Despite the devoted championship of a few critics (outstandingly, Jonathan Rosenbaum), Jacques Rivette's films have yet to establish themselves as a major presence in modem cinema and the development of film culture. (1) Yet I have come to believe with steadily increasing conviction that they deserve such position, and that every effort should be made to secure for them more adequate distribution and a wider public. That most of the films remain largely inaccessible (either not available or rarely screened) is a great pity for the simplest, most basic reason: that, approached with the right expectations (or, at least, without the wrong ones), the films have an enormous capacity for giving delight. The excitement (both intellectual and emotional) with which my students responded to L 'Amour Fou and Celine and Julie Go Boating was unmistakable: it was the kind of direct response that has nothing to do with trying to please a professor. The films spoke to them immediately and engaged them totally; there was the sense, in ensuing discussion, that whole new potentialities had been opened up.

Those potentialities are intimately bound up, certainly, with the films' narrative freedom and formal innovativeness, but they also transcend the level of formalist experimentation. And, seeing the films again, I must bear witness to a certain dissatisfaction with the emphases of Jonathan Rosenbaum and other Rivette champions (hoping that this does not appear ungrateful: without that championship my attention might never have been drawn to the films). The effect of those emphases has been to create a somewhat abstract impression of the films, repeatedly stressing the importance of their play with narrative but leaving the reader with the feeling that it doesn't matter much what the narratives are about: any old narrative would do, as long as its illusionist domination of the spectator is undermined and its classical-realist logic disrupted. Rosenbaum insists at moments on the political force of Rivette's procedures. For example, in his introduction to the indispensable British Film Institute booklet (Rivette: Texts and Interviews) we come upon such overt provocation as ", . . there is a great deal more political import to be found in these films-directly addressing the manner in which spectacle and spectator conspire to produce or deny meanings-than in the collected works of such purely unreflexive illusionist directors as Fassbinder, Herzog, Pakula or Rosi." Yet (leaving aside the various tempting questions this raises, as to the felicity of that rather odd list, or the validity of the label stuck indiscriminately on its four incongruous bedfellows) the overall effect of the essay is to reinforce the notion of Rivette as a peripheral formalist experimenter. A Film Comment interview (by Rosenbaum, Lauren Sedofsky and Gilbert Adair; Sept.-Oct. 1974) manages to discuss Celine and Julie at some length without at any point engaging with it as, centrally and predominantly, a film about women and one of the most radical and positive feminist statements the cinema has yet achieved. (Adair informs us that the film is "open to as many interpretations as there are spectators," a not unfamiliar form of critical abdication, especially in relation to the myth of the "open text"; in this case demonstrable nonsense.) The aim of this article, then, is not only to keep alive the battle for the greater accessibility of Rivette's films, but also to attempt to restore to them the social-political significance and force which Rosenbaum's advocacy tends to obscure.

As a preliminary, one must briefly confront the films' most obvious eccentricity and commercial stumbling-block, their length, (L'Amour Fou is 4 1/4 hours, Celine and Julie 3 1/4; and Out One: Spectre, made between them, is- even ignoring the existence of an almost 13-hour version-longer than either.) Why must the films be so inconsiderately long? The simplest, and perhaps most important, answer, is that there is no socially acceptable reason: none of the usual pretexts taken to justify the length of, say, Gone with the Wind. 1900. or The Deerhunter operates here. Clearly, the length of our cinema program is closely bound up with the more obvious conditions of the current phase of consumer-capitalism: alienated labor, the five-day week, the 9-5 job, the nuclear family, a norm common to all levels from employee to executive, necessitating that (weekends apart) "leisure" be packed into a 2-3 hour slot between the time the kids are put to bed and the "early night" required by the next day's toil. Otherwise, in one of those ugly and brutal, but entirely taken-for-granted, phrases that characterize our culture, "time is money"; and if we are going to surrender four hours of our "money" to watching a spectacle, we must be repeatedly reassured that the spectacle we are buying was extremely expensive, that we are purchasing a visibly valuable commodity. Rivette's films, on the contrary, are perceptibly cheap-despite their length, they are transparently "low budget" (Celine and Julie even shot on 16mm). Nor is the length of the films validated by complexities of plot, large numbers of characters, "epic" events. What is the plot of L 'Amour Fou? Sebastien tries to produce Andromaque; Claire goes mad. Rivette could easily have told it to us in fifteen minutes and spared us the superfluous four hours.

The "unjustified" length of the films, then, represents an act of cultural transgression. The question, "Why this length?," should immediately provoke a reciprocal one: Why the standard length? Why should we automatically expect our movies to last between 90 minutes and two hours, feel cheated if they are less and demand particular justifications if they are more? Rivette not only demands our time, he demands our patience. We have to sit through long stretches of film in which nothing much (in terms of the conventional narrative expectations for which mainstream cinema has conditioned us, leading us point by logical point through the traditional process of order--disturbance-restoration/replacement) seems to be happening, or much the same thing seems to go on happening or to happen again. Yet few people who have sat through Rivette's films from beginning to end wish them any shorter. (Some reject them altogether, but even they seem not to question that the length is somehow part of the point, an essential component of the films' "signifying practice," their function within culture.) It is the very fact that the length is unnecessary that renders it indispensable. The apparent longueurs (the films seem shorter every time one re-sees them) are the most effective means of forcing us to question the experiences mainstream cinema offers us, the positions in which it places us. As with the question, "Why this length?," the question "Why are we being asked to watch this?" raises a reciprocal one: Why are we "normally" asked to watch the particular patterns that "this" transgresses? Further questions immediately follow: What makes a narrative "interesting"? Are we conditioned to apply only very limited criteria of "interestingness"? Can we learn to be interested by other narrative procedures? The issues are radical, and as political as you want to make them.

But I am not, so far, keeping my promise. I have said little more than Jonathan Rosenbaum has said many times: that Rivette's films amount to a radical and disruptive intervention in mainstream cinema on the formal level. In order to go beyond this, and explore questions of what the films are actually about and how this relates to (demands) their formal innovativeness, I want to present certain fundamental notions about our culture and the function of art within it-notions that, as soon as they are stated, appear self-evident, yet, like so much that is self-evident, are seldom squarely confronted. (It is only by refusing to confront the self-evident that our culture averts total and immediate revolution.) Our Western culture is, and has been since the beginning of written history, patriarchal: which is to say not just male-dominated but dominated by the figure of the Father, to be understood both literally and symbolically. The father is head of the family, the Pope is father of the church, God the Father presides over all. Our institutions, that both represent and create the social order, are dominated by father-figures (president, judge, chief-of-police, chairman of the board, head of company, etc.) and embody (even if they are occasionally women) the principle of patriarchal authority. The social ordering, then, is the execution of the Law of the Father, and consists of assigning all subordinate groups (women, children, gays. . .) to their "correct" places. Obviously, the specifics of our current cultural situation add to this considerations of race, property and class: the privileged "father" of our society is the adult white bourgeois heterosexual male.

By and large, art, through the ages, has been the illustration, reduplication and rein-forcement of this order on the level of ideology; it has also (hence its endless ambiguities) precariously contained the various rebellions against it, the containing dangerously involving an acknowledgement of their existence. Our heritage of art and literature represents (taken en masse, as methods of containment) an overwhelming reduplication of patriarchal structures. The traditional notions of art-as-representation and art-as-order can no longer be taken for granted: we have to ask precisely what is being represented (not simple "reality," but a reality culturally defined, from a particular and analyzable perspective) and what order is being imposed or reproduced (not the order of nature, but the order of culture). This reveals a whole new collection of fathers: the artist, the author, the film director: their self-appointed function that of "representing" and "ordering." It is not to belittle the achievements of women in the arts (especially the great women writers-Austen, Bronte, Eliot) to say that, on the symbolic level, the familiar figure of the omniscient author is male. The male author knows everything from the outset, and leads the reader (viewer) by "logical" stages to the privileged position from which he/she can share that knowledge: the knowledge operating both on the narrative/hermeneutic/suspense level of what happened? (traditionally, who got married and who died) and on the level of value structures (who got rewarded and who got punished, and how we are to view these events).

If it is the male who represents, it is the female (though not of course exclusively) who is represented: the artist/model relationship is archetypally a male/female relationship. It would be silly to argue that the whole traditional function of art has been to inscribe women in their places under patriarchy, but this is clearly one of its traditional functions, and a major one. With it goes, of course, a whole structure of assumptions about sexuality and sexual relations.

The narrative procedures of progressive, pre-New Wave directors like Renoir and Ophuls (both greatly admired by Rivette) can be seen as tentative steps in that hesitant and uneven, two-steps-forward-one-step-back progress towards liberation that civilization must make if it is to survive. I want to argue that L'Amour Fou represents a more decisive step, fusing those procedures, and that Celine and Julie is the triumphant celebration of an arrival: albeit an arrival that looks ahead to the further roads to be travelled.

The overthrow (or voluntary abdication) of the male author/director is at once the subject of L 'Amour Fou's narrative and enacted in its formal strategies. Crucial to this is its systematic foregrounding of the processes of representation. Its structure resembles a Chinese box, or one of those Babushka dolls that Claire purchases and dismantles in increasingly desperate fascination; at the same time, the business of representation is no longer the prerogative of a single omnipotent patriarchal figure but divided among a number of representers operating in different media and on different levels. Rivette persuaded Jean-Pierre Kalfon to conceive and rehearse a production of Racine's Andromaque; he then persuaded Andre S. Labarthe (at that time a leading Cahiers critic) to make a TV documentary (16mm) on the rehearsals, interviewing the actors about their roles and experiences. Both Kalfon and La-barthe were allowed complete autonomy in producing and directing, while their endeavors remained within the boundaries of fiction: that is to say, there was never serious question of the production of the play actually being staged or of the documentary being shown on television. We have, then, Rivette making a film of Labarthe shooting a documentary of Kalfon producing a play by Racine reinterpreting a Greek myth. It is true that the top level of representation-Rivette's 35mm film-is not called into question, i.e., its mechanics are not foregrounded in the Godardian manner (the showing of cameras, clapperboards, etc., the acknowledgement of the actors as actors), the other levels (where the mechanics are foregrounded) being contained within a "realist" fiction. But the film consistently promotes such awareness of representational media that awareness of the film as itself a construct becomes inescapable, simply the logical culmination of the Chinese box structure. The problem of representation (how does one produce Racine? how does one direct actors? how does one film a rehearsal?) is an important component of the film's thematic; on the formal level, the frequent cutting between 35mm and 16mm footage ensures that the spectator's awareness of filmic means is never allowed to lapse for long.

Central to the film is the analogy suggested between Sebastien (Kalfon) as producer and Rivette as filmmaker. Sebastien's crisis of confidence in himself-his increasing uncertainty as to what he is doing and his right to do it-is precipitated by Claire's refusal of direction and her withdrawal from the play: each stage in his progressive abdication as controller of the performance is counterpointed with the deterioration of their relationship and Claire's withdrawal into insanity. Sebastien, throughout the film, vacillates between a desire to surrender authority (over the production, over the text, over Claire) and a fear of surrendering it: one might cite his ambivalent attitude to the pop-up toy with which he and Claire play in bed, where he both resents the fact that the toy won't pop up and contemplates (positively) the possibility of a female pop-up toy! His own descent into madness (or a state resembling it) is left ambiguous in its motivation: on one level he voluntarily enters insanity to join Claire, as his only means of retaining contact and helping her; on another, the descent appears a logical consequence of his abdication, a personal necessity, partly a self-punishment (expressed most clearly in the scene where he slashes his clothes and lacerates himself with a razor blade). I know of no other film that so powerfully communicates the terror moving out of one's ideologically constructed, socially conditioned and ratified, hence secure, position and identity, into. . .what? Which' is precisely the question with which the film leaves one: a political question if ever there was one.

For a moment, it looks as if both the Claire-plot and the Sebastien-plot are going to be resolved. In so far as his sharing of her madness is an act of therapy, it is successful, but in an unpredictable and ironic way: when she emerges from her insanity it is to the realization that she must leave him. Whereupon he phones his assistant and calls a rehearsal. We are reminded that, whereas he abdicated from a clearly defined position as patriarchal organizer, she had nothing much to abdicate from. But, just as Claire's journey on the train (with which the film begins and ends) has no defined goal, so the first performance never happens. Sebastien (actor as well as producer, hence indispensable to the production) lies slumped on the floor of the apartment, listening to fragments of the record of disintegration on Claire's tape recorder, while the theater audience waits.

In a scene that asks to be read as confessional, on the part of Rivette as well as Sebastien, the latter talks of the intention and status of the production of Andromque. He has reached the conclusion that Racine should be spoken as "conversation," and this emerges in the context as not only a perception about Racine's text but a perception about actors: that each should speak as her/himself, instead of having a uniform style of delivery imposed by the producer. In the event, the production that finally evolves becomes a striking equivalent for the overall method of the film, its idiosyn-cratic balance/tension between naturalism and formalism: the actors' "conversation," without forfeiting its naturalness, becomes ritualized by the use of a battery of percussion instruments to punctuate the statements.

Sebastien's second conclusion, connected to the first, is that the purpose of the production should be the satisfaction of the participants, rather than of the audience. This is clearly crucial to Rivette's cinema, and appears a direct affront to many traditional "democratic" ideas about art and entertainment: that the prime function of art is communication, that the artist should "respect" her/his audience, that the first consideration is to give the audience what it wants. But Rivette would, I think, reject only the last of these (which marks, in any case, the point where "democratic principle" merges into capitalist exploitation). Indeed, his respect for his audience is shown precisely in the way they are left free to accept or reject what is communicated.

This last point requires clarification. There is a sense in which every spectator is always free: nothing prevents him/her from getting up and leaving the theater, the individual intelligence can say "No" to any effect, any strategy, any moment. Yet most films are not made for the free choices of the individual intelligence. To widely varying degrees (from, say, Renoir and Hawks at one end of the spectrum to Hitchcock and Kubrick at the other) they pre-suppose a position which they also construct, from moment to moment, from cut to cut. It is very difficult to imagine what totally non-manipulative art might be: supposing that the artist has something (not necessarily a "message" or even a defined world-view) that he wishes to communicate (and why else be an artist?), he/she will always seek ways to establish points, to guide perceptions, to determine judgments. Yet the spectrum of possible degrees is very wide: the position the film defines for the spectator does not have to be (as so much recent film criticism tends to suggest) one of total and helpless passivity, the filmmaker does not have to conceive of his primary function as the determining and limiting of the spectator's response.

L 'Amour Fou strikes me as among the least manipulative films I have ever seen. Its length and longueurs are one aspect of this. A central principle of traditional art has been economy: the assumption that the "meaning" of the work is given in certain climactic or integrative moments of particular richness and complexity, which the rest moves up to or away from, but to which everything in the work should be seen to be relevant and necessary. (The moment in The Deer Hunter where the "one shot" of Robert de Niro's hunting and charismatic mastery becomes the bullet that Christopher Walken fires into his own brain might stand as a superb example). This principle is not lacking in L 'Amour Fou (one could clearly point to certain scenes, certain events, certain phases in the narrative, which are of obvious structural/thematic significance, where "things come together"), yet its dominance is significantly challenged. The structuring "moments" are not signalled in the ways to which we are accustomed, and there might be more than usual disagreement as to precisely which moments they are. We are allowed an unwanted degree of freedom in terms of where we find significance, what weight we give to various actions, speeches, gestures, how we distinguish between the important and the trivial. (The great exception to this, Sebastien's walk through the streets towards the end of the film, marked by subjective shots and overwrought music, culminating in his self-confrontation in a mirror-window, has always troubled me, its rupturing of the film's overall tone and method seeming arbitrary and unproductive).

L 'Amour Fou, despite the fact that Sebastien and Claire are given roughly equal screen time, remains essentially centered on the male. Throughout, it is the male who acts, the woman either being passive and helpless or acting ineffectually and futilely (e.g., the attempts to buy or steal a basset). The inequality is reflected in the film's formal procedures: the sequences involving Claire are necessary 35mm footage directed by Rivette (however much interpretative freedom was given Bulle Ogier), with no equivalent for the foregrounding of representation in the rehearsal scenes. The pattern of male action/female passivity is broken only by Claire's two negative decisions: to leave the play at the beginning of the film, and to leave Sebastien at the end. (The decisions can be seen as positive in so far as they are necessary steps towards Claire's liberation, but both take the form of rejection rather than affirmation.)

L 'Amour Fou is structured on the Sebastien/Claire opposition: the two central characters are opposed throughout, the central conflict is between them. Celine and Julie Go Boating is structured on a very different set of oppositions which unite the two central characters and pit them against the weight of the past, the weight of ideology. It is not coincidental that the central characters of L 'Amour Fou are male and female while those of Celine and Julie are both women. The film might be seen as fulfilling, and at the same time surpassing, what we may deduce that Ophuls would have done with La Femme de Paul: here, women become fully autonomous, and it is they who decide and determine the action and its outcome, both as characters and as actresses.

The credits attribute the scenario to Juliet Berto (Celine), Dominique Labourier (Julie), Bulle Ogier, Marie-France Pisier and, finally, Rivette, with di Gregorio acknowledged as dialogue writer: the women worked out their own roles and between them decided the whole progress of the film, Rivette providing only a suggested starting-point. The distinction between actresses and characters is continually blurred-we often have the sense that Celine and Julie are constructing the film out of their imaginations. In place of the traditional reading process whereby we decipher a previously constructed work in order to arrive at and share the author's privileged position of knowledge, we share here, to an unwanted degree, in the process of construction, the division between that and the process of reading becoming narrower than in any earlier fiction film I can think of. At the same time, the traditional reading process is foregrounded in Celine and Julie's attempts to decipher the story within the "house of fiction": they become readers of a novel, spectators at a play, the audience in a movie theater (debating, at one point, whether or not there should be an intermission, and deciding against it)-an audience miraculously given the power to intervene in the action and change its predetermined outcome.

The film is structured on a systematic series of oppositions between the Celine/Julie story and the story within the house; simply to trace these oppositions is to make explicit the film's meaning.

1. "Reality" vs. Fiction. The film creates a "reality" that is more fantastic than the events in the house of fiction": the paradox is central to its significance. Its "reality" easily accommodates the supernatural, the fanciful, the inconsequential; its fiction, though presented in fragments, is governed by the laws of classical realist narrative (plausibility, character consistency, etc.) and has no room for the illogical or for magic (except in so far as the characters in the house are ghosts from the distant past). The fragmentation (we, along with Celine and Julie, at first get only flashes of the story, and have to piece it together like a jigsaw) effectively destroys the illusionist power of realist fiction, while the magical nature of the film's "reality" undermines the dominance of ideological norms based on assumptions and "common sense." The events within the house are not merely fictitious but "literary," adapted from works by an illustrious male author, Henry James, a novelist noted for formal perfection, conscious and deliberate artistry, and the systematic suppression of an overt authorial viewpoint; Celine and Julie (Berto and Labourier) invent their own fiction/reality, foregrounding the process of construction.

2. Present vs. Past. The characters in the house -- apart from the child Madlyn -- are the "living dead" (the parallels with Romero's zombies may not be immediately obvious but, once noted, are extremely suggestive). They represent the past that hangs over the present (its structure of values, as recorded in its literature): the legacy of ideology. Celine and Julie create their present against this past, in the process destroying its power. The characters in the house (all of whom wish the child dead and are her potential murderers) are governed entirely by the inherited assumptions of patriarchy; Celine and Julie (whose single-minded endeavor is to save the child) reject those assumptions and look towards a future of liberated, autonomous beings -- necessarily much less clearly defined.

3. Spontaneous vs. Staged. Rivette has made it clear that, by the time shooting began, the film was thoroughly scripted. Nonetheless, it sets up a clear distinction between the free, impulsive, spontaneous behavior of Celine and Julie (whose scenes are consistently played to give the impression of improvisation), and the extremely stylized, "acted" behavior (literary dialogue, theatrical movements and delivery) of the people in the house. Celine and Julie are marked as "free," able to invent their own lives; the "house" characters are imprisoned in a predetermined performance, their lines given them by an author, their attitudes and actions determined by the dominant ideological assumptions. This is underlined by the repetition of lines, gestures, scenes: they are actors trapped in a perpetually repeated play, ghosts doomed forever to reiterate their life's misdeeds, puppets of ideology.

4. Freedom vs. Entrapment. Ce1ine and Julie have free movement: they come and go as they please, the apartment is simply a place to eat, sleep and be together. Sophie, Camille and Olivier, on the other hand, never go outside the house: their lives are circumscribed by its walls. The house, all the inhabitants of which plot against each other, compete with each other, essentially hate each other, recalls Sartre's Huis-clos; it might also be seen as itself the symbol for the dominant ideology ("the house we live in" -- or wish to move out of and see demolished).

5. Autonomy vs. Marriage. Celine and Julie reject marriage, and with it the assumption that a woman is defined, given her meaning, through her relationship to a man. In one sense, Celine takes Julie's place when Julie's childhood sweetheart Guilou returns to resume their "ideal" relationship. Guilou wants every-thing a man is supposed to want of a woman: a wife/mother (they are both dressed in virginal white) and a whore (the obscene dialogue). Celine (as Julie) leaves him literally with his pants down. In the house, on the other hand, marriage is the supreme aim: the entire "house of fiction" plot is centered on the two women's efforts to capture Olivier as husband, their only means of securing an identity for themselves.

6. Female-centered vs. Male-centered. Celine and Julie control their own lives, driving out all the manifestations of male domination. The scene in which Celine takes Julie's place to annihilate Guilou is balanced/answered by the scene in which Julie takes Celine's place in the night club and denounces (in a moment surely derived from Dance. Girl. Dance. but carried further) the managers and customers who want to objectify women as sexual images for the male gaze. In the house, although the actions are mostly initiated by the women, it is the "passive" male who determines them and gives them their aim. Olivier dominates everything not through any force of personality but simply by his position as patriarch, the women's motivation residing solely in the desire to become his next wife. Camille and Sophie habitually construct themselves as objects for his gaze, in their modes of dress and make-up, their calculated poses, their facial expressions; their activeness is repressed and perverted into intrigue and manipulation. The Celine-Julie relationship, on the contrary, represents a celebration of women's activity, of unrepressed, open and directed energy. Their beauty is of a quite different order from that of Camille and Sophie, marked as natural, non-constructed-the beauty of women being themselves, for themselves.

7. Telepathic Communication vs. Suspicion and Distrust. If Celine takes Julie's place to "ruin" her chances of marriage, and Julie Celine's in order to lose her the chance of a job, we are not to read these actions in terms of underhandedness or betrayal. Rather, we are led to believe that each is simply executing the other's real wishes, because she knows what those wishes are and is better able to fulfill them: each frees the other from a possible entrapment in the dominant patriarchal order. Repeatedly in the film the existence of a magical, telepathic communication is confirmed. We may wonder whether Celine and Julie represent two sides of the same personality, or simply the togetherness, and instinctual mutual understanding of needs, of revolutionary women sick of the tyranny of the male.

In the house (the imaginary "real" world of bourgeois ideology), on the other hand, there is no direct communication at all, only subterfuge, deception, manipulation, paranoia. Everyone plots against everyone, all interpersonal communication (which requires, as its basis, the desire to communicate) is blocked. By juxtaposing the action in the house with the Celine/Julie relationship, Rivette realizes the potential of James's fiction to be read as a devastating critique of heterosexual relations under patriarchal capitalism.

8. Friendship vs. Enmity. Celine and Julie, not without hesitation and difficulty, establish as close a togetherness as is compatible with the preservation of personal autonomy. In the house, the women can never be friends: they are divided forever by their competition for possession of the male, their only access to power. It is the ideological dominance of the male that divides them, makes friendship impossible-as it has sought to do to women through the centuries, a point repeatedly confirmed throughout the classical cinema.

9. Childishness vs. "Maturity. " This is perhaps the film's most problematic aspect. Throughout, Celine and Julie behave like children: how do we, as adults, accept them as embodying a new (even if tentative) ideal? But the house presents the alternative: "maturity" is an ideological concept, built on the norms of patriarchy. According to the ideal those norms are meant to embody, the "mature" man takes possession of the woman, becomes in his turn the father; the "mature" woman acquiesces in the patriarchal structure, becomes the wise, kind mother. The very concept of "maturity," that is to say is an aspect of the oppression of women (and of gays) in our culture: "maturity" as we know it is heterosexual-male-defined. Olivier, the patriarch, complacently passive at the center of things, is mature; Camille and Sophie, desperately competing for his favors, ready to murder a child to achieve their glorious end, hating each other (and really themselves) while pretending affection, are mature. On the other hand, Celine and Julie, unattached, not needing man for their self-definition, loving and accepting, the rescuers of the child the mature people wish to kill, are themselves children. Both sets of characters play games, in somewhat different senses: Celine and Julie play for fun and intercommunication, from a delight in togetherness; the "games" of Camille and Sophie are surreptitious (emblematized by the traditional game of "Grandmother's Footsteps" at the party). The implications of all this are enormous and profoundly disturbing. How (if we accept the principles of women's liberation, or simply of "liberation," to which the liberation of women is obviously crucial) are we to relate to the past and its works, which are the products of patriarchy and necessarily structured by its values? The use of Henry James in the film suggests one way, at least, in which the works of the past can still be used (other than as "academic" curiosities or historical documents reiterating discredited patterns): for their exposure of the impossible tensions, the impossible strain on human relations, that the patriarchal order entails.

And one traditional figure of long standing is at once recuperated into the film and partially re-thought: the Romantic image of the Child as symbol of the future, of hope, of new life, of possible transformation (a figure whose literary history is cogently documented by Peter Coveney in his Image of Childhood). Madlyn acquires complex connotations in the course of the film. She is to be seen, first and foremost, as a real child, treated by her "loving" relatives as a disposable pawn in their power game, treated by Celine and Julie as a person, without condescension or sentimentality. But she is also linked to the childhood of both women. In physical appearance she strikingly resembles Celine, and could easily be taken for Celine-as-child; Julie's conversation with her old nurse reveals that the child in the house next door who became sick and disappeared (Madlyn in the past) was just the same age as Julie. Madlyn, then, is the child within the two women -- the healthy, growing part of themselves which they rescue from entrapment within patriarchal ideology.

I want finally to consider the film's limitations and the nature of its achievement. I am not certain whether, at this stage of the evolution of society (and of Feminism), one should speak of "limitations," or simply of the conditions that make the film possible. One objection that I have encountered (interestingly, from men), that the depiction of Celine and Julie draws lavishly on thoroughly traditional myths of "the feminine" (the association of women, explicit or implicit, with cats, witches and children), can, I think, be turned to the film's advantage. The argument of Feminism is not that there are no such things as "feminine" characteristics, but that (a) under patriarchy they have been misrecognized and undervalued and (b) they should not be regarded as the exclusive prerogative of women, but as potentially common-perhaps to differing degrees-to both sexes. The aim would be, then, not simply to reject the myths, but to cleanse them of their pejorative connotations and question the use to which they have been put-the construction of Woman as "the other." Men, for example, have traditionally claimed "reason" for themselves, established it as the superior quality, and compensated women (condescendingly) by according them "intuition." One may certainly argue that men lose as much as women from the insistence upon such a dichotomy. The answer, then, would not be to deny the existence of "intuition" altogether, but to revalue it. In the film, Celine and Julie consistently use their intelligence, but never at the expense of magic. The film's supreme moment triumphantly endorses "magic" over the "moronic logic" of the laws of narrative, by finally breaking down the boundaries between "reality" and "fantasy": the moment where, after the last "dream" of rescuing Madlyn from the house, Celine and Julie wake up to discover that the child is indeed there with them in the apartment. It is one of the great liberating moments of cinema: the moment that allows Celine and Julie to "go boating," and to pass the ghosts, now stiffened into total paralysis, drifting downstream (back to the past) as the women row Madlyn upstream towards the future.

One may be more legitimately worried by the film's treatment of men and of the possibility of heterosexual relationships; yet, again, perhaps it is simply acknowledging a particular phase of social evolution. Women have always been oppressed, but now (though the oppression continues, with minimal "liberal" modifications) they know they are oppressed. The difficulties of the heterosexual relation within the existing economic/ideological structure (a structure which depends on women's subordination) have become (as most of my women friends repeatedly testify) almost inseperable: even men who can intellectually grasp and accept the notion of woman's autonomy gradually reveal enormous problems with it emotionally. Our culture may have to go through a phase (when the current period of general but uneasy and unconvincing reaction ends and the next wave of Feminism gathers power) wherein the close ties are same-sex, until new modes of male-female relationships are hammered out. Yet this does not satisfac-torily excuse or explain the film's total withdrawal of sympathy from the male: all the men in it are permitted to exist only on the level of parody, denied even a moment of grace, forgiveness or understanding. (Interestingly, this trait is reflected in certain recent pop-Feminist movies such as Private Benjamin and Nine to Five).

On the other hand, the film's treatment of Lesbianism is curiously equivocal. With her (male and female) friends in the bar, Celine explicitly rejects any suggestion that she and Julie relate sexually ("She's not a dyke"); yet it is unmistakably implied that the two women share the same bed. One has the choice of interpreting Celine's remarks as just her way of dealing with casual acquaintances who "wouldn't understand," or of seeing the women's relationship as pre-sexual, the affectionate romping and cuddling of schoolgirls (not that that, of course, is without sexual connotations -- the question is of the extent to which these are acted upon). Having dared so much, the film might have dared further: to give us one of the first positive images of Lesbianism in the history of narrative cinema -- an option it leaves for Bruce Beresford and his remarkable The Getting of Wisdom.

Nonetheless, the film constitutes an extraordinary and forward-looking achievement. Before it was shown in my film study course, the students were debating as to whether radical (as opposed to liberal) Feminism were compatible with "entertainment," and the general tendency was to conclude that it isn't: the codes that govern entertainment, the long-established habits of patriarchy, with their traditionally defined "ways of seeing" and their reliance on culturally constructed images of masculinity and femininity, would directly conflict with everything a Feminist filmmaker wished to achieve. This diagnosis seemed confirmed by Laura Mulvey's seminal article in Screen. "Narrative Form and Visual Pleasure," and also-somewhat depressingly-by the practice of the Mulvey/Wollen Riddles of the Sphin. Riddles -- and especially its sustained center-piece, "Louise's story told in thirteen shots" -- is challenging and theoretically interesting, but our notions of "pleasure" would have to be redefined beyond recognition before the film could be claimed as "entertainment." The impression one took from the Mulvey/Wollen intervention was of a choice between mindless surrender to the addictive and reprehensible pleasures of mainstream cinema on the one hand, and sheer cerebral hard work on the other. Celine and Julie -- in the context of the progressive narrative tradition I have described -- at the very least suggests ways of breaking this impasse. This is not to assert that the film is accessible within mass culture. Yet, while never pandering to ingrained habits, while systematically deconstructing and exposing the dominant codes of narrative, characterization and presentation, Celine and Julie engages its viewers in a narrative play that permits identification, fun, suspense, emotional involvement; it allows, indeed encourages, pleasure; one can describe it as intensely entertaining, without stretching the customary meaning of the word beyond breaking-point.

To conclude -- and to consolidate this note of hope for the future of narrative -- I want to return briefly to the comparison I threw out earlier with Dawn of the Dead. Romero's film is formally more conservative, more traditional, than Rivette's, though extremely audacious in the context of the contemporary commercial cinema (it was, of course, produced independently). The logical corollary is that it also cannot proceed as far in its thematic development, in terms of a revolutionary break with the past. Nonetheless, beside today's weary rehashes of the attitudes and narrative procedures of the Hollywood past, it stands as a reminder of the degree to which traditional narrative can be bent and stretched into progressive forms. (2)

The parallels between the two films are striking. Like the occupants of the "house of fiction" -- though in a less elaborated and complex way -- Romero's zombies, mindlessly gravitating to the shopping mall, represent the habits of the past from which the living characters must strive to extricate themselves (the make-up of the ghosts in the last part of Rivette's film, as they lose their potency, is almost identical). The film never reaches anything comparable to the positive image of women in Celine and Julie, but it follows Fran's development as she progressively casts off the entrapments of patriarchy: notions of the woman as inferior, as helpless, as irrational, as passive; the acceptance of herself as an image for the gaze of the male; crucially, the idea of marriage as norm. At the end of the film the emblems of male power are either transferred to the woman or surrendered (Fran flies the helicopter, the surviving male relinquishes the rifle to the zombies). The film even offers an equivalent for the rescue of Madlyn in Fran's unborn child, carried out of the clutches of the past towards an uncertain and precarious future. There is also, throughout the film, a strong element of play, in the stylized comic-strip violence, in the film's awareness of itself as genre movie, as fantasy: the audience is invited to participate in a macabre and bloody game that manages to remain, by virtue of the stylization, at once fundamentally serious and good-humored.

It is clear that the most audacious and radical films will continue to be made outside the commercial mainstream. Yet the co-presence in seventies cinema of Celine and Julie and Dawn of the Dead suggests the possibility of significant parallel developments in different spheres of independent film-making, at different removes from the mainstream. Our delight in story-telling and story-hearing, while it needs to be carefully scrutinized, does not have to be abandoned.