what is being shown in the scene; Scott is right in that regard. But by calling the audience’s attention to the filmic discourse—how the scene is being presented—Wright also calls attention to the authorial agent behind this presentation. Again, audiences cannot attribute this long take to Briony, even in retrospect, but by emphasizing the construction of the narrative by some controlling author, Wright prepares viewers for the revelation that the content of the film, as well as the discourse, is being manipulated.

This emphasis on synthetic cinematic elements pervades the film, but a few examples will suffice to demonstrate the larger point. Immediately following the long unbroken shot on the beach, the camera cuts to the inside of the bar, where Robbie searches for something to quench his thirst among a unruly mob of soldiers waiting for evacuation. Making his way to the back of the bar, he discovers yet another surreal detail. In what appears to be the basement, a French film is being screened for an unseen audience. (Geraghty identifies this film as Le Quai des Brumes (Port of Shadows), a 1938 film directed by Marcel Carné [2009, 100].) Having finally found a still, private place, Robbie climbs down into the basement behind the screen and, with the black and white faces of two lovers obscuring him from one audience and silhouetting him for another, he breaks down. Covering his face in his hands, Robbie quietly comes apart as the horrific tableau of Dunkirk hits home. (See fig. 3.)

The juxtaposition between the diminutive figure of Robbie in the foreground and the close-ups in the film behind him is thematically appropriate, reminding both Robbie and the viewers of the love he is desperate to return to in England; having finally reached Dunkirk, Robbie is tantalizingly close to the future he envisions with Cecelia, symbolized in the film by the image of a seaside cottage on a postcard from Cecelia that he carries with him. But in addition to this thematic parallel between Le Quai des Brumes and Robbie’s yearning to return to Cecelia, there is a synthetic parallel, as well. By so explicitly foregrounding this film-within-a-film, Wright calls special attention to the gap between Le Quai’s fictional ontology, represented by a pair of lovers embracing, and the storyworld of Dunkirk, characterized by death and separation. For reconfiguring viewers, the relationship between the fantasy of Le Quai and the reality of the Dunkirk beach is precisely the relationship between this same representation of Dunkirk, now understood to be Briony’s romanticized revision of history in which Robbie is destined to return to Cecelia, and the grim reality of the fact that he will be dead within hours and will never see her again.

A more complex example of Wright highlighting the synthetic affordances of cinema occurs a few scenes later. After searching the ruined town for shelter, Robbie and Nettle finally locate an intact basement to take refuge in for the night. The room is crowded with other soldiers waiting for the evacuation, but Robbie is able to find a place to lie down. As he drifts off to sleep, he lights a match to look one last time at Cecelia’s postcard. As Robbie falls asleep, the match dies, leaving the audience with him in darkness, punctuated only by the sound of waves lapping on the shore. Then suddenly the audience is presented with three scenes they’ve seen before—Robbie and Cecelia fighting over the urn and a shard falling into the water, a close-up of Robbie’s hand offering Briony the wrong letter to deliver to Cecelia, and an extreme close-up of Robbie’s typewriter inscribing the fateful word “cunt” on this same letter—but each of these scenes is presented in reverse. The shard drifts up through the water, Briony hands the letter back to Robbie, and the letters t, n, u, and c are lifted miraculously off the page. Then, just as suddenly, viewers see Robbie as a soldier, walking alone through a field of poppies, then a crowd of men staring directly into the camera and singing The White Cliffs of Dover, an iconic WWII song of hope. The following shot brings us back to Robbie behind the movie screen as he lifts his eyes to the figures projected on it, suggesting that this choir of men is actually Le Quoi’s audience. At this point, Robbie’s voiceover narration joins the soldiers’ chorus as he repeats his promise to Cecelia to “find you, love you, marry you, and live without shame.” Finally, the audience again witnesses the police taking Robbie away on that summer’s night in 1935—again shown in reverse, Robbie exiting the police car and walking backward to Cecelia, only to start moving the correct direction in time so that the audience hears Cecelia whisper to Robbie, “I love you. Come back. Come back to me.” At this point, the audience is returned to 1940 and the Dunkirk basement, where Nettle is trying to quiet Robbie, who has been shouting in his sleep. Nettle shares the good news that the evacuation will begin in a few hours; it appears that the men are saved. The scene ends with Robbie drifting back off to sleep, assuring Nettle, “You won’t hear another word from me. Promise.” The next image is a hospital corridor with the subtitle “London / Three weeks earlier,” and the audience is thrust from Robbie’s story into Briony’s.

This sequence is striking in the way that, like the long take on the beach, it accentuates the particular affordances of cinema—in this case reversing time by literally reversing the film—in order to bring to the foreground the agency of a controlling author. It is perhaps more emphatic in this way than the long take, because while rare, long takes are easily legible to film audiences. While the absence of cuts is unusual in the beach sequence, no reinterpretation of the storyworld is required to account for it. This is not the case for the reversed pieces of film here; audiences must in some way account for this unusual facet of the discourse. In this particular case, Wright offers uninitiated audiences an easy answer by bookending the reversed film with shots of Robbie, first falling asleep, then waking up; this expressionistic presentation is then easily naturalized as Robbie’s dream, focused as it is on literally reversing the critical events of that day in 1935 and returning to Cecelia’s arms, even if it means walking backward. Reconfiguring audiences, however, can replace this explanation with another: this is not a representation of Robbie’s dream, but of Robbie’s death. For in retrospect, the audience is aware that Robbie never leaves this basement, and that this moment of synthetic play is also the precise moment when Briony chooses, in her novel, to purposefully replace the bleak realism of Robbie’s history with her own fiction of his survival. In this sense, Wright’s reversal of the film represents not only Robbie’s desire to rewrite the past, but also Briony’s.¹⁷ Even though she is not directly responsible for the filmic discourse, here the expressionistic reversal of the film acts as a thematic parallel to Briony’s act of creation, though it initially is legible only as a representation of Robbie’s dreaming subconscious. For reconfiguring readers, however, the reversal of the film’s visual track acts as a filmic analogue to the retrospection and reversal of Briony’s narrative.

One final example of Wright highlighting the cinematic will suffice to demonstrate the general pattern—this one replacing Connolly’s letter in the narrative progression. In this scene, Briony bravely walks away from Luc’s bedside through a ward crowded with the sleeping wounded. Debussy’s Clair de lune plays quietly on the soundtrack as the film cuts from Briony to documentary footage of the Dunkirk evacuation. Black and white images of soldiers disembarking from ships, grinning as they devour sandwiches, and sharing cigarettes fill the screen without explanation. After a series of these images, the film cuts to Briony’s face with a bright light coming over her shoulder; she is watching, with other hospital staff and patients, a news reel showing footage of the Dunkirk evacuation—another image of an audience watching a film. Almost immediately after viewers have reoriented themselves to this discovery, the news reel changes to footage “on the home front,” in which Queen Elizabeth visits a chocolate factory in the north of England. The factory is owned by none other than Paul Marshall, Lola’s rapist from 1935, and as the announcer informs the audience of their impending marriage, they appear on screen, interacting with the queen in the documentary footage. Prompted by seeing Lola, Briony walks away from the screen; the next scene builds on her momentum, as she marches through London streets to witness the Marshalls’ wedding.

Because this footage replaces Connolly’s letter in the narrative progression, it also complicates the letter’s function as a motivation for Briony’s revisionist history. Rather than focusing on Luc’s death and Connolly’s letter as the seeds which grow into Briony’s rejection of bleak realism, this sequence foregrounds instead the injustice of the Marshalls’ marriage set against representations of Robbie’s compatriots at Dunkirk as the immediate precursor to Briony first observing the wedding, then finding Cecelia to tell her of her plans to come forward and confess. In addition, Wright’s use of documentary footage here serves much the same function as McEwan’s inclusion of the Imperial War Museum as source material. The film’s audience here is asked to conflate a variety of ontological levels here, as documentary footage is used as a film-within-the-film, which then turns out to itself be a fiction within the larger narrative. Moreover, embedded within this historical footage are fictional characters; the presence of the actors playing the Marshalls—digitally inserted with CGI—further contribute to the ontological fluidity of Wright’s usage of this footage.

While other examples exist throughout the film, these suffice to demonstrate Wright’s solution to the problems posed by McEwan’s novel. Because he cannot rely on allusion and revision to plant clues to Briony’s role as intermediate author—indeed, because Briony no longer is the intermediate author—Wright instead uses the discursive elements of his film to create diegetic flux and to highlight the synthetic affordances of cinema. This strategy cues audiences to pay attention to the signs of a manipulating authorial agent in the film, and while this agent cannot be retroactively identified as Briony herself, reconfiguring viewers can read these instances as an analogue to her intervention. But what consequences do these different strategies for the ethics of both narratives’ endings? In my last section, I will examine the ethical shift that results from the opportunities and limitations afforded to Wright by the cinematic form.

8. A final act of kindness

For McEwan, the ethical question posed by “London, 1999”—is Briony’s revisionist history an act of legitimate atonement?—is undercut by the form her confession takes. Because McEwan presents her confession in the form of a diary entry, the earnestness and honesty of her appeal to “a stand against oblivion and despair” is somewhat undercut (2001, 372). In fact, as Briony notes, because of the inevitable libel suit the Marshalls would bring against Briony and her publisher, the novel will never be published while any of the relevant parties are alive—thus assuring not only that the Marshalls escape condemnation, but that the novel’s author will, as well.¹⁸ As a result, when Briony says that she would “like to think that it isn’t weakness or evasion” to rewrite history (ibid.), McEwan’s readers are right to note that she protests too much; in fact, evading the compounded consequences of her childhood lie is precisely what she has done by choosing to withhold the novel’s publication and confess only in the privacy of her diary. Thus when Briony claims that “[i]t’s not impossible” to imagine “Robbie and Cecelia, still alive, still in love, sitting side by side in the library, smiling at The Trials of Arabella,” McEwan’s implied audience is invited to read this statement as the escapist wish of a woman who has taken a step toward atoning for her crime, but an insufficient one. Yes, it is possible to imagine Robbie and Cecelia’s unfulfilled future together, but the simple act of imagining it—as Briony has done in her novel—doesn’t make it so, nor can this private act of imagination serve as penance or compensation for their terrible loss.

In addition, the conflation of Briony’s imaginative escapism with her childhood drama cues readers to see her revisionist history not only as an evasion of her ethical responsibility, but also an aesthetic failing, a reversion to the immature artistic priorities of Arabella that her younger self so clearly rejects:

Rather than retain this manifesto as the first principle of her art, Briony reveals in her confession the extent to which she has reverted back into the facile comfort of fairy tale logic; instead of remembering “how easy” it was for Briony to carelessly ruin Robbie and Cecelia’s lives, both her confession and her composition reveal how easy it is for her to imagine their presence beside her. For readers that have been invested both in Robbie and Cecelia’s relationship and Briony’s development as an artist, her ethical and aesthetic misjudgments cue the audience to reject her Arabella fantasy as both a sympathetic but inadequate atonement and an unserious aesthetic effort.

Wright’s investment in the affordances of cinema produces a quite different coda, both because an ending featuring an old woman writing in her diary would be decidedly uncinematic, and because the shift in authorial responsibility from Briony to the implied Wright leaves no room for a consideration of Briony’s aesthetic development. Instead, Wright opts for a different solution to the problem, and in so doing, shifts his audience from ethical critics of Briony’s revision to accomplices to her crime. In the film, the coda begins with Briony riding on the tube away from her sister’s flat in 1940, determined to make a public confession—the same scene which ends Part 3 in McEwan’s text. The clacking of the train’s wheels morphs into the clacking of typewriter keys, and the screen goes suddenly black, a new and disembodied voice asking, “I’m sorry, could we stop for a moment?” The next image is a bank of television monitors, all showing the same image of an elderly woman sitting across from a man. The woman rises to walk out of frame, and the footage rewinds on each monitor, just like in Robbie’s dream. Starting again, the man says, “Briony Tallis, your new novel, your twenty-first, is called Atonement,” and Briony, now an old woman, lowers her face into her hands and repeats her request that they pause. We see Briony compose herself under the harsh buzz of a dressing room’s fluorescent lights, and then return to what is clearly the set of a television interview.

The interviewer begins again, and Briony corrects him that Atonement is not her twenty-first novel, but her last. The camera begins over the interviewer’s shoulder in a classic shot-counter-shot set-up, but as Briony describes the illness that will end her literary career, the shot slowly closes in on her, framing out the interviewer, and the set dressing behind Briony goes dark, leaving her speaking against a black void. She discusses how long the novel took to write, but when the interviewer asks if the problem was its autobiographical nature and the fact that no names have been changed—precisely the specific problem in McEwan’s novel—she answers with a terse no. Instead—as dramatic extradiegetic string music fades in on the soundtrack—Briony confesses that the problem was that “the effect of all this honesty was rather pitiless, you see. I couldn’t imagine any longer what purpose would be served by it … by honesty, by reality.” One by one, then, Briony comes clean about her novel’s evasions, as we see depicted on the screen cinematic images of the historical truth she narrates: Briony typing her novel in the hospital instead of going to her sister’s flat; Robbie’s lifeless body in the Dunkirk basement, clutching Cecelia’s postcard; water rushing in slow-motion down the steps of the Balham tube station to drown those sheltered there, Cecelia’s body drifting through the water like the reversed shard of broken porcelain. The film then returns to the site of the interview, but with a difference; the camera has shifted slightly so that it is now facing Briony straight on, capturing her confession as she meets the viewer’s eyes (see fig. 4):

These are the last words of the film, but the camera then cuts to a shot of sunlight glittering off a receding wave as the last strains of the string arrangement dissipate and a piano theme takes its place. As we watch, Robbie and Cecelia, “still alive, still in love,” walking together down the beach, playfully pushing each other into the water and spinning hand-in-hand as the handheld camera spins with them. The two climb up the dunes to a white house overlooking the ocean and the Dover cliffs; the film ends with this shot, the physical incarnation of the image on Cecelia’s postcard (see fig. 5).

The cinematic affordances of Wright’s film create a coda, then, that is fundamentally different from McEwan’s in two significant ways. First, by shifting the form of Briony’s confession from a diary entry to a televised interview, Wright has done away with the issue of her confession’s legitimacy as decisively as Briony dismisses the interviewer’s suggestion that retaining the actual names was the reason for the novel’s long gestation. Briony’s confession is here direct and decisive, acknowledging not only her childhood crime but also her fictional obfuscation not only to the interviewer and, by extension, a widespread television audience, but also directly to another potentially aggrieved party: the viewers themselves, who have been misled by her fiction. As such, the entire question of the seriousness of Briony’s confession is eliminated; whether viewers judge her attempt to atone as adequate or just, the public nature of her confession suggests that it is not a weakness or an evasion.

This point leads us to the second way in which film as a medium shifts the ethics of Wright’s coda from those of McEwan’s. In the novel, the vision of Robbie and Cecelia alive, together, and supporting Briony and her art by attending the premiere of Arabella is filtered directly through Briony’s consciousness. Readers have no problem attributing this wish solely to Briony and reading it as another facet of her desire to escape her responsibility by taking refuge in fiction, as well as a reversion to her childhood artistic practices. By contrast, Wright’s final scene of Robbie and Cecelia on the beach occupies the same place in the progression, but like the other discursive manipulations discussed above—the diegetic play on the soundtrack, the long take, the expressionistic rewinding—there is no way for audiences to definitively attribute these images to Briony herself. Combined with Briony’s calm, direct confession which breaks the fourth wall, this implicates the audience in her fiction, and by including these happy images of Robbie and Cecelia, the implied author backhandedly offers the audience exactly what they wanted all along, without affording them the opportunity to completely pass off this narrative desire on Briony as the author of an ethically flawed narrative. Instead, the same images that the audience anticipated and hoped for as the resolution of Atonement’s narrative momentum are painful and hollow—as I quoted at the start, “about as substantial and authentic as the diffused, air freshener-ad light that keeps bathing the characters in rays of synthetic sunshine.”

By presenting these images without explicitly locating them as the product of Briony’s ethical or aesthetic efforts, Wright reminds viewers that they are also the product of our own wishes for Robbie and Cecelia, which turn out to have been a cinematic Arabella: facile, simplistic, unrealistic, and escapist. While McEwan’s coda also invites readers to recognize their own narrative desire to turn away from bleak realism and toward a fairy tale happy ending for Cecelia and Robbie, it offers a scapegoat for this desire; McEwan’s audience can view themselves as taken in by the manipulations of an author who, in the coda, reveals herself to be weak and evasive. But while McEwan’s coda invites the audience to critique Briony for her desire to escape into the fictional conventions of romance and happy endings, Wright’s text offers no analogous way out, as this Briony is neither responsible for the visual text which produced the same audience desire, nor is she the weak and evasive author of McEwan’s text, confessing without consequence to her own diary. Rather, Wright’s ending turns that ethical critique on the viewers themselves by giving them the happy ending that they wanted, but empty, drained of any fulfillment it could have offered.

Thus Wright manipulates the affordances of cinema to set up the coda’s surprise in a way that is similar to McEwan’s use of allusion and revision, relying on vastly different filmic means to accomplish the same end: a radical ontological reconfiguration. But this manipulation also produces a very different ethical valance to the film’s twist ending, turning the critique from Briony to the audience; rather than McEwan’s invitation for the audience to consider Briony’s ethical failings, Wright’s ethics challenge the audience to consider their own complicity in an impossible Arabella fantasy. But in doing so, Wright’s adaptation also acts as a response to the ethics of its source material, suggesting that perhaps we shouldn’t be so hard on Briony’s wish, despite her crime, to have everything work out alright in the end, to reject “the bleakest realism” and find instead that “the lovers survive and flourish,” happily ever after, like fairy tales say (McEwan 2001, 371). After all, like Briony, we wished it, too.

9. Recuperating fidelity

In closing, it is important to note that the kind of analysis we have been doing of McEwan’s and Wright’s endings would be impossible under either old formalist conceptions of fidelity or under the current critical orthodoxy that forbids any talk of fidelity. By focusing on the relationship between the implied authors and implied audiences, as well as on the kinds of affective and ethical experiences these audiences are invited by authors to have, we have been able to avoid the kinds of formal comparisons that were characteristic of early adaptation studies and that poststructuralist theorists rightly criticized. In addition, by recognizing that different means can be used to produce similar ends, that similar means can be used to produce different ends, and that narratives can be deployed by different authors for different purposes, we have been able to examine the difference between these two texts without assuming that Wright’s formal or ethical departures from McEwan’s text constitutes an aesthetic or moral failing on the part of the filmmaker. In fact, by approaching the problem of fidelity with means and ends in mind, we can see that transposing McEwan’s narrative into a new medium even requires Wright to shift away from McEwan’s authorial strategies, allowing us to read his adaptation choices not as a betrayal of the original text but simply as a natural consequence of the transformation into a new medium.

Thinking about fidelity and adaptation from a rhetorical perspective has also allowed us to engage with the discourse of intermedial reference which is central to contemporary adaptation studies, exploring the allusiveness that pervades both McEwan’s novel and Wright’s film. However, we have done so without neglecting the central intermedial relationship at stake here: the relationship between source and adaptation. The current state of adaptation theory would have us map the labyrinth of cinematic allusion and resonance, and I don’t mean to suggest that this focus is unproductive or uninteresting. But it is emphatically not a substitute for discussing the choices an adapter makes and to what extent these choices produce a text that is similar to or different from the source material. As I hope my analysis here has shown, looking at fidelity from a rhetorical perspective can not only illuminate the adapter’s choice, but also the source material; by understanding Wright’s solutions to the problem created by McEwan’s novel, we have come to better understand the construction of McEwan’s novel, as well. This kind of analysis is only made possible by recuperating fidelity as a theoretical concept and reintroducing it into the conversation of adaptation studies. This work must be cautious but rigorous, and it must proceed from an awareness of the pitfalls fidelity has already led the discipline into—but it must begin.

Finally, this analysis answers the question we began with: “[a]re categories which are non-specific with regard to media useful for disciplines which often define themselves by the peculiarity of their media-related subject?” My reading of McEwan’s and Wright’s Atonements shows that rhetorical theory allows us to identify the medial affordances that offer both constraints and opportunities to Wright as an adapter of McEwan’s narrative and to productively analyze the rhetorical consequences of these affordances with regard to the ethics of the twist ending. But in doing so, this reading has also demonstrated that a media-specific analysis can be done using rhetorical tools which are not themselves media-specific. Indeed, the transmedial nature of adaptation studies requires an approach that can both be specific about medial affordances and can also address itself to those elements of narrative, including rhetorical effects, that transcend any one specific medium. The rhetorical method I have laid out here is a step toward developing such an approach.

Bibliography

Apstein, Barbara (2003). “Ian McEwan’s Atonement and ‘The Techniques of Mrs. Woolf’”. In: Virginia Woolf Miscellany 64, pp. 11-12.

Atonement. (2007). Dir. Joe Wright. Screenplay by Christopher Hampton. Perf. James McAvoy, Keira Knightley, Romola Garai, Saoirse Ronan, and Vanessa Redgrave. Focus Features. [DVD 2008].

Barthes, Roland. (1977). Image – Music – Text. New York. [English Translation Stephen Heath].

Beresford-Howe, Jeff (2007). “Review of Atonement, dir. Joe Wright”. In: FilmThreat.com. Hamster Stampede, 10 Dec 2007. URL: http://www.filmthreat.com/reviews/10451/ (19 Mar 2011).

Bluestone, George (1957). Novels into Film: The Metamorphosis of Fiction into Cinema. Berkeley.

Boerner, Margaret (2002). “A Bad End” [Review of Atonement, by Ian McEwan]. In: The Weekly Standard (29 Apr 2002), p. 43.

Bordwell, David (2008). “A Cinema of Flourishes: Decorative Style in 1920s and 1930s Japanese Film”. In: Poetics of Cinema. New York, pp. 375-93.

Crosthwaite, Paul (2007). “Speed, War, and Traumatic Affect: Reading Ian McEwan’s Atonement”. In: Cultural Politics 3.1, pp. 51-70.

D’Angelo, Kathleen (2009). “‘To Make a Novel’: The Construction of a Critical Readership in Ian McEwan’s Atonement”. In: Studies in the Novel 41.1, pp. 88-105.

Geraghty, Christine (2009). “Foregrounding the Media: Atonement (2007) as an Adaptation”. In: Adaptation 2.2, pp. 91-109.

Giles, Jeff (2002). “A Novel of (Bad) Manners”. In: Newsweek (8-15 Apr 2002), p. 94.

Grmelová, Anna (2007). “‘About suffering they were never wrong, the old masters’: An Intertextual Reading of Ian McEwan’s Atonement”. In: Litteraria Pragensia 17.34, pp. 153-57.

Hidalgo, Pilar (2005). “Memory and Storytelling in Ian McEwan’s Atonement”. In: Critique 46.2, pp. 82-91.

Leitsch, Thomas M. (2003). “Where Are We Going, Where Have We Been?” In: James M. Welsh et al. (Eds.), The Literature/Film Reader: Issues of Adaptation. New York 2007, 327-34.

Macfarlane, Robert (2001). “A Version of Events” [Review of Atonement, by Ian McEwan]. In: Times Literary Supplement (28 Sept 2001), p. 23.

Moore, Caroline (2001). “A Crime of the Imagination” [Review of Atonement, by Ian McEwan]. In: Sunday Telegraph (16 Sept 2001), p. 12.

Phelan, James (2007). Experiencing Fiction: Judgments, Progressions, and the Rhetorical Theory of Narrative. Columbus, OH.

Stam, Robert (2000). “Beyond Fidelity: The Dialogics of Adaptation”. In: James Naremore (Ed.), Film Adaptation. New Brunswick, NJ, pp. 54-76.

Wolf, Werner (2005). “Intermediality”. In: David Herman et al. (Eds.), Routledge Encyclopedia of Narrative Theory. New York, pp. 252-256.

Matthew Bolton, Ph.D.
Assistant Professor of English
Concordia College - New York
171 White Plains Road
Bronxville, NY 10708
E-mail: Matthew.Bolton@concordia-ny.edu

Bitte zitieren Sie nicht die HTML-Version, sondern ausschließlich die PDF-Datei / Please do not cite the HTML version but only the PDF file:

1 I use the term “intertextuality” here to reflect the poststructuralist debt of contemporary adaptation scholars, but “intermediality” more properly expresses what these scholars mean when they refer to intertextuality in the context of adaptation. Werner Wolf expresses the difference succinctly: “[Intermediality] is still sometimes confused with ‘intertextuality’, especially if ‘text’ is used as an umbrella term covering all semiotic systems. […] If, however, ‘text’ is used more narrowly, referring to verbal texts only, the distinction between intertextuality and intermediality is straightforward. Thus understood, intertextuality is a variant of ‘intramediality’ and refers exclusively to ‘homomedial’ relations between verbal texts or text systems. Intermediality, in contrast, applies in its broadest sense to any transgression of boundaries between media and thus is concerned with ‘heteromedial’ relations between different semiotic complexes or between different parts of a semiotic complex” (2005, 252).
Pace their objection to the very notion of fidelity, I will risk some confusion here by describing the poststructuralists’s approach to adaptation as faithfully as possible, including their use of “intertextuality” in the broadest semiotic sense to refer to any textual relationship, whether the two texts are of the same medium or not; readers will be best served here by reading “intertextual” and “intermedial” as interchangeable.

2Indeed, one prominent adaptation scholar argues that the emphasis on intertextuality “seeks to dethrone the English Department’s traditional emphasis on literature, the existing canon that deserves close study and faithful adaptation, and replace it with literacy, the study of the ways texts have been, might be, and should be read and rewritten” (Leitch 2003, 332).

3Because scholarship about fidelity has historically included language that suggests that an adapter’s deviation from source material is a moral failing—a “violation,” for example—it is worth specifying what I mean by “ethics” here. In the tradition of the rhetorical approach shaped by Wayne Booth, James Phelan, Peter Rabinowitz, Martha Nussbaum, and Adam Zachary Newton, I see an audience member’s encounter with a narrative text as one that has an ethical dimension that necessarily includes judgments both about the story told (judgments about characters and events) and about the telling of the story (judgments about authorial purpose and narrative discourse). As such, I argue that the ethics of McEwan’s and Wright’s endings are inextricably bound to the medial affordances they work within; in other words, the ethical judgments that audiences make about Briony’s attempts at atonement are guided by the way McEwan and Wright use the limitations and opportunities of the medium in which they tell Briony’s story. However, I emphatically do not mean to judge Wright’s choice to adhere to or deviate from McEwan’s text as an ethical one, nor do I intend to inflect my use of the word “fidelity” with moral (or moralizing) connotations. My goal is to compare the way that Briony’s representation in two different media shapes two different ethical reactions from two different implied audiences, not to judge Wright’s filmmaking itself as an ethical act.

4It is perhaps a sign of Wright’s fidelity to his source material that these complaints echo reviews of McEwan’s novel. Caroline Moore called McEwan’s ending “frustrating” (2001, 12), and Margaret Boerner goes farther to say that “[i]n a kind of lunacy […] McEwan destroys the structure he has set up” (2002, 43).

7 As Briony records in the coda, her second draft—the first to include an account of her life—wasn’t written until seven years after Connolly’s letter; perhaps the ease and accuracy with which he imagined the truth caused this delay in a young woman not yet ready to face it.

9 Crosthwaite comments on this authorial crossover but is less interested in the ontological rupture caused by it and focuses more on the way “McEwan’s artistic predicament as well as the development of some of his (partial) solutions are inscribed into the text itself” (2007, 62).

11 This isn’t to say that Wright’s film isn’t intertextual. In fact, Christine Geraghty (2009) identifies a variety of cinematic references in the film, including David Lean’s Brief Encounter (1945), Carol Reed’s The Third Man (1949), Frank Laudner and Sidney Gilliat’s Millions Like Us (1943), and Basil Dearden’s The Bells Go Down (1943). This focus on identifying an adaptation’s intertexts aside from the source material is a typical example of the form most contemporary adaptation scholarship takes. The point here is simply that in the case of the film this intertextuality is evidence of Wright’s filmic sophistication, unrelated to the question of Briony’s prose authorship.

12 In this sense, these flourishes fulfill Bordwell’s decorative function of style. As opposed to denotative, thematic, or expressive functions, “style takes narrative denotation or an expressive quality as an occasion for exhibiting perceptual qualities or patterns”—in this case, the exhibition of cinema’s affordances (2008, 377).

14 Briony pointedly asks her sister, “What do you think it would feel like to be someone else?” In one sense, this is perhaps the fundamental question of any fiction, and is thus thematically appropriate at the beginning of what audiences will later discover to be Briony’s fictional attempt to feel like Cecelia and Robbie.

15 This is particularly true of Atonement’s UK audiences. A 2005 BBC study found that the hymn—an excerpt from John Greenleaf Whittier’s The Brewing of Soma set to a tune by Hubert Parry—polled second on a list of favorite hymns, after “How Great Thou Art” (“The nation’s favourite hymn”).

16 The New Yorker review of the film is similarly dismissive, calling the shot “overkill” (Lane 2007, par. 4). Lane also points out that the Dunkirk evacuation takes place five years after 1935, “not four, as the film innumerately tells us” (ibid.). Audiences quick enough to catch the mistake are likely to dismiss it as just that—a mistake—but it is possible to read this as an analogue both for Briony’s “thousand-ton bomb,” an error pointing towards her responsibility for the story (though not the discourse), as well as for the achronological appearance of Auden’s “In Memory of W. B. Yeats” discussed by Grmelová.

17 Though not critical to my point here, there is a long association between an audience’s experience of watching a film and the experience of dreaming, suggested as early as 1916 by Hugo Munsterberg and elaborated by Suzanne Langer (1953), Parker Tyler (1971), and Colin McGinn (2005), as well as given cinematic form in films from Sherlock Jr. (Keaton, 1924) to Inception (Nolan, 2010). If “a movie is a dream idealized […] a dream as we wish we had them,” then this link provides an opportunity for Wright to merge our narrative desires with Robbie’s romantic longings; like Robbie, we yearn for an ending that undoes the mistakes of the past (McGinn 2005, 168). As I argue below, it is precisely this desire that becomes the object of ethical critique in the coda.