Introduction

Film in the First Person, Film as “Act of Love”

The film of tomorrow appears to me as even more personal than an individual and autobiographical novel, like a confession, or a diary. The filmmakers will express themselves in the first person and will relate what has happened to them: it may be the story of their first love or their most recent; of their political awakening; the story of a trip, a sickness, their military service, their marriage, their last vacation . . . and it will be enjoyable because it will be new and true. . . . The film of tomorrow will resemble the person who made it and the number of spectators will be proportional to the number of friends the director has. The film of tomorrow will be an act of love.¹

In 1957, Truffaut, then a young film critic at the journal Cahiers du cinéma, soon to be a young filmmaker, wrote a manifesto intended to launch a new era of filmmaking and film criticism. The quotation above comes from that essay, a piece as prophetic as it is enigmatic. There are several points worth pursuing in this short passage, particularly since Truffaut’s vision of the “film of tomorrow” turned out to be more than the overwrought outpourings of an impassioned young artist. His ideas seem to have struck a spark and taken fire with the other critics, reviewers, filmmakers, and filmgoers who would collaborate to create the art-cinema movement. The “film of tomorrow” he envisions, then, becomes the film of today, or at least a particular genre of film today: art cinema, films directed by recognized film authors. This genre depends upon precisely the idea of a person—an author—and claims the author’s preeminence just a decade before Roland Barthes proclaims the death of the author, in a famous essay of 1968.² What Truffaut describes here is a kind of filmmaking in which the author communicates something of himself to his audience, projects a self, a life, and creates a relationship.

For a moment, in this context, it is important to consider Barthes’s argument because it emerges from the political situation of 1968 as an antiauthorial (antiauthoritarian) statement: readers, Barthes argues, must liberate themselves from the tyranny of the idea of an author’s authority over textual meaning. Meaning, he maintains, is something produced not by the author, nor by the author’s text standing alone, but by the reader. Release from textual authority in Barthes’s argument runs parallel to liberation from religious (scriptural) and political authority. More specific to literary criticism is Barthes’s injunction to scholars to leave behind centuries of rooting in authorial biographies for “clues” to the meaning of the text. The biological author of the text is dead (if not literally dead, then at least absent and usually unavailable for comment). His remains (in the form of his recorded life’s events) do not constitute a relic that might be consulted by readers in the quest for true meaning. But Truffaut (writing nearly a decade before Barthes’s essay) introduces an interesting twist to the status of the author in relation to the reader (or in this case, the viewer). His author is imagined as very much alive and in possession of a life story that is constructed in relation with the viewer. As “friends,” the author and the reader are imagined as always already understanding one another, perhaps in an act of mutual recognition: the emblematic events Truffaut describes as the stuff of the “film of tomorrow” (first love, military service) could potentially be instances in which the spectator would see not only the author’s life but also his (or her, in some cases) own. But more than that, Truffaut argues for a form of address in film that begins with an “I,” the first person, and that “I” is identified with the author. One might respond to Truffaut’s declaration by saying that it merely revives a Romantic, reactionary notion of authorship as creative genius, inviting the spectator into the web of biographic investigation. But I would argue that his idea of the spectator as friend and cinema as an act of love leads in a different direction, one that reflects the collaborative nature of filmmaking itself. By holding fast to the idea of “I” within cinematic discourse, Truffaut introduces the possibility of a collaborative, interactive selfhood, with the cinematic screen offering a field for its projection. The reader is born not through the death of the author but, as the cinematic author tells his story, projects a self.

The film critics and directors of the Cahiers du Cinéma of the 1950s and ’60s, and the generation of critics that follows their lead, develop the concept of film authorship in a way that make acts of cinematic self-projection a possibility, and for art cinema, even a necessity. The art-cinema movement (supported by a scholarly and popular critical apparatus, national cinema funding in Europe, and art-house distribution) promoted in viewers an understanding of narrative film as an authorial project. As a result, a recurring aspect in the work of some art-house directors is explicit autobiographical reference, accompanied by meditations on film authorship, self-representation, and, ultimately, what it means to be or have a “self.”

When one reads Truffaut’s manifesto with care, it seems that he is talking about some kind of cinéma vérité, with a first-person voice-over narrative explaining footage of family vacations or marital struggles. How else can one understand his idea of confessional/diaristic film that is “more personal” than an autobiographical novel? How does one interpret the notion of film “expressed in the first person,” unless through the vehicle of the voice-over narrative? And what could it mean to say that a film “resembles” the person who made it? In what way does a film resemble a person who may or may not appear on the screen? And finally, the most intriguing question: Does Truffaut really mean that the film’s spectators and the director’s friends are one and the same? That the film, in addressing the director’s friends, performs an act of love?

To begin to think about these questions, we can return to the film Truffaut directed, once he had delivered his proclamation. Truffaut has described The 400 Blows as an episode from his childhood, in which (to summarize roughly) he was truant from school, stole a typewriter, and was subsequently shipped off to a detention center for juvenile delinquents. The release of the film caused an uproar in Truffaut’s family; his mother and father reacted so strongly to what they felt were negative depictions of their family life that Truffaut agreed to publish, on the day of the film’s opening, a denial that its material was autobiographical. Up to that point he had said frankly in interviews that the film depicted events from his childhood. In order for viewers to understand nondocumentary films like Truffaut’s as autobiographical, then, there must be someone who is understood as the author of the film, and, in most cases, there must be significant knowledge of what literary critic Gérard Genette calls the paratext.³ A paratext, broadly put, consists of elements that surround and inform the body of a text, either in the materials accompanying it (in the case of film, credits, dedications, epigraphs, etc.) or sources that pertain to it (interviews, commentary, critical reviews or essays, etc.).⁴ An important paratext for Truffaut’s work generally is his biography, where we find the following comment about The 400 Blows: “Though the narrative was autobiographical down to the last detail, Truffaut wanted to present it as a fiction.”⁵

I think it important to consider the idea that “Truffaut wanted to present [his story] as a fiction.” The autobiographical form that art cinema transmits is necessarily fictional, for reasons I will explore more thoroughly below. It would seem on the face of things that Truffaut’s insistence in his manifesto that the “films of tomorrow” should be “true” runs afoul of his intention to create fiction. But in this instance, one needs to distinguish between the factual and the true. Fiction provides the field for mimesis, or “resemblance,” as in “the film of tomorrow will resemble the person who made it.” And the creation of mimesis in cinematic terms involves relationship, both between the Paris of Truffaut’s childhood and the Paris of the time of the film’s making, and between the actors in the film and the people of Truffaut’s childhood. The work of creating a self-image will take place through resemblance-relationship, and the first relationship, we might say, exists between the director and his other. Instead of documentary footage, there is an actor, Jean-Pierre Léaud, who is given the task of resembling the young truant through the vehicle of the character Antoine Doinel.

For some theorists of autobiography, it is precisely the insertion of a figure like Antoine Doinel, played by a young actor who is not (yet in some sense resembles) Truffaut, that complicates the matter of cinematic autobiography. Philippe Lejeune set a benchmark for autobiography studies in the 1970s with his insistence on an “autobiographical pact,” a kind of contract between the autobiographer and the reader that promises an absolute identification between the writer, the person whose name is given on the title page, and the subject of the autobiography.⁶ In a subsequent attempt to define a cinematic form for autobiography, Lejeune leaves fictional films aside and turns instead to the documentary form, narrowing his view to those films that focus the camera on the filmmaker at the time of narration. The filmmaker may then relate episodes from his or her past that appear framed within the documentary as public or personal archival footage.⁷ Given the photographic foundation of film, it is clear why Lejeune feels he must craft his definition in this way. If a cinematic autobiographer does not use archival footage but instead stages an event from the past in which, for instance, the filmmaker appears as a child, it is clear that the photographed subject is not the same as the autobiographer’s past self. Consequently, nondocumentary film cannot help but break Lejeune’s autobiographical contract.

There are indeed documentary films that satisfy Lejeune’s requirements; one thinks of films like Ross McElwee’s Sherman’s March (1986) and Jonathan Caouette’s Tarnation (2003), to name just two. (See Jim Lane’s Autobiographical Documentary in America from the 1960s to the Present for a detailed history of the genre.)⁸ And Truffaut’s manifesto would seem to be in perfect alignment with Lejeune’s requirements, but we see in The 400 Blows that when he makes a film, he interprets these requirements rather differently, choosing fiction over documentary. One might argue, in fact, that the films in my study are allied most closely with autobiographical fiction, and that a category for discussing these films thus already exists in the realm of literary theory. But my work on photography and autobiography leads me to view cinema as a special case; it is in fact precisely photographic representation that poses a problem for Lejeune and opens up an avenue of discourse for me.⁹

In Picturing Ourselves: Photography and Autobiography, I argue that the invention and proliferation of photographs led to new ways of thinking about and representing selfhood and life histories. A few points from that book bear repeating in this context because film, after all, is a photography-based medium (I am leaving aside for the moment the question of digital cinema). The first is the issue that creates problems for Lejeune, namely, that photographs are a trace of a person’s physical presence in space and time. But because each photograph represents a specific place and time, it cannot stand for the entire person across time and space, as the pronoun “I” is made to do in textual autobiography. Lejeune’s autobiographical pact argues for identification between the author, the narrator, and the protagonist; photographs insert a wedge between the personae. One of the outcomes of this separation of selves is the possibility of regarding the self as another, in a material way. So the photograph is both a true trace of a person and the true sign of the present self’s separation from its past.

If we reflect on what this might mean for cinema as a medium of self-projection, a couple of ideas emerge: first, cinema offers the possibility of creating the illusion of presence. In the moment that we watch a film, the figures on the screen are present to us as people who have lived (Barthes, in Camera Lucida, says that they have been there). Ing-mar Bergman (who is not alone in this observation) writes that he feels he is watching ghosts when he watches a film. The presence on the screen is spectral (a shade, a trace, a shadow) of something that was material. Cinema offers the opportunity to reenter (with some of our senses) a reality of the past, in other words, and in that sense it produces the potential for a powerful addition to autobiographical discourse, which attempts to re-create a past world in language. Unlike language, a photograph seems to give us the world (or at least some aspects of the world) with a kind of immediacy and traces of physical substance. Theorists of textual autobiography have pondered the connection of language to world, making it clear that one of the significant engines driving autobiographical discourse is the conviction in the possibility of that connection.¹⁰ From my point of view, photography both strengthens and complicates the connection between the world and the represented self, while film, adding the qualities of “motion” and sound, pushes us further in seeing some direct link between world and representation. In the films I will discuss, we hear the voice of Ingmar Bergman, we see Truffaut and Almodóvar and Allen. Two of these men are dead. And all of them will be. But their recorded images and voices continue to reach audiences and will presumably continue to reach audiences when all of them are dead. This is perhaps the power that draws certain filmmakers into an act of self-projection.

At the same time, the films of these directors express awareness of the fatal flaws in this idea. Cinema does not give us the world as it was or any immediate “reality.” The photographic property of allowing a person to see him or herself as an Other finds full expression in cinema, where the film author works collaboratively, often with a host of collaborators, to produce self-projection. If film is imagined as an act of self-representation, that self disperses and takes up residence in the bodies and senses of others (actors, technicians, audience). Still, this dispersal offers an alternative model for thinking about selfhood and self-representation, which is why I part ways with Lejeune on his notion of what an autobiographical cinema might be. Narrative cinema, fictional cinema, and, in particular, cinema produced by directors who are identified as “authors” create a platform for exploring the collaborative and compromised and polyvocal act of self-creation. Theorists of textual autobiography such as Susanna Egan, G. Thomas Couser, Sidonie Smith, and Gillian Whitlock have explored the ways in which writers formulate identity and life stories intersubjectively, often with an aim toward expressing a communal identity of some kind (gender, political, disability, postcolonial, etc.).¹¹ So the notion of the self created through the presence of others is not original or unique to cinema. In addition, these theorists all remark on the central importance of a representation of the body as key to self-representation and self-formation, so the material presence of the body is not only significant in photographic or cinematic media. But there is a difference, which is the viewer’s sense of direct access to the body when viewing a photograph or a film. Not only that, but as photography theorists Roland Barthes and Alan Trachtenberg both hint (in the face of reason and logic), we as viewers get the sense that the photographed person might be able to view us.¹² The possibility of opening up a circuit between the body of the author and the body of the viewer is one of the linchpins of my study. I want to argue that the films I analyze here move beyond the realm of self-representation and attempt to engage a closer encounter with the viewer.

The work of another theorist, Elizabeth Bruss, offers a hint as to how we can proceed. Bruss argues that it is the truth-telling nature of photography, its testimonial power, that gets in the way of autobiographical discourse in narrative film: we can no longer pretend that the “I” narrating (behind the camera) is one with the “I” narrated (on-screen).¹³ The process of fictionalizing or constructing a self exists just as fully in textual autobiography and autobiographical fiction, but the division between narrating and narrated selves becomes graphic—material, visible—in fictional cinema’s staged, reenacted past. Bruss’s essay, however, unlike Lejeune’s, does allow for the consideration of nondocumentary films; in fact, the ones she mentions are all works by well-known directors of fictional films. She contemplates the question of why some nondocumentary art-house films are understood as autobiographical and wonders in conclusion whether, rather than closing off the category of autobiography to exclude such films, we ought to reconsider the evolving nature of autobiography and selfhood as they are transformed through this genre of cinema.¹⁴

Auteurs as Autobiographers

One sign of the auteurist film’s construction of a palpable subjectivity (a subjectivity tied to the identity of the director) is the frequent employment of autobiographical gestures on the part of art-cinema directors, and here I will examine those autobiographical moves more closely. As I proceed in my analysis, and the study of cinematic autobiography becomes a study of cinematic of self-projection, we will see that it is not so much my point that particular auteurist films are autobiographical (though they are widely understood as such); it is more to the point that autobiographical elements and gestures within a film indicate the construction of a personalized cinematic subject, a subject that becomes recognizable to the viewer and seems to want to move into the viewer’s space through the apparatus of cinema. Beyond the use of autobiographical elements, then, we will see in auteurist films a number of strategies that perform what I call authorial self-projection. For instance, there might be repetitions of aesthetic elements or narrative topoi or actual lines or scenes that lead viewers to recognize a directorial signature; that is, signs that indicate the presence of a particular author. In other films, the director actually appears within the narrative, either as him or herself or as a fictional character. Another form of self-projection can occur through the repeated employment of a particular actor or group of actors, sometimes including persons attached to the director in “real life” (family members, spouses, etc.). Finally, metanarrative references to the work or apparatus of filmmaking point to the presence of a director behind the scenes, even as the evocation of the apparatus seems to suggest at the same time that it is the machine, and not the director, that makes the film. In this study, I will trace the various signs that indicate directorial presence, claiming them as an extended autobiographical form, one that pushes against the boundaries of autobiographical discourse by extending the model of self-construction and self-representation into an act of self-projection. I accept, in fact, many of Philippe Lejeune’s arguments about why a narrative film (like The 400 Blows or Fanny and Alexander) cannot be an autobiography, no matter how autobiographical.¹⁵ Narrative films with autobiographical content are limited in scope because of the boundaries imposed by typical narrative film length, so at best we will see only a short episode from the subject’s life. Truffaut’s remarks about how the “film of tomorrow” will tell the story of the director’s “first love” or “military service” underscore this point. Narrative focalization functions differently in films and texts; while Truffaut proclaims that the film of tomorrow will be “narrated in the first person,” it is not always clear how the cinematic spectator will understand where that person is located. The viewer’s attention moves frequently among various subjective positions established by the camera or the narrative, shifting away from the strict attention to “I” typical of textual autobiographical narrative. The communicative link between author and viewer takes on a different form in cinematic discourse, particularly in the discourse of art cinema. As Lejeune notes, we have a problem of vocabulary: “autobiography” is not quite right, even if we sense the presence of an autobiographical subject in narrative film.

“Self-projection” as a term offers more than just the obvious pun on cinematic screen projection. The term represents a shift in vocabulary, from a focus on life story and self-representation to the use of the cinematic apparatus (actors, projector, screen, cinematography) as a means of creating the sense of an author, and as a means for that author to reach out, to project to spectators. Though I do not want to describe cinematic self-projection as a defense mechanism necessarily, or to diagnose any particular neurosis or other condition in art-cinema directors, it can be instructive to allow for some resonance with psychoanalytic theory. For if we think metaphorically, we can see how projection—described as the ascription of one’s own traits, feelings, ideas, and beliefs onto others (as well as onto an actual screen)—might offer some insight into how autobiographical self-representation works for an art-cinema auteur. In Freudian terms, projection occurs in order to relieve anxiety by allowing expression of forbidden desires without having to claim them consciously as one’s own. But in an artistic setting, an arena or theater is created for the expression of those impulses and desires that is removed from present time-and-place relations. Not only the author, but the actors, the spectators, and even the cinematic apparatus itself engage in the performance of a projection that yields a vision ascribed to an author; ultimately, the image is projected back onto the author, or rather, the constructed author. In performing this projection, something becomes clear: that the presence of the Other is necessary for the construction of the self. This is true in psychology and philosophy as it is true for textual as well as cinematic self-representation (see Freud, Lacan, Buber, Levinas), but in cinema the act of employing others (particularly actors, but also spectators and the apparatus) as a means of projection becomes particularly visible.

One might claim that the auteurist directors, and I in analyzing them, fall prey to what is no more than an illusion, not unlike the one experienced by the characters in Andrei Tarkovsky’s science-fiction film Solaris: the auteurist’s cinematic subject, the director as recognizable author. (In Tarkovsky’s film, the hapless voyagers on a space station are taken in and held captive by illusory others, projected from their own psyches.) But in looking more closely at the films I will study, one finds that it is more complicated than that. The directors often present an autobiographical connection, only to undo it, undermine it, complicate it. The films emphasize the fluidity of identity, its intersubjective construction (that is, its dependence on others), the instability of identity, role-play and masks, the elusive quality of memory, competing narratives, and so on. Even as particular groups of films under a particular director’s name seem to produce a recognizable image—a Bergman film, a Fellini film, a Herzog film—the connection between the films and their directors is prone to break, reconnect, and break again. Under these circumstances, how do we perceive a film as autobiographical?

At the conclusion of Louis Malle’s Au revoir les enfants (1987), a male voice-over comes in as a kind of oral footnote, and asserts that the narrative we have just followed is a true story from the unidentified and invisible speaker’s childhood. Following Kaja Silverman’s thoughts on male voice-over, we can propose that though we have no image of the director, or perhaps precisely because we have none, the power of the voice-over suggests the voice of authority. Silverman contends that the “sequestered” male voice in cinema—that is, the voice on the soundtrack that is not uttered by a figure on the screen—“works to align the male subject with potency, authoritative knowledge, and the law,” and she goes on to aver that in fact “this disembodied voice can be seen as ‘exemplary’ for male subjectivity, attesting to an achieved invisibility, omniscience, and discursive power.”¹⁶ In the case of Malle’s film, the connection between the body and voice does not occur onscreen, but it does occur very powerfully off-screen, in the paratext surrounding the film. The marketing and distribution and reception and criticism surrounding the film (including interviews packaged with the DVD version) announce that the story of the film comes from Malle’s childhood, so that the viewer familiar with this background can connect the voice not to just any body, but to the director’s body. In this way, the paratext supports a reading of the film as autobiographical in a way that viewing the film on its own does not. When we consider other ways in which the filmmaker might have delivered the “true story” information (giving it as a title screen with his name attached, for instance), we can see that he plays a game of hide-and-seek with the viewer. Malle both does and does not wish to present this story as his own, which might in part relate to the focus of the narrative: the attempted rescue of Jewish children by a Catholic priest, and the betrayal of them by their community (and ultimately, their nation). It is, in other words, both autobiographical and not an autobiography, a story from Malle’s life, but not the story of his life. At the same time, he gives the story his imprimatur in the form of his voice, which operates as a coda: look at the film da capo, viewer, and see what difference it makes when we know it is “true.”

One of the central figures in Pedro Almodóvar’s La mala educación (Bad Education, 2004) is Enrique, a filmmaker who draws his film narratives from newspaper stories: “real life.” As the film opens, he receives a visit from a man purporting to be his childhood friend (and first lover), a person who had fallen prey to a sexually abusive priest while he and Enrique were in boarding school. This visitor (who is an impostor; he is actually the brother of the now-deceased friend) wants Enrique to produce a film based on a short story he has written—“The Visit”—which is in fact based on a “true story” Enrique should recognize: the story of his own early sexual experiences and the abuse by the priest. As this short summary begins to show, Almodóvar’s narrative produces a tangle of interwoven, partly fictional, partly “true” stories (fictional or true within the narrative, that is). And in fact, the underlying concern of this narrative tangle, with its layers of disguises and repeated betrayals, is the relation of cinematic storytelling to “real life.” Given that one of the protagonists of the film is a gay filmmaker, who, as the final title card of the film tells us, continues to make films passionately, it does not take a great leap of the imagination to identify the film as autobiographical; the Spanish director’s films project an image of sensuality and preoccupation with sexuality.¹⁷ And Almodóvar’s production company, after all, is called El Deseo (Desire). Indeed, discussions around the film just after its release centered on the possibility of its autobiographical nature, with particular reference to the possibility that the Spanish director was sexually abused by a priest while a young student. In response to questions about the film’s relation to his life, Almodóvar explains in a commentary that “Bad Education is a very personal film, but not exactly autobiographical; that is, I want to say that I am not relating the story of my life in school.” But he goes on to remark in the same interview that “everything that is not autobiography is plagiarism. . . . The film is autobiographical, but in a deeper sense; I am behind the characters, but I am not telling the story of my life.”¹⁸ And in answer to the interviewer’s direct question about whether he was abused, he asserts, “Me da igual”—“It’s all the same to me.” Thus we can say that Almodóvar offers yet another version of the autobiographical Fort-Da game: now it is, now it is not, about him.

Woody Allen’s Annie Hall (1977) involves a romance between Alvy Singer, a stand-up comedian played by Allen (who was a stand-up comedian before he made films) and Annie Hall (played by Diane Keaton, who had a real-life romance with Allen before the making of the film and whose original name was Diane Hall). It is with performances like Allen’s that the issue raised by Silverman’s argument gets brought into play—namely, whether the masculine voice in cinema maintains its power through a separation from the image of the enunciating body. Allen’s self-representation on both corporeal and vocal planes indeed seems to point toward impotence, but as an ironic turn, since in fact the narrative is dominated by and constructed through his presence. But I will say more about Allen’s self-representations in a later chapter. The humor of Annie Hall turns upon the supposedly irreconcilable cultural differences between the two protagonists (one a Jew—Singer, the other a WASP—Hall). While the film exaggerates the ethnic markers for comic effect, they are essentially biographically correct for Allen (born Alan Konigsberg into a New York Jewish family) and Keaton, who was raised a Methodist. Not surprisingly, many viewers understood Annie Hall as an autobiographical film, though Allen has stated explicitly in interviews that it is not.

Within the narrative of Annie Hall there is a humorous allusion to authorial authenticity communicated through an author’s body: when Alvy Singer, standing in line at a movie theater, becomes annoyed by a conceited self-styled intellectual’s flawed explication of Marshall McLuhan’s theories, Alvy produces McLuhan himself, who assures the offending intellectual that “You don’t understand my ideas at all.” Thus Allen allows Alvy to legitimate his own interpretation of McLuhan’s theories by producing McLuhan, indicating the strength of authorial presence for the production of truth, even as Allen (in the film’s paratext) vigorously denies the authorial significance of his own body situated in what appears to be his story. It is here, and in the other denials and half-denials I have cited, that an impression of the complexity of the auteur’s relationship to his cinematic self-projection begins to take shape.

Even when a director frankly acknowledges the autobiographical content of his narrative, complications seem inevitable. In Ingmar Bergman’s documentary, Dokument Fanny och Alexander (The Making of Fanny and Alexander, 1986), we watch Bergman painstakingly direct the young Bertil Guve, who plays Alexander in the film. It seems a rather trivial scene: the boy is playing idly with a toy theater stage; his head rests on one outstretched arm and is visible to the film camera behind the open stage curtains; the other hand moves the paper puppets desultorily. Bergman sits close by and watches intently, giving minute instructions for moving, coughing, positioning the body and arms. At one point he says, “Feel, like, sick” (“Må, liksom, illa”). At another point he observes, “You want your mother.” These seem odd directions from a director to an actor, but Bergman is in fact working to produce an inner state in the young actor that he himself knows intimately. The intertitle that precedes this sequence in the documentary reads: “I attempt to re-create a few moments from my childhood.” And the completed sequence, the boy’s face looming large behind the toy stage, makes up the opening moments of Bergman’s film Fanny and Alexander (1982).

The documentary The Making of Fanny and Alexander seems a clear instance of paratext, a kind of commentary on the feature film from behind the scenes. But in watching the documentary, a question arises as to how separately one ought to view Fanny and Alexander and The Making of Fanny and Alexander. The latter was filmed simultaneously with the former, using different camera angles and capturing different parts of the action: the sequences that appear in the feature film and the sequences of Bergman directing them. It is of course possible to watch Fanny and Alexander without reference to the documentary, but once one has seen the documentary, it is as if the double nature of the feature film is suddenly revealed: it is both by and about Bergman. The intertitles tell us this explicitly, and we see the relationship between Bergman and Alexander enacted. In the sequence described above, in which Bergman directs Bertil Guve as his childhood self, he cautions the camera operator for Fanny and Alexander: “Keep me out of the shot.” Strewn throughout the documentary are moments like these, in which a dance of cameras allows the parallel films to be recorded independently of one another. Another paratext for Fanny and Alexander is Bergman’s autobiography, Laterna Magica (The Magic Lantern), which appeared the same year as The Making of Fanny and Alexander and includes accounts of numerous incidents that made their way, more or less unchanged, into the film Fanny and Alexander. So it would seem that Bergman offers ample support in his self-crafted paratexts for the reading of his epic feature film as a kind of autobiography.

And yet—Alexander of Fanny and Alexander is not Ingmar Bergman for a number of reasons. One of the most obvious is that he is called Alexander, which already breaks with Philippe Lejeune’s rule of author/ narrator/protagonist name identification for the autobiographical contract. Bertil Guve is not the young Ingmar Bergman. And then Bergman places his fictional film in the year 1907, eleven years before his birth. Alexander is, therefore, more than twenty years older than Bergman, if we are imagining both of them as real people living in real time, which they are, obviously, not. And there are many other deviations: Alexander has one sister; Bergman has a sister and a brother. Bergman’s mother and father were not theater people (unless one counts a pastorate in the Swedish Church as a theatrical profession, which Bergman sometimes is inclined to do). This is leaving aside the film’s supernatural events, which would remove it in any case from the verification process that Lejeune describes as a necessary distinction in the reading of autobiography. But now we can see that Bergman’s film fits the pattern I have been tracing: he places significant paratextual clues leading to the reception of his films as autobiographical, yet makes it impossible to read the film as an autobiography.

What seems to emerge in auteurist art cinema is both a stubborn adherence to embodied self-agency (that of the author/director) and the notion of a boundless or permeable self, a self that includes and absorbs others, even as the very definition of the auteur wants to claim individuality and uniqueness for the director. This accounts, perhaps, for the tension expressed in all of the above cases between the admission of an autobiographical pact and its denial. Elsewhere I have argued that the scene between Ismael and Alexander cited above offers an image of Bergman’s relationship to the Swedish playwright August Strindberg, whose work Bergman frequently staged.¹⁹ By explicitly situating the narrative in 1907 (eleven years before Bergman’s birth, the year in which Strindberg’s A Ghost Sonata had its premiere), and by naming the madman Ismael (one of Strindberg’s autobiographical aliases), and by openly identifying (in various places) Alexander’s experiences as events taken from his own childhood, Bergman seems to propose in this scene a merging of minds between Strindberg (who died before Bergman’s birth) and himself as a child. The presence of Strindberg in the film, which becomes more explicit as the film goes along, brings home the idea that the film is not the product of a single imagination, but a merged imagination.²⁰ By pulling Strindberg into his life narrative (or allowing Strindberg to appropriate him as amanuensis), Bergman extends his life beyond the parameters of his birth and death years, beyond the experience of his individual body.

Auteurist autobiographers engage other bodies in their acts of self-representation. Truffaut, for instance, consistently uses the same avatar for his life story: the actor Jean-Pierre Léaud. In The Adventures of Antoine Doinel, Truffaut writes that the figure they create together through a suite of autobiographical films, Antoine Doinel, becomes an amalgam of the two men—a merged selfhood. And all the actors (and unseen actors, such as technicians and producers) involved in an authored film become in some sense emanations of the author’s vision, extensions of the director’s thought. Despite the merging of self and other(s), something that would seem to disrupt ordinary understanding of what autobiographical discourse performs, the audience, usually with the aid of some hint from reviews or other aspects of the film’s paratext, receives the narrative as autobiographical. There exists, in other words, a compact between auteurist filmmakers and their audience that parallels the autobiographical pact that Lejeune described. And the division of the autobiographical subject into two parts, one in front of the camera and the other behind the camera, may prove not to be an absolute obstruction to understanding narrative film as autobiography. Instead, it can be a way to open up a new way of thinking about autobiography, selfhood, and film spectatorship, one that evolves through the technology of cinema.

One might well ask why it is important to consider the autobiographical elements of art-house films. Certainly it is possible to enjoy, interpret, and understand a film like The 400 Blows or Fanny and Alexander without any knowledge of the director’s life. But I would argue that art-cinema culture desires a spectator with knowledge broader than an acquaintance with a single film; it begs for a familiarity with the entire oeuvre of a given director, and with the works of other directors (thus the cross-referencing between art-house directors), and with the paratext surrounding the author’s work. While any one of the narratives of this study can stand alone as a text, they were created and distributed with this richer tapestry of knowledge in mind, and the references to the directors’ lives and persons scattered throughout these films act as signposts pointing toward that shared pool of knowledge and relationship. The spectator, in other words, plays an important role in the construction of the author. With the viewer’s cooperation, the idea of an intersubjective model of selfhood takes on flesh in the films of these authors, as they explore what it means to be a collaborative and constructed self.

Boundless Selves

Kanske är vi samma person? Kanske har vi inga gränser? Kanske flyter vi genom varandra, strömmer genom varandra, obegränsat och storartat!

Maybe we’re the same person? Maybe we have no boundaries? Maybe we flow through each other, stream through each other, boundlessly and magnificently!²¹

In the scene cited above, the boy Alexander finds himself locked in a room with the “dangerous” madman Ismael (or is it a madwoman? the male character is played by a female actress). The man/woman Ismael has asked Alexander to write his name on a piece of paper, but when the boy reads out what he has written, the name on the paper is not “Alexander Ekdahl,” but “Ismael Retzinsky.” Somehow the boy has written the name of the imprisoned madman, a name he did not know and cannot even pronounce, which moves Ismael to speculate that they are “the same person,” that there are no boundaries between them. In staging this scene, Bergman repeats a motif that forms the central focus of another of his films, Persona (1966), in which two women seem to merge into one. Ismael regards the merging of two selves as “magnificent,” and certainly his power to move somehow into Alexander’s mind and body in order to produce his own name strikes the viewer as marvelous and magical, but also more than a little dangerous. Ultimately this merging of minds between Ismael and Alexander seems to generate a telepathic power that reaches out and kills Alexander’s enemy, the Bishop. As Ismael says, “[Making a voodoo doll and sticking pins in it] is a rather clumsy method, when you think of how swiftly an evil thought can move.” And the merging of selves in Persona creates a kind of horror film, replete with screeching soundtrack. There can be something deeply disturbing in the loss of self, the loss of agency implied in that merging. And there is something implied as well about the art of filmmaking, a particular kind of filmmaking that demands that the actor (and the viewer) surrender him or herself to the overpowering vision of the film’s author. Yet Ismael’s view, that the merging of selfhood can be something wonderfully powerful and liberating, obtains as well; it is the murder of the Bishop, apparently effected by Ismael/Alexander, that frees Alexander’s mother from the prison house of her marriage and allows her to return to her beloved family.

Merged or mirrored selfhood is a theme that runs through Bergman’s films over decades, in Smultronstället (Wild Strawberries, 1957), Ansiktet (The Magician, 1958), Persona (1966), Vargtimmen (Hour of the Wolf, 1968), Viskningar och rop (Cries and Whispers, 1972), and Fanny and Alexander (1982), to name the most obvious examples. But this topos of the boundless self is not unique to Bergman’s films; it takes various forms in the work of other art-cinema directors. Pedro Almodóvar’s fascination with transplants and cosmetic surgery, for instance, points toward a transgression of the body’s “natural” limits, with both terrifying and liberating results. Andrei Tarkovsky’s Solaris (1972), based on the science-fiction novel by Polish author Stanislaw Lem, represents a model of self-projection that in some respects resembles the cinematic model, in which the fears and desires of the film’s characters take shape as embodied hallucinations of people the characters have known. In this way Tarkovsky’s film, much more concerned about the human psyche than any interplanetary alien life form, meditates on what it is we experience when we encounter an Other, and what it is that comprises a self and its boundaries. These examples come from the diegetic plane; another arena in which we can see a merging of selves takes place in the direction of actors. In auteurist cinema, we might argue that the first part of the autobiographical contract takes place between the filmmaker as director and the filmmaker as spectator. To understand this, we can return to Ingmar Bergman as he directs the actors of Fanny and Alexander in The Making of Fanny and Alexander.

Bergman’s focus on the boy during the opening sequence discussed above seems like a violation and is more than a little frightening to watch. He aspires to place within the child’s body and mind the sensations of his own childhood—“feel ill,” “you want to go to your mother”—in a visceral way. And not only does directorial agency enter into and move the actor’s body; the actor is meant to represent emotions and sensations in order to reproduce them (or an impression of them) in the viewer. There is transference, in other words, between the director’s memory into the boy’s body that is meant to reach the spectator. But in directing the boy, Bergman becomes a spectator as well. He is the first audience, the one who will judge whether the sensations have been accurately captured, whether they can convey his inner state to the imagined viewer.

This same process repeats throughout The Making of Fanny and Alexander, perhaps most strikingly in the telepathic episode between Alexander and Ismael, in which we see Bergman direct two actors, the boy Bertil Guve and Stina Ekblad. What happens between Ismael and Alexander in Fanny and Alexander can be understood as a representation of a kind of filmmaking in which the relationship between director and actor is symbiotic, a merging of consciousness and vision that has a real impact on a third party, though Ismael (who performs in this scene as a kind of director) ascribes that vision and the emotion that produced it to Alexander (in this scene Ismael’s “actor”). “You are thinking about the death of someone,” says Ismael, as he looks into Alexander’s mind. Step by step, Ismael leads Alexander from the thought of the Bishop’s death to a staging of the Bishop’s death, which takes place, as the crosscutting of the film tells us, at the same time as the mind-merge. The spectator perceives the sequence of film depicting the events leading to the Bishop’s death as the “film” of Ismael and Alexander’s shared vision. A question emerges subsequently in Fanny and Alexander as to whether their vision represents Alexander/Ismael’s willing the Bishop to die (that is, the events are generated by what we ordinarily would call a supernatural force) or whether there is a “rational” explanation for what happened. A police officer comes to clear up the case after the Bishop’s death, since some suspicion fell not on Alexander but on Alexander’s mother, who had given the Bishop sleeping powder that prevented him from saving himself in the fire. The policeman’s alternative explanation of “unhappy coincidence” urges the viewer to ask: What did I see exactly? What was its relationship to “reality”? Given, there is a way in which that question reflects a hopeless naïveté. There is no reality in the film—it is “only a movie.” But by setting an eccentric image of film production within the narrative (Ismael’s “production” of the sequence depicting the Bishop’s death), Fanny and Alexander hints at a vexed relationship between human consciousness and film reality. What is it that produces a film? A human vision—a collaborative one. Who or what steers the collaboration? How do we describe or imagine that subject?

The questions become compounded when we add The Making of Fanny and Alexander to our analysis. The documentary includes a sequence in which Bergman, sitting close to cinematographer Sven Nykvist so that he can look through Nykvist’s lens, directs the scene I have just described. In minute detail he describes to Ekblad and Guve how they should position their bodies in the image frame, the tone of voice to be used, and so on. But when the action begins for a final filming, he falls silent and watches intently, intensely. The documentary camera moves in, Bergman-like, close to his face and focuses tightly on his emotion. We can hear the dialogue on the soundtrack: “I am erasing myself, I enter you …” Bergman’s concentration is so focused and strong that he seems to be acting in the scene as well, his lips move as if he is mouthing the lines, his body is taut with restrained motion.

The intertitle preceding the scene reads: “Our concentration is total.” Watching this sequence of Bergman watching his actors, we can see that Bergman is both viewer and actor in the scene. He is separated from his actors behind the camera while they stand in front, but he attempts to control them through sheer force of will to produce the “film” of his imagination: we can see that this is so. How can we see what is in Bergman’s mind? It is projected for us in The Making of Fanny and Alexander, when the camera moves in so closely to Bergman’s face that we imagine we can see his thoughts. But then this, too, is a performance. The film camera is there, filming Bergman, and he must surely be conscious of it. The relationship of film to reality and to inner vision remains a matter of inquiry and investigation, rather than a given, clear line between auteur and film. But in these shots of Bergman directing, something comes clear: the line connecting director and spectator happens in the first instance when the director is the primary spectator of the film.

Scholarship on film spectatorship has long focused on the spectator’s state of consciousness, often likened to a dream, the viewer’s identification with the personae constructed through the film narrative and projected on the screen. Laura Mulvey’s “Visual Pleasure and Narrative Cinema,” a foundational essay on the gendered nature of film spectatorship, and André Bazin’s seminal What Is Cinema? both deal with the idea of the embodied film viewer’s relationship to cinematic narrative. But the author can also function as a powerful point of identification, imagined in a number of ways: as a person whose life and life story appear to be communicated in cinematic works issued under the director’s name; as a theoretical, cultural, and economic construct that is imagined to be readable in the media and across a body of films; as the imagined voice and consciousness speaking through these films. Ultimately the author begins to take the form of another person through the viewer’s imagination.²²

An art film often places a demand of empathy or intellectual involvement on the viewer, something that urges us to attribute a mental state to the vision and narrative of film, which is most easily and humanly understood as another human’s imagination and will. In particular, films that aim to create “higher meaning” (art films, auteurist films) have as one of their projects the representation of what it means to have a self. Film uses figurations of people to bend us to the perception (one we gladly embrace) that there are other humans out there, and that we are human, too. But what can happen in our interaction with those projected humans (or that projected humanness) is an erasure of boundaries, a confluence of the person imagining, the images of that imagination, and the viewers entering into the imagining, emotionally, mentally, and physically. “Maybe we are the same person? Maybe there are no boundaries?”

In an essay on Alfred Hitchcock’s trademark appearances in his films, D. A. Miller guides the reader through a series of cameos that are increasingly subtle: first the typical appearance of Hitchcock as an unnamed figure in the frame, then of a character in the film holding a book written by Hitchcock in which the author’s picture is just visible, then a shot of the book with the picture hidden, etc.²³ As the references to the auteur become ever more difficult to perceive, Miller feels himself more and more directly addressed. He notes that the appearance of Hitchcock in cameo, the one that every audience member with any knowledge of Hitchcock’s films awaits expectantly, elicits knowing chuckles from the audience; but he argues that this initial recognition of the author is too cheap, too easy, too broadly advertised, for the spectator who sees it to feel addressed directly as a Hitchcock connoisseur. Instead, it is the less obvious clues—the book half-hidden in the actor’s hand, the barely visible image of Hitchcock’s face on the book’s cover, the typography of the book when the face is no longer visible at all—these are the clues that only the most diligent and knowledgeable viewer could hope to perceive. That is, someone like Miller, who not only sits watching like a hawk on a telephone wire for tidbit traces of Hitchcock to race across the screen, but feels that these delicious moments were placed there precisely for him (and perhaps for him alone!) to find. It could be maintained that Miller is simply an eccentric viewer, one who inserts much too much of his own autobiographical persona into his scholarship; certainly a look at his articles on Roland Barthes and sexuality, or on Hitchcock’s Vertigo, reveal a vigorous employment of an autobiographical subject. But I would hold with Barbara Johnson in saying that the autobiographical position Miller adopts is not mere self-indulgence but is precisely central to his argument, much the same way I would argue that the autobiographical position adopted in auteurist film is central to understanding the ways these films engage the question of selfhood and subjectivity.²⁴

What Miller’s argument reveals is the delight, the pleasure taken when the viewer imagines that there is someone in the film, a person, who addresses the viewer, touches the viewer personally. The pleasure of believing that there is a knowledge, a (re)connaissance, a handshake, a recognition, and not only the viewer’s recognition of the auteurist subject, but the auteur’s “recognition” of the viewer, a recognition expressed through address. While this argument might seem to verge on the superstitious, it aligns with the (superstitious) way in which photographs can be and have been read: in Camera Lucida, Roland Barthes insists upon the “look” of the photograph, that is, that the photographed subject looks at the viewer, and Alan Trachtenberg, in his scholarly study of American photography, claims that the photographs of American slaves look at their viewer, arguing that “if we reciprocate their look, we have acknowledged what the pictures most overtly deny: the universal humanness we share with them.”²⁵ It should be noted that the difference here is that both Barthes and Trachtenberg locate the “look” of the photograph in the gaze of the photographed subject, and Miller and I are thinking instead of a “something” personalized as a someone, the director, that regards us and addresses us through the agency of the film’s narrative images. But Trachtenberg’s claim contains a germ of what I have been claiming for cinematic projected subjectivity; namely, that it is a shared or collaborative subjectivity, one that moves through the agency of the cinematic medium between auteur, actor, and spectator. A spectator such as Miller (or myself, or anyone who works with the idea of the auteur as subject) has to assume that the auteur has imagined a viewer who will view with the auteur, who will pick up on what the auteur has himself first projected and seen.

In writing about one of the films in my study, Andrei Tarkovsky’s Mirror, Natasha Synessios notes, “I saw Mirror in my native Athens, in the early 1980s. At the time I knew nothing of Russian language, culture or history and had no context in which to understand the deeper significance of some of the episodes in the film. Yet I . . . felt that this was a film about me.”²⁶ She says that when she went over the voluminous correspondence from the film’s viewers to its director, “the refrain, echoing through all the letters, was ‘this is a film about me.’”²⁷ In order, however, for the film to be “about me,” about the viewer, it seems that it must first be imagined as being also about someone else, the subject who speaks through the images and the cinematic narrative and can be most readily identified as the auteur. Thus one can understand the letters directed to Tarkovsky, explaining that the film he had directed had indeed performed the function of mirroring. Though the title “mirror” might originally have referred to the way in which the film reflected the auteur’s life (filmed in part in his childhood village in a reconstructed replica of his childhood home), it becomes a mirror for his audience, even for spectators like Synessios, who did not share the material reality of his Russian childhood. Synessios went on to learn Russian and was one of the translators of an edition of Tarkovsky’s collected screenplays, thus becoming Tarkovsky’s projected voice.

I have discussed in some detail how auteurist film posits a form of communication between a director and a spectator, a kind of self-projection on the part of the director that expresses itself in particular terms to the viewer. But this is not to forget that at the same time there is a constant reference in the films and their paratexts to the fact that the films in fact cannot and do not form any such direct line of communication between auteur and viewer. It is a fascinating if perplexing characteristic of auteurist filmmaking that the filmmakers seem to undermine the identification between themselves and their work on the one hand, and on the other hand participate wholeheartedly in the construction of themselves as auteurs, creating an off-screen and sometimes on-screen persona linked to their own bodies.

About This Book

In this study I will restrict myself to the work of a small group of film authors, not because they are the only ones to grapple with the problem of selfhood and self-representation in cinema, but because their use of autobiographical material and acts of self-projection offer particularly rich grounds of exploration. It is imperative that self-projecting auteurs have a reputation sufficient to allow for broad recognition of their names in connection with their art. The group I have chosen are internationally celebrated auteurs, and the paratexts surrounding their films are accordingly extensive and significant. In order for self-projection to occur and be understood, a contract must be in place that states that the cinematic work has a point of origin in a recognized artist’s imagination, and the number of directors who have received this international level of recognition is limited. Still, there are directors I excluded but could have discussed to good effect: Jean-Luc Godard, Rainer Maria Fassbinder, Akira Kurosawa, Jane Campion, Quentin Tarantino, to name just a few. Other scholars have written extensively on the self-projecting character of their work. And earlier film directors would certainly provide insight into the problems I address here: Buster Keaton and Charlie Chaplin spring readily to mind, in part because of their use of their own bodies as vehicles of projection. But I want to limit myself to a handful of directors from a limited span of film history (1960 to today, roughly) with the idea that the reader will be able to supply other productive examples, filling in the blanks from his or her own store of cinematic experience.

It seems important at this juncture to mention the questions of gender and race. The focus of my analysis will be on the work of white male auteurs, and one must ask why this is the case. I will first turn to the matter of gender. Women filmmakers—Agnès Varda, who began her career during the art-cinema surge of the French New Wave in the 1960s, Chantal Akerman, Belgian avant-garde director, Margarethe von Trotta, a prominent director of the New German Cinema movement of the 1970s, Jane Campion, the New Zealand director who has achieved international recognition for her unique cinematic vision, Claire Denis, one of the most provocative film authors working in France today, Susanne Bier, the Danish director who has crossed over to Hollywood, and a good number of others from around the world—are artists working at a high standard and can certainly be understood as film authors. They are described as auteurs by scholars as well (often as “female auteurs”), though as British filmmaker Sally Potter notes, “‘auteur’ is much more readily used as a term for male directors than female directors,” and she adds, “people don’t concern themselves with the profound collaborations that men have. But as soon as a woman has a collaborator it’s thought of as ‘oh it’s not really hers then.’”²⁸1 would agree with Potter that “auteur” much more frequently designates male film authors, and that women are more readily imagined as collaborators, while men assume more easily the reputation of lone Romantic genius. Yet there are other reasons men are more readily marked as auteurs, and my study will highlight some of these reasons.

The definitive issues for me in selecting auteurs for this study will reveal something of why the primary focus is on men. The spectator’s role in receiving a film as an act of self-projection demands that the auteur be highly recognizable as auteur—that there be a structure of production and reception in place that allows or persuades filmgoers to “see” the author in the work; to go see the film because the author will be “in” it. When I write of production, I mean that literally; it is striking that filmmakers who achieve auteur status often go on to form their own production companies. Of the women cited above, Campion and Varda formed production companies with somewhat limited success. More successful were François Truffaut, Ingmar Bergman, Woody Allen, and Pedro Almodóvar, each of whom developed their own production companies, often assisted by family members (Truffaut’s father-in-law, Bergman’s wife, Allen’s sister, Almodóvar’s brother). Practices of apprenticeship in film directing have also favored men (when women, like Denis, have become successful, they often served as assistants first to male directors), as have the mechanisms for funding (whether state, as was often the case in Europe, or private) and granting awards for cinematic work. In short, the power required to marshal sufficient economic and personal forces to attain auteur status resided largely with men from the outset, and it is only recently that women directors have achieved the kind of status required for instant name and face recognition. It was not until 2009, for instance, that a woman, Kathryn Bigelow, won the Academy Award for Best Director, and that was for The Hurt Locker, a war film dominated by the stories of men. The Cannes Film Festival has given its award for direction to a woman precisely once, in 1961, to Yulya Solntseva, a Soviet filmmaker with very little international recognition today. In her autobiographical documentary, The Beaches of Agnès, Varda comments wryly that when her film Cleo de 5 à 7 (1962) was nominated for the Palme d’Or at Cannes, all cameras were trained on the beautiful blond star, Corinne Marchand. No one noticed the small, dark director standing nearby. But the situation was quite different when François Truffaut and Jean-Luc Godard stepped forward as auteurs; though their stars also garnered attention, they stood at the center of the limelight as directors. One of the most important reasons to focus on the male auteur, then, is a level of prominence in the cinematic cultural landscape that allows him to be recognized, discussed, and studied by a significant international audience. The publicity machine of auteurism has always functioned most effectively when the subject was male.

But economic clout and name recognition are perhaps more symptomatic than essential to my main argument about the problematic of self-formation and self-projection. When Sally Potter muses on the fact that it is culturally more acceptable to see women in a collaborative position, the “profound collaboration” she ascribes to men might actually be of a different nature than that often ascribed to women. The word we might use for the work of filmmaking and identity formation within the films of this study might be closer to “appropriation” than “collaboration.” The idea that a film’s vision emanates from a single artistic consciousness implies the overshadowing or appropriation of other artistic subjects working on the film project: the cinematographer (who wields the caméra-stylo, after all), the set designers, costume designers, composers and musicians, and certainly not least, the actors, who become avatars for the director’s creative mind. This happens when women direct as well, certainly, but the degree to which films are received as being “by” and “of” a male auteur emphasizes the way in which auteurism elides and subsumes the other participants in the collaboration.

Within the narratives of the films discussed here, a kind of acknowledgment of the vampiric appropriation of others often surfaces: in Bergman’s Persona most prominently, perhaps, but also in Truffaut’s The Wild Child, Almodóvar’s Talk to Her (in which two men project their respective fantasies on to two women, both in a coma), Werner Herzog’s Nosferatu and My Best Fiend and Grizzly Man (the protagonist/filmmaker’s girlfriend is erased from the narrative and is ultimately eaten by a bear, along with the filmmaker/protagonist), Lars von Trier’s The Five Obstructions (the story of a duel between two directors for control), and many others. It is in fact significant how many films by the auteurs featured in this study derive part, if not most, of their narrative tension from this type of soul-appropriation topos, which I take to be part of their complex response to the problems of self-formation and self-representation in the cinematic medium. I will return to this subject, particularly in the chapter on auteurs and their actors, where a libidinal relationship often produces palpable energy.

And on the subject of gender formation, to sketch things rather broadly: while concerned about questions of self-construction, Truffaut, Fellini, and Allen (of the auteurs of this study) tend not to stray from the idea of masculine selfhood. They maintain to a large degree the idea of “women’s problems” as something occurring outside their own subject positions, with hetero-normal relationships the primary focus of attention. Bergman, Tarkovsky, Almodóvar, and von Trier, on the other hand, all make female subjectivity (or the problem of gendered subjectivity) a major narrative concern. Their films make a point of entering into a female subject position, problematizing it, appropriating it, often wrestling with the idea of embodied personhood through the female position first of all. Scholars (such as Marilyn Blackwell, writing of Bergman) have noted that this apparent concern with feminine subjectivity represents nothing more than another facet of self-construction, and in most cases the female position represented in the works of the filmmakers listed above is a highly vexed, if not tortured, position (one thinks of Bergman’s Cries and Whispers and von Trier’s Breaking the Waves or Dancer in the Dark). And in that torture, I think, the transgressing act of using other bodies and minds toward one’s own creative and economic ends might find a form of representation.

Scholarship on female auteurs, in contrast, can throw light on some of the issues I explore here in the work of male filmmakers. Catherine Grant, in an essay that surveys the arc of feminist theory on women directors, notes that “the theoretical pendulum has swung back . . . ‘with a difference,’” meaning that from a position of regarding the auteur as textual, disembodied, and/or unconscious, scholars have begun to think of female authors as historical, embodied, conscious agents.²⁹ Feminist film scholars, despite a strong affiliation with the notion of gender as construct as put forward by theorists like Judith Butler, rejection of biological essentialism, and firm grounding in poststructural debunking of the biographical author as authority, found themselves returning to the idea of the female film author as a historical and biological person. Judith Mayne, for instance, in writing of American director Dorothy Arzner, calls her work “a study in portraiture, in the literal and figurative senses of the term.”³⁰ It is difficult to write of “women auteurs” without reference to an embodied subject, after all. Kaja Silverman, in discussing female film authorship, writes that texts must be discussed “in relation to the biological gender of the biographical author, since it is clearly not the same thing, socially or politically, for a woman to speak with a female voice as it is for a man to do so, and vice versa,” though her study gravitates toward acts of enunciation rather than connections to biographical persons.³¹ This marks a repetition of a point I mentioned earlier; that there is an apparently unavoidable oscillation between positions of arguing against embodiment and inevitably turning back toward the body. In feminist criticism, the stakes are different in cultural and sociological terms, but they speak to the same problem that resonates through the films of this study, whether by men or women. The central problem is subject formation, with gender one of the crucial aspects of selfhood. In the analyses of the films that follow, I will attend to questions of gender as they arise, with particular awareness of how the idea of appropriation versus collaboration can shed light on masculine strategies of self-formation.

Similarly, films in which author-directors formulate selfhood as part of a minority discourse exhibit a tension between embodied essentialism and constructed identity, with an emphasis on the question of race. Spike Lee, for instance, has created a compelling body of work that in many ways conforms to the requirements I set up for cinematic self-projection: his films are connected by a recognizable aesthetic and thematic vocabulary, he often writes his own scripts, he employs a number of actors who appear repeatedly in his films (including his sister and father), his narratives often carry autobiographical traces, and he has appeared prominently as an actor in a number of his films. Like several of the directors I will analyze in these pages, Lee founded his own production company: Forty Acres and a Mule. But even the name of the production company defines Lee as a director with something different at stake in his film work. The name alludes to the promise issued by Civil War General Sherman that freed slaves would be allotted “forty acres and a mule” with which to support themselves, a policy enacted only briefly and in a limited area in the Southeast. President Andrew Johnson’s revocation of Sherman’s order led to association of the phrase “forty acres and a mule” with the failure of Americans to offer adequate (or any) reparation for the work of enslaved Africans and their descendants. Lee’s use of the term marks the production company as his “stake,” but also as his call to arms. His purpose in projecting a self, in other words, has at least as much to do with the need to formulate a collective self-image for African Americans as it does with an individual artistic self-projection. This in no way discounts Lee’s importance as a filmmaker, nor even places him outside the parameters I would employ to define filmmakers as self-projectors. But it does mean that a thoroughgoing analysis of Spike Lee’s performance of self-projection would entail the introduction of a larger theoretical apparatus on constructions of racial identity. Whiteness, too, is a constructed racial identity that could be discussed in connection with the directors of this study in productive ways. This seemed to me beyond the bounds of what I was ready to attempt in these pages, but I hope that others might be inspired to consider how racial and ethnic identity might figure in the act of directorial self-projection.

Having attempted an explanation of how I chose my authors, I should remark briefly on how I chose which films to analyze within each author’s body of work. My choices may not strike the reader as obvious: certainly one would expect to find 8½ among the films of Fellini, but why not Amarcord, which is the most explicitly autobiographical, perhaps, of his works? If I discuss Bergman’s intimate and working relationship with Liv Ullmann, why not Fellini and Giulietta Masina? Or Woody Allen and Mia Farrow? Why is Lars von Trier represented by two of his most peripheral and lighthearted pieces, while a masterpiece like Melancholia not only reached a much broader audience but offers the potential for discussion of von Trier’s own struggle with depression? It seems apparent as well that Melancholia fits much more easily into what a viewer would recognize as von Trier’s profile, as it can be traced from Breaking the Waves to Dancer in the Dark to Dogville to Antichrist. My answer is that I chose films that illuminate various aspects of self-projection, as I formulate that concept. But I do not mean to say that other films could not be considered as acts of self-projection, any more than I want to say that the authors chosen for this study are the only ones to practice self-projection as I define it within these pages. On the contrary, I would hope that the reader will be able to identify readily other authors and works that could and should be discussed in the context of self-projection, since it is part of my argument that self-projection constitutes a kind of modality of art cinema.

Then to offer a summary of the book’s content: chapter 1 deals with the director as performer in his films. Here self-projection is performed not through narrative (or not solely through narrative), but through the presence of the auteur’s body, and the necessary relation the body creates between the world projected and the world in which the director lives. In an elaboration of the first chapter’s argument, chapter 2 looks at directors performing as directors in documentaries, mockumentaries, and narrative films. In these films the viewer is offered a supposed “behind-the-scenes” look at the auteur at work, only to find that the scene behind the scenes is yet another screen of projection. The director and actor as merged selfhood is the topic of chapter 3, as the auteur continues to unravel and reassemble in new guises, in the form of the actors that perform as avatars of auteurist self-projection. Relationships between auteurs and actors achieve an intimacy that parallels the kind of intimacy urged upon the cinematic spectator, a hap-tic quality that involves the bodies as well as the minds of the auteur, the avatar, and the audience. Finally, chapter 4 examines the auteur’s deployment of the cinematic apparatus as topos, with the implication that the parts of the apparatus (projector, screen, camera, and so on) function as prosthetic extensions of the auteur’s body and vision.

In the chapters that follow, I will pursue the director’s act of self-projection from his use of his own body as actor within the story, to his creation of a directorial person in documentary films about filmmaking, to his use of actors as avatars, to his deployment of the cinematic apparatus, cinematic technology, as a means of self-projection. In each instance, a new problem of self-construction and self-representation will emerge, and in each case I will discuss what it is that art cinema in particular contributes to discussions of autobiographical discourse and self-formation. Throughout the study, the reader will encounter a discussion of people who either have lived or are living: Allen, Bergman, Fellini, Herzog, Tarkovsky, Truffaut, von Trier. The fact that they were or are living people with histories in particular languages, landscapes, cultures, and times is important to my understanding of why it is they we, as viewers, feel we can form a relationship with them, and why they, as artists, seem to call a relationship with their viewers into being. At the same time, it should be understood that we have no real access to them as people. We are talking about a field of discourse here, in which invisible authors project an image to an invisible audience.