A groundbreaking study of how modern American filmmakers are using the "art film" model to explore the power of nature versus the power of art.
What do contemporary American movies and directors have to say about the relationship between nature and art? How do science fiction films like Steven Spielberg's A.I. and Darren Aronofsky's π represent the apparent oppositions between nature and culture, wild and tame?
Steven Dillon's intriguing new volume surveys American cinema from 1990 to 2002 with substantial descriptions of sixty films, emphasizing small-budget independent American film. Directors studied include Steven Soderbergh, Darren Aronofsky, Todd Haynes, Harmony Korine, and Gus Van Sant, as well as more canonical figures like Martin Scorcese, Robert Altman, David Lynch, and Steven Spielberg. The book takes its title and inspiration from Andrei Tarkovsky's 1972 film Solaris, a science fiction ghost story that relentlessly explores the relationship between the powers of nature and art. The author argues that American film has the best chance of aesthetic success when it acknowledges that a film is actually a film. The best American movies tell an endless ghost story, as they perform the agonizing nearness and distance of the cinematic image.
This groundbreaking commentary examines the rarely seen bridge between select American film directors and their typically more adventurous European counterparts. Filmmakers such as Lynch and Soderbergh are cross-cut together with Tarkovsky and the great French director, Jean-Luc Godard, in order to test the limits and possibilities of American film. Both enthusiastically cinephilic and fiercely critical, this book puts a decade of U.S. film in its global place, as part of an ongoing conversation on nature and art.
- Chapter One. Tarkovsky's Solaris and the Cinematic Abyss
- Chapter Two. Steven Soderbergh's Tinted World
- Chapter Three. Aronofsky, Sundance, and the Return to Nature
- Chapter Four. Mulholland Drive, Cahiers du cinéma, and the Horror of Cinephilia
- Chapter Five. Spielberg's A.I.: Animation, Time, and Digital Culture
- Chapter Six. Cinema against Art: Artists and Paintings in Contemporary American Film
- Chapter Seven. A Plague of Frogs: Expressionism and Naturalism in 1990s American Film
- Chapter Eight. Situating American Film in Godard, Jarmusch, and Scorsese
The image, sir, alone capable of denying nothingness, is also the gaze of nothingness on us.
Godard, In Praise of Love (2001)
Film Theory, or Film as Theory
Psychoanalytic theory rose to such prominence in film studies because film seemed to demand an analysis of fantasy. Flowing across the screen, projected through the dark, come these dream images—idealized desires and ego-building identifications. Yet even as we watch and inhabit these fantasies and dreams, we have the underlying consciousness that these dreams are not true, that they are constructions, that "this is only a movie." Lacanian theory, with its centralized concept of "lack," seems perfectly fitted to address itself to this dream screen, a screen both full of desire and perpetually absent. Dozens of models for cinematic fantasy have been offered over the past forty years, from the description of the "cinematic apparatus" by Jean-Louis Baudry (1970), which emphasized the ideological delusions of film, to Marc Vernet's Figures de l'absence: De l'invisible au cinéma (1988). Even Richard Allen's recent critique of Lacanian film theory, Projecting Illusion: Film Spectatorship and the Impression of Reality (1995), still finds the category of fantasy useful in treating the problem of cinematic illusionism. So many models of cinematic fantasy have been offered, indeed, that one can easily sympathize with a vehement call to end such studies: "Film theory should be less a theory of fantasy (psychoanalytic or otherwise) than a theory of the affects and transformations of bodies." Here Steven Shaviro, in a brilliantly polemical book, would like to announce the end of Lacanian theory and the beginning of a criticism inspired by Gilles Deleuze and Georges Bataille.
The present book intends, nonetheless, to continue the investigation of cinematic fantasy. The simultaneous fullness and emptiness of the cinematic experience seems fundamental, and worthy of further critical discussion. Movies themselves repeatedly stage this problem and always have. But instead of turning to Freudian models and language to aid our understanding, I will turn to films—above all, to Andrei Tarkovsky's Solaris (1972), one of the most profound cinematic dreams ever conceived. Taking Tarkovsky's film as a model for the beautiful and infernal absence created by cinema will substantially augment what psychoanalytic analyses have offered so far. To show that Solaris has something important to say about cinematic experience is to underscore both the power of cinematic self-reflexivity and the complexities of cinephilia. Psychoanalytic explanations tend to reject cinematic self-reflexivity, as we shall see, and to refuse to occupy the place of the cinephile. By contrast, Solaris presents a self-reflexive narrative of cinematic love that does not reduce the film to a fetish.
I will argue, furthermore, that Tarkovsky's film has a particular relevance to American film of the 1990s, where amidst a cultural landscape of virtual reality, fantasy takes on an unprecedented significance. Recent films like David Lynch's Mulholland Drive (2001) and Steven Spielberg's A.I. (2001) turn out to be contemporary versions of Tarkovsky's masterpiece. And it is no coincidence that Steven Soderbergh made his own version of Solaris in 2002. Chapter Two will show that Soderbergh's famously varied career has been directed toward Solaris for some time.
Marc Vernet begins his book Figures de l'absence: De l'invisible au cinema with snippets from three of the most important French writers on film:
Mitry: L'image n'apparaît pas comme "objet," mais comme absence de réalité.
Bazin: la présence-absence du représenté
Metz: la signifiant imaginaire
These three phrases emphasize what may be a fundamental aspect of the cinematic image. The cinematic image is both present and absent. But different kinds of cinema mark this phenomenon in different ways. Classical Hollywood cinema is typically characterized by "invisibility" and "transparency," by a continual refusal to acknowledge that the film is actually a film. After modernist filmmakers like Godard took the lead, ideological criticism from the 1970s vigorously argued that film was required to identify itself as film through gestures of self-reflexivity. Vernet's Figures de l'absence circles back around, making the case that even classical Hollywood film is, after all, full of devices that call attention to the cinematic apparatus. Vernet studies devices such as "the look at the camera," superimpositions and dissolves, and the use of offscreen space in order to argue that all kinds of film, not just Godard's, continually remind the spectator that "this is only a movie."
But I want to insist on differences. Clearly there is a felt difference between the self-reflexivity that occurs in Minnelli and Hitchcock on the one hand and in Godard on the other. Vernet wants to level the field counterintuitively, a project that provides a fresh, new look at certain films, but may blur the distinction between what is self-conscious in some films and not others. Self-reflexive films overtly break cinematic illusionism in ways that most classical Hollywood films do not. Cinema's "absent presence" is therefore still a problem, an ongoing question, which in my view not all contemporary American films face up to equally. It is this fundamental and paradoxical experience of cinematic presence that I call the Solaris effect and that I wish to explore in this book. And so I begin with films instead of psychoanalysis.
Self-reflexivity stands as a crucial reason for my turn away from psychoanalysis. Psychoanalytic approaches to film tend not to be very good at processing cinematic self-reflexivity. As Richard Allen argues, the viewing subject in Lacanian studies of film is treated essentially as if he is unconscious. If self-reflexivity shocks consciousness or leads to self-consciousness, then a psychoanalytic explanation of such an effect is going to have some difficulties. In The Imaginary Signifier, for example, Christian Metz spends a good deal of time working through the shadowy forms of reality that are captured on the screen. Here is a typical characterization by Metz of cinema's penchant for absence:
What is characteristic of the cinema is not the imaginary that it may happen to represent, but the imaginary that it is from the start, the imaginary that constitutes it as a signifier (the two are not unrelated; it is so well able to represent it because it is it; however it is it even when it no longer represents it). The (possible) reduplication inaugurating the intention of fiction is preceded in the cinema by a first reduplication, always-already achieved, which inaugurates the signifier. The imaginary, by definition, combines within it a certain presence and a certain absence. In the cinema it is not just the fictional signified, if there is one, that is thus made present in the mode of absence, it is from the outset the signifier.
Thus the cinema, "more perceptual" than certain arts according to the list of its sensory registers, is also "less perceptual" than others once the status of these perceptions is envisaged rather than their number or diversity; for its perceptions are all in a sense "false." Or rather, the activity of perception which it involves is real (the cinema is not a phantasy), but the perceived is not really the object, it is its shade, its phantom, its double, its replica in a new kind of mirror. It will be said that literature, after all, is itself only made of replicas (written words, presenting absent objects). But at least it does not present them to us with all the really perceived detail that the screen does (giving more and taking as much, i.e. taking more). The unique position of the cinema lies in this dual character of its signifier: unaccustomed perceptual wealth, but at the same time stamped with unreality to an unusual degree, and from the very outset. More than the other arts, or in a more unique way, the cinema involves us in the imaginary: it drums up all perception, but to switch it immediately over into its own absence, which is nonetheless the only signifier present.
This is a long, complicated passage, but a good example of how Metz works through the presence and absence of cinema with his semiotic and psychoanalytic tools.
But for Metz all films are "fiction films," which inevitably embody these complex psychoanalytic forms of signification. Thus Metz only barely touches on attempts to "defictionalize" film, that is, on gestures that interrogate or reflect on the cinematic medium in the course of the film. Metz's framework, in fact, cannot tolerate self-referentiality, or it does so by simply absorbing it back into the fictive pool of everything. For Metz the laws of cinematic expression come from outside the cinema, and whatever a film could say about itself would stand only as one more element of its fiction. But it seems to me that films can quite usefully talk about themselves and about the nature of cinema, and in terms that are just as rewarding as those offered by Metz. Moments of cinematic self-reflexivity are not necessarily obvious or obviously praiseworthy; it still takes an interpreter to identify moments of self-reflexivity and evaluate their content. Metz seeks out rules of cinematic signification, yet without much interest in what the cinema itself might have to say about those rules. It is very odd to note that although The Imaginary Signifier was written at the height of Godard's radical period (1973-1976), there is not one reference to him in the entire book.
And self-referentiality is equally cannibalized by Deleuze. Godard is now, to be sure, a central hero for Deleuze, and the turn between the "movement-image" (Cinema 1) and the "time-image" (Cinema 2) occurs right at the French New Wave, with Godard leading the way. Yet "modern" cinema in Deleuze has to do with "direct time," not with self-consciousness or self-reflexivity. Indeed, in Deleuze's terms, self-reflexivity is again pushed to the brink of impossibility. If "the essence of cinema has thought as its higher purpose, nothing but thought," then how is cinema to refer to itself? Deleuze's scatterings of the object continually reject both the "self" and the "reference" that would be required to make up this cinematic "self-reference." Deleuze's deconstructive monism, which rejects the conceptual oppositions of body and thought, presence and absence, will not linger long over films that may or may not talk about themselves.
Nonetheless we will frequently take heed of one of Deleuze's most important contributions, which is to reverse the potentially tragic separation of photography and world by affirming that break as the beginning of creative thought.
The link between man and the world is broken. Henceforth, this link must become an object of belief: it is the impossible which can only be restored within a faith. Belief is no longer addressed to a different or transformed world. Man is in the world as if in a pure optical and sound situation. . . . Only belief in the world can reconnect man to what he sees and hears. The cinema must film, not the world, but belief in this world, our only link. The nature of cinematographic illusion has often been considered. Restoring our belief in the world—this is the power of modern cinema (when it stops being bad).
Many descriptions of cinematic absence have an elegiac or melancholy tone. But since Deleuze rejects the absence and loss of psychoanalysis, his whole attitude and tonality is decisively different. His writings on cinema make for an instructive counterweight to the usual nostalgic and melancholy forms of cinephilia. Yet Tarkovsky has much to say, as well, about nostalgia, melancholy, and love.
Whether following Lacan or Deleuze, most general theories of film have not proved consistently useful; they are spectacular, or fascinating, but students of film these days usually do not have a theory. Even more so, general theories of art are always too full of inconsistencies and omissions to seem very convincing. No general remarks seem applicable to all films or all art. To avoid unnecessary generalizations, therefore, I will discuss art and film on a much more limited scale, by aiming Tarkovsky's Solaris at one recent period of American film. This book studies self-reflexivity in American film from 1990 to 2002. Western art (painting, sculpture) was inarguably and unapologetically self-reflexive for almost the entire twentieth century. Self-reflexive art comments on its medium, on what art can possibly be. Self-reflexive art also places itself in a historical context, in among other artworks. My argument is that late twentieth-century American film needed to acknowledge its medium, its place in cinematic history, or else risk failing as an aesthetic object. Such an aesthetics is openly prescriptive, but most aesthetics are. What I mean by cinematic and historical self-reflexivity will be explained by the dozens of examples that make up this book.
One might think that contemporary American film would be the last place to look for art and self-reflexivity, but there are many important directors who have sustained ideas about style and expressivity. Indeed, American films specialize in artifice, and contemporary directors often comment cinematically on their use of technology and their stylistic choices. Realistic films may be an option elsewhere in the world, although even films like Abbas Kiarostami's Taste of Cherry (1997), which shows us the cameras and crew at the end, or Carlos Reygadas's Japón (2002), which makes the main character an artist, foreground their cinematic medium. Yet because of the way the film industry works, and because of its overdetermined place in film history, realism is very rarely a plausible approach for contemporary American film. For instance, almost all American movies these days are financed with stars; a producer who can get a star or two has a chance of getting some money. But with a star in a picture the film already looks unreal; what ever else the movie might seem to be about, it is certainly about celebrity and performance. Many American films take this aspect into account, therefore, and do not try to pretend that Jack Nicholson or Nicole Kidman are not actually there. American films from the 1990s that pretend that stars are not there, when they are, do not make sense.
Similarly, American films made in the 1990s ought to have acknowledged that they were, indeed, films. American culture during 1990s was more technology conscious than it had ever been before, and digital technology slowly took over the film industry. An American film that does not want to admit that the camera is even present thus lacks minimal credibility. Often, the stylistic alternatives open to American films are given as a choice between artificial Hollywood and more naturalistic or more honest independent film. But naturalism has almost always failed in recent American film. The choice is really between different kinds of artifice. The emptiness of blockbuster Hollywood films, filled wall to wall with digital effects and GameBoy characters, gives artifice a bad name. Just as Godard once said, too provocatively no doubt, that he preferred watching a James Bond movie to the philosophical pretensions of Antonioni, action movies by John McTiernan make more cinematic sense, in many respects, than the naturalistic pretensions of John Sayles. McTiernan at least makes movies, but Sayles makes movies that would rather not be movies.
Godard, we should also point out, not only makes films, he makes theories of film. Godard's example shows that films can think about film, and can start us thinking about film, as rewardingly as psychoanalysts, philosophers, or novelists. Godard regarded his early films as extensions of his film criticism, and his work has consistently obliterated the distinction between theory and practice. His films show us what the problems of cinema are; they teach us what to look for. Most film historians would agree with this description of Godard's critical significance, and in The Solaris Effect I will invoke Godard as often as Tarkovsky.
But I will also treat Tarkovsky as others would treat Godard, as someone whose films rise to the level of film theory. I want to insist on the novel idea that Tarkovsky investigates the possibilities and problems of cinema as deliberately and as searchingly as Godard. Tarkovsky's Solaris is an extraordinary meditation on cinema, and we can use that film as a model for critical interpretation. Whereas most critics see Godard and Tarkovsky as emblematic opposites, the ironic modernist versus the naive metaphysician, both share an uncommon seriousness, an extraordinary refusal to compromise, and a passionate need always to explore. Godard's longer career has left us many more artifacts of exploration, but Tarkovsky's seven films all answer repeated viewing. Each time we see Solaris, for example, we are shown again what cinematic seeing looks like and feels like. Are there other film theorists whose programs and descriptions carry such memorable conviction?
Tarkovsky's Solaris and the Coldness of Time
"I am not Rheya."
"Then who are you?"
"Rheya. . . . But I know that I am not the woman you once loved."
"Yes, but that was a long time ago. That past does not exist, but you do, here and now. Don't you see?"
Stanislaw Lem, Solaris
Andrei Tarkovsky's Solaris provides a detailed and complex model for cinematic illusion. Solaris is not explicitly about cinematic illusionism, as are films like Buster Keaton's Sherlock, Jr. (1924) or Woody Allen's Purple Rose of Cairo (1985), but it is a film implicitly about film. Interpretation makes the implicit explicit, and convinces because it works, not because it is accurate. Freudian readings or Marxist readings of literature and art are not correct, but they are often productive. Because they seem to do interpretive work, we learn more about a particular work of art and art in general by deploying Freudian or Marxist categories. Instead of turning to Freud or Lacan, I will draw on Tarkovsky's Solaris for some interpretive categories and ideas for this introductory chapter and the remainder of the book.
Solaris (1972) is Tarkovsky's third feature film, following Ivan's Childhood (1962) and Andrei Rublev (1969). The science fiction world of Solaris allows Tarkovsky to pursue what he called "poetic cinema," a cinema that proceeds by intuition and association in contrast to cause-and-effect narrative. The premise of the film is that, under the influence of the planet Solaris, the dreams of astronauts in a space station come into the reality of waking life. Solaris centers on the consciousness of the astronaut Kris, who meets once again his long dead wife, Hari. For Kris, the appearance of his resurrected wife is at once a beautiful dream and a horrific agony. His agony is that he knows, every moment, that she is a figment of his imagination. They embrace, even make love, but she is not real. Solaris is a ghost story, but with a peculiar sort of ghost. What Kris sees is exactly his wife, but the figure before him was never his wife. The relationship between Kris and his dead, perfectly real wife I take to be the archetypal relationship of audience and screen at the cinema. There is photographic reality, sensual and emotional immersion, but also a concurrent knowledge that the reality is all along an artifice, a constructed hallucination.
To use Solaris as a model for cinematic hallucination, I do not necessarily need to show that Tarkovsky intended such a reading, yet there are indications that he might go along. Tarkovsky is often characterized as a cinematic naif, the antithesis of self-conscious, self-reflexive, and sophisticated Godard. We imagine Tarkovsky strolling through the woods, talking about immortality and God. We read his lyrical and ecstatic prose, and he comes to appear indeed as someone who wants only to get transcendental beauty and emotion on camera, without any thought of the camera itself.
Yet I would argue that Tarkovsky is a deeply self-conscious director, consistently and overtly aware of expressing himself through the medium of film. Although Tarkovsky rejects the technological realism of Kubrick's 2001: A Space Odyssey (1968) in Solaris, the sci-fi framework still allows Tarkovsky to represent an ongoing consciousness of technology and communication. Television monitors, video players, audiotape recorders, and photographs play important visual and narrative roles in Solaris. Tarkovsky does not use the science-fictional bric-a-brac to think about the future of manned space flight, but he does arguably use it to think about cinematic illusion. That this impossible love takes place in a spaceship full of television monitors and video recorders helps us see that Kris's love for Hari is like our impossible desire for cinema. What we desire is contact and communication, but both the monitors and Hari ironically emphasize our solitude. We would recognize our solitude less, perhaps, if we were simply alone.
In his book The Cinema of Andrei Tarkovsky, Mark Le Fanu reads Solaris as a "film-within-[a]-film" that emphasizes the permanence of memory.
From the early sequence where the youthful Burton looks back at his older ruined self from the widescreen wall monitor; to the middle scene in which the video of the dead Gibaryan addresses Kelvin "from the far side of the grave"; to the central and profoundly moving sequence where Kelvin's own "home video" shows Hari the image of herself as she was on earth near the snow-covered dacha—everything combines to demonstrate that memory need not be extinction; and that on the contrary we live in significance to the extent that we are prepared to embrace the shadows of our loss.
But isn't this also, really, the metaphysics of film itself? Derrida calls cinema the "science of ghosts." Those actors on screen (the big screen, not just the screen-within-a-screen), aren't they also present to us and absent at the same time? And isn't this in fact what makes the cinema often so poignant? Its present tense is so often also a past tense.
As I will argue below, one of Tarkovsky's most important contributions is a reevaluation of this present absence, of this poignancy. For Le Fanu, Tarkovsky's cinematic self-consciousness is not so much self-reflexivity, a meditation on cinema, as it is a "meditation on immortality." In this view, Solaris is about love, memory, and death, and it uses cinema to get there.
But I would balance Le Fanu's examples of perfect photographic memory with the equally numerous examples of Lethean forgetfulness. These are images that empty out their presence. Hari, for instance, is an amnesiac. She is often confused as to who she is and where she came from. We ourselves are never sure what happened to the couple in the past. We know that Hari committed suicide, but we are not sure what kind of lives these characters lived before that. Indeed, from this evidence, we might well conclude that film has access only to brief moments in the past, but not to anything substantial in the end. Yes, Hari is "immortal," continually reviving herself, yet she revives with only blurred memories. In a later chapter, I will discuss the idealizing of cinematic memory through digital technology. But Tarkovsky's Solaris does not make these idealized gestures. In one of Soderbergh's most substantial annotations of Tarkovsky, his Kris says, "I was haunted by the idea that I remembered her wrong." This addition is very much in line with the spirit of Tarkovsky's earlier version.
And Hari is more than a beautiful amnesiac, for she is not even a woman at all. She is one of the most complicated aliens in all of science fiction. Neither a bug-eyed monster nor a human-like cyborg, Hari is neither machine nor human being. Tarkovsky's Solaris makes her scientific status purposefully uncertain. Her presence raises all of the same philosophical questions about identity and consciousness as those that loom around the cyborgs in Ridley Scott's Blade Runner (1982). But the problems surrounding her go deeper than those attached to robots and artificial intelligence. For her identity and her otherness are radically unclear. She has some of the same characteristics as a machine, since she is almost impossible to kill (she drinks liquid oxygen and survives). She exhibits superhuman strength when she tears right through a door of the spaceship. And the other crew members all treat her as a thing, indifferently. They call her "a mechanical reproduction, a copy." She is a machine on the edge of the human—"I am becoming a human being," she says—yet she is more like a ghost or a dream. Like cinema itself, she is a copy, a reproduction, an alien, a ghost.
It might be argued that the Solaris effect is the essence of self-conscious film. Film self-consciousness is signaled not only through self-reflexive postmodernism, but also through the insistent foregrounding of ghosts. I would encourage the reader to think back on how frequently films narrate analogies to this problematic of screen existence. This book is not a history of European film, nor does its survey of American film stand or fall on its theory, but I will remind the reader from time to time of other films that work through this same disorienting ontology. This is why thematic studies of cinematic ghosts and monsters are so important, and why the genres of horror and science fiction are so central to our understanding of film. In contemporary American film, recent works by M. Night Shyamalan, especially The Sixth Sense (1999) and Signs (2002), are not just ghost movies, but are also about the ghostliness of film representation. It is no coincidence that the great movie lover and independent director Jim Jarmusch recently made films called Dead Man (1995) and Ghost Dog: The Way of the Samurai (1999), films which are not only about death, but about the liminal, spectral nature of movies as well. Films as different as Ghost World (Zwigoff, 2000), The Man Who Wasn't There (Coen brothers, 2001), and The Blackout (Ferrara, 1997) all enact commentaries around the idea of the film screen as an imaginary signifier. Hal Hartley's No Such Thing (2001) ends the moment that his monster (borrowed from Cocteau) dies.
World cinema can often be read similarly through this kind of self-conscious elaboration of presence and absence. Akira Kurosawa's repeated theme, that the world is illusion, overlaps with an idea that the screen is there and not there. Surely the impression of Rashomon (1950) or Throne of Blood (1957) is one of life as a temporary ghostliness. The ghostliness of life becomes partly an existential, philosophical claim, but is also a commentary on the condition of film. When, at the end of Antonioni's Blow-Up (1966), the photographer vanishes amidst a vast lawn of green, we realize once and for all that we have been watching a powerful dialogue between something and nothing. The genius of Kiyoshi Kurosawa's Cure (1997) is found in its creation of a villain who is purely zero, an embodied absence. François Ozon's Under the Sand (2000) is another Solaris film, in which a woman (Charlotte Rampling), after the loss of her husband, lives her life in desperate fantasy. Hitchcock's Vertigo (1958) is yet another telling of Solaris, avant la lettre, and we will come across it more than once in the pages that follow.
The tone of Solaris is particularly significant. Solaris ought to be a tragedy, a dramatic retelling of the story of Orpheus and Eurydice. But even Cocteau's genre-disrupting Orpheus (1949) has more straightforward dramatic and tragic signals than Solaris. Le Fanu finds that the most substantial difference between Lem's novel and the film is that Tarkovsky foregrounds "human sorrow, which has only a minor place in the book." But where do we see such sorrow in the film? In moments like this: "What profound melancholy and suffering is implied in [Hari's] inexpressive subtle gaze." Although this is a dramatic situation, to say the least, Tarkovsky has deliberately avoided theatrical forms of drama. "Inexpressiveness," indeed, becomes a keynote.
Think how passively these characters respond to their remarkable circumstances. For a long time Kris is too stunned by Hari's return to register what is going on, and Hari is likewise quietly confused. Kris appears to be "emotionally dead" for most of the film, which is not a mistake on Tarkovsky's part, but a deliberate avoidance of dramatic tragedy. Electronic music surges in and out, underscoring the hauntedness of things, but not the particular crises. The plot crises (when Kris sends Hari off on a rocket ship; when she commits suicide) are thus evened out and neutralized. The felt torture and despair of the ship's inhabitants is pressed out and flattened by the inexorable weight of time.
What Solaris gives us as a model is a different way of understanding and feeling the relationship of cinematic absence and presence. Instead of psychoanalytic identification, Solaris emphasizes existential solitude. Instead of tragic loss, or Deleuzean affirmation, Solaris emphasizes ongoingness and waiting.
In any case, my mission is accomplished. What next? Return to earth? Gradually everything will return to normal. New interests, new friends. But I won't be able to concentrate on them fully. Never. Have I the right to refuse even the remotest possibility of contact with this giant with which my race has tried to establish understanding for dozens of years? Or to stay here? . . . On the station, among the things and objects we both have touched? Which still remember our presence? To what end? In the hope that she'll return? But I do not have that hope. I must wait.
Time is neither idealized nor demonized. The divisions in time caused by death or catastrophe are divisions that we cause in ourselves in the present. The ghosts in Solaris are of the dreamers' own making, a product of both love and betrayal. Cinematic ghosts do not haunt us because they are dead, but because we have betrayed them by outliving them.
Tarkovsky's Solaris models not only our perception of the cinematic object but also the status of the cinematic object as art. Besides being cinematically self-conscious in general, conscious of their identity as products of camera and photography, Tarkovsky's works are aesthetically self-conscious, conscious of their place in a field of artistic objects. Paintings occur in many of Tarkovsky's works, and Tarkovsky's characteristic long take seems to call attention to itself as a species of painting. The long takes, the artworks, and the classical music may strike the viewer as signs of pretension—look, my film too is an artwork! Yet when paintings in museums and studio galleries incorporate other paintings, this is simply called self-reflexivity, and is recognized as an appropriate option for art. Surely film, too, can provide its own preliminary cultural annotations. When Kubrick uses classical music, he does not claim for his own films the cultural status of a waltz by Johann Strauss or a sarabande by Handel. Similarly, a painting "quoted" by Tarkovsky does complicated work. Tarkovsky never simply confirms the "timeless" nature of a masterpiece in his terrifically time-conscious cinema.
Tarkovsky sets his film down amid art most deliberately in a late sequence. The three male astronauts and Hari gather together in the ship's library. Almost exaggeratedly, the library contains all kinds of representative artifacts from earth's culture. The room is cluttered with shelves of books, a bust of Homer, a small Venus de Milo, framed photographs, a stained glass window, African masks, and a violin on the wall. A beautiful chandelier hangs from the ceiling, and the room is filled with lit candles. The characters philosophize over a copy of Don Quixote. Most strikingly, a series of Brueghel paintings stands in the background, a miniature gallery. The whole room is clearly representative, not real, and frames the dialogue as surely as a frame around a picture. We are going to talk about art and culture in here, the movie says, in this room that goes out of its way to look timeless and even anachronistic (what are those candles doing here?).
And it is during this sequence, and in these confines, that Hari cries out that she is becoming human. It is here that the men call her a "matrix," a "copy," and a "mechanical reproduction." When Sartorius accuses her in these words, the Venus de Milo stands out in the background. Hari's identity does not just waver between human and inhuman, between reality and hallucination, but between art and technology. How should we categorize her existence? What should we call her artifice? Is she like a painting? Is a movie like a painting? Is a painting like a person? She weeps in despair and Kris tries to console her. The other two astronauts leave. Kris walks Sartorius back to his room.
When Kris comes back to the library, he finds Hari sitting at the table, absorbed in thought. Tragedy has dissipated to meditation. She is smoking a cigarette, turned away, and the smoke rises out of the top of her head. We turn around to look at the front of her, at her absorbed eyes, and then the film cuts to a close-up of Brueghel's Hunters in the Snow. We are plunged into a long sequence in which the camera pans all over the picture, as if to mimic Hari's eyes. A few sounds re-create images from the painting (dogs barking, bells ringing), but weird electronic music is foregrounded in the sound track. Through the music, the painting becomes haunted, ghostly. The visual idea is devastating in itself, as Hari, the mechanical reproduction, broods over a copy of Brueghel's painting. The ghost is haunted by a picture. If Tarkovsky here aligns his film with art, then, movie with painting, he draws a parallel between the ability of each picture to haunt us with its images. Art in this instance is surely not about masterpieces, but about ghosts.
In their commentary on Solaris, Tarkovsky scholars Vida Johnson and Graham Petrie idealize the artworks. The spaceship has undergone a loss of gravity at this moment in the film, and Kris and Hari now levitate in each other's arms. Johnson and Petrie say that the couple is "at last content"; they float in front of the paintings and are happy. Johnson and Petrie say that Tarkovsky's "nostalgia for the earth" explains why Brueghel's paintings so appeal to him at this moment. Yet this description will not do. Hari and Kris are not expressing contentedness; they actually look rather blank. They are never able to sink into their contentedness, for the unreality of their situation is always all too apparent. The levitation is beautiful but temporary. Their time together is really one manner of disorientation after another. And the paintings around them are strangely lit; they seem to have been reproduced on glass, and to be backlit. This scene does not imply the "naturalness" and "timelessness" of art, but instead the ghostly artifice of art. And then the next sequence begins with the revelation that Hari has drunk liquid oxygen, further eroding any idealistic reading of the levitation. Altogether, we see Hari standing before a replica of the Venus de Milo and meditating on a copy of Brueghel. She floats before the painting and then tries to kill herself. In this sequence, neither the copies nor the suicide provides an idealizing framework for art.
Tarkovsky not only situates film's relationship to art, but also works through film's relationship to nature. In Chapter Three I will examine how recent independent American film expresses an overt or secret yearning for nature as an alternative to special-effects Hollywood, or as a ground to a decentered virtual reality. But I will also express serious doubts about this return to nature. After all, the medium itself is immersed in technology and artifice. Tarkovsky, of course, has been similarly and severely criticized for his naive invocations of the natural world. Fredric Jameson has probably stated the case most succinctly:
The deepest contradiction in Tarkovsky is then that offered by the highest technology of the photographic apparatus itself. No reflexivity acknowledges this second hidden presence, thus threatening to transform Tarkovskian nature-mysticism into the sheerest ideology.33
But Tarkovsky's representation of the natural world is as self-consciously contextualized and as subtle as his references to art. Solaris is a complex discussion of art, nature, and film, in comparison to which contemporary American films often appear naive and unselfcritical.
The first section of Solaris takes place in a rustic setting, a clear contrast to the space laboratory to come. But the apparent contrast of natural earth to outer space is also complicated at once, since the waving, watery images with which the film begins recall ahead of time the swirling waters of the planet Solaris. Nature and imagination, or earth and outer space, are therefore not opposites, but clearly related pairs. As we will see, many American films, such as π and Minority Report (Spielberg, 2002), undermine themselves by celebrating a feature-length rush of artifice and then returning to nature in conclusion. By contrast, Solaris ends by ambiguously blending together the planet Earth and the planet Solaris, as the camera rises up spectacularly to reveal the family's country house now settled into the middle of a churning Solarian sea.
The natural world is framed self-consciously throughout this first section of the film. The very first image shows us a leaf floating right to left across water, water that reflects and shimmers. There are already layers to the water, and no clearly idyllic associations. Soon we meet Kris, in a pan from his feet to his head; he holds some sort of metal case in his hand, a tool or even a camera. Nature is displayed in fragments in these opening sections; space is dislocated, and we see Kris in clearly different landscapes. A horse runs back and forth, but in no coherent spatial relationship to Kris. There is, however, a clear sense of editorial arrangement, selection, and framing, with the consequence that nature seems parceled out. Hence we do not have to wait until the scenes set in the science-fictional spaceship to see machinery present in the world. Man has already imposed his technological eye on the landscape.
When the visitors arrive at the house, the boundaries of nature and culture are once again emphasized and confused. The open door makes the interior of the house seem continuous with the exterior, as if the house is quite happily missing a wall. Yet inside we find a birdcage, wooden beams showing the grain, an arrangement of flowers—all signs, in other words, of man's conquest over nature. A bearded bust of Plato or Homer continues this theme, as does a framed picture of a balloon (perhaps a reference to Tarkovsky's previous film, Andrei Rublev). Immediately the talk is about the space station, and after a short rain shower, they start watching television. The landscape is beautiful, it seems, and the house idyllic perfection. But neither landscape nor house is presented in simple terms. Kris's father says, "It's so pleasant here. This house reminds me of my grandfather's house. I really liked it. So we decided to build one just like it." Even the house, so simple seeming, is built out of desire—a copy, a duplication. As a beloved copy, this house is the very first exhibition of the Solaris effect.
The Solaris effect stands in between psychoanalysis and Baudrillard. In Baudrillard, contemporary culture can no longer be analyzed by psychology or sociology, because the alienation of subject and object has given way to the smooth screen of the network. For Baudrillard, "in our virtual world, the question of the Real, of the referent, of the subject and its object, can no longer be posed." In my own thinking, Baudrillard's description of the simulacrum, the pure simulation with no relation to reality, applies to most Hollywood films. By comparison, the Solaris effect will be evidenced in these pages by films that split through the simulacrum by referencing the real.
Reality can break into a fiction film through a documentary effect, as it does so often in Godard. Reality also gathers around certain accidental features, such as the prosthetic hands of Harold Russell in The Best Years of Our Lives (1946). In Tarkovsky's Solaris, one might attribute a hovering reality effect to the long sequence in the first third, in which we ride in an automobile through tunnels and across highways. As in Godard's Alphaville (1965), we are simply supposed to pretend that this obviously contemporary world belongs to the future. This lack of pretense ensures that reality will remain an ongoing question. Above all, however, it is self-reflexivity that makes manifest the materiality of film, by breaking through the simulacrum. The Solaris effect keeps open the question of the real, and makes work with subject and object continue to remain useful.
The Solaris Effect is a selective survey of contemporary American film. What follows will generally emphasize independent film at the expense of Hollywood film. There were hundreds of U.S. films released in the 1990s, of course, but I will talk substantially about only sixty or so. But within the limits of this survey I hope to address some of the most important debates that have sprung up around contemporary American film. In some ways, this book is old-fashioned and humanistic. It does not go into much detail about the economics or sociology of moviemaking and moviegoing. Despite my sometimes severe judgments, this book also celebrates a substantial cross-section of recent U.S. films. It does not take the approach of "The End of Cinema" at the end of the twentieth century. And while The Solaris Effect also offers an ontology of film, it is one less apparently complicated than those deriving from Lacan or Merleau-Ponty. I use the word "reality" from time to time, for example, in a rather innocent sense. Nonetheless, my hope is that, by the end of the book, my use of Tarkovsky's Solaris will have borne fruit, and that its conclusion can stand as a contribution to the debate on contemporary film aesthetics and value.
The Solaris Effect continues in the next chapter with a look at the career of Steven Soderbergh, often regarded as the most independently minded director in Hollywood. Soderbergh began his career with a landmark of cinematic self-reflexivity, sex, lies, and videotape (1989), and every subsequent film has been, among other things, a meditation on the surface of the screen. Just as Soderbergh goes back and forth between independent filmmaking and Hollywood, shuttling from no-budget movies like Schizopolis (1996) to glamorous star vehicles like Out of Sight (1998), Soderbergh's films circle back and forth between reality and artifice, refusing both Hollywood's simulacrum and the naturalism of John Sayles. Although critics and Soderbergh himself always emphasize how different Soderbergh's films are, one from another, all of Soderbergh's films work through the absent presence of cinema; they are always, with different emphases and tonalities, performing the Solaris effect. Thus Soderbergh's career has been aimed towards a remake of Tarkovsky's Solaris since the very beginning.
Following the chapter on Soderbergh, the next six chapters will take different approaches to contemporary American film and the Solaris effect. Chapter Three will look at films by Robert Redford, in addition to films shown at Robert Redford's Sundance Festival. Redford's films, such as A River Runs through It (1992) and The Legend of Bagger Vance (2000) are transparent paeans to nature, in line with the overt environmentalism of the Sundance community. But even very urban entries in the Sundance Festival, such as Darren Aronofsky's π (1998) and Marc Levin's Slam (1998), often reject artifice and attempt to ground themselves in nature. Chapter Four studies the dream aesthetic of David Lynch, by emphasizing the constructedness of Lynch's dreams. Lynch's films emerge self-consciously as dream theater, with red curtains over the portals amid the trees. All of Lynch's films also give us the opportunity to study the intense, impossible love of film known as cinephilia, so central to the Solaris effect. Here we will compare the ramifications of the Solaris effect, where Kris dreams back his alien beloved, to various examples of critical cinephilia offered by the great French film magazine Cahiers du cinéma.
In Chapter Five I will focus primarily on Steven Spielberg's A.I. Artificial Intelligence, reading it both as a poignant performance of the Solaris effect and as a self-reflexive commentary on digital cinema. Digital technology changed cinema forever in the 1990s—or did it? Against certain theorists who maintain that digital cinema is absolutely postphotographic cinema, I argue that the Solaris effect still holds, whether we are looking at computer-generated imagery (CGI) or 35 mm film. Chapter Six directly examines artists and art by describing films that are explicitly about artists and their paintings. Directors such as Robert Altman and Stanley Kubrick practice an aesthetics of meaninglessness, in which cinematic art is that which has not more meaning, but less. This meaninglessness often sits on the edge of an abyss of horror, which leads to a discussion of art in the contemporary serial-killer film. Chapter Seven consists of a broad survey of directors who work explicitly under the banner of artifice, directors like Gus Van Sant, Harmony Korine, and Todd Haynes. These directors are contrasted to those who want, in the name of nature, to opt out of the artificial. In my argument, the pretension to naturalism makes no more sense than Hollywood's pretension to global conquest.
My final chapter will most polemically argue against this pretension to naturalism. In the age of empire, American film must identify itself as ephemeral or collaborate with the globe-trotting artifacts of imperialism. Hollywood films, like cartoons, radiate omnipotence, whereas many independent films ground themselves in the immortal truths of nature. But contemporary American films must align themselves with the humility and ghostliness of the Solaris effect, or else contribute to the American effect, the McDonaldization, or the Matrixization, of everything. In my description, many American films do embrace the Solaris effect, offering themselves, in many different ways, as examples of cinematic transience.
Tarkovsky's Solaris is the perfect encyclopedia for the study of contemporary American film. Like Spielberg's A.I. and Aronofsky's π, it is a science-fiction film. Like Lynch's Twin Peaks: Fire Walk With Me and Lost Highway, it is a horror film. Solaris analyses the visual representation of the natural world, like Todd Haynes's Safe and Van Sant's Gerry. Solaris examines the role of painting in relation to cinematic art, like Altman's Vincent and Theo and Scorsese's Age of Innocence. Above all, Solaris tells the tale of impossible love that is at center of all cinema, the love of that which is both present and absent, alien and human. I argue that American film must either retell this ghost story or else tell dangerous stories of immortality. Tarkovsky is a serious director, often maligned for poisoning film with his metaphysics and his ethics. Movies ought to be fun! But when film, and the capital that follows film, is so obviously a part of U.S. cultural imperialism, an ethical response to U.S. film seems called for. So Tarkovsky points us also in the direction of an ethical response by reminding us that if movies can be so important, then they need to be judged carefully and critically.