LOLA home                        HOME        CURRENT ISSUE 

The Broken Trilogy:
Jacques Rivette's Phantoms   

Adrian Martin





At its extreme, film is the rejection of film, its contradiction (its ‘anti-film’): only the milestones remain, the tokens of its ‘passage’; forever past/future.

– Jacques Rivette, 1969 (1)

  1. Jacques Rivette in the collective text ‘Montage’, in Jonathan Rosenbaum (ed.), Rivette: Texts and Interviews (London: British Film Institute, 1977), p. 73.

A proposition: the cinema of Jacques Rivette (1928-2016) is profoundly psychoanalytic. It is psychoanalytic through and through, on every level, and in at least two major ways. First, it practices a wild psychoanalysis – arising from the many forms of improvisation, play, psychodrama, encounter, desiring impulse and acting out. Second, it practices a reflective, secretive, inward-turning psychoanalysis, attuned to the silent working-through of trauma and its after-effects of both mourning and oblivion. This second psychoanalysis is captured well in the poignant koan at the heart of the director’s final film, 36 Views of Pic Saint-Loup (aka Around a Small Mountain, 2009), uttered by Jane Birkin in the central role: ‘My curse is to remember … my curse is to forget’.


Many commentators have intuited this deep connection between cinema and psychoanalysis in Rivette. Jonathan Rosenbaum writes of films that ‘we may remember afterward like shards of unfathomable dreams’. (2) John Hughes titled a 1975 interview with Rivette ‘The Director as Psychoanalyst’, and later suggested that he ‘educates and excites his actors into a kind of conscious dream-state which enables him to film the Unconscious’. (3) Rivette himself – affected by key experimental screen narratives of the mid-1960s including Ingmar Bergman’s Persona (1966), Robert Rossen’s Lilith (1964) and Carl Dreyer’s Gertrud (1964) – superbly evoked the way that certain films, even if they do not ‘function formally as a dream … nevertheless also prescribe an “oneiric” reading: at once the telling of a dream and an analysis (an analysis in which the roles are unceasingly changing …)’. (4)


It is a critical commonplace that Rivette’s films interrogate the conditions of performance – in film, in theatre, in life, in love, in politics. In almost every case, a story of everyday, ever-widening connections between people is juxtaposed with intimations of a hidden conspiracy that sometimes turns out to be whimsical (as in Up Down Fragile, 1995) and oft-times sinister enough to justify the paranoia and breakdown it triggers in individual lives (Out 1: Noli me tangere [1971], Out 1: Spectre [1974], Secret défense [1997]). Sometimes, the ‘top secrets’ behind Rivette’s fictions refer not to some shadowy structure of state power but a familial tangle, thus ‘bringing it all back home’ as many Rivette films manage, each in their diverse ways, to do.


The pattern of remembering and forgetting in Rivette’s cinema – cursed to do both, dwelling in the ‘forever past/future’ – relates not only to characters and their fictions, but the swirls and eddies of the filmic œuvre itself, the gestures it makes in an artistic and cultural sense, and the psychic investments of the principal auteur in his own difficult, unfolding work.



2. Jonathan Rosenbaum, ‘Rivette’s Rupture’, Chicago Reader, 27 February 1992.

3. John Hughes, ‘Autodialogue’, Film Comment (May/June 1978), pp. 70-71.

4. Rivette: Texts, p. 86.











Don’t fail me, or you’ll lose the very memory of me.

– Emmanuelle Béart in Story of Marie and Julien (2003)


In October 1987, upon the release in France of his film Wings of Desire, Wim Wenders was asked to guest-edit a special edition of Cahiers du cinéma. Wenders chose as his theme the dream projects (he called them ‘submarine films’) of other directors: testimonies to films they have always wanted to make, films they have been prevented from making, films still or perpetually in script development. (5)


Some of the entries in the issue are sad, and none more so than Jacques Rivette’s sparse, forlorn page. (6) He retells the story of why his project Story of Marie and Julien (Histoire de Marie et Julien) was not made in 1975. Rivette was, in September of that year, in the middle of what he later described as a ‘mad idea’: the breakneck shooting, back-to-back, of four interrelated feature films. (7) The narrative worlds of the films are not the same or continuous (in fact, they are starkly incommensurate), and each film was to nominally belong to a different genre: love story, mystery-thriller, pirate-adventure, musical. Apart from the elaborate formal experimentation involved in the project (more on this below), only one element really united the four pieces: the decisive presence, in each plot, of female phantoms. The initial title of the series was Les Filles du feu (‘Daughters of Fire’), later renamed Scènes de la vie parallèle (‘Scenes of Parallel Life’) – a transformation of the titles of Honoré de Balzac’s literary series ‘Studies of Manners’, comprising the ‘Scenes’ of Private, Provincial and Parisian Life.

  5. Wim Wenders, ‘Introduction’, Cahiers du cinéma, no. 400 (October 1987), p. 6.  

6. ‘Jacques Rivette’,
Cahiers du cinéma, no. 400 (October 1987), p. 42.  

7. Hélène Frappat and Jacques Rivette,
Trois films fantômes de Jacques Rivette (Paris: Cahiers du cinéma, 2002), p. 13.

Due to logistics, the films had to be shot out of order. With the second (Duelle [une quarantaine]) and third (Noroît) in the can, Rivette began work on the first in the series: Marie and Julien, the love story. It starred the well-known British actor Albert Finney and French-American musical legend Leslie Caron. But the semi-improvisatory process – there was no script as such, merely a scene breakdown – ended two days in, when Rivette collapsed from nervous exhaustion (according to co-writer Eduardo de Gregorio, ‘he cracked’). (8) His recovery took the better part of two years, during which time he edited and released Duelle and Noroît (sometimes respectively referred to in English as Twhylight and Nor’wester) in 1976. Marie and Julien was abandoned, as was the complete plan for a tetralogy.


In the Cahiers issue edited by Wenders, Rivette speaks about the Marie and Julien incident – a full twelve years later – in the blank tones of a difficult, post-traumatic mourning: he describes the memory as ‘more than a regret, a true remorse’. (9) Nothing exists of the project, ‘no film or stills’; he claims to have even forgotten what its 1975 conclusion was meant to be, since ‘we hesitated between several endings and left it to the shoot to decide’. (10) Marie and Julien had effectively vanished from the face of the earth – leaving only a hole in its principal maker’s memory. The only visual testimony he can offer is an inventive collage so true to the spirit of this love story between a human and a phantom: stills from two ‘completely unconnected’ films, Alan Parker’s Shoot the Moon (1982) and Nanni Loy’s Head of the Family (1967), arranged so that Finney and Caron (respectively) look across at each other, ‘face to face, yet separate’. (11) Rivette ends this piece with an appeal to his lost actors:




8. Hélène Frappat, Jacques Rivette, secret compris (Paris: Cahiers du cinéma, 2001), p. 152.  


9. ‘Jacques Rivette’, p. 42.  


10. Frappat, Jacques Rivette, p. 152.


11. ‘Jacques Rivette’, p. 42.  

  I hope that Leslie and Albert, if these lines pass under their gaze, will realise that this still-born film is the greatest regret of my life as a filmmaker – and that to have abandoned them, at the start of our shared adventure, is the greatest remorse of the director. (12)  


12. Ibid.

This unmade Story of Marie and Julien was, however, to have a long, largely secretive and protracted afterlife – until its full-blown resurrection as a film starring Emmanuelle Béart and Jerzy Radziwilowicz in 2003. According to de Gregorio, Rivette tried, at various times, to resuscitate the project with other male leads cast alongside Caron: Michel Piccoli, the director Maurice Pialat (who refused), and even himself – a not insignificant detail. (13) It made another sort of return, in this half-life period, in a small book lovingly assembled in 2002 by critic-novelist-screenwriter Hélène Frappat: Trois films fantômes de Jacques Rivette (‘Three Phantom Films’).


Films about phantoms, phantom films: the semantic slide offered by Frappat is apt, and inescapable. In her 2001 study Jacques Rivette, secret compris (‘Understood Secret’), Frappat posits Marie and Julien as one of the ‘missing films, the phantoms whose traces we seek’ in Rivette’s career (of course, she wrote this before Marie and Julien finally became a real film). (14) She does not merely mean that we can seek the traces of their production documents, such as design sketches or scripts; we can also – if we are faithful auteurists – seek the echoes, allusions, reworkings of these phantom texts in all the works that Rivette did manage to achieve subsequent to 1976.



13. Frappat, Jacques Rivette, p. 152.





14. Ibid., p. 150.




More and more I think that there is no auteur in films and that a film is something that pre-exists in its own right. […] you are trying to reach it, to discover it, taking precautions to avoid spoiling it or deforming it.

Rivette, 1968 (15)



15. Rivette: Texts, p. 31.

Auteurism – and I here I invoke this critical method at its height of its intellectual inventiveness, not the diminished caricature of dime-store Romantic ideology that so often makes the rounds as a straw man in theoretical commentary these days – is fascinated by the many ways in which a director’s works can be seen to speak to each other and form diverse networks. What auteurists seek are the ways in which the films – successively, or in more displaced, circuitous patterns – answer, extend, invert, fulfil, critique or even destroy each other. Hence the non-linear, non-chronological arrangement of titles and chapters that the best auteurist studies often pursue, as in the sophisticated, book-length treatments of Alain Resnais by François Thomas or Jean-Louis Leutrat and Suzanne Liandrat-Guigues, of Abel Ferrara by Nicole Brenez, of Jerry Lewis by Chris Fujiwara, or of Rivette by Hélène Frappat .


Understandably, filmmakers themselves sometimes scoff at this mode of critical reverie: they know, in the nitty-gritty day-to-day of the movie business (whether commercial or art movies) that their next film will not necessarily be one they planned or hoped to do, but rather the one that  through diverse, often surprising paths of opportunity or chance – manages to become a viable, concrete project.


This is why the development or evolution of a film artist’s work, at least in the milieux of feature-length, relatively expensive narrative cinema, often takes an ‘indirect aim’ (as Raymond Bellour once wrote of Hitchcock), ongoing obsessions or investigations investing themselves, often only partially, whenever or wherever they can within the framework of any given project. (16) Rather than the ticking-off of simple, repeated themes, stories and character-types, it is the wayward progression of such drives and investments that an auteurist analyst intuits, and which he or she sets out to decipher; it is only the inner logic of this indirect aim, in its frequently bumpy progression, that gives an œuvre its identity and richness. Such a mode of decipherment calls for a different mode of cine-psychoanalysis than the scenario-type most commonly used in film interpretation.



16. Raymond Bellour, ‘Alfred Hitchcock’, in J.L. Bory and C.M. Cluny (eds), Dossiers du cinéma: Cinéastes I (Paris: Casterman, 1971), pp. 117-21.  

Rivette’s films offer themselves to the auteurist quest in a special way. Several key films in Rivette’s career are – on the model of a process-based art – laboratories, open-ended experiments that feed the variations to come, in various explicit or disguised ways, in subsequent work. Thus, producer Stéphane Tchal Gadjieff remarks of the thirteen-hour Out 1 that it ‘constituted for Rivette a kind of treasure chest. I believe a part of the films he subsequently made take up and develop elements that he wanted to work with beginning with Out’. (17)


Conversely, especially in the final decade of Rivette’s work, there was a marked involution, a striking return to precisely his projects of the 1970s. After realising Story of Marie and Julien, Rivette was able to make Don’t Touch the Axe (2006), a close adaptation of the ‘Duchess of Langeais’ section of Balzac’s History of the Thirteen – which was, in a far looser way, the ‘master text’ generating the complex, conspiracy intrigue of Out 1.

  17. Frappat, Jacques Rivette, p. 149.

Indeed, these two films also return to an even earlier Rivette work – L’Amour fou (1968), the veritable inauguration of his experimental phase, which (as two French interviewers intuited) was the only previous occasion on which the director had allowed himself to explore ‘the complexity, the instability of a couple’s connections … the same, very naked pain, tied to love’ – to which the interviewee replies, after a long silence, simply ‘yes’. (18) And 36 Views of Pic Saint-Loup is a comedy-drama of exquisite agony, recalling many of the director’s previous works – and centred very exactly on the remembering and working through of a traumatic incident.


How to account, in a psychic sense, for these swirls of forwards-projection and backwards-return, of remembering and forgetting, in Rivette’s career? That is my central question in this essay, with particular emphasis on Story of Marie and Julien as both the culmination of a broken trilogy and the gesture that poignantly creates the room for a new life, a second chance – just at the moment that the life-force itself is draining away. This interpretive plunge will necessitate (as sophisticated auteurism has often been driven to do) some degree of biographical speculation into the life of a rather secretive and fiercely private man.



18. Jean-Marc Lalanne and Jean-Baptiste Morain, ‘L’art secret’, Les Inrockuptibles, 30 March 2007. English translation.







Why four films at the same time? In the first place because (since the filmmaker does not enjoy the same status in relation to his characters as the Balzacian novelist does) it is the only way to establish a specific ‘circulation’ between these films with certain characters and certain décors reappearing from one to another under different lights, contradictory or complementary.

Rivette, 1974 (19)




19. Rivette: Texts, p. 89.  

Duelle, Noroît and Story of Marie and Julien form a strange kind of trilogy, to say the least. First, it is a trilogy by default, since it lacks the fourth part of the envisaged tetralogy. Second, it is a trilogy accomplished after a gap of almost thirty years. Third, Marie and Julien, as eventually realised, both is and is not presented as part of a series. Shorn of the ‘Scenes of Parallel Life’ title, it is surely a stand-alone work in many viewers’ minds.


There are other broken trilogies in cinema history – for example, Lars von Trier’s as yet uncompleted set of Dogville (2003), Manderlay (2005) and Wasington [sic]. There are also striking examples of trilogies completed with a substantial lag between the second and third entries: Dario Argento’s Mother of Tears (2007) made 27 years after Inferno (1980), George Miller’s Mad Max: Fury Road (2015) made 30 years after Mad Max Beyond Thunderdome (1985) or, most spectacularly, José Mojica MarinsEmbodiment of Evil (2008) made 41 years after This Night I’ll Possess Your Corpse (1967). There is something undeniably heroic about all these efforts, however we might evaluate the results: it is as if the director’s Schopenhauerian Will has at last asserted its personal, artistic mission over all the impossible, frustrating exigencies of the movie-making industry, and thus bent Chance to Destiny.


However – in every one of these cases, but especially in Rivette’s – other factors, some external to the filmmaker and some internal, also intervene in this process of delay and give it a particular historical and cultural meaning. At both the simplest and most complex levels, times change: the world in which the conclusion to a series was conceived turns out to be very different to the one that eventually greets it. And thus the gesture of completing the work and placing it in this new context is inevitably a complex one that is going to call upon the entire history of the director’s work and its changing, sometimes combative relationship with the times (and places) it has crossed.





Yes, I think that’s the basis for everything: to treat the text as material which plays a role exactly similar to the other materials in the film: the actors’ faces, their gestures, the photographic texture.

Rivette, 1973 (20)



20. Ibid. p. 52.

In 1974, Rivette composed a short statement of intent for the Filles du feu series. In it, he specifies several open-ended intentions about the purpose of shooting four films end to end:



-          That it would ‘mainly’ be a ‘what happens’ experiment in seeing how four very different films, if shot in this way, ‘might be modified (accentuated, influenced, transformed) by this interplay’. (21)

-          That the films will have their narrative basis in a semi-invented mythology (inspired by the Celtic tradition) involving, each time, the confrontation of Sun and Moon Goddesses over ‘the forty days of Carnival’. (22)

-          That each ‘block-sequence’ of the four films would be ‘subjected to a method designed to break down … conventional dramatic techniques’, establishing ‘an écriture based on actions, movements, attitudes, the actor’s “gestural”’. (23)

-          That live music – meaning the presence of musicians actually playing along with the actors, both off and on screen – would be crucial to all the films, in order to create an interrelationship of three ‘spaces’ or parameters: the body’s space created through its movement; the space imposed by décor and the camera’s field; and the ‘simultaneous musical space’. (24)

-          That the four films would be marked by a formal evolution, a ‘progressive accentuation’ in the radical work on mise en scène: from Marie and Julien where is to be ‘an element of dislocation and strangeness within a dramatic construct still following the rules of romantic fiction’ through to Noroît (initially posited as fourth in line, later made third) in which all the ‘various aspects are to be driven to paroxysm’. (25)



21. Ibid., p. 89.  


22. Ibid.  



23. Ibid.  



24. Ibid., p. 90.  



25. Ibid.  

In other words, the series was to enact a disintegration of film language and narrative form from its starting to ending points. In the event, the step from Duelle to Noroît alone bore the burden of acting out this progression. For those who value these films, it is the two-step of the ‘dislocation or strangeness’ of Duelle then giving way to the far more extreme and disorienting avant-gardism of Noroît which endows this pair of closely interlinked films with their enduring charm, disquiet and fascination.

There is something else which Rivette has always been more or less silent about: no films in this director’s career are more extremely camp or outrightly queer than Duelle and Noroît, with their outrageous costumes, histrionic performances, pervasive gender switches (Bernadette Lafont takes the leading male pirate role in Noroît) and general riot of gaudy locations and vibrant colours. This is a tendency in Rivette that first exploded in Céline and Julie Go Boating (1974) – and was only rarely to reappear, and only in very tempered forms, past 1980. The queer critical appreciation of the series, then and since, is well caught in David Ehrenstein’s tribute to Duelle – ‘Surely it’s about death. And drugs. And style – in a way that hasn’t been seen on screen since the heyday of Sternberg and Dietrich’ (26) – and in John Hughes’ en passant remark that ‘the German S&M leather underground … links Lang to Rivette (and Godard)’. (27) This brings into the reception of the films (quite validly, in my view) a whole 1970s counterculture of gender, drug and lifestyle experimentation that otherwise remains discreetly off-screen.










26. David Ehrenstein, ‘Duelle, Senses of Cinema, no. 43 (2007).

27. Hughes, ‘Autodialogue’, pp. 70-71.

At the same time, the two films marked a decisive break in Rivette’s career. Although there was much controlled improvisation across the diverse elements of the mise en scène – the interplay of camera, actors and live music – there was also, necessarily, a fairly rigid structure imposed on the narratives, each film being comprised of ‘some fifteen block-sequences […] divided into three main sections, three acts, corresponding to the three lunar phases’. (28) This was a surprisingly classical gesture coming, in the mid-1970s, from the maker of Out and Céline and Julie! Juliet Berto (who was a last-minute replacement in Duelle for the Brazilian actress Norma Bengell, star of Mario Bava’s Planet of the Vampires [1965]) has testified to the fact that, in contrast to the improvisation which had been ‘too intense’ on Céline and Julie, her role in this film was based upon ‘an extremely worked-out text; we had only to play on the transformation of our facial expressions and, with more money, we were able to go further on this level: costumes, vocalisation …’. (29)




28. Rivette: Texts, p. 89.  


29. Jean-Claude Moireau, ‘Entretien: Juliet Berto’, Cinéma, no. 314 (November 1985), p. 20.




There’s an abyss between you and me, and I don’t know how to cross over.

Story of Marie and Julien


In light of the curious mid-1970s conjunction of the almost nostalgically classical and the most outré avant-garde leanings in Rivette’s cinema, Rosenbaum offers an intriguing, art history-inspired periodisation for the Filles du feu series: Duelle and Noroît ‘could be called transitional works between the modernism of Rivette’s first six features and the postmodernism of his last six’. (30)


After Noroît, Rivette’s cinema will never again be so experimental, daring or rule-breaking. Did something more than the director’s health crack in that moment of crisis in 1975? Did his artistic resolve also take a battering? And did that particular crack trigger, or come to associate itself, with other cracks in the life and times, even less accessible to us? Whatever the case, it is undeniable that there is a strong turn to classicism in the post-1980 films, especially beginning from L’amour par terre (1984).



30. Rosenbaum, ‘Rivette’s Rupture’.



François Thomas has noted an equally important stylistic-formal index of this shift. Rivette abandons the experiments in ‘wild’ direct sound that are taken to the extreme in the Filles du feu films (sound recording itself becoming part of the adventure of shooting, as in Robert Altman or Jacques Rozier), and opts for a far cleaner, minimal soundscape (listen to the clocks in Marie and Julien), with its sonic backgrounds carefully equalised in post-production mixing – whereas in Duelle and Noroît virtually every cut ushers in a violent re-setting of direct-sound atmospheres. (31) Indeed, the 1980s inaugurates a certain ambience common to virtually all Rivette’s latter films: in contrast to the often densely populated social and cultural worlds-in-a-frame marking the 1960s and 70s, there is a notable depopulation (he loves filming in the deserted Paris of August). More generally, the camera and figure movements within a given environment that create mise en scène – an art at which Rivette has long been a supreme master – take on what can be regarded as a more motivated nature (the camera following the characters) and a more traditional staging purpose (to underline and express emotions and interrelationships), both traits associated with filmic classicism; whereas the interplay of movements and décor in Duelle and Noroît is altogether more wayward, playful and unpredictable at virtually every moment of the fiction.


31. Personal conversation with the author, Cinesonic conference, Melbourne, 1999.  


Let us now place Marie and Julien – both its initially planned and eventually completed versions – within this periodisation and its ramifications. In its spot as the projected first film of the Filles du feu series, it was to be the most seemingly normal of the tetralogy. It would have functioned, in the mid-1970s context – immediately after the most visible phase of Rivette’s modernist experiments in L’amour fou, Out 1 and Céline and Julie – as a kind of demonstration of conventional ground rules, a classicism between quotation marks.


By 2003, naturally, this context has entirely altered. There is some continuity of personnel (William and Nicole Lubtchansky, respectively cinematographer and editor), and an appearance by Nicole Garcia from Duelle, but other key collaborators (producer, screenwriters) have changed greatly. The experimentation with direct sound recording has disappeared altogether as a parameter; so has much of the camp romance surrounding the appropriation of Celtic mythology (the central female figure is no longer anything so grand as a Sun or Moon Goddess, merely a phantom). Generally, the quality of excess is gone from the mise en scène’s constant, fluid modulation. Is this a capitulation to neo-classicism by an ex-radical, of the kind we see in the careers of Roman Polanski, Claude Chabrol or Wim Wenders?


Yet classicism can never simply return, whole cloth, after modernism: it comes freighted with doubt, with question marks, with grace notes of ambiguity and destabilisation. Rosenbaum emphasises that, alongside the graceful and extended long takes – hallmark of a Bazinian aesthetic and ethic that clearly left a deep mark on Rivette since his youthful days at Cahiers – there is also, consistently in Rivette’s later career, the jarring elisions enacted by the often startling cuts both within and between scenes, reminiscent especially of the modernist principles which Rivette propounded in the 1969 collective text ‘Montage’, and evident in his work, embryonically, right from his feature debut in Paris Belongs to Us (1961). (32)


Seen in this context, Marie and Julien reaches out from within the limits of its own fiction, to touch and possibly unlock a larger secret or mystery in its director’s œuvre.





32. Jonathan Rosenbaum, ‘The Choice Between Art and Life’, Chicago Reader, 31 January 1992.




The laws of the phantom world escape us.

Rivette, 2002 (33)

  33. Frappat & Rivette, Trois films fantômes, p. 15.

In Klimt (2006), Raúl Ruiz structured the biopic of an artist in an unusual, striking way. Much of the film hinges on a moment of chance coincidence: the very instant that the word ‘Paris’ is uttered in Klimt’s presence, a window or mirror also happens to break. And, from that moment, the two things are welded, perfectly illogically, in his psyche for the rest of his life, determining memories, sensations, associations: Paris and breaking glass, which undergo (individually and together) many wild poetic transmutations in the course of the action. Two signifiers, thus, with no clear or obvious signified, and no necessary semantic connection, but carrying (almost comically) the enormous weight of an individual’s identity and destiny in their mad, unstoppable, signifying path.


Watching Klimt can easily make one reflect: all our lives are held in strong but meaningless patterns like these, chance coincidences that trigger unfathomable signifying chains. No conventional, redemptive, thematic meaning, only abandonment to what the art critic Edward Colless calls the ‘error of our ways’. (34) In 36 Views of Pic Saint-Loup, this is given a stunning crystallisation: the trauma experienced by Birkin’s character, and the working-through it achieves, comes down, quite literally, to the successful tearing of a single, flimsy piece of paper within the circus ring: this tiny, almost weightless little signifier which, nonetheless, holds so much personal and collective weight.


Part of the psychoanalytic depth of Rivette’s cinema comes from this primacy and force of the signifier – an idea which sounds very 1970s, but needs to be resuscitated in an age when tidy signifieds once again lazily rule so much commentary on film. In 1973, he spoke admiringly of films that





34. Edward Colless, The Error of My Ways: Selected Writing 1981-1994 (Brisbane: Institute of Modern Art, 1995).  

  impose themselves visually through their monumentality. […] What I mean is that there is a weight to what is on screen, and which is there on screen as a statue might be, or a building or a huge beast. And this weight is perhaps what Barthes would call the weight of the signifier […] [there is] an element of violence, of affirmation without evidence, of erotic power, which I’m trying to express when I talk of monumentality […] Knowing [narrative] will reappear, one might as well try to have it circulate as much as possible, to use Barthes’ phrase […] to have the signifieds that are present be caught up and carried in the general movement of the signifiers. (35)  




35. Rivette: Texts, pp. 49, 52.

To watch Duelle and Noroît today is to be instantly transported back to the moment in world film culture when the signifier, grasped and theorised in this way, vitally and urgently mattered. The films literally creak under the weight of their signifiers; every footstep, every camera movement, every change of light or colour in them is palpably felt by the viewer. Rosenbaum testified, in that mid-1970s era, to the ‘breaks in legibility, ruptures of tone, momentary disorientations or encumbrances that we usually skip over or skim […] In Duelle, however, we must pass through them’. He evokes the process of ‘returning each of these sign systems to a purer state’, thereby allowing viewers to ‘witness the primal birth of meanings and sensations when some of these ingredients link up’. (36)




36. Jonathan Rosenbaum, ‘Duelle: Notes on a First Viewing’, Film Comment (September-October 1976), p. 28;




For me, the most powerful pleasure in cinema – and this is something that interests me more and more, and I don’t know if it can be related to this cinema of signification, of monumentality, that we were talking about – is connected with terror and anguish.

Rivette, 1973 (37)




37. Rivette: Texts, p. 53.

In the eventually realised version of Story of Marie and Julien, this magnificent flight of the signifier has, in a sense, gone deep inside the film, been internalised as a symptom, a quiet source of perturbation. It no longer plays on the surface so histrionically but, rather, silently unmakes all the linkages of sense and certainty. There is an immense anguish in the film – not something purely locatable or localisable in the characters and their fictional feelings: an anguish that circulates.


In a remarkable text on his last three films, Jean-Marie Samocki suggests that 36 Views of Pic Saint-Loup allowed Rivette, at the end, to liberate himself from the dark energies unleashed by both Marie and Julien and Don’t Touch the Axe. As twin-films exploring the amorous passion of the couple, they erect opposing poles: where the former is devoted to the ‘positive, magical exaltation of the absolute’, the latter is its ‘exaggerated, imprisoning’ expression. (38) Both films dance, in a truly agonised and agonising way, around the Romantic (in all senses) dream of total fusion between a man and a woman: where Marie and Julien crosses the abyss of non-relation in order to arrive at a moment of supernatural transcendence (thus inserting itself into a vast history of such supernatural love stories in cinema, from Peter Ibbetson (1935), The Enchanted Cottage [1945] and The Ghost and Mrs Muir [1947] to On a Clear Day You Can See Forever [1970], Lovers of the Arctic Circle [1998] and Le Pont des Arts [2004]), Don’t Touch the Axe insists on the impossibility of fusion to the point of madness and death.





38. Jean-Marie Samocki, ‘Jacques Rivette, après l’absolu’, Trafic, no. 72 (Winter 2009), p. 30.

Like 36 Views of Pic Saint-Loup, the gesture of Story of Marie and Julien as a work comes ultimately to rest upon what is itself, literally, a small, fleeting but decisive gesture. We can grasp it through an attention to the way in which Rivette calls upon and dismantles genre – for, in this regard, Marie and Julien is a risky, high-wire mix. The film plays out at the same crossroads theorised by Stanley Cavell, between a ‘comedy of remarriage’ and a ‘melodrama of the unknown woman’, two generic forms that can often seem like X-ray reversals of each other. Eduardo de Gregorio described the project as developed in the 1970s as ‘a sort of variation on Vertigo [Hitchcock, 1958], the story of a man who lives alone, surrounded by the memory of a woman he loved, who then encounters a woman identical to her – and he lives with her until he discovers she is actually dead’. (39) This is a description that swiftly nudges the film into a bracket of contemporaneous (and quite explicit) reworkings of Vertigo, including Philippe Garrel’s Wild Innocence (2001), Paul Schrader’s Forever Mine (1999) and Chantal Akerman’s The Captive (2000). But Rivette’s own view of the initial project, at least from the vantage point of 2002, is rather different, and suggestively so:





39. Frappat, Jacques Rivette, p. 152.  

  The principal motif was a variant on the old romantic theme of the ‘dead lover’, who must fall in love with a mortal being in order to lift the curse that prohibits her from entering the world of the dead. The other ambition of the project was to tell a story of amour fou between a man and a woman who are in their 40s: ‘At Long Last Love’, as Cole Porter put it. (40)  


40. Frappat & Rivette, Trois films fantômes, p. 40.

This reference to Porter – or, in the final moments of the finished film, Blossom Dearie singing the Hilliard/Garson standard ‘Our Day Will Come’, an irresistible echo of the use of Peggy Lee’s ‘Senza Fine’ at the close of Va savoir (2001) – indicates the leap that Story of Marie and Julien will make, right at the end of its narrative, from a tearful moment of fantastique sorrow (Marie disappearing, obliterating Julien’s memory of her, and then crying) and the ultimate upbeat, unexpected event, when the tears fill her wrist wound, triggering a flow of mortal blood … and Marie answers Julien’s sleepy enquiry of who she is and why she is sitting there with the perky ‘Give me a little time’.


(Give me a little time: this cry echoes for us at the – almost literally unbelievable – moment of Out 1’s triumphant return to the world of cinema in 2016, on DVD/Blu-Ray and in numerous showcase screenings. A film so often described, in its 1970s era, as being about post-1968 exhaustion, disillusionment, descent into psychosis, individualistic solipsism, death of the radical-political dream … and too often re-described that way, still today, frozen in its ‘first and final instance signified’ of history and place, trapped in this endless repetition of a sad dead-end. When, quite to the contrary, we should be seeking to liberate it, in all its experimental spirit, from the shackles of its moment, taking it for a walk along new lines of light and thought.)


Could, indeed, Rivette’s ‘hesitation between several endings’ for Marie and Julien in the 1970s have given arise to this ingeniously modernist double ending of 2003, in which both genre and tone are completely switched around in a split-second? Whatever its inspiration, this ending manages to seize, in a fragile but commanding way, exactly what it explicitly asks for: a little time … an extra moment or plateau of phantasmic time for cinema, and for those who invest their lives in producing its material imaginary.




  Nothing definitive should be said here, however: the tale is to continue soon on another stage and with other phantoms.
–   Jean-André Fieschi, 1972 (41)
  41. Jean-André Fieschi, ‘Jean-Marie Straub [& Danièle Huillet]’, in Richard Roud (ed.), Cinema: A Critical Dictionary (London: Secker & Warburg, 1980), p. 877.

Claire Denis’ splendid documentary Jacques Rivette, The Watchman (1990) alludes several times to the extreme privacy of the director’s private life, and the fact that even his closest collaborators seemed scarcely to know him in any conventional, social sense. So I provisionally close this essay with three biographical details. Even if I were not able to vouch, in each case, for their truth, I would insist on their significance.

It was told to me that, by the time Rivette was ready to film 36 Views of Pic Saint-Loup (a project he prepared over several years), he had already begun to enter a fast-escalating state of mental deterioration, unable to remember, from day to day, what had already been shot. He lived for another seven years with Alzheimer’s, and this was the reason – usually unspoken in public – why there were no more interviews, no appearances by him on proliferating DVD extras, and no more movies. In this circumstance, the extended family of regular collaborators on 36 Views gathered around the director and the script outline to bring it (his shortest film) to completion, using all the resources of play and improvisation that are so evident in the final result. Knowing that the film is, in this sense, a homage to a literally vanishing auteur brings an exceptional poignancy to its viewing.

Guillaume Depardieu, who suicided not long after appearing in Don’t Touch the Axe, testified to the joy he felt being around Rivette and his companion on the set: Véronique Manniez (author of a book on Secret défense in 1998), who stayed with him to the end, nursed him in their home, and married him (becoming Manniez-Rivette) – a second marriage for the director, three decades after the end of his collaborative relationship with Marilù Parolini (1931-2012). At long last, love.

In 2010, it was announced that Rivette’s archive – documents, notes and traces pertaining to his career, right back to the interview transcripts with great directors he typed out for Cahiers in the 1950s – had just been donated to France’s Bibliothèque du film. All those thousands of pieces of paper, doubtless as weighty and as weightless as the piece torn apart in the crucial moment of 36 Views of Pic Saint-Loup

An earlier version of this text, composed in 2010, appeared in Claire Perkins & Constantine Verevis (eds), Film Trilogies: New Critical Approaches (Palgrave, 2012).


from Issue 6: Distances


© Adrian Martin, December 2010 / 29 January 2016.
Cannot be reprinted without permission of the author and editors.