Screen/Memory: Rape and Its Alibis in Last Year at Marienbad

Last year at Marienbad 1961

 [fig. 1, 2] 

Although it enjoyed a succes de scandale when it was released in 1961 and quickly became a cult film, Last Year at Marienbad, directed by Alain Resnais from a script by Alain Robbe-Grillet, is still considered a difficult film. This is largely because its story remains stubbornly undecidable: both verbal narration and visual montage are systematically disjointed, preventing the viewer from piecing together any single coherent narrative. Many archetypal stories are suggested—Cinderella, Orpheus and Eurydice, a bargain with death —but none is complete, and each subverts the others. This indeterminacy is quite deliberate: Resnais has said about the film that it is "open to all myths," and one critic has likened the film to a Rorschach blot.1 What can be said is that the film turns around a question, what, if anything, happened last year? The three principal characters are named in the script only as A (a woman, played by Delphine Seyrig) M (perhaps her husband or her doctor or…?), and mother man, X. The setting is an elegant spa or perhaps a sanatorium. X narrates almost the entire film, during which he attempts to convince A that the two of them had a love affair last year and that she had promised to go away with him. A does not consent to his amorous advances, however, and she also resists his story, claiming that they have never met before.

In spite of the film's many games and game structures, the critic's task is not simply to decide which of the competing plots "wins," any more than we are asked to judge which of the characters in Akira Kmosawa's Rashomon is telling the whole truth. Instead of weighing the relative merits of various understandings of the film, therefore, I want to add one significant meaning to those already proposed by Marienbad's many interpreters. Then I will investigate why that interpretation has been overlooked, explained away, or denied by critics and even foreclosed by the film itself. More specifically, I plan to argue that one of the film's potential plots is that of a rape and a cover-up.

In his introduction to the script, Robbe-Grillet calls Last Year at Marienbad "the story of a persuasion.”2 Although, Robbe-Grillet, like Resnais, invites critical attention by claiming that the film is open to many readings, his word "persuasion" unobtrusively places a limiting frame around possible meanings. Accordingly, critics have seen the film as a fantasy, a memory on a false memory), a game, a dream, an instance of hypnotic suggestion, or some combination of these. But what either took place or didn't is always a love affair. Marienbad remains a gentle and mannered story, and its hints of violence (aggression, fear, even pistol shots, either remain outside the frame of interpretation altogether or are locked into cliches of rivalry, adultery, and jealousy. While Robbe-Grillet's own statements have been this simplistic, I find the film itself much more subtle and challenging. Perhaps as a result of Robbe-Grillet's invitation, interpretations, while multiple, have not been infinite and in fact remain strictly circumscribed.3

Clearly, an approach that favors multiple over univocal reading can still exclude certain meanings, de-authorize some approaches. Moreover, within an ideology of polyvalence, criteria for limiting interpretation can remain invisible: a text can appear open-minded but still retain a frame of assumptions that excludes certain important meanings. When one considers that what might be excluded from Last Year at Marienbad is the possibility of rape, polyvalence itself begins to look less like an openness to interpretation than an extremely potent smokescreen. Such limitations on the proliferation of meaning are especially weighty in light of the fact that the same rhetorical maneuvers are used by X himself, by the film as a whole, and even beyond it, as I will show.

Viewers sympathetic both to feminism and to postmodernism must thread our way between the Scylla of univocal readings and the Charybdis of infinitely proliferating indeterminacy. The possibility of rape makes it especially urgent that we avoid both positions: a theocracy of a Single Truth is profoundly antidemocratic, on the other hand, real people (nearly always women) get raped, and they do not want to hear that rape is only one among an infinite number of possible meanings of their experience. When critics, a scriptwriter, and a character in a film want to persuade me not to pursue a reading, I can't help noting that the strategies used to keep viewers from seeing rape inside this text are those used outside it as well, which is why Marienbad is a particularly useful text for exploring the discursive binds into which anyone who wants to make rape visible is put. Examining the complex knot of images, themes, and narrative manoeuvres that give rise to the possibility of rape in Marienbad, we will thus eventually be forced to confront the highly problematic intersection of postmodern narrativity with feminist interpretation.

But first, I want to show how Last Year at Marienbad dramatizes strategies used in fiction, in criticism, and in life to deny the existence of rape and to create more acceptable, alternative, and unfortunately often more readily believed "alibi" narratives. The film does not tell (indeed, film cannot tell, according to Robbe-Grillet and other film theorists) what "really" happened in the past, but it does show how discourses about the past are constructed, suppressed, and rewritten. We will also be able to discern how power comes into play in the construction of such discourses.

Is There a Rape in This Text?

The first part of my title—"Screen/Memory"—refers, then, to the split between what occurs on the screen, now, and what may have happened in the past. This is the apparent divergence between fiction and truth, between the verifiable present and the reconstructions of memory. As spectators, we receive the film as a conflict among various versions of something that may or may not have happened "last year." Even this understanding is invalidated by Robbe-Grillet's and other film theorists' observation that filmic images are always experienced as present and that the past can only be evoked through the use of narrative conventions. Such theories shift attention to the present: whatever event exists is happening right now before our eyes.4 Similarly, except through arbitrary filmic conventions, internal mental or psychological states (such as intent, for example, or consent) cannot be portrayed. Since rape leaves no concrete, intersubjectively verifiable evidence to prevent the construction of multiple and contradictory narratives,5 rape is a perfect crime for a film.The specific difficulties of "proving" that what occurred was a rape, framed within the possibilities and limitations of filmic representation, add up to stage (even invite) the discursive disappearance of a crime.

Rewriting rape as another story is already part of the cultural discourses about rape and even in one of our most potent narratives for interpreting the past: psychoanalysis. In his controversial book, The Assault on Truth, Jeffrey Masson documents his discovery of Freud's abandonment, under pressure from the medical establishment and the public, of his "seduction theory" of hysteria.6 Early in his career (e.g., "On the Aetiology of Hysteria," 1899), Freud expressed his shock at the number of his female patients who remembered, under analysis, having been "seduced" as small children. Over a period of time and working this time from his observations of infantile sexuality, Freud concluded that his patients fantasized these childhood seductions. Freud and the story of psychoanalysis thus provide a prototype of the sort of narrative reworking I am identifying in Marienbad. Freud too revised the narrative of rape to call it seduction (romance, "persuasion") and then fantasy. And Masson's book retraces the steps in the erasure of rape from the text of psychoanalysis, with the subsequent emergence of another developmental narrative, the Oedipus complex and the theory of infantile sexuality on which all subsequent psychoanalytic theory depends. Feminists today are examining the narratives of psychoanalysis as themselves a sort of hysterical discourse, seeking the feminine perspective that has been repressed. Its peculiar ambivalence about rape may help explain why feminists have found psychoanalysis useful for reinterpreting the past while at the same time maintaining a careful critical distance from its discourse.

Without the slash, my title refers to another sort of screen: that of the "Screen Memories" of Freud's 1899 essay.' In that essay, Freud problematizes the interpretation of memories using the image of a memory screen whereon are projected fragments, metaphors, even inverons of actual events. As in his study of dreams, here too Freud argues that the past may be repressed, censored or transformed by memory and that it is only through reading the metaphorical discourses of the present that we can have access to an event. So while in film theory the screen can represent a memory, in psychoanalysis memory can be a screen or metaphor. In both cases, past experience is accessible only as a metaphor, mediated by rhetorical conventions of representation. So we must contend, hear and elsewhere, not with unproblematic representation of (a) rape, but with rape itself as rep-resentation.

The second part of my title, "Rape and Its Alibis," refers to my contention that Last Year at Marienbad can be seen as a rape story narrated by a rapist. According to such a reading, in the absence of any represented rape event, the film as a whole functions like an alibi. I am using the word "alibi" here in its etymological sense: as a claim to have been "elsewhere" at the time a crime was committed. Thus defined, alibi reveals its affinities with the metaphorical screens already mentioned. Like metaphor, alibi substitutes an image or narrative here and now that replaces and in some sense represents another virtual one elsewhere. I prefer the term "alibi," because it suggests that the substitution is motivated and that power (for our purposes the unequal distribution of power between men and women) is at issue.

What I am describing, then, is a pervasive phenomenon: in fiction and life, rape is a special kind of crime in relation to narrative It differs from other violent crimes in the kind of alibis it permits. To prove his innocence, someone suspected of murder must show he himself was elsewhere or that the murder was committed by another person. He can rarely claim that no crime occurred. Murder is not a crime whose noncommission can be narrated. Rape, on the other hand,e can be discursively transformed into another kind of story. This is exactly the sort of thing that happens when rape is rewritten retrospectively into "persuasion," "seduction," or even "romance." It happens, for example, in Jean Renoir's A Day in the Country, where we witness a rape and its subsequent reinterpretation by the young woman (and the film itself) into a nostalgically remembered romantic moment. A rape defense case can rest on the claim that what occurred was not a rape and so the question is not who committed the crime, but whether a crime occurred at all. It is thus not surprising that Rashomon —that prototypical examination of conflicting testimonies about a past event—is about a rape and not some other kind of crime.

Although no critic has taken seriously the possibility that rape is an issue in Last Year at Marienbad (in fact several have wound circuitous paths to avoid it),8 ironically, Robbe-Grillet's script actually includes a silent and "rather swift and brutal rape scene" (p.146), which he described elsewhere as "a rape, 'realistic' in the style of Punch and Judy, full of exaggeration and theatricality.”9 This scene was removed by Resnais during filming. Of course that excision makes it easier for me to argue that rape is an issue at all in a film where no actual rape occurs. But such a scene is not needed. Quite to the contrary, it is the way this absence/presence is orchestrated that gives Marienbad its postmodern resonance. And it is the absence of any rape event in the film that shifts the emphasis to discursive processes, furnishing a motive for X's narration. It is in this sense that the entire film can be seen as an alibi: not as reference to a hidden past truth, but as the creation of a story in the present—a story that excludes another story while inscribing it nonetheless. Rape is thus not the "secret" of what happened last year. Rape in Marienbad is neither remembered not forgotten; rather, it is shown. While not described, it is nevertheless inscribed, but rendered in comprehensible because fragmented and scattered about the film in inconsequential details, leaving a hole in the centre.10 We will thus do better to think of rape not as a past event but rather as a present threat, a possibility among others, a condition of meaning.11

In Marienbad, since X himself is present on the screen, narrating the film, it is the crime itself that is "alibi", that is, it is not where we might expect to find it in the story, but rather it is fragmented, displaced, metaphorized, repressed from the narrative, and ultimately inscribed "elsewhere" (but still, on the screen, in the present). If rape as event has been suppressed from the story, it is present as discourse, dispersed in multiple thematic codes. It is represented symbolically by a series of broken things: a glass, A's shoe, later a balustrade over which X escapes detection by M. It is present in a theme of penetration (into rooms, into thoughts). It can be seen in the fear A's face reveals, suggesting inner experience and memory; it is present in her repeated and increasingly frantic refrain: "I beg of you, leave we alone” It is there in the various manifestations of X's pursuit and A's flight. It is visible in the actress' self-protective gesture (arm held diagonally across her torso) that becomes her character's signature.

Sexual violence is also implicit in one juxtaposition suggesting that A has been shot: scenes of M and other men with pistols engaging M a marksmanship game cut to a shot of A lying crumpled on the floor of her bedroom. As the camera moves closer, we see her opened eyes and her finger placed coyly on her mouth. The combined effect of the two scenes is that of a violent event transformed into an erotic one. Similarly, violence is suggested, by A's startling and otherwise inexplicable scream when, in a series of rapidly intercut shots, she drops a glass simultaneously in a bar and in her bedroom as she hears X approaching. The fact that the scream seems to come from nowhere and X explained away as a vague "malaise" is especially suggestive in light of Lacan's post-Freudian view that hysterical symptoms reveal themselves as signifiers for repressed traumas by then seemingly excessive or misplaced (displaced) affect. Ironically, it is X's insistence on last year that serves as an alibi: his desperate and repeated efforts to turn the discussion to a "formerly" and an "elsewhere"—last year, perhaps in Frederiksbad or Baden-Salsa—avert attention from the discursive crime that is happening before our eyes: the reinscription of rape as a love story.

Eisenstein's theories of montage, with which New Wave film. makers experimented extensively, can help us understand how these examples work. According to Eisenstein, juxtaposition of images in montage permits dialectical emergence of meanings that are not present in any of the images alone.'[ So where Marienbad's story is most visibly a "persuasion," the juxtaposition of its images (in the rapid montage and multiple jump cuts for which Resnais is famous) creates a space in which fragments of an obscured rape narrative can be glimpsed and, more important, in which we can unmask the narrative procedures whereby rape is reemplotted as persuasion. Thus in Marienbad, we are able to see the mechanisms whereby rape is always put elsewhere and made unrepresentable. It is always other to Truth, always alibi.

The most obvious moment of revision occurs when two alternative versions of a scene are juxtaposed. Each of the sequences apparently follows from X's ominously symbolic earlier statement "I penetrate into your bedroom" (p. 110, my translation). In the first version, X advances along a corridor toward A's bedroom. Hearing his approach, A retreats, seems trapped, makes self-protective gestures, is obviously frightened (see figure I). Then—cut—the scene begins again. In a crescendo of desperation, X cries, "No, no, no! (violently:) That's wrong.... (calmer:) It wasn't by force" (p. 147).13 This time, in a series of overexposed and rapidly cut shots, A advances to meet X, whom she greets with a smile and opens arms (see figure 2). Fear is thus re-imaged as welcome, terror as joy, and rape as romance.

What is extraordinary here is not that the story is revised—that a brutal approach and metaphorical "penetration" into A's room are recast as a warm welcome and that, even in the initial scene, resistance is depicted as erotic. In everyday parlance, no means yes.)

What is extraordinary is that both versions coexist in the finished film. Instead of the product of revision, then, what we are given is the process itself. But of course the second scene does not replace the first. Watching the second version, we, as spectators, remember what happened last time. We are witnesses (accomplices?) to the construction of an alibi.

It can be and has been claimed that the idea of a rape in the film is controverted by the charm and seductiveness of X. Okay, maybe a rape is theme here, but does he rape her or does she rather desire or imagine it? The apparent absence of any motive for violence on X's part can be invoked in support of this claim, as can A's hysterical fear or her imputed guilt feelings at desiring or having sex with him. It has even been argued that we see the mere film through A's point of view and that the story occurs inside her head, that the film is her mindscreen.”14

This is not a defensible interpretation. While we do occasionally enter into A's visual mindscreen, it is crucial to notice, as critics have failed so far to do, that throughout the film both narrative and visual authority are clearly male. In the scenes described above, X is both the narrator and the angle of vision through which we know A. As can be seen in figure 1, the view from behind his head conflates the camera, the character, and the spectator, implicating all three in the revision of the scene and the production of an alibi narrative.15 The film is narrated exclusively by X. A is an object of vision, of exchange, of desire, and of narrative. In one scene, for instance, A lies on her bed idly arranging photographs of herself in a configuration that repeats that of a game in which X and M engage throughout the film. She thus explicitly displays herself as the token in their game, thereby contributing to her own construction as object of exchange between men. She is never a subject of discourse except to voice her own rejection of the story X proposes. In treating her scream as a "malaise" and herself as an invalid, the characters and the film as a whole "invalidate" her subjectivity and her point of view.

A's desire for narration and interpretation—in short, her manner of reading or seeing—are also invalidated. Here again, the film thematizes the issue, inscribing its own viewers. A statue in the castle's garden portraying a man and a woman is an important mise-en-scene (or interior duplication of the text's own functioning) because it is the excuse for a disagreement between X and A about interpretation.Where absence of event highlighted discursive processes, absence of confirmable knowledge (of the statue, of the past) spotlights interpretation. X arouses and exploits A’s curiosity about the sculpted figures' identities because he knows that the goal of interpretation is to prolong itself: "To say something, I talked about the statue. I told you that ..." he tells A, and then he proceeds to elaborate multiple alternative explanations of the statue. A is a different sort of reader; she wants to know whom the stone figures represent, and she makes several suggestions, as reported by X: "Then you asked me the names of the characters. I answered that it didn't matter.—You didn't agree with me, and you began giving them names, more or less at random, I think ... Pyrrhus and Andromache, Helen and Agamemnon.... Then I said that it could just as well be you and I ... (A silence.) Or anyone" (pp. 63-65). X's in. response to A's desire to know is his modus operandi: "Don't give them any name ... They could have had so many other adventures," he says.

The statue's importance as an interior duplication or meta-commentary16 derives from X's desire for multiple meanings, an approach to interpretation that clearly applies to the whole film. Several points are worth noting about this important scene, because they open a distance between the film and the viewer's options for interpretation. First, there is the paradox, central to the film, that X's discourse is double, even duplicitous. He is telling A what supposedly happened last year and describing their conversation as it happens: his narration is simultaneously descriptive and performative. When A laughs, for example, X incorporates her laughter into the "remembered" scene, a manoeuvre that puts his past tense in doubt and suggests he is inventing the past from the materials of the present. It is thus also clear that X's method of interpretation is not innocent. His (and the film's) continued existence and his project of persuasion depend on persevering in his dialogue with A. To determine a final meaning for the statue would put an end to the possibility of discussing it (as indeed M does, when he intervenes to "identify" the statue, and might even arrest meaning in an interpretation inimical to X's motives and desires. A, who has less to lose from definitive interpretation, suggests two possible identities for the couple depicted by the statue: Pyrrhus and Andromache, Helen and Agamemnon. Her suggestions are not chosen "at random," as X testifies; each of the couples is a case of abduction and rape. In addition to being a characteristic of the text as a whole, resistance to definitive interpretation is therefore also one of X's most powerful strategies. This is where the text's polyvalence ceases to be neutral and starts to look like a cover-up. X doesn't care what the interpretations are, as long as they are multiple. He has much to gain from a method of reading that prefers (almost) infinite semiosis. Infinite minus one. Moreover, critics have colluded with X by assuming him to be a reliable narrator. Where Marienbad shows us how rape can be discursively deconstructed by the very character who has the most to gain from such a strategy and where the woman's perspective is systematically invalidated is where postmodern narrativity threatens to become problematic, even antithetical to feminist interpretation. Last Year at Marienbad constructs an incompatibility between feminine subjectivity and the plural text.17

Without sacrificing the openness and polyvalence of the text, we might begin to work our way out of that impasse by remembering to ask: plural for whom? and against whom? What has been consistently overlooked in discussions of Marienbad and what makes the film extremely useful for exploration of this problem is the fact that here interpretation is clearly gender-coded. X's approach, the deconstructive one, the one the text itself prefers, is both positive and masculine. It takes the initiative in presenting itself as seductive and desirable. A's approach, on the other hand, is both retrograde and coded feminine; it is what needs to be mastered or overcome in order to accede to modernity. Borrowing Roland Barthes' terminology, we could say that X sees the statue as a "writerly text," A sees it as a "readerly text." In a conflict between two kinds of readers, what is important is not that A is a bad reader, but that bad reading is coded feminine. The woman's reading is thus easily invalidated. Rape, here, constitutes a kind of limit to postmodern interpretation, making it impossible for the feminist critic to say that these gender implications are coincidental.

Moreover, in an epistemological conflict about how the past can be known and indeed what can be known about it, the dispute between A and X about interpretive strategies must also be seen as a conflict between the text and its viewer. Marienbad thematizes its viewer via mise-en-abyme and filmic suture so that it becomes a thoroughly self-reflecting text. Thus questions of X's authority are also questions of the film's and Robbe-Grillet's and the New Novel's and New Wave's (relationship to its public. Critical arguments that erase rape from the film are exactly those used to invalidate real women's accounts of rape: It could have been many other things. Are you sure? Are you sure you're sure? 19 Rape becomes an interpretive "malaise": cared for by the authority of her doctor/husband, A, along with her experience and her point of view, are, as noted above, invalidated.”20 A's own lack of authority cancels out the possibility and invalidates the desire for definitive knowledge. The viewer is caught in a contradictory identification, thematically with A in her frustrated desire for understanding, but filmically with X as we see through his eyes and narrative. Viewers sympathetic both to feminist analysis and to postmodernism find themselves in a double bind. To read like (with) X is to collude with him in erasing the possibility of rape from the film. To read with (like) A is to reject postmodern textuality and interpretation.

Ultimately what is at stake for X and A, for the film as a whole, for (feminist) criticism, and even in rape cases outside the cinema is narrative authority and how it is engendered.. Discussions of narrative authority have rarely taken gender into account and just as often have bracketed questions of power. To take one of the most productive recent examples, Ross Chambers' Story and Situation: Narrative Seduction and the Power of Fiction describes "narrative seduction" as that which keeps the reader or viewer interested. In terms strikingly similar to Robbe-Grillet's, Chambers proposes that a text's seductive power

is definable as the power to achieve authority and to produce involvement (the authority of the storyteller, the involvement of the narratee) within a situation from which power is itself absent. If such power can be called the power of seduction, it is because seduction is, by definition, a phenomenon of persuasion: it cannot rely on forcer institutional authority ("power"), for it is, precisely, a means of achieving mastery in the absence of such means of control.22

The alternative to seduction is thus boredom: loss of interest on the part of the reader/viewer and loss of authority on the part of the narrative.

But it would seem somewhat dubious to posit, even hypothetically, an absence of power; neither in fiction nor in life do texts and narrators exist in a power vacuum (and power differences between the sexes is a form of "institutional authority"). Rather, it seems to me that the other mentioned alternative has to be reckoned with: authority ("persuasion," seduction) is an alternative to its lack, to be sure, but given the existence of power differences, failure of seduction can also lead to force. It is interesting that in a discussion of persuasion, authority, and seduction in and of narrative, Chambers never examines rape as a conceptual category. Again, we must ask: persuasion (seduction) for whom? The two polar alternatives to seduction are failure (impotence?), as Chambers argues, and also rape. And so like X, the filmmaker has to deny (or excise) the possibility of force—"No, no, no. That's wrong. It wasn't by force."—in order to maintain the film's authority as an interesting text.

But at another level, Chambers' approach is perfectly correct. The more a text has only itself to rely on (in Chambers' terminology, "narrational authority" and not simply the "narrative authority" of conveying information), in other words, the more it depends on fascination, seduction, and persuasion, the more rape is impossible to represent. Or the more rape comes itself to represent a failure of narrative authority and persuasion and seduction. (I will return shortly to the use of rape as metaphor.) Narrative and other seductions depend on the distancing of boredom, but also on the distancing of force. My essay has thus been about the consequences and the necessity of that absence of rape from narratives like Last Year at Marienbad. I am now compelled to argue that a text (or film) that wants to remain self-conscious and self-reflecting cannot portray rape; it must speak in terms of seduction and persuasion because it is a text. Such texts (and culture itself to the extent that culture is composed of self-conscious narratives) depend on denying or repressing any awareness of rape. This is why feminism must take post-modern textual self-awareness into account. As a first step, we must turn our attention from the text's power to represent to its power to un-represent. In this light, A's desire for definitive interpretation and X's multiplication of interpretations are ironic. Neither character favors truly polyvalent interpretation: A proposes only two identities for the statue (Pyrrhus and Andromache, Helen and Agamemnon), but X privileges a single reading by excluding it. (“No, no, no. That's wrong. It warn't by force.")

If, as Craig Owens claims, postmodernism is primarily to be understood as "a crisis of cultural authority" and a loss of seductiveness and mastery,23 Marienbad's self-referential aspects and its deconstruction of all claims to authority foreground the postmodern character of the film. If we add to this list of metaphorical losses a loss of potency—both sexual and narrative—we can begin to understand why both rape and its absence become necessary elements in the struggle. We can then see another of Marienbad's plots as being postmodernism's attempt to (re)establish authority )prove its masculinity). X wants to prove both that he could have raped A and that he didn't, just as someone who has an alibi could have committed a crime but didn't. Similarly, postmodern texts want to foreground not an inability but a refusal to master: 'The story of a persuasion."

Postmodernism is also characterized by what I called hysterical discourse above, after Freud's explanation, in his analysis of Dora, that one of the defining characteristics of hysteria's "an inability to tell coherent stories" because of what has been repressed.24 It is not by reducing such stories to readability, however, but only through reading their very incoherence that their particular rhetoric can be recognized. If, as some have claimed, Last Year at Marienbad is one of the earliest important examples of postmodemist fiction,25 it is also another instance in which culture in this case postmodern culture) is founded on the repression and rewriting of an act of rape.26

As I have just demonstrated, rape can easily become a floating signifier, available for the elaboration of metaphor. Marienbad has been seen through many metaphors: as the story of the film's seduction of its viewers, of the specificity of filmic narrative, of voice and its mastery over the image, even of the relationship between filmmaker and scenarist. In my turn, I am reading the film as a drama of narrative mastery and control, of postmodernism's mastery over modernism, and (as I am showing elsewhere) of France's identity crisis over Algerian independence. In all these scenarios, Marienbad is the story of a threatened failure of (narrative) authority and how easily it might become force. This accounts perhaps for an impression frequently held by readers/viewers of postmodernism that we are being assaulted by the text.

It is also possible that the ease with which such metaphorization occurs is one of the causes and consequences of the phenomenon I am describing. One of the most foolproof ways of hiding something, after all, is to call it something else, so that while rape metaphors can be used to illuminate both terms compared (as I hope I am doing here), if the metaphor is automatically and from the start a catachresis—that is, if rape is never seen as literal at all—rape metaphors themselves can serve, paradoxically, to cover up rape. Critics, busily constructing alibis along with the film itself, can thus all too easily buy into the claim that real meaning is elsewhere. In other words, in life, fiction, or criticism, rape can be rewritten by those who have the narrative authority to do so.

I want to conclude by asking whether, by ostentatiously pointing to what is absent and by its visible revisions, Marienbad can be called a liberating film, even a feminist film. I don't think so. The film's positioning of the spectator is consistently complicitous with X's narrative perspective. The questions we are invited to ask about it resemble but fall short of moving beyond the ways we are encouraged to respond to rape in life: Is she leading him on? Was her door open? Is she denying a sexual encounter by calling it rape? If there was an event last year, all the film itself proposes was an encounter that A refuses to remember. In short, the film repeats the myths and representations of rape in nonliterary discourse, but does not demystify them. In fact, it offers the spectator the same old alibis.

The task of demystification is still left to the feminist viewer, who must first identify rape as rape (not simply and automatically as persuasion or seduction or metaphor), second begin to untangle the discourses that surround it and contribute to its disappearance, and third show how an interpretive strategy that valorizes multiple readings itself needs to be interpreted. We cannot claim that Marienbad is a rape story to the exclusion of other stories, but only how the other stories serve to cover up rape by rewriting it. And perhaps show how easy it is to hide rape, and thus how dangerous.

So then what of the dilemma of the critic caught between the postmodern text and feminist interpretation? Are the two, as Alice Jardine worries, oxymoronic?27 If so and if keeping rape literal is our feminist goal (as it most be in nonliterary life), we have to be anti-postmodern. If, on the other hand, is is in our feminist interest (and I think it is) to understand how rape is metaphorized, becomes discourse, even a(nother) founding event of (postmodern) culture, then there are many serious questions we can ask of texts like Marienbad. We need to notice, for example, how mastery and authority are so often coded masculine and what is to be mastered (through authority, seduction, or violence) as feminine. We can imagine ways in which they could be coded feminine or not gender-coded at all. We might examine in what ways limiting frames can be imposed and meanings can be excluded even in texts acknowledged to be polyvalent or indeterminate. And if it is in our interest to know how rape can be un-represented, either through denying its existence entirely or through causing it itself to represent something else, then we had better take an interest in postmodern textuality.

Lynn A. Higgins

Notes

I am gainful to Carla Freccero, Marianne Hirsch, Lawrence Kritzman, Albert LaValley, Amy Lawrence, Neal Oxenhandler, Brenda Silver, and Steven Ungar, who read versions of this essay and offered valuable suggestions.

1. Andrew Sarris, Interviews with Film Directors (New York: Avon, 1967) p. 436 ( John J. Michalcryk, The French Literary Filmmakers (Philadelphia: Art Alliance Press, 19801, P. 111.

2. Alain Robbe-Grillet, Last Year at Marienbad, trans. Richard Howard New York: Grove Press, 1962), p. 10. All quotations here are from the English translation, unless otherwise noted. The script of L'Année dernière à Marienbad was originally published as a "cine-roman" (Paris: Editions de Minuit, 1961).

3. A few examples will give a sense of the astonishing variety of interpretations. Bruce Morrissette, Les Romans de Robbe-Grillet (Pans: Editions de Minuit, 1963), suggests that the principles of hypnosis provide one way to explain the characters' behavior. Francois Weyergans describes A's internal conflict between the pleasure principle and the reality principle: "Dans le dedale," Cahiers du Cinema 21(123):22-28. Claude Ollier, in "Ce Soir a Marienbad," Nouvelle Revue Francaise (October and November 1961), nos. 106/107, pp. 711-719 and 906-912, analyzes a similar struggle between reason (in the person of M) and irrational obsession (X). James Monaco decides that ultimately the film is "about storytelling. It is X's job to convince, A's job to resist: the primal relationship between storyteller and audience": Alain Resnais: The Role of Imagination (New York: Oxford University Press, 1978). Jean-Edern Hallier sees the film as a story about the desire for immortality, with M playing the role of Death: "Toute une vie a Marienbad," Tel Quel (1961), 7:49-52. These and other critics generally concede that the film invites multiple interpretations, of which their reading is only one. An exception is John Ward's Alain Resnais, or the Theme of Time (Garden City, N.Y.: Doubleday, 1968). Ward declares that "what exactly took place the year before" is that "M kills A and X is left alone to mourn" (P. 39).

4. Robbe-Grillet, For A New Novel: Essays on Fiction, trans. Richard Howard (New York: Grove Press, 1965), pp. 152-153. See also Roland Barthes, Image, Music, Text, trans. Stephen Heath (New York: Hill & Wang, 1977), pp. 15-51.

5. For a fuller discussion of this problem, see Frances Ferguson, "Rape and the Rise of the Novel," Representations (Fall 1987), 20: 88-112.

6. Jeffrey Moussaieff Masson, The Assault on Truth: Freud's Suppression of the Seduction Theory (New York: Farrar, Strauss & Giroux, 1984).

7. Sigmund Freud, "Screen Memories," Collected Papers, ed. James Strachey (London: Hogarth Press, 1950), 5:47-69.

8. For example, Allen "Miller, The Cinematic Muse: Critical Studies in the History of French Cinema (Columbia: University of Missouri Press, 1979,) p. 174, identifies A's room as "the locus upon which the narrative quest is centered, for it is here that the full range (my emphasis) of hypotheses are developed, ranging from rejection to death to joyous acceptance." Ward claims that X imagines several possible endings, including suicide, accident, and rape (p. 50). On the other hand, in his introduction to the screenplay Robbe-Grillet speaks of "fantasies of tragedy in the heroine's mind: rape, murder, suicide ..." (p. I I). In these and other examples, rape is mentioned briefly, if at all, and then never resurfaces. A second kind of circumvention is more interesting: rape appears just as fleetingly, but already metaphorized from the start. Gaston Bounoure's Alain Resnaise (Paris: Seghers, 1974,) for example, cites Robbe-Grillet as saying that X introduces the past by force into a closed world ("quant au passé que le heros introduit de force dans ce monde clos... ," p. 75). And Thiher's entire analysis is based on an unexamined metaphor that would see the film as "a seduction ... not only of the unknown woman whom the narrator pursues throughout the film, but also of our vision" (p. 166). Finally, Robbe-Grillet is reported to have said, enigmatically, "I would describe my relationship with Resnais as the rape of Resnais by Robbe-Grillet". Melinda Camber Porter, Through Parisian Eyes: Reflections on Contemporary French Arts and Culture (New York: Oxford University Press, 1986), p. 83.

9. Andre Gardies, Alain Robbe-Grillet (Paris: Seghers, 1972,) p. 118 (my translation). Note that Robbe-Grillet's comment is already a first rewriting —of a "brutal" rape into a comical one.

10. Robbe-Grillet is well known for generating ambiguities by means of a gap in the story. For example, his 1955 novel Le Voyeur (Paris: Editions de Minuit) revolves around a brief lapse in the protagonist's memory and his fear that he has committed or will be accused of the rape and murder of a young girl. For an interesting discussion of alibis in Le Voyeur, see Jeffrey Kittay, "Alibi: On Handwriting, Rewriting and Writing Rhythms and Le Voyeur," Romanic Review (1980,) 7111): 57-74. About Marienbad Robbe-Grillet said in an interview: "What happened—if something did happen once upon a time—constantly produces sort of a gap in the story.... Everything, up to the 'hole,' is told—then told again after the hole—and we try to reconcile the two edges in order to make this annoying emptiness disappear. But what happens is the exact opposite: it's the emptiness that overruns, that fills everything" (Sarris, Interviews, p. 451).

11. Is it necessary to point out that this is how every woman is socially positioned? For a discussion of how rapability defines women, see Dianne Herman, "The Rape Culture," in Jo Freeman, ed., Women: A Feminist Perspective pp. 20-38 (Palo Alto: Mayfield, 1984), and Catherine A. MacKinnon, "Feminism, Marxism, Method, and the State: Toward Feminist Jurisprudence," Signs: Journal of Women in Culture and Society (1983), 8(4):635-658.

12. Sergei Eisenstein, Film Form: Essays in Film Theory (New York: Harcourt Brace Jovanovich, 1949), p. 49.

13. Earlier in the film, X calmly explains, "... finally ... I took you, half by force," and then, "Oh no ... probably it wasn't by force. […] But you're the only one who knows that" (script, pp. 115-116).

14. The term is from Bruce Kawin, Mindscreen: Bergman, Godard, and First-Person Film (Princeton: Princeton University Press, 1978), p. 82.

15. In fact, the scene illustrates perfectly the way the cinema constructs the viewer as male, as theorized by Laura Mulvey, "Visual Pleasure and Narrative Cinema," Screen (Autumn 1975), 16(3) and E. Ann Kaplan, "Is the Gaze Male?" in Women and Him: Both Sides of the Camera, pp. 23-35). New York: Methuen, 1983).

16. Resnais has called the film a "documentary about a statue" (Sarris, Interviews, p. 451).

17. This is, of course, a problem that is frequently discussed by feminist critics: I am not the first to point out how feminism's construction of the female subject conflicts with postmodern deconstruction of the subject. See, for example, Alice Jardine, Gynesis: Configurations of Woman and Modernity (Ithaca: Cornell University Press, 1985). Craig Owens' view that feminism is part of postmodernism is helpful, as is his description of feminism's challenges to the reassuring stability of (male) mastery of meaning. See his "The Discourse of Others: Feminists and Postmodernism," in Hal Foster, ed., The Anti-Aesthetic: Essays on Postmodem Culture, pp. 57-82 (Port Townsend, Wash.: Bay Press, 1983).

18. Roland Barthes, S/Z (Paris: Editions du Seuil, 1970).

19. Ferguson, p. 89.

20. Think of Charlotte Perkins Gilman, The Yellow Wallpaper (Old Westbury, N.Y.: Feminist Press, 1973).

21. I use the term "engendered" to mean both "generated" and "given gendered meanings," following Teresa de Lauretis, Technologies of Gender: Essays on Theory, Film, and Fiction (Bloomington: Indiana University Press, 1987).

22. Ross Chambers, Story and Situation: Narrative Seduction and the Power of Fiction (Minneapolis: University of Minnesota Press, 1984), p. 212.

23. Owens, "The Discourse of Others," p. 57 and passim.

24. The specific narrative difficulties are relevant as well: in hysteria "communications run dry, leaving gaps unfilled." Sigmund Freud, Dora: An Analysis of a Case of Hysteria (New York: MacMillan, 1963), pp. 30-31. See also note 10 above.

25. Thiher, The Cinematic Muse, p. 170; Kawin, Mindscreen, p. 198.

26. Scholars in a wide variety of fields have explored other instances of founding rapes. For example, much feminist analysis has been devoted to rereading the rape of Lucrece as the founding event of the Roman Republic. See, for example: Coppelia Kahn, "Lucrece: The Sexual Politics of Subjectivity," in this volume; Nancy J. Vickers, "This Heraldry in Lucrece's Face," in Susan Rubin Suleiman, ed. The Female Body in Western Culture: Contemporary Perspectives, pp. 209-222 (1985; reprint, Cambridge: Harvard Univer-sity Press, 1986); and Stephanie H. Jed, Chaste Thinking: The Rape of Lucretia and the Birth of Humanism (Bloomington: Indiana University Press, 19891.

27. Jardine, Gynesis, p. 22.