Sound is not voice. The desire for it to be so, however, seems to lie at the heart of much compelling art, music, and film. How we feel about this desire - that to be human at all is to thoroughly take that desire for granted or, conversely, that to live in post-Enlightenment (much less postmodern) culture is to see that desire as romantic in the worst possible sense - is a question visited upon audiences with the most uncanny and disconcerting force in Lars von Trier's film Dancer in the Dark. When the film was first released in May of 2000, it provoked violently divergent responses from its audiences; even as it won the Palme D'Or at the Cannes Film Festival, the audience, as one reviewer reported, "erupted with an indecipherable storm of cheers and catcalls."

^1. Brian D. Johnson, "Singin' in the brain: Bjork hits an ethereal note as a day-dreaming martyr," Maclean's (October 16, 2000), p. 74. Further references are in the text. a Some viewers walked out of the film visibly shaken and in tears, while others just walked out - halfway through the film. This nearly unprecedented polarization carried through to the reviews themselves, which ranged from the awestruck to the dismissive and merely nasty. Some praised the film lavishly: "You've never seen anything like Dancer in the Dark," one wrote; another called it "a work of thrilling originality."

^2. Peter Travers, Rolling Stone, no. 851 (October 12, 2000), p. 99; Johnson, p. 74. b At the other end of the spectrum, a reviewer for The Nation complained that the film was about "seeing how much of the preposterous he [von Trier] can get you to swallow without gagging"; more pointedly still, another labeled it "a genuinely infantile work," an "ugly, self-indulgent folly."

^3.Stuart Klawanis, The Nation vol. 271, no. 12, p. 34; Jonathan Romney, New Statesman, vol. 129, no. 4504 (September 18, 2000), p. 44. c Even reviewers who supported the film felt the need to disavow it; while admitting that "the power of Dancer in the Dark is undeniable," David Ansen, in Newsweek, called it "a magnificent sham" riddled with "emotional sadism."

^4. David Ansen, "Light and Dark," Newsweek, September 25, 2000, p. 66. d

What is going on here?

To begin an answer to that question, we need to get a fix on what this weird and iconoclastic film is up to, how to approach it, what sorts of generic expectations we may bring to it. That, however, is only the point of entry into the much more complicated question I will take up later: the question of what we might call the film's ethical project, and how that project might well be the source of its audience's and reviewers' hysterical reactions (a term whose appropriateness will become clear, I hope, in due course). For both of these reasons - reasons of genre and reasons of ethics - I'm going to have to give a resolutely theoretical reading of this film, not because I get my kicks inflicting theory on people, (well, ok, I do get my kicks inflicting theory on people) but because I don't think we can really begin to understand what is going on in this moving, exasperating, and altogether original film without recourse to theory, in whose absence we are left with the generalizations, fulminations, and sputterings that have made the reviews - pro or con - such a waste of time.

For starters - to take up the question of genre first - we have to understand that for Dancer in the Dark, any hint of "reality," "character" in the usual sense, verisimilitude and the like are, for the purposes at hand, the merest - and I do mean the merest - vehicles for the film's deeper concerns. Here it will suffice, perhaps, to simply register the shameless melodrama of the plot: the incredibly innocent Selma (played by the pop phenom Bjork, in what nearly everyone agrees is a stunning performance), who is slowly going blind, sacrifices her own life so that her ten-year old son Gene (Vladica Kostic) may receive an operation that will save his sight from the ravages of the same congenital disease. Dancer in the Dark, in other words, is no more satisfactory, fulfilling, or compelling in terms of plausibility and Aristotelean necessity than, say, The Marriage of Figaro - and that is precisely the point. In fact, the film's power is in a profound sense inseparable from what many viewers will see as its "absurdity," if one wants to put it that way - precisely in the way that the absurdity of opera (its melodrama, its hyperbole, its staginess - all of those qualities that make people either love or hate opera too) is in fact absolutely central to its philosophical and ethical project. To put it as bluntly as possible, Dancer in the Dark, like opera, isn't about "reality"; it is about what "reality" turns away from, occludes, and denies. And in that, it is (again like opera) more real than reality - but more about that in a moment.

On the question of genre, it needs to be said, I think, that even though Dancer in the Dark invites us to take it in the genre of the Hollywood musical, this is ultimately a blind alley. The film is not a musical. While the musical insists upon as a constitutive feature the seamless continuity of the world inside and outside the musical numbers themselves - characters engaged in "realistic" dialogue amongst passers-by who (realistically) pay no attention to them suddenly break into song, and the passers-by suddenly join in - Dancer goes to great pains to insist on the radical split between the world of Selma's fantasy, in which the musical sequences take place, and the world that the film itself in broader terms constructs and inhabits. Still, the film certainly does situate itself in relation to the Hollywood musical, its conventions, and what they signify. And in this light, Dancer might be viewed as an intensification of the stakes of the Hollywood musical, taking it more seriously than it ever did itself (which may be in part what so irritates those who despise this film), and, at the same time, a deconstruction of the musical's way of imagining those stakes - a posture that the film achieves by insisting on the clear distinction between the world of the main character, Selma, and its own. In these terms, Dancer in the Dark would force the question, "would Gene Kelly be willing to die to dance with his umbrella in Singin' in the Rain? Fred Astaire put a gun to his head for the sake of dancing with his mop?" It is as if the problem were not that the musical as a genre is so preposterous that no one can sit still for it anymore, but rather that it isn't preposterous enough - which is to say that it no longer takes seriously enough, with enough extremity, its own claims.

To do so - to be that preposterous and that serious at the same time - is to move by way of thumbnail definition from the realm of the musical to the realm of opera, and to realize that on the most basic level those claims have to do with life and death, and with sound and vision as figures for experiencing the world and the loss of the world - all of which are related to the question of film as a medium, and how this film relates to the limits of that medium. From this vantage, we would do better to think of Dancer in the Dark as a kind of postmodern opera rather than a "musical." Here, Stanley Cavell's work on these questions - spanning by now several books, from the early study The World Viewed: Reflections on the Ontology of Film, through two books on Hollywood genre films, to the collections Themes Out of School and the more recent A Pitch of Philosophy - can be of some help. For Cavell, the philosophical and ethical significance of film and of voice in opera are structured by the larger problematic that occupies the whole of his work: namely, the problem of philosophical skepticism. After Descartes and Kant, skepticism names not just an epistemological problem but a more profound and deeply ethical "loss of the world" that is coterminous with Enlightenment modernity itself, in which the modern condition is to be "homeless" in the world, permanently doomed to "haunt" it rather than inhabit it, as Cavell sometimes puts it. For Cavell, the significance of film and of operatic voice are located at what he calls the "crossing" of the lines of skepticism and romanticism - that is to say, the juncture at which our desire for contact with the world of things and of others, our need to believe that what we know, experience, and love is of the world, is crossed by our knowledge that we are profoundly and permanently isolated, locked (as Cavell's mentor Emerson puts it) in "a prison of glass."

The most famous version of the "settlement" with skepticism, Cavell argues, is probably Kant's in The Critique of Pure Reason, which argues that (1) Experience is constituted by appearances. (2) Appearances are of something else, which accordingly cannot itself appear. (3) All and only functions of experience can be known; these are our categories of the understanding. (4) It follows that the something else - that of which appearances are appearances, whose existence we must grant - cannot be known [the famous ding an sich or "thing in itself"]. In discovering this limitation of reason, reason proves its power to itself, over itself. (5) Moreover, since it is unavoidable for our reason to be drawn to think about this unknowable ground of appearance, reason reveals itself to itself in this necessity also.

^5.Stanley Cavell, In Quest of the Ordinary (Chicago, University of Chicago Press, 1988), p. 30. Further references are given in the text. e

The dissatisfaction with Kant's settlement with skepticism is readily imaginable, of course, but what is less clear - and even more important to Cavell - is the "companion satisfaction" that is "expressed in Kant's portrait of the human being as living in two worlds, in one of them determined, in the other free.... One romantic use for this idea of two worlds lies in its accounting for the human being's dissatisfaction with, as it were, itself...as if the one stance produced the wish for the other, as if the best proof of human existence were its power to yearn, as if for its better, or other, existence. Another romantic use for this idea of our two worlds is its...insight that the human being now lives in neither world, that we are, as is said, between worlds" - a condition Cavell characterizes as the endemic "worldlessness" or "homelessness" that is of a piece with the modern condition (Quest 31-2). All of which makes philosophy - in a phrase that has obvious resonance for the character of Selma in Dancer in the Dark - "a philosophy of immigrancy, of the human as a stranger" (Pitch xv), in light of which Selma's encroaching blindness in the film might be read as a figure for the inevitability of the general human condition of being "in the dark," wandering in a world of shadows and specters, never at home but merely, sometimes, at rest.

For Cavell, then, the skeptic is neither a "knave" nor a "fool" but rather "foregoes the world for just the reason that the world is important." In fact, Cavell argues, far from "foolish," it is "one way to describe the tragedy King Lear records."

^6.Stanley Cavell, Must We Mean What We Say? (New York: Scribners, 1969), pp. 322-3, qtd. in Stanley Cavell, A Pitch of Philosophy: Autobiographical Exercises (Cambridge: Harvard University Press, 1994), p. 114. Further references to this latter text are given internally. f If we understand that "the idea of the voice, or rather of its silencing," is "the goal of skepticism," then we can see why, "as an answer, voice comes too late; Desdemona is suffocated; Cordelia is hanged, and the last thing Lear looks upon, as it was about the first thing we knew he cared about, is her mouth" (Pitch 116). Along the same lines, the philosophical import of voice in opera, then, is that it communicates that "we may leap, as it were, from a judgment of the world as unreal, or alien, to an encompassing sense of another realm flush with this one, into which there is no good reason we do not or cannot step, unless opera works out the reasons. Such a view," Cavell continues, "will take singing, I guess above all the aria, to express the sense of being pressed or stretched between worlds - one in which to be seen, the roughly familiar world of the philosophers, and one from which to be heard," a world "to which one releases or abandons one's spirit," and which "recedes when the breath of the song ends" (Pitch 144).

The resonance of this formulation for the character of Selma is clear enough, I think, and it is only sharpened by Cavell's suggestion that "Kant's vision of the human being as living in two worlds" (Pitch 141) corresponds roughly to "two general matching interpretations of the expressive capacity of song: ecstasy over the absolute success of its expressiveness in recalling the world, as if bringing it back to life; melancholia over its inability to sustain the world, which may be put as an expression of the absolute inexpressiveness of the voice, of its failure to make itself heard, to become intelligible." This last - abandoning one's spirit to and giving voice to a world that no one will hear - is "evidently a mad state," Cavell adds, and it is one that "seems to be reserved for the women of opera" (Pitch 140), and in the case at hand, of course, for Selma. Here, Cavell is responding to Catherine Clément's assertion in Opera, or the Undoing of Women, that "opera is about the death of women" - that is to say, it is about the "countless forms in which men want and want not to hear the woman's voice...to know and not know what she knows about men's desires" (Pitch 132) - a claim that Cavell will modulate into the rather different (and, shall we way, more strictly philosophical) assertion that woman's singing "exposes her as thinking, so exposes her to the power of those who do not want her to think" (146), in which case she becomes, for Cavell, a figure for "that philosophical self-torment whose shape is skepticism, in which the philosopher wants and wants not to exempt himself from the closet of privacy, wants and wants not to become intelligible, expressive, exposed" (132).

The stakes of this revisionist relationship to Clément's thesis are, perhaps, apparent enough for a film that ends with Selma's death by hanging - does she die because she's a woman? Because she thinks? Because she sings from a world that imagines the two might coincide? But the stakes of that revision are complicated by Cavell's surprising suggestion that the "mad state" reserved for women in opera usually takes place "only after their words can treat some difficulty internal to their marrying," as if "skepticism is narratively figured as an assault on marriage" (Pitch 140-1). What we find in the woman's operatic voice is then exposure "to a world of the separation of the self from itself, in which the splitting of the self into speech is expressed as the separation from someone who represents to that self the continuance of the world...in whom one's expectation of intelligibility has been placed, and collapses" (Pitch 151). Moreover - and this extends Cavell's contention above about Lear, that voice "always comes too late" - this thematization is redoubled in Hollywood film in the theme of remarriage (Cavell has written a whole book about it, Pursuits of Happiness: The Hollywood Comedy of Remarriage), which suggests "that the validity of the bond of marriage is assured...by something I call the willingness for remarriage, as a way of continuing to affirm the happiness of one's initial leap. As if the chance of happiness" - the chance of continuing to sing and dance, to hope, in the face of skepticism - "exists only when it seconds itself."

^7.Stanley Cavell, "The Thoughts of Movies," in Themes Out of School: Effects and Causes (San Francisco: North Point Press, 1984), p. 13. Further references are in the text.g

Now what is interesting about this aspect of Cavell's thesis, of course, is that in Dancer in the Dark, what cannot be missed is the matter of Selma's conspicuously absent husband and her equally conspicuous rebuffs of her suitor, Jeff (played by Peter Stormare). And while she permits a certain amount of intimacy with her neighbor and landlord Bill (David Morse) (they talk late into the night about going to musicals when they were kids, and so on), we are to understand that this is possible only because Bill is married - as if, in Cavell's terms, Selma's ability to continue to believe in the world that is rapidly receding from sight resides not in the possibility of her (re)marriage (hence her repeated rejections of Jeff's overtures in the film's most important musical number, "I've Seen It All") but in her handing down the gift of (continued) sight to her son, as if the only way of ensuring the continued existence of the world is not marriage and what it signifies but rather the rejection of marriage in an act of sacrifice that might be characterized as radically feminine in its rejection of the nuclear heteronormativity that (at the very least) lurks in the background of Cavell's speculations on (re)marriage - questions I will return to in some detail below.

This relationship of the two worlds - of vision, associated with epistemology and sense certainty on the one hand, and of voice, associated with the loss of the world under skepticism and the hope of its recovery on the other - is complicated by Cavell's contention that the ethical and philosophical project of opera was at a certain point taken over by film - a contention he bases on analogizing "the camera's powers of transfiguration to those of music, each providing settings of words and persons that unpredictably take them into a new medium with laws of its own" (Pitch 136, 137). Just what those "settings" are may be clarified by Cavell's adaptation of Heidegger's famous thesis on the broken tool. Film, for Cavell, "is a phenomenon in which a particular mode of sight or awareness is brought into play" by "a disruption of what Heidegger calls the 'work-world,' a disruption of the matters of course running among our tools, and the occupations they extend. It is upon the disruption of such matters of course (of a tool, say by its breaking)" that we find, to use Heidgger's phrase, "the worldhood of the world announcing itself" in all its conspicuousness and obstinacy, its thereness (Themes 174). "We have here to do," he continues, "with something about the human capacity for sight" - and here the link with the issue of skepticism becomes clear - "or for sensuous awareness generally, something we might express as our condemnation to project, to inhabit, a world that goes essentially beyond the delivery of our senses" - in view of which one may read Buster Keaton in The General, for instance, "to exemplify an acceptance of the enormity of this realization of human limitation, denying neither the abyss that at any time may open before our plans, nor the possibility, despite that open possibility, of living honorably" (175). More to the point for our purposes, the same might be said of Fred Astaire's dancing, which, far from being "escapist" (as is usually charged with the Hollywood musical) "is meant as a removal not from life but from death," as "facing the music, as a response to the life of inexorable consequences" (Themes 23) - a reading that would seem to apply quite poignantly to Selma's musical fantasies in the face of blindness and, eventually, of death itself.

Keaton's comedy, Astaire's dancing, and Selma's musicals all "face the music" of skepticism in the same way that the aria does in opera, but the difference is that film "democratizes the knowledge, hence at once blesses and curses us with it," by telling us that it is as available to all "as the ability is to hold a camera on a subject, so that a failure so to perceive, to persist in missing the subject, is ascribable only to ourselves." The philosophical and ethical problem of inhabiting "a world that goes essentially beyond the delivery of our senses" is only intensified, in other words, in film by virtue of its very medium, and our sense of film's specific relation to this problem will be sharpened if we pay attention to Cavell's distinction between painting and photography (and film, for him, as a mode of the latter). After the advent of photography, what painting wanted, he suggests, "was a sense of presentness - not exactly a conviction of the world's presence to us, but of our presence to it. At some point the unhinging of our consciousness from the world [as in the "fall" into the truth of skepticism after Kant] interposed our subjectivity between us and our presentness to the world. Then our subjectivity became what is present to us, individuality became isolation. The route to conviction in reality was through the acknowledgment of that endless presence of self."

^8.Stanley Cavell, The World Viewed: Reflections on the Ontology of Film, Enlarged Edition (Cambridge, Mass.: Harvard UP, 1979). p. 22. Further references are given in the text.h Photography, on the other hand - and with it film - "overcame subjectivity in a way undreamed of by painting, a way that could not satisfy painting, one which does not so much deflect the act of painting as escape it altogether: by automatism, by removing the human agent from the task of reproduction.... To maintain conviction in our connection with reality, to maintain our presentness, painting accepts the recession of the world. Photography maintains the presentness of the world by accepting our absence from it. The reality in a photograph is present to me while I am not present to it" (The World Viewed 23).

So it is that film for Cavell has a kind "magical" ability to capitalize on "the idea of and wish for the world re-created in its own image"; film meets the threat of skepticism "not by literally presenting us with the world, but by permitting us to view it unseen," "as though the world's projection explains our forms of unknownness and of our inability to know" (The World Viewed 40). There is an important reversal here; in fact, two reversals. If music and voice as we find them in opera met the loss of the world under skepticism by an assertion that we nevertheless miraculously exist - in this sense, in another register, music and song come to the rescue of language after skepticism, as Cavell sometimes puts it - it did so only at the price of acknowledging that the world of things was always already lost, gone. In photography and film, on the other hand, the existence of the world is miraculously affirmed via automatism, but the price we pay for the world's recovery is that it no longer exists for us, it just exists. We can't know or touch the world precisely because it manifests itself unbidden, without our help - film is what the world looks like when we're not there.

What is most interesting here - especially for the purpose of discussing a film such as Dancer in the Dark, in which the relationship between the visual and the auditory is so important - is Cavell's insistence that while "we don't know how to think of the connection between a photograph and what it is a photograph of" - "The image is not a likeness," he rightly insists - "one might wonder that similar questions do not arise about recordings of sound" (The World Viewed 18). "Is the difference between auditory and visual transcription," Cavell asks, "a function of the fact that we are fully accustomed to hearing things that are invisible, not present to us, not present with us? We would be in trouble if we weren't so accustomed, because it is the nature of hearing that what is heard comes from someplace, whereas what you can see you can look at.... [W]e are not accustomed to seeing things that are invisible, or not present to us, not present with us.... Yet this seems, ontologically, to be what is happening when we look at a photograph" (18). The idea here is that with the visual, the lines of determination, if one wants to put it that way, run from the subject to the object, to what we "look at," and hence the magic of the photograph and of film is that our role in so making the world manifest is suddenly removed from the equation. With sound, on the other hand, the lines run from the object - "where sound comes from" - to the subject, so that a corollary magic would involve our insertion into the equation, as if we had to actively listen, just as we actively direct sight, to hear anything at all.

What I want to suggest is that something like this reversal is exactly what happens in Dancer in the Dark, with profound implications for how the film stages the relationship between the auditory and the visual and, within that, the relationship of both of these to the project of film as a medium. There are two dynamics at work here, and it is crucial to disarticulate them: namely, Selma's drama and its philosophical and ethical significance, and what the film, from a quite different vantage, does with that drama. It is here, I think, on the strength of this disarticulation, that we can begin to sense some of the limits of Cavell's work - the extent to which it is, we might say, Selmacentric. Let us return briefly to Cavell's account of visual versus auditory transcription in The World Viewed. "Suppose one tried accounting for the familiarity of recordings by saying, 'When I say, listening to a record, 'That's an English horn,' what I really mean is, 'That's the sound of an English horn'; moreover, when I am in the presence of an English horn playing, I still don't literally hear the horn, I hear the sound of the horn. So I don't worry about hearing a horn when the horn is not present, because what I hear is exactly the same...whether the thing is present or not.' What this rigmarole calls attention to is that sounds can be perfectly copied, and that we have various interests in copying them" (368).

What is interesting here, of course, is how Cavell's discussion of visual versus auditory transcription would appear to take its place as part of that film theory that "has assured us," as Kaja Silverman puts it, "that there is no difference between recorded and prerecorded sounds - that the apparatus is miraculously capable of capturing and retransmitting the profilmic event in all its auditory plenitude," so that "with each new testimonial to the authenticity of recorded sound, cinema seems once again capable of restoring all phenomenal losses."

^9.Kaja Silverman, The Acoustic Mirror: The Female Voice in Psychoanalysis and Cinema (Bloomington: Indiana University Press, 1988), p. 42. Further references are in the text.i As she points out, however, it is relatively easy to demonstrate that every acoustic event is inseparable from the space in which it occurs, and that in sound recording as in image recording, the technological apparatus in question is always highly selective, isolating and intensifying some features and ignoring others. Indeed, as Douglas Kahn has exhaustively shown in his study Noise Water Meat: A History of Sound in the Arts, reviewed by Allison Hunter in ebr the kinds of qualifications voiced by Silverman lie at the very heart of the huge body of work in sound art in the twentieth century, much of which foregrounds the technological mediation and environmental embeddedness of sound as a medium (think, for example, of John Cage or Alvin Lucier, to name only two well-known examples).

^10.Douglas Kahn, Noise, Water, Meat: A History of Sound in the Arts (Cambridge, Mass.: MIT Press, 1999). Kahn's chapter, "Drawing the Line: Music, Noise, and Phonography" is particularly instructive in this connection.j In this light, what is interesting and symptomatic is that Cavell's remarks on auditory versus visual transcription would seem to reintroduce the very kind of phenomenological plenitude that film "automatically" delivers in Cavell's account, but without the attendant clarification that that plenitude is the product of fantasy only - that it is "magic" to imagine that the world automatically registers itself, without our help.

To linger over this moment in Cavell's work is to realize that there is a crucial and altogether symptomatic aporia - in fact a double aporia - at the heart of Cavell's understanding of voice in relation to sound. As for the first, remember that for Cavell sound and voice are in the deepest sense not continuous but opposed: voice aligned with the subject (it takes over the function of the Word after language has been subjected to the withering force of skepticism), and sound with the object world (as that which comes from the world to the subject, as it were unbidden). But it is difficult to see how the difference between sound and voice can be maintained as a constitutive ontological difference, how the interiority of voice as expression can be quarantined from the exteriority that is its medium and condition of possibility. To put it as concisely as possible, voice and sound exist along a continuum, not a divide, which is simply to say, in another register, that one person's voice is another person's noise (indeed, what else would most contemporary music seem to be about?) - a point hardly laid to rest by appeals to the generic norms of opera or any other art form.

More important than this, however, is the second aporia, of voice itself. As Cavell explains it, film "reverses the acsension in theater of character over actor"; in theater, the emphasis is that "this character could (will) accept other actors," which thus figures "the fatedness in human existence, the self's finality or typicality." In film, the actor, not the role, is predominant, and this is a vehicle for film's (democratic) emphasis on "the potentiality in human existence, the self's journeying." In opera, however, the relation of actor/singer and role "is unimportant beside the fact of the new conception it introduces of the relation of voice and body," in which "this voice is located in - one might say disembodied within - this figure, this double, this person, this persona, this singer, whose voice is essentially unaffected by the role" (Pitch 137). There are doubles and there are doubles, however, and what makes Cavell's account here fascinating is its radical ambivalence about the voice "disembodied within" - but within what? Here, the Cavellian voice would seem legible - and it is not at all clear that Cavell would disagree - as a variety of what Slavoj Zizek calls "the Cartesian subject in all its abstraction, the empty punctuality we reach after subtracting all its particular contents" - a kind of aesthetico-philosophical expression of the subject of "abstract" democracy, without regard to race, or sex, or religious preference, and so on, in which what we might call the "principle" of voice is "disembodied within" a subject whose contingent features are unimportant, as in Cavell's "this figure, this double, this person, this persona" - in short, this etcetera.

^11.Slavoj Zizek, Looking Awry: An Introduction to Jacques Lacan Through Popular Culture (Cambridge, Mass.: MIT Press, 1991), p. 163. Cavell himself makes the connection, in his way, in his analysis of Emerson's rewriting of Descartes "proof" of the self in In Quest of the Ordinary (Chicago: University of Chicago Press, 1988), p. 109. Emerson, of course, is Cavell's preeminent philosopher of the democratic promise. See also here Cavell's contention, immediately after the passage just cited, that "A Cartesian intuition of the absolute metaphysical difference between mind and body...appears to describe conditions of the possibility of opera" -- a contention that, unfortunately, Cavell leaves telegraphic at best (Pitch 138). k What this suggests, I think, is that voice in Cavell is a figure for presence, but a presence that - as in Descartes and Kant - should not be confused with substance and is, indeed, based on the transcendence of substance. In more general terms, then, the apparent opposition of sound and voice in Cavell - the first aporia I touched on above - is subtended by the more fundamental continuity of presence that links them. To put it another way, in Cavell voice and sound break along the divide marked, as it were, by a skepticism that does not question presence itself but merely distributes it to either the subject or the object in a homogeneous epistemological economy in which possession of presence by one is balanced by the deficit of the other.

It is here, I think, that the Lacanian schema of the subject I have already invoked by way of Zizek and Silverman, with its interweaving of the two "sides" of voice and sound, Symbolic and Real, and so on, may be of help. Silverman, for example, in her pathbreaking study The Acoustic Mirror: The Female Voice in Psychoanalysis and Cinema, insists, following Lacan, that meaning and materiality, subject and object, are always coimplicated and interwoven in a symbolic and psychic economy of imbalance constituted by the lack at the center of the subject, who can be subject only insofar as he has acceded to the dictates of a Symbolic order not his own, in what she characterizes as a "pre-Oedipal castration" of "a subject who is structured by lack long before the 'discovery' of sexual difference, a subject whose very coherence and certitude are predicated on division and alienation" (16). This interweaving of presence and absence has for her particular and direct implications for reading the engendering and embodying of voice in film. Classical cinema, Silverman writes, "requires the female voice to assume similar responsibilities to those it confers upon the female body," where it operates as a fetish "filling in for and covering over what is unspeakable within male subjectivity. In her vocal, as in her corporeal, capacity," she continues, "woman-as-fetish may be asked to represent that phenomenal plenitude which is lost to the male subject with his entry into language," though she is "more frequently obliged to display than to conceal lack - to protect the male subject from knowledge of his own castration by absorbing his losses as well as those that structure female subjectivity" (38-9).

Here, of course, we can't help but recall Cavell's contention that opera is about the "countless forms in which men want and want not to hear the woman's voice...to know and not know what she knows about men's desires" (Pitch 132). But it is just as crucial here to remember that in Cavell's account, "women's singing exposes them to death" is rewritten specifically in terms of "exposes her as thinking, so exposes her to the power of those who do not want her to think" (Pitch 146) - in which case she becomes for Cavell a figure for "that philosophical self-torment whose shape is skepticism, in which the philosopher wants and wants not to exempt himself from the closet of privacy, wants and wants not to become intelligible, expressive, exposed" (132). In making this turn, however, Cavell would seem to take away with one hand what he has given with the other, and that is, of course, the specificity of female embodiment in relation to voice. It is true that in Cavell's reading of both film and opera the body serves the crucial function of "tethering" (to use J. L. Austin's term) back to the human, to the question of being human, the words that have, under the pressure of skepticism, floated free of the realm of the "ordinary" and "everyday." This gets handled in any number of ways in Cavell's work - for example, in his insistence that in film, this specific body of this specific actor overtakes the character played by the actor. On this issue of voice, however, as we have already seen, the problem is not so much that the subject is put under a rigorous (democratic) abstraction by voice, but rather that that abstraction and its opposite - the materiality and contingency of the subject - remain too pure in their opposition.

A similar double gesture is at work in Cavell's use of Freud's distinction between orality and vocality (in Freud's essay "Negation") to account for the at once "primitive" (or "bodily") and "sophisticated" (or "performative") power of the voice. Cavell wants to capture the interlacing of "the spectacular vocality of opera in its aspect as orality and in its aspect as exposure or display, sometimes named seductiveness" (Pitch 145); for Cavell, the power of "voice in opera as a judgment of the world on the basis of, called forth by, pain beyond a concept" is itself rooted in the "the oral, primitive basis of judgment," as explained in Freud's theory in "Negation," where introjection and expulsion from the body are the origins of affirmative or negative judgments, so that, for Cavell, the very drawing and expelling of the breath in singing enacts a kind of ur-dialectic between bringing the world nearer (overcoming skepticism) and then pushing it away in a transcendence that is also a mourning (Pitch 148). But here, one would simply want to point out, by way of Lacan and his inheritors such as Silverman and Zizek, that the drives (including orality) are always already denaturalized because they are accessible only retroactively by means of the Symbolic itself. From this vantage, the fundamental issue with the voice's power is not whether it can be tethered, via the body ("orality"), to the world of the Cavellian ordinary and everyday (thereby ensuring us that the pain of the operatic voice remains real and not, as it were, merely epistemological). Rather, the body itself is already denaturalized and "derailed" (to use Zizek's term) by the Symbolic order, so that the "primitive" basis of voice (the drive), rather than "coming first" as in Cavell, is instead a retroactively determined and "excessive" product of the Symbolic, of desire, in a psychic economy characterized above all by imbalance. All of which is to say that the suggestive correspondence between the Lacanian theory of the split subject of desire and Cavell's reading of "singing as (dis)embodied within the doubleness of the human" and "the splitting of the self into speech" (Pitch 151) is and should remain only that - suggestive.

^12.Though I cannot pursue the matter here, it should be pointed out that this is taken up rather explicitly by Cavell (in his way) in A Pitch of Philosophy, pp. 145-151, where he discussed Clément's debt to (and waning interest in) Lacan and -- even more importantly -- attempts to make use of the Freudian distinction between "primitive" orality and "sophisticated" vocality. l

Meanwhile - to clarify the stakes of some of this for the film itself - it is clear enough that Selma is doubly marked by figures of castration (indeed, by the most canonical such figures there are) in her encroaching blindness and in her death by hanging. But the question - if we have turned now from Cavell's terms to those of psychoanalysis - is the nature of this castration, what it is supposed to signify. Is this, as Silverman might suggest, about killing off the feminine and maternal body in the services of phallic disavowal of pre-Oedipal castration, in which Selma is sacrificed for those losses she is made to bear? Or is something else going on here? Indeed, what is most interesting and most important about Selma's castration, if we want to call it that, is not that it robs her of agency, but - quite the reverse - that it makes her the film's maximum example of agency. Moreover, the force of her agency would seem to increase in direct proportion to her growing loss of vision and the increasingly melodramatic "absurdity" of her situation and how she responds to its mounting crisis. This is made clear in any number of ways, not least in her steadfast refusal of the otherwise advantageous romantic overtures by Jeff, her assertion, in the film's most powerful musical number, that "I've seen all I need to see," even though he suggests that by marrying him, seeing Niagara falls, having grandchildren, and so on, that the world will, in Cavellian terms, be restored to her in all its heterogeneity. (Most important here, perhaps, is the crucial motif of Selma's absent husband - a point to which I will return in a moment.)

We can clarify the status of castration in relation to the feminine and the Symbolic in the film most readily, perhaps, by recourse to Silverman's fascinating discussion of how in Hollywood "castration is not the only trope through which dominant cinema conflates the female voice with the female body" (63). Here, she takes issue quite pointedly with Michel Chion's formulation in La voix au cinéma that "In much the same way that the feminine sex is the ultimate point in the deshabille (the point after which it is no longer possible to deny the absence of the penis), there is an ultimate point in the embodiment of the voice, and that is the mouth from which the voice issues" (qtd. Silverman 50). In Silverman's estimation, Chion here simply reproduces on theoretical terrain Hollywood's conflation of "the female voice with the female body" which organizes "female sexuality around the image of...'the insatiable organ hole'" that may be figured as either mouth or vagina (63). With this turn, "the interiority which Hollywood imputes to her has nothing whatever to do with transcendence or Cartesian cogitation. On the contrary, that interiority helps to establish the female body as the absolute limit of female subjectivity.... [W]oman's psyche is only a further extension of her body - its other side, or, to be more precise, its inside" (64). What this means, for Silverman, however, is that "the yawning chasm of a corporeal interiority" that "is posited as a major port of entry into her subjectivity" in better viewed, instead, as "the site at which that subjectivity is introduced into her," with the voice "the preferred point of insertion" (67). In short, the female voice and with it the mouth from which it issues are the point of entry for the phallus, the Law, and the Symbolic into female subjectivity.

Here, however - valuable as it is for exposing some of the problems with Cavell's work on voice, opera, and film - we glimpse something like the limit of Silverman's thesis for understanding Dancer in the Dark - or perhaps we should say that we begin to understand how radically Dancer departs from the Hollywood conventions critiqued by Silverman. For what is unmistakable in Selma's drama is the unmasking of the Law as a senseless, contingent machine, constructed utterly by self-instantiation across a void. The film makes this clear in any number of ways, from the adjacent drama of Bill, her landlord policeman who betrays Selma and steals her money to pay off bills run up by his free-spending wife, to the almost sado-masochistic courtroom drama and the construction of Selma as a murderer, to the fact that "justice" and death by hanging for Selma are determined in the end not my justice but by money, and so on. In light of all of this, we might give a rather different interpretation than Silverman's to the altogether unavoidable matter of Bjork's performative relationship to the mouth and tongue as site of the female voice; at key moments in the musical numbers, it erupts into a kind of fleshy protuberance or wall, with the tongue blocking entry into the interiority of the female subject as the voice soars. Here, the performative use of the mouth and tongue uncannily expresses not the "entry" of the Symbolic and the phallic Law into the feminine subject via the "organ hole" of the mouth and voice - not "the site at which that subjectivity is introduced into her" (Silverman 67) - but rather its rejection and blockage, which coincides with the raising of the female voice itself to its highest registers.

^13.As I suggest near the end of this study, Judith Butler's theorization of "the lesbian phallus" in Bodies That Matter would be very much to the point here, in light of which Bjork's tongue performatively signifies the "phallic" rejection of the Symbolic, of marriage and the need of a husband, etc.--all that is thematized by Selma's assertion that "I've seen all I need to see."m

I would like to take this striking performative punctuation of the film as an index of the fact that there is another, more profound sense of "the feminine" at work in Dancer in the Dark - a sense that perhaps accounts for the wild ambivalence and hysteria that have greeted the film so far. Here, Zizek's work on sacrifice, suicide, and "the act as feminine" will help us understand that there are two different aspects of the "feminine" at work in the film. The point of agreement between Silverman and Zizek would no doubt be what distinguishes both from Cavell - their insistence via Lacan that any relationship to the world of the object, the Thing, the body, the drives, and so on is always riven with difference and denaturalized. They would disagree, however, on the ethical ramifications of this fact vis a vis the question of the feminine. Where Silverman would find in the phallic regime of Hollywood film the displacement of pre-Oedipal losses onto the feminine body and voice, Zizek would identify the phallic itself with such losses and would therefore locate "the feminine" at the very core, and as the very truth of, the phallic. This is so in Zizek's reading because the phallus in Lacan as the "origin" of desire is not "natural," not given as such, but is instead a signifier - which is to say that desire and the phallus that constitutes it are socially produced and culturally determined, so that the Real (of the so-called "drives," the biological, the body, and, of course, the feminine body in contrast to male cogito) becomes accessible only by being retroactively posited as "original" and "natural" by the contingent and diacritical system of the Symbolic itself. As Zizek puts it, the phallus-as-signifier thus operates - against the clichéd notion of the phallus as "the siege of male 'natural' penetrative-aggressive potency-power" - as "a kind of 'prosthetic,' 'artificial' supplement; it designates the point at which the big Other [the Symbolic], a decentered agency, supplements the subject's failure," its "lack of co-ordination and unity."

^14.Slavoj Zizek, The Plague of Fantasies (London: Verso, 1997), pp. 135-6. Further references are in the text.n

Zizek explores this theme in any number of registers, including romanticism's commonplace of "madness as the positive foundation of 'normality'" (which "clearly announces the Freudian thesis that the 'pathological' provides the key to the normal"). Most interesting for our purposes, perhaps, is his example of the Enlightenment idea that it is blindness itself that provides the key to understanding the logic of vision, in the same way that, in Malebranche, "the 'pathological' case of feeling a hand one does not have" in fact "provides the key to explaining how a 'normal' person feels the hand he actually possesses." In "strict analogy" to Lacan's claim that "A madman is not only a beggar who thinks he is a king, but also a king who thinks he is a king" - that is, "who directly grounds his symbolic mandate in his immediate natural properties" - Malebranche claims that a madman is not only he who feels his missing hand without having one, but also he who feels the hand he really has, "since when I claim to feel my hand directly, I confound two ontologically different registers: the material, bodily hand and the representation of a hand in my mind, which is the only thing I am actually aware of" (Plague 142-3). And this, in turn, is strictly analogous to the status of the phallus itself as prosthetic, since it too is referenced to the "natural" body and yet can only be experienced through mediation by the regime of the signifier and the Symbolic.

There is an important point of contact here between Zizek's account of the Lacanian phallus and the set of terms that cluster in Derrida's work around the prosthesis, the supplement, and so on - a point to which I will return in a moment. For now, however, what needs to be registered for us to understand the status of the feminine in relation to the film's ethical project is that the truth of the phallus is the truth that the subject is always already a prosthetic subject, always in need of the prosthetic supplement provided by precisely that which is castrating in the first place, thus generating - in contrast to Cavellian skepticism - a constitutively unbalanced psychic economy driven by what Zizek calls "the loop of (symbolic) castration" (Plague 135). The prosthetics of subjectivity are registered and thematized in all sorts of obvious ways in the film, of course, and most obvious of all is the conspicuous fact of Selma's failing eyesight and the various strategies used to supplement it: the crib sheet she uses at her visit to the eye doctor, for example, which she memorizes so that she can pretend to read the eye chart and keep her job; the Coke-bottle eyeglasses she shares with her son Gene like a prosthetic mother-son bond, and which fall to the floor in a cut shot at the moment of her hanging, as if to suggest that only in death does one escape the prosthetics of subjectivity; and the fact that the "natural," originary state of being sighted can only be achieved for Gene by means of surgical intervention - a kind of weird literalization of the Freudian notion of retroactive causality. Other examples abound. Most fascinating of all, perhaps, is the scene in which Selma and Kathy attend the movies to watch a Hollywood musical. Here, however, "watching the film" takes the following form: Kathy tells Selma in a verbal blow-by-blow what is going on on the screen that she cannot see, only to have that linear account interrupted by a running argument which erupts with another patron a few rows up who is irritated by her talking. For Selma, watching the film consists of seeing nothing and hearing a soundtrack, overwritten by a verbal account, derailed by a shouting match - all of which, it should be added, she gleefully takes in.

This scene invokes, of course, the central prosthetic thematization of the film - how with failing sight the realm of sound becomes more and more Selma's way to "bring the world nearer" (to use the Emersonian phrase invoked by Cavell) - a fact painfully evoked in Selma's jail cell on death row as she desperately presses her ear to the ventilation grate in the deafening silence, trying to hold onto to one last aural thread of the world around her. At this precise juncture, however, it is crucial to insist on the difference between what the prosthetic relations of vision and sound mean to Selma and what they mean to the film, the better to understand the ethical project that drives the film's use of Selma as a character and vehicle. For what cannot be missed by any viewer, I think, is the striking, even jarring, difference between how the film is shot "inside" and "outside" Selma's fantasy musical scenes, with the former in vivid color and carefully - in fact remarkably - choreographed and edited (with footage taken, reportedly, from 100 digital cameras used to film each sequence), and the latter presented in washed out sepia tones in the best cinema verité or documentary style.

Now the point here is not, of course, some untenable distinction between the "cooked" and the "raw," the "artful" and the "authentic," the mediated and the umediated, and so on, but rather the film's startling and principled insistence on this visual difference. All of which, we must remember, is framed by a question of genre: namely, why the Hollywood musical as the mode of Selma's fantasies? The most succinct answer, I think, is to say that for Selma the Hollywood musical uses music, song, and voice to prosthetically assume the functions of "cognitive mapping" (to use Fredric Jameson's well-worn phrase) usually reserved in the ontological tradition for the visual, in which a world - one should properly say at such a point the world - presents itself in evidence, as it were, before the gaze of the "centered" subject around which the world of tables and chairs (or in Gene Kelley's case, umbrellas) coalesces. To put it in psychoanalytic terms, in the Hollywood musical it is as if the fantasy structure of "normal" vision itself is laid bare. For Lacan, "if I am anything in the picture, it is always in the form of the screen...the stain, the spot"; "in the scopic field," he continues, "the gaze is outside," it belongs to, as Stephen Melville puts it succinctly, "not the (small 0) other but to the Other - language, world, the fact of a movement of signification beyond human meaning."

^15.Lacan is quoted in Stephen Melville, "In the Light of the Other," Whitewalls 23 (Fall 1989):18-20.o

In the Hollywood musical, on the other hand, there is no point from which I am seen that I cannot see. In Lacan's analysis, opacity rather than transparency constitutes the structure of visuality, but is not the world of the Hollywood musical above all a world that is not opaque, a world of transparency where objects - like Fred Astair's mop in the famous dance number - are immediately meaningful and obey our every whim, where the infirmity and foreigness of the body itself is suddenly transcended as we "dance dance dance!" (to borrow the lyrics from the film's musical number "Cvalda")?

As we have seen, it is precisely this willingness to hope against hope and believe against belief that is invoked by Cavell in viewing the musical and its dance as "an escape from death," as "facing the music" of skepticism. But what I want to suggest is that part of the genius of the film - and certainly crucial to its emotional torque - is that it allows Selma's romantic, indeed melodramatic (indeed, operatic) deployment of this prosthetic "solution" to the loss of vision as a means of ensuring the world's consistency (an understandable solution to a desperate problem) while at the same time deconstructing that solution - specifically, in how the film "outside" of Selma's fantasy world is shot. For what the film insists upon, rigorously and systematically, is the difference between that aural "vision" of the world and the subject's place in it via fantasy and what the world looks like when those fantasies and identifications are suspended - when they are, as it were, subjected to analysis. And in so doing, the film uses the "pathological" fact of Selma's blindness and the compensatory prosthetic strategies it generates to disclose a properly deconstructed notion of the visual, very much along the lines of Zizek's gloss on Malbranche, with the point being not that "only those who cannot see can see," but rather "even those who cannot see cannot see."

The film's most important musical number, "I've Seen It All," would seem to register this theoretical point about fantasy and identification very much in the terms discussed by Zizek in his gloss on MUDs (Multiple User Domains) on the Internet - that the point of the Lacanian notion of the "split" or "decentered" subject is not "there are simply more Egos/Selves in the same individual" with which one might identify, but rather that this decentering of "the void of the subject" itself is derailed and constituted by the Symbolic and the phallus as signifier - its "hollowing out," by the signifier, as Lacanians like to say - in relation to its "content," to "the bundle of imaginary and/or symbolic identifications." Here we might revisit, I think, Selma's repeated rejections in "I've Seen It All" of the further identifications held out to her by Jeff (of wife, of grandmother, and so on); in her repeated insistence that "I've seen all I need to see," the film registers the fact that "the subject's division is not the division between one Self and another, between two contents, but the division between something and nothing, between the feature of identification and the void" (141) - hence the film's rigorous insistence, which borders on a kind of fatalism, that further identifications ("seeing more" in the song's terms) will not change this basic existential equation of the prosthetic subject, whose structural void and distance from all possible "selves" are revealed and thematized by Selma's encroaching blindness and punctuated, one might say, in the musical number as Selma wraps her arms around herself and falls to the ground in a fetal position, dangerously (suicidally?) near the passing train, as an "answer" to Jeff's repeated calls for further, other identifications.

To return, however, to how the film is shot: what makes the film postmodern, if one wants to use that term, is that it allows the investment in a more or less traditional fantasy of vision through its thematization in the story of Selma, but at the same time divorces visuality from transcendence, Identity, the Ego, the cogito, and so on in its cinema verité camera work. The apparent "realism" of how the film is shot, however, operates not to put us in touch with some unmediated relationship with "the way things are" - that, as we saw with Silverman's critique of aural and visual transcription, is not in the cards - but instead is calculated to insist that "this is what the world looks like - it doesn't look 'like' anything at all," a fact that may be referenced in turn to the contingency (that is, lack of necessity) governing the camera work itself. This jarring but crucial contrast is prepared for from the film's opening moments, in the juxtaposition of the operatic overture and its painterly affirmation of the subject of vision (and this in Cavell's sense), followed immediately by the mundane local audition for The Sound of Music, filmed in nearly distracting hand-held style. What all of this suggests, I think, is that we are to take the fantasy scenes, with their vivid coloration and careful choreography (as in the visually stunning "I've Seen It All" or the complexly woven "Cvalda"), in contrast to the devil-may-care and palid shooting of the everyday scenes - as more real than the "documentary," "real life" scenes from which they supposedly depart - and that is precisely their problem. In psychoanalytic terms, the world paradoxically "comes to life" only through fantasy, but it is the subject's very fantasy itself that bars her from "what is really going on" in the world of the object, which obeys its own laws and doesn't "look like" the subject's desire (or anybody's else's). This doesn't mean that fantasy is being disavowed, as if one could escape it - only that it is being carefully situated, and in a way quite specific to the medium of film.

If we want to think of this in deconstructive rather than psychoanalytic terms, the film might be said to enact what Laura Oswald has called the strategy of "cinema-graphia," which identifies cinema with "those traces of non-presence" such as the splice, the cut, or the frame that draw attention to "the endless production/deconstruction of the meaning and subject of film discourse across the film frame " In so doing, cinema-graphia "shatters the mirror in which the subject is held as a unity by defining the image as