There are films about painters, films that feature paintings in the plot, and there are films about particular paintings. In the first category, the centenary has given us several van Gogh movies (directed by Paul Cox, Robert Altman, Maurice Pialat), and in Derek Jarman's Caravaggio we had the anti-myth to the myth of the creative genius tormented by his Art. In all of them, what remains, one way or another, is the "agony and the ecstasy," whether embodied by Kirk Douglas, Tim Roth, or Nigel Terry.

Paintings, and especially painted portraits abound in what has been called the women's paranoia cycle of Hollywood melodramas from the 1940s, but they also star prominently in some celebrated "films noirs" of the 1950S: one thinks of Rebecca and Suspicion, Laura and Woman in the Window, The Two Mrs Carrolls and Strangers on a Train. Hitchcock, as one can see, is particularly fond of them, but so are Germanic directors like Lang and Preminger. Such portraits activate a host of associations, partly historical (they often connote a period setting and a genre: the Gothic), partly social (in a world of objects and people, a painting is always extravagant, excessive in that it is both object and person), partly economic (whoever owns a painting has surplus value to display, which means it also often functions as a signifier of class), and finally, the connotations are inescapably sexual (Beauty and Fatality, Perfection, Woman, the Unattainable Object of Desire). Sometimes they are the very epitome of patriarchy, as Joan Fontaine's father disapprovingly looks down on her choice of Cary Grant as husband in Suspicion.

Films featuring series of paintings are mostly "European," and they seem to belong to the 1980s: Godard's Passion, Raul Ruiz' Hypothese du Tableau Vole, and -- stretching the term painting a little -- Peter Greenaway's The Draughtsman's Contract. In each case, what is explored are tableaux vivants, though to different ends. Greenaway sees social hierarchies mirrored in the pictorial geometries, both of which fail to contain the more elemental or anarchic forces set free by (female) sexuality (of which the moving image becomes an ally); in Ruiz, the tableaux vivants tell of all the narrative possibilities -- all the possible movies, in other words -- locked up in static images, and of interpretation games far more devious but also more interesting than the analytic master-narratives of Marx or Freud. For Ruiz, the relation between cinema and painting raises the question of pictorial realism generally, meaningful only if read as allegory -- a point to which I will return. In Jean-Luc Godard's Passion finally, the stillness of the tableaux is not only juxtaposed to the machine noise on the factory floor and the noise cluttering up personal relationships; 'the scenes taken from Velasquez, Rembrandt, Ingres, Goya -- even while depicting violence and destruction -- suggest the possibilities of existential confrontation not afforded any of the protagonists. Painting, it seems, provides a vanishing point from which to view a world in the process of disintegration, but at the price -- as in Kafka -- of excluding its protagonists from both.

It is not immediately clear what category Jacques Rivette's La Belle Noiseuse belongs to. Attention seems equally divided between the artist, the portrait that gives the film its title, and the painting as material artifact and commodity. Emphasized as in no other film is the process of creation itself, the artist's labor-intensive hard grind (or "scratch," since it is pen-and-ink-on-paper we mostly hear), and the bone-crushing, limb-twisting postures the model is subjected to. Rivette's story is simple enough: A famous painter, Eduard Frenhofer (Michel Piccoli), suffering from a prolonged fallow period, is persuaded by his dealer, Porbus (Gilles Arbona), to take up again a canvas which he, Frenhofer, had abandoned ten years earlier -- the "Belle Noiseuse" of the title -- with the help of a new model, Marianne (Emmanuelle Beart), the girlfriend of Nicolas (David Bursztein), an aspiring young painter. The couple happen to be visiting the area where Frenhofer has made his home, not least because Nicolas wants to know what the great master has been up to in his country refuge. Outraged at first, Marianne consents to be Frenhofer's sitter, and over a space of five days, the painting is completed. Hovering in the wings is Frenhofer's wife, Liz (Jane Birkin), his erstwhile favorite model and the original "Belle Noiseuse," who is both eager for Frenhofer to get over his creative block and afraid of being replaced. But she knows that "Frenhofer est un gentleman," and indeed, he has no sexual interest in Marianne, except that the casual, sometimes brutal and in the end quite sadistic regime he inflicts on his model during the sittings do seem to unnerve the young woman, her poise and cool temporarily breaking under the strain. Both couples go through an emotional crisis, deeper and possibly more serious for the older couple, since it seems to convince Liz that not even the completion of the painting will release either of them from their living death. Porbus, however, wants to celebrate "La Belle Noiseuse," and a picnic in the grounds of Frenhofer's estate serves as a kind of coda, with the young couple also departing, though not before Nicolas tells Frenhofer that he is not too impressed by the new work. What he does not know is that the painting on display is not at all "La Belle Noiseuse," but one which Frenhofer had done in one all-night session, not even bothering with the girl as model. In voice-over Marianne tells the audience that what happened after they got back to Paris is another story.

Arguably, Rivette means us to take his drama as just that: Frenhofer's struggle to conquer his anxiety, to wrestle with his muse and angel, in order to bequeath to posterity some essence of his vision, the work that says it all. "Faster, faster, mach one, mach two" he explains to Marianne, as if death was already too close for any mere terrestrial motion towards a goal. Cloistered away in his somber studio, while outside Nature is vibrating to a Mediterranean mid-summer heat, the tragic irony would then be that in spite of subjecting himself and those around him to the most intense pain and sacrifice, there can only be a masterpiece that nobody sees, and one that brings neither redemption nor transcendence. Frenhofer is then the modernist after modernism, the antithesis of Cezanne (who would not have tried painting the view from Frenhofer's tower and balcony), but also the opposite of Picasso, for whom the painter and his model/lover/wife became the emblem of how to renew his art by dramatizing through this relationship every conceivable vital and venal, violent and voyeuristic impulse. Just such a sketch evoking Picasso can be found among Frenhofer's discarded canvasses stacked on the studio wall, perhaps because, as Rivette mentions in an interview, his friend Claire Denis bombarded him with Picasso postcards depicting this motif while Rivette was hesitating whether to undertake the project at all. In the end, Frenhofer is more like Beckett's comment on Bram van Velde in Four Dialogues with Georges Dutuit: "nothing to paint, nothing to paint with, and yet nothing left to do but paint."

The success of La Belle Noiseuse -- winner of the Grand Prix at Cannes in 1991 -- and the fact that it is by far Rivette's most accessible film since La Religieuse, makes it plausible that audiences see a qualified but nonetheless comforting reaffirmation of the values not only of art with a capital "A", but also of the European art cinema. Indeed, perhaps Rivette (until now mostly an enigmatic outsider even in his own country) wanted to try his hand at the genre better known through Eric Rohmer. La Belle Noiseuse could well be one of those "contes moraux" which have become Rohmer's trademark and quality guarantee. Rivette's tight plotting, the film's many formal symmetries and neat ironies, the division into clearly felt scenes and acts, the respect for the unities of French classical drama all recall Rohmer, both theme and setting making one think of La Collectioneuse or Rohmer's homage to Matisse, Pauline a la Plage. The opening of La Belle Noiseuse, with the young couple pretending tourists from England is pure Rohmer, if it wasn't also vintage (compareable to the opening of L'Amour par terre where a solemn group of men and women is led through back streets and courtyards up several flights of stairs into a Paris apartment to become eavesdropping witnesses to the infidelities and domestic complications of an executive with a wife and a mistress, before the spectator realizes that these are down-at-the-heel actors who have invented not street theatre but apartment theatre).

To be familiar with other Rivette films certainly helps to make sense not only of this opening; quite naturally the temptation is to regard La Belle Noiseuse in the light of the director's other films. If there are still any auteurists out there, here is a chance to practice the old skills of recognizing personal themes and formal obsessions, of spotting allusions and putting together the cross-references and inter-texts: in other words, precisely, to salute the artist and his inimitable signature. What in the earlier films had been the structuring principle, namely to use the theatre, a performance to be rehearsed, a show to be put on, in the course of which the characters find out some -- inevitably painful -- truth about themselves is here the function of painting, in each case setting off art versus life, the classical versus the vagaries of personal relations, formal order vs. the anarchies of l'amour fou, the destructiveness of self-obsession. Rivette himself has called this principle "la vie parallele," and virtually all his films take one text and overwrite or underlay it with another.

La Belle Noiseuse is unquestionably an auteur's film, but the very fact that it advertises this status so insistently suggests that we may have to regard it as something that no longer can be "taken as read." Perhaps the very principle of "la vie parallele" has assumed another meaning, and makes its own contribution to a particular polemic: A film about an artist (what more overdetermined a choice of actor for this part than Michel Piccoli?), a female star (Emmanuelle Beart, fresh from her success as "Manon des sources"), and the difficulties of fixing a representation. It also counts as a move in what seems to be a "Kulturkampf" raging in France over the meaning and definition of French cinema. In this cultural battle, "painting and cinema" has become a kind of code, though it is, at least for an outsider, far from clear exactly where the lines are drawn. This is certainly no longer the line between the "tradition de qualite" and the "Nouvelle Vague." From Truffaut to Rohmer, from Tavernier to Pialat, including even Carax and Besson, directors seem to have made their peace with a fairly capacious version of the "great tradition" of the French cinema, preferring to rework its stock situations and stable constellations, rather than inaugurating radical breaks. What, then, is the evidence for assuming that a film about painting made in 1991 may have a special topicality? Since the mid-1980s there has been a steadily increasing number of (excellent) books about cinema and painting by leading French film critics and academics: Rivette's scripterwriter, Pascal Bonitzer's Decadrages: peinture et cinema (1985), two scholarly conferences devoted to the topic at Quimper and Chantilly in 1987, Jacques Aumont's L'oeil interminable (1989), Marc Vernet's Figures de l'absence (1990), Raymond Bellour's collectior of papers from one of the conferences, Cinema et Peinture: Approches (1990), his own L'Entr'Images (1990), presented at a seminar and a lecture at the NFT in 1989, and finally, in 1991, an international colloquium at the Paris Louvre, or "Le portrait peint au cinema," documented in a special issue of the journal IRIS.

While not mentioning either a battle or even a crisis, many of these essays raise the point whether one should not look at the cinema from the vantage point of painting? This, in a sense, takes one back to the art-historical or filmological debates of the 1940s and 1950s, to Elie Faure and Andre Bazin, (1) whose implicit question was: is the cinema an art, and if so, how does it relate to the other arts? Instead, should one not assume this battle to have been won? Is it not time to reverse the angle, and look at painting from the vantage point of cinema? Aumont calls this an "analecture," a retrospective reading of (the history of painting) in the face of the existence of the cinema and its impact or pictorial questions of spatial disposition, framing, expression, lighting, the representation of time (the "pregnant moment"), and above all, the spectator' role and place in front of a "view." Cynics may say that this is merely a rather arcane debate over the direction of academic film studies, in the wake of disenchantment with cine-semiotics and psycho-semiotics: a swing of the pendulum away from the literary-linguistic foundations of "serious" film analysis to an equally respectable "art-history" discourse, with the cinema still looking for a pedigree. One might even contrast what is happening in France unfavorably with the situation here. In Britain it was, among others, John Berger's Ways of Seeing (2) (and its polemics with Kenneth Clark) that helped fuel a debate about the boundaries between fine art and popular culture. Against the history (of capitalism) that both were seen to be implicated in, an alliance emerged which led in the 1980s to the confluence of art history, film and TV studies, feminist theory, merging in the cultural studies courses at universities, art colleges and polytechnics. In France, the terms of the debate, at least from the focal point of the cinema, do not appear to be high culture versus popular culture, nor does it look as if any overtly political agenda has made inroads in the curriculum and emerged as something akin to cultural studies.

Most instructive for bringing some of the undercurrents to the fore was Raymond Bellour's (3) lecture in London just cited, where he compared what is happening in the realm of cinema and the image to the "revolution in poetic language" of which the writer Mallarme was the messenger, when he lectured in Oxford in 1894 on the theme of "on a touche au vers," meaning the breakdown of meter in French poetry and thus of the radical difference between prose and verse. Bellour, half-jokingly, half-seriously suggested that he too had a message to bring to London: "on a touche a l'image." The divide for him seemed to run between "cinema" and the "nouvelles images," the latter itself a complex historical phenomenon, obliging us to see Nam June Paik taking the first Sony porta-pack into the street, Christian Metz (4) writing his first film-semiological essay, and Godard making Le Mepris as all belonging to the same moment in time. In other words, the "crisis" which in Britain since the mid-1960s concerned the fate of popular culture, the avant-garde, high culture, and high tech, working its way through the debates as the issue of representation, consumption and "spectatorship" was in France a debate about the material, linguistic and psychic support of cinema -- all driven by the fact that the photographic image could no longer be taken as the medium's self-evident basis, and therefore doing away with any indexical relation between reality and the image. It is against this background that in a fundamental sense, painting could become a metaphor for the cinema (as the medium associated with the history of photography) in contrast to the electronic or the digital image. Given its longer history, painting was to provide a certain vantage point on this rupture.

Thus, behind the equation "cinema and painting" other (dialectically intertwined or deeply antagonistic) pairs are lined up: cinema and architecture, cinema and video, cinema and television. What "painting and cinema" seems to signal is not necessarily where one stands in the divide, but rather indicates how one proposes to go about articulating that stand: perhaps "reculer pour mieux sauter," or putting together an inventory. Aumont's book is typical in this respect: it takes a historical view for the cinema according to him, is quintessentially the 19th century reaching right into the middle of the 20th, and he sees his book as a kind of janitor's job, tidying up after the show is definitely over, making sure the building is secure and everything is in its proper place.

To this one could add a more local issue, the struggle over the critical heritage of Cahiers du cinema, and the right to interpret the history of its influence. A two-volume chronicle, Antoine de Baecque's Les cahiers du cincma, l'histoire d'une revue (5) apparently sent many ex-contributors and collaborators to their word-processors for rectifications, amplifications, justifications. In short, almost all aspects of French film culture seem to be involved in a major film-cultural stocktaking. One of the most brilliant Cahiers du cinema critics of the late 1970s and early 1980s, Serge Daney, gave up his job at the magazine in 1987 when he became media critic of the daily Liberation, discussing television, advertising, commercial video with an erudition and critical wit not seen before in French journalism or criticism. (6)

Daney is perhaps the most radical among those who think it is time to repay the cinema its due: it has taught us how to look at the world, now we have to learn how to look at the other arts and media through the lens of the cinema, but a cinema so naturalized, so culturally internalized as to be nowhere in particular and yet everywhere. In its near-hundred year history, it has become a kind of truth, namely our truth. What Daney had in mind was strikingly confirmed in a recent BBC2 Moving Pictures item devoted to the memory of Jean Vigo. Bernardo Bertolucci, asked to talk about L'Atalante, quite spontaneously described the film not as a film, but as a reality existing in its own right, a reality existing next to other realities. He ended up talking about what he called "liquid cinema," a notion especially suggestive. Bertolucci's way of celebrating Vigo contrasted with a no less enthusiastic Lindsay Anderson in the same program, for whom there were masterpieces like Zero de Conduite which had inspired him in If... and, even more importantly, there was an "artist" with a "personal vision," an entity to which he admonished all young filmmakers to remain true. Bertolucci not only paid homage to Vigo by reworking a scene from L' Atalante he particularly liked in Last Tango in Paris. At the Canal St. Martin, a life-belt with "L' Atalante" written on it is tossed to the couple in the water, only to sink like a stone. By giving the cinema a dense materiality, Bertolucci's ultimate compliment to Vigo was to speak of his film as existing in the real world, like a building or the Canal St. Martin itself, landmarks we can all visit and inspect. Martin Scorsese, also has this exact attitude. Gone are the days when love of cinema meant talking about "film as film" -- works with their own aesthetic texture, structure and textuality (as the first generation of film scholars, say, Robin Wood (7) or Victor Perkins (8) had to do in order to legitimate studying the cinema at all). Like Bertolucci or Daney, we may need to treat films as events that have happened to us, experiences that are inalienably ours, and thus as material facts. The cinema has helped carry the burden of history, or has given the illusion of carrying it, but it has also bequeathed a kind of double or parallel life, shadowing another, perhaps ever more shadowy life, as our culture's real past become its movies.

That Rivette's films -- "scenes de la vie parallele" -- appear to hold in many ways key positions in the more specifically French debate is not in itself surprising when one remembers his beginnings. Probably the most intellectually precocious of the young Turks around Bazin and the early years of Cahiers du cinema, Rivette was nonetheless -- along with Godard -- one of Bazin's more unruly sons when it came to deciding whether he belonged to those who "believed in reality" or those who "believed in the image." Championing Hawks and Hitchcock, and also Fritz Lang, Rivette always oscillated between the classical cinema of Wyler and Preminger beloved by Bazin, and a more offbeat Hollywood. He preferred the "improbable truth" (the French title of Lang's Beyond a Reasonable Doubt) to Rossellini's "things are there -- why tamper with them"?)

The reference point, then, for La Belle Noiseuse may well have to be Godard's Passion, which proved to be a key film of the 1980s. As so often, Godard sensed the tremors announcing the landslide earlier than most, and in Passion; and the accompanying television program, Scenario de Passion he began to redefine his cinema, but maybe also the modem European cinema generally (Wenders certainly seems to follow in Godard's footsteps, though in a grandiosely overblown manner, in Until the End of the World). Godard, precisely, went back to painting. But in Passion, cinema is the vanishing point between painting on one side, and the video screen and monitor on the other: it is a "film" shot with a big Mitchell camera that the Polish director is unable to finish, and it is perhaps no accident (for the genesis of Rivette's project) that Passion features Michel Piccoli as the patron and patriarch, lording it over not only his employees, but his wife, who is the lover of Jerzy, playing the director (and used by Godard because he was the "Man of Marble" from Andrzej Wajda's film), now unable to muster the "solidarity" needed to still make cinema.

Seen as part of this dialogue of French film culture with itself, we have to assume that La Belle Noiseuse's "classicism," its well-lit sets and carefully composed shots, its "logical" editing rhythm and shot changes, its balanced alternations between indoor scenes and the dappled outdoors, it day-times and its night-times, and thus its apparently solemn affirmation of the spiritual values of great art, is less a polemical restatement of the "politique des auteurs," and a rather more subtle or nuanced intervention in present-day cultural politics. Rivette has made an auteur's film, but one in the full knowledge that it has to be an auteur's film, for reasons of survival, not only as bulwark against the anonymous output of TV, but also so it can be shown at Cannes. Festivals are the places where films financed by television receive the world's endorsement that they nevertheless still count as cinema, by a process that Godard has called "giving a film its passport" -- a phrase already used several years ago by Peter Wollen, describing his own work as "films without a passport."

A number of distinct aspects of Rivette's film come to mind in support of reading it as just such a statement about the impossibility of the auteur and yet the necessity of being one. First, there is the question of authenticity and the original. What at first sight is curiously old-fashioned if not naive about La Belle Noiseuse is its belief in authenticity and the original throughout most of the movie, until Frenhofer, coolly and methodically, fakes himself by hiding the "Belle Noiseuse" forever, while passing off a painting quickly daubed during one late night session as the fruit of ten years' creative agony. Apart from the three women in the know (who do not speak out), nobody so much as suspects the fraud and the substitution.

Second, an argument about Authorship, Style and Signature. If La Belle Noiseuse is indeed the parting self-portrait of the author as a cultural icon of authentic art, signature itself substitutes for style. In Rivette's film, this signature comes from beyond the grave. One is reminded of Wenders' The American Friend, where the importance of the paintings sold by Dennis Hopper was that they were forged, but by the painter himself, who was played by Nicholas Ray, the film auteur par excellence of the Cahiers group, and of whom Godard once wrote: "if the Hollywood cinema were to disappear, Nicholas Ray would single-handedly reinvent it." In one of the most dramatic scenes of the Rivette film, Frenhofer's wife enters his studio at night, looks at the painting of "La Belle Noiseuse" (which we never see -- except like a fetishist spying a piece of thigh, we catch a glimpse of carmine red, when the covering sheet is accidentally lifted for an instant). Frenhofer's wife, evidently shocked by what is on view, walks round the painting, and next to his signature on the back, she paints a cross, as if to confirm that this has been painted by a ghost. The gesture turns La Belle Noiseuse into something close to a horror film, halfway between the gothic tales around painted portraits mentioned in the beginning, and Roger Corman's Tomb of Ligea or Fall of the House of Usher.

Third, an argument about craftsmanship, labor and duration. Much of the film is taken up with the act of painting itself. The fact that in La Belle Noiseuse all that effort, all that painful scratching of pen on paper, the sketches, the posing, the crucifixions that the model's body undergoes, seems in the end to have been produced merely to be hidden forever, is perhaps a more oblique comment than one at first assumes, on what can be the relation between the labor that enters into a work, and its value or effect. The discrepancy between labor and value was already the subject of Whistler's argument with Ruskin, and thus stands as a crucial debate at the threshold of the modem era, signaling the end of correlating the value (exhibition or social use) of a work of art with the labor (read: personal pain or mental anguish) invested in producing it.

Fourth, the contest between cinema and painting over "representation." The artist-painter's torment in the cinema is always slightly ridiculous, because it is betrayed by the cinema's facility in rendering what the painter is striving after -- this particular quality of light, that particular painterly effect, this particular likeness. Hence, any canvas actually shown in a film invariably turns out to be either bad art or a fake, the cinema always seeming to mock painting at the same time as it defers to its cultural status. One of the most shocking moments in John Berger's La Belle Noiseuse was when he walked up to a Botticelli in what looked like the National Gallery, took out a Stanley-knife and cut a sizeable square out of the priceless canvas. After such knowledge, no painting can survive its representation in cinema: one more reason why "La Belle Noiseuse" must remain hidden at the end.

Finally, an argument about different kinds of time. Well before German directors like Syberberg and Reitz opted for length to make themselves heard, and cut a sizeable chunk of time out of the media landscape of television, Rivette produced monsters of extended time, from L'amour Fou to Out One. It is, apart from anything else, a response to the need of European art films to counteract the blockbuster media-blitz of Hollywood. Yet the length of Rivette's films also foregrounds the spectator's place and the experience of viewing -- not excluding boredom. Rivette is a more experimental director than most, opening his films to varying degrees of attention and attentiveness, and by making painting his subject he is able to enact a certain kind of viewing: contemplation, exploration, negotiating distance and proximity, occupying a different space, and yet "entering into a picture." The emphasis on both process and product reinforces this parallel, so that over long stretches of the film, the spectator is, as it were, alone with his thoughts, "watching paint dry" -- itself an aesthetic statement in the age of media-instantaneity and electronic images.

But La Belle Noiseuse also enacts this different form of spectatorship concretely, carving a "spatial form" out of the time it takes to view it. Length becomes one of the auteur's weapons in his battle against so-called "dominant cinema": the film lays and splays itself across television's time slots and scheduled evanescence, as well as breaking down a first-run cinema's two or three evening performances. La Belle Noiseuse is four hours, carefully segmented internally into dramatic acts, but also externally, by a break that the film itself announces. Thus, when the model in the film gets giggly from exhaustion, and several times grabs for a cigarette, the film advises patrons in the cinema to take a break as well and come back for the next sitting. La Belle Noiseuse is nothing if not aware of the kind of special occasion contract it has with its audience, and although TV's archetypal moment of disjuncture, the commercial break, might fit just as well, one wonders how the two-hour TV version manages to convey this double articulation of duration.

All this may be no more than saying that La Belle Noiseuse can be and must be read as allegory, or rather, as that particular form of allegory known as mise-en-abyme. This is perhaps the more surprising, since the film is, in its narrative as well as its mise-en-scene, one of the most "classical" films imaginable, respecting at all times the ground rules of cinematic realism. But rather than being conformist, this classicism functions as an act of resistance. Looking at Rivette's oeuvre, it is possible to argue that his films have always anticipated another technology -- that of video, of the video recorder and of electronic images, especially in their obsession with parallel realities, with going into dream-time and paranoia-time, with layering one text with another, confronting theatre and life in a modulated commentary on Anna Magnani's question in Renoir's Le Carosse d'Or: "where does the theatre end and life begin"? But the tension and pathos of Rivette's films, their quality of clairvoyance and hyper-alertness actually depended on the resistance which the medium "film" offers and imposes on both filmmaker and viewers in differentiating and resolving this layering: these strainings after representing twilight states, these superimpositions of parallel worlds need a realist medium, need the solidity of celluloid. Similarly, in La Belle Noiseuse, the pay-off of all this labor comes when Frenhofer takes out the unfinished canvass of ten years' earlier, in order to paint the new picture over the old, seemingly obliterating his wife-as-model by the young woman, the two merging and mingling, the face of the first gradually but only partially, hidden beneath a veil of blue crayon. It might be a video-effect, and yet it crucially must not be a video-effect.

Ultimately, it is this capacity to be a "realist text" and allegorical at the same time that makes La Belle Noiseuse contemporary, and to my mind, an "intervention" rather than a conservative restatement. One might cite Borges and Roland Barthes: Rivette is rewriting a classical (readerly) text as an allegorical (writerly) text. As in the case of Barthes most famous allegorical rewriting of a realist text, Balzac's novella Sarra sine in S/Z, so the realist text of Rivette is also a novella by Balzac, Le chef-d'oeuvre inconnu. From it, Rivette takes the initial situation where the young Poussain offers his mistress to the master Frenhofer as a model, in order to spy on Frenhofer and get a glimpse of the one painting Frenhofer refuses to put on show, "La Belle Noiseuse," reputed to be a masterpiece, the chef-d'oeuvre inconnu. In Rivette's film, the title of the painting is itself thematized, by what may well be no more than a piece of folk etymology: Noiseuse comes from noix, nuts, and in Quebecois slang, it means a woman who is a "pain in the ass."

Deliberately and bluntly, La Belle Noiseuse parades a world of men who enter into a kind of bargain or exchange whose object is a woman. Not only is the young painter's girlfriend offered as bait or gift, she is also intended as a substitute for Frenhofer's wife, regarded by Porbus as the cause of Frenhofer's creative block. But the twist and thus the film's central allegory, or re-reading of Balzac, is that Rivette makes of Frenhofer the Minotaur, a creature both powerful and baffled, half-man, half-beast. It brings the Frenhofer figure once more close to Picasso, for whom the Minotaur was a central reference point. But more importantly, it emphasizes the different role the young woman has as the sacrificial victim, offered by the men to appease the man-god/man-beast of artistic genius. And the question which the film raises in that last voice-over is whether Marianne is in fact a kind of Ariadne, venturing forward so that the wily, but also cowardly Theseus can follow, to slay the Minotaur, or at any rate, to take away his power. What in Balzac is an Oedipus story becomes in Rivette a Theseus myth, or rather an Ariadne story.

This suggests two things, by way of conclusion. I would see in La Belle Noiseuse Rivette's decided plea for cinema, but not as a simulacrum of painting, nor of its cultural status or commodity value: rather the plea for a cinema where the virtual realities and parallel worlds are created by the fact that you can believe in what you do not see, in contrast to a Hollywood cinema where you can see what you cannot possibly believe (thanks to special effects), and a television which can do neither, and only asserts. Yet it also suggests that the change from the photographic image to the digital image is more than a change in technology, or delivery system, but will entail a long and protracted struggle not only over the interpretation of this or that film, but over the meaning of the cinema altogether. This debate, hardly begun, seems figured in the allegorical mise-en-abyme into which "La Belle Noiseuse" so definitively disappears. The painting is finally what one suspected it to have been: a mirror, but a mirror standing for "the visual," through which our civilization seems destined to step, rematerializing on the other side as something quite different. Rivette, perfectly agreeing with Daney's analysis, might take exactly the opposite position. Once the painting has vanished and the guests have departed, we hear the sounds of the village: a baby crying, the voice of a woman answering, the noises of people at work -- a world outside once more coming alive, though now perhaps we have the inner eye to finally see it.