Georges Menager/Paris Match Archive/Getty Images

Robert Bresson and the donkey that starred in Au hazard Balthazar, 1966


When asked why he had made so few films—thirteen features over a period of forty years—Robert Bresson invariably answered that it was hard to get funding for the sort of work he wanted to do. “Money,” he memorably said, “likes to know everything in advance.” Bresson, by contrast, wanted to know almost nothing in advance. He was drawn to the idea of “wonderful chances, those that act with precision.” “The things we bring off by chance—what power they have!” He wanted to surprise reality in its lair, and was willing to wait until reality wasn’t posing too carefully. We can see why the sources of money might be worried; and we may enjoy the wry joke lurking in the title of Bresson’s last film, released in 1983: L’Argent.

Bresson was born in 1901 and died in 1999. He made two early films with actors—Les Anges du Péché (1943) and Les Dames du Bois de Boulogne (1945)—and after that renounced the use of professionals altogether—although some of his nonactors did go on to have acting careers. Pauline Kael was not alone in thinking that Bresson’s next film, Journal d’un curé de campagne (1951), was “masterly,” and his reputation as the hero of a new French cinema was secured with Un condamné à mort s’est échappé (1956) and Pickpocket (1959). These last two are among Bresson’s most accomplished works by any standards, and although amazing films were still to come, there are question marks hanging over all of them. They include Procès de Jeanne d’Arc (1962), Au hasard Balthazar (1966), Mouchette (1967), and Lancelot du Lac (1974).

Bresson himself was the first to raise questions about what sort of films he made. Of Au hasard Balthazar, currently ranked sixteenth in Sight and Sound’s list of greatest films of all time, he said it “had some happy moments along with some imperfections.” He meant moments that worked well for him as pieces of film; we wouldn’t look to him for any happiness in his characters, and least of all in the central figure of this film, an unfortunate donkey whose life we follow from burden to burden and from beating to beating. A happy moment would occur for Bresson when a fragment of disregarded life was perfectly caught by the camera, when he succeeded in following his own instruction: “Make the objects look as if they want to be there.”

“Images and sounds like people who make acquaintance on a journey,” he wrote in Notes on the Cinematograph, “and afterwards cannot separate. In an early interview, from 1943, he spoke of the advantages of “une maladresse supérieure” in making films, an awkwardness that is better than skill. For him a film “must be raw,” and no sooner has he described his Joan of Arc film as “possible and realistic” than he throws in the opposing terms as well: “or…impossible and unrealistic.”

“There are horses, knights in armor, a tournament,” he says of Lancelot du Lac, “all as anachronistic as possible.” “I wanted there to be constant anachronism.” There is, but of a kind that reverses our usual sense of the term. Bresson’s stately, near-parodic Arthurian images do not mistake or confuse times, they measure time itself, tell us how late we are and how little we have changed. The magnificent final sequence of the film shows the aftermath of a battle, the end of a world. Riderless horses race through a forest; the camera pauses on a heap of dead knights that looks at first sight like a mound of scrap metal. A wounded Lancelot staggers clanking into the frame, and falls into the heap to die.

This is what it all came to, epic, religion, romance, chivalry. “It’s probable,” Bresson remarked in an interview, “that our mystical knights killed, pillaged, and burned without hesitation.” He also said that “the temptation of modern life was constantly with me” as he was making the film, and when Guinevere tells Lancelot that “God is not an object that one brings back,” she can be heard as commenting not only on her lover’s failure to find the Holy Grail but also on a great many other misconceived quests, medieval and modern.

These thoughts and images recall Susan Sontag’s astute comment, made ten years earlier than this film: “He does not intend his characters to be implausible, I’m sure, but he does, I think, intend them to be opaque.” By the time of Lancelot Bresson seems to be going for the implausible as well, because he believes it may catch the truth that mere plausibility only simulates or evades. Either way the characters remain opaque—the perfect word—and the effect as applied to a whole film was well caught by Marguerite Duras, when she said of Au hasard Balthazar that she both was “overwhelmed” by the work and at the same time couldn’t say “exactly what it was that I was watching.” She is in good company since Jean-Luc Godard, Louis Malle, and François Reichenbach, in the same interview session, all assert eloquently that Bresson is entirely different from any other director who ever lived but fail to tell us why.


Bresson on Bresson collects interviews given over forty years, includes discussions of all of the director’s films, and has sections on adaptation, on sound, and on Bresson’s book of aphorisms, Notes on the Cinematograph (1975). The works that get the most attention are Procès de Jeanne d’Arc and Au hasard Balthazar.

From the moment of its publication, Notes on the Cinematograph became indispensable to all kinds of people, of very different persuasions, who wanted to think about film as an art. It’s not a handbook or a guide; it’s too fragmentary and tendentious for that. It’s more like a version of Pascal’s Pensées directed at the cinema. Bresson is writing half for himself and half for us, keeping his thoughts on his own narrow way and letting us see his discipline at work.

Most of the thoughts are dated 1950–1958, and are followed by a short set of “further notes” dated 1960–1974. In an interview given in the year of the book’s publication he said he thought he might at some point write “a follow-up…, maybe in a less laconic form,” but it’s hard to imagine a Bresson who isn’t laconic. Indeed, the interviews collected in Bresson on Bresson often quote or anticipate the notes.

In both cases he likes to borrow phrases, let them do the talking. In Notes alone, for example, he cites model sentences or images recorded by Baudelaire, Cézanne, Chateaubriand, Corot, Dante, Da Vinci, Debussy, Sévigné, Dostoevsky, El Greco, Milton, Montaigne, Montesquiou, Mozart, Pascal, Proust, Purcell, Racine, Auguste Renoir, Vivaldi, and a French general signaled only by the initial M.

The effect is to make us feel that Chateaubriand or Mozart or Pascal already had Bresson in mind, or indeed that he has invented them. When Chateaubriand remarks of certain poets that they don’t lack naturalness, they lack nature, he has briefly announced almost everything Bresson has to say about practiced performers on stage and screen: they can act naturally but it is only an act. Bresson notes a similar slippage between what’s truthful (véridique) and what’s true (vrai). When Mozart says of certain works of his that “they are brilliant…, but they lack poverty,” he is close to the heart of Bresson’s aesthetics. And when Pascal offers the following account of a dual perception, he is distinguishing between what Bresson sees as the blandness of the conventional cinema and the unruly precision of the cinematograph: “A town or countryside at a distance is a town, a countryside; but as one approaches, those are houses, trees, tiles, leaves, grasses, ants, ants’ legs, to infinity.”

One of the few sustained arguments in Notes, reworded on almost every page, concerns this distinction. Bresson wants us to think of the cinematograph as an alternative to what he calls the cinema, that is, essentially filmed theater. He likes the etymology of the second half of the first word, and constantly refers to film as a form of writing. “Cinematography [le cinématographe] is a writing with images in movement and with sounds.” Historically, the name of the Lumière brothers’ machine (le cinématographe) was shortened to the name of an art form (le cinéma), and Bresson is asking the long version to name the unsquandered, noncommercial working of the art. The gesture produces some very odd effects in his prose and his speech, and the translators of both books are probably right to smooth them out, most of the time rendering cinématographe as cinematography, and occasionally producing a person, the cinematographer.

Sometimes it is clear that the machine has become an art and couldn’t be a person: “Un acteur est dans le cinématographe comme dans un pays étranger.” (An actor in cinematography might as well be in a foreign country.) “Toute une critique ne faisant pas de démarcation entre cinéma et cinématographe.” (A whole gaggle of critics making no distinction between cinema and cinematography.) At other times a person seems a plausible option, although the balance of Bresson’s logic is lost: “Le cinéma puise dans un fonds commun. Le cinématographe fait un voyage de découverte.” (Cinema draws on a common fund. The cinematographer is making a voyage of discovery.) We do see, though, that Bresson is trying to teach us, through his sentences, what the new meaning of an old word is, and why he needs it: the cinematograph is what the cinema should be.



What we might call the Duras effect has many versions among viewers of Bresson’s work, and two of them seem especially worth pausing over. The first is that the films as we see them and as we remember them are often astonishingly different from each other; they appear almost unrelated. My experience of watching Mouchette, for example, was one of unrelieved gloom, because I felt Bresson’s only achievement was to have perfectly conveyed the dingy horror of his fourteen-year-old main character’s existence. By the next day the film had changed shape in my mind. It was full of varied questions about the world that had seemed monotonous, and I kept thinking of the heroine’s sullen courage—she was a heroine for me now, and not just a sufferer.

Mouchette, based on Georges Bernanos’s novel Nouvelle histoire de Mouchette, is set in rural France, the Vaucluse to be precise. Rural, not rustic: Bresson makes sure the loud noises of passing trucks keep invading the village interiors. Mouchette’s mother is dying (and dies) and, apart from the frame where she poses a brief question at the beginning of the film (“What will they become without me?”), is never seen out of her bed. There is a small, often crying, baby to look after, Mouchette’s father treats his daughter as a defenseless creature he can push around, and her elder brother is doing what he can to turn into a copy of his father. Her schoolmates despise her because she is poor and lonely, and she has only one flickering moment of happiness, when she gets to ride in a bumper car at a fair.

Her father soon puts a stop to this and sends her home. The main action of the film concerns Mouchette’s accidentally witnessing what seems to be a murder in the woods: the local tramp tries to kill the local gamekeeper, and as far as the tramp and Mouchette and the audience know at that point, he succeeds. Mouchette likes the tramp and agrees to say nothing. His response to his own act and her kindness is first to get drunk, then to have an epileptic fit, then to rape her. She doesn’t recover from this violence—even after she has managed to get some sleep she is still weeping—and kills herself by rolling into the river. The gamekeeper didn’t die.

Mouchette shows us,” Bresson says, “the cruelty and misery we ordinarily choose not to see.” But he also says, “I was afraid I’d make something atrocious” (“Je craignais l’affreux”). The context of the remark—he says the same thing in another interview too—suggests that he thought the bleakness of the world represented in the Bernanos novel might be too much for any film to show. The answer to his own question (“Could I make Mouchette tolerable without making her sweet?”) is yes, but only just. Or yes, but not while we’re watching the film.

In memory, though, certain curious suggestions keep coming back. There is the contrast between Mouchette’s unworried laughter at the fair and her usual grimace of anger. There are the repeated scenes of her lurking across the road from her school so that she can throw stones and mud at her unfriendly class. There is the extraordinary dignity with which she refuses, after the rape, to answer the questions of one of the three women—the “terrible” women, Bresson says—who are pretending to be nice to her in order to get all the fodder they can for gossip. Well, she does more than refuse. Asked to confirm that the tramp made her drunk before assaulting her, she draws herself up proudly and says, “Monsieur Arsène is my lover.” She will not be a victim in the eyes of the woman who is asking. It’s less clear how she sees herself.

Bresson says that her suicide is not the result of despair, and that he is following Bernanos in this. “For her, death isn’t an end…it’s a beginning.” However, this is Bresson’s religion talking (“I believe in the soul, in God”) and not the film. What we see are Mouchette’s three attempts to die. In the novel she “slides” into the water, and here she wraps herself in some clothes she has been given and rolls down a short slope only to be interrupted by a passing tractor and then stopped by a bunch of reeds. On the third try she makes it, but we don’t see her enter the water, only the disturbed water she has entered. “This vanishing: that’s death,” Bresson says.

Of course there is nothing in this ending to contradict the idea of an afterlife if we wish to believe in one. But what I most remember is Mouchette’s determination to accomplish the action she has settled on. In the same way that she converts a rape into an affair, however drab and implausible, she also takes firm charge of her death. No one pushes her into it, and she looks, as she repeats her descent, oddly like a child playing a game.

The other version of the Duras effect that I’d like to pause over is this: descriptions of Bresson’s films, even in his own words, make them sound hopelessly abstract and thesis-ridden, whereas they are exactly the reverse, intensely concrete and saying (as distinct from showing) almost nothing. The perfect case here is Au hasard Balthazar, a great film that sounds, when anyone talks about it, like the worst film ever made. Would you rush to see a film its director describes as involving “a character that might resemble Charlot in Chaplin’s early films, but that is also an animal, a donkey, present in all its purity, its tranquility, its serenity, its sanctity”? I wonder what the other donkeys thought when they saw the movie.

Bresson, so averse to the idea of symbolism in his Joan of Arc film (“There is no symbolism there. I don’t like symbolism”), talks of almost nothing else in relation to Balthazar. The name is that of one of the Magi—the complete rhyming title is the motto of an old French family, and registers, of course, Bresson’s own devotion to chance—and Bresson says he was thinking of the donkeys in “both Testaments.” Even when he says he “wouldn’t explain” the film, he explains at once: “There is a donkey representing candor, simplicity, acceptance, and then there are people representing pride, greed, cruelty, etc.—our vices.”

The film has two storylines. A donkey is born, becomes a children’s pet, passes from master to master, one of them played by Jean-Claude Guilbert, who is the tramp in Mouchette, and each one more unpleasant than the last. The animal’s life includes a spell in a circus, where it is taught to count, and a spectacular death on a hillside among a crowd of sheep. Jean-Luc Godard asked Bresson about this ending, suggesting that it seemed “more improvised” than other moments in the film. “Maybe in the beginning you imagined only three or four sheep?” Bresson says, “You’re right about the improvisation, but not about the number. In fact, I had imagined three or four thousand sheep. I just didn’t have them.”

The other storyline involves Marie, “a lost little girl,” as Bresson calls her. “Or rather, the girl who loses herself.” We see her with her childhood friends, the loyal rejected suitor and the glamorous delinquent she seems to fancy. Later, in desperation, she gives herself to an old man in return for food and shelter. She resembles Mouchette in certain ways, although she is older and not so poor, her father is austere rather than hostile, and she is both more charming and more passive. Toward the end of the film she is given the marvelous line “J’ai toujours voulu m’enfuir,” “I always wanted to run away.” That past tense has a terrible wishfulness about it. Marie does finally run away, but only after being raped and beaten by the delinquent beau and his friends, and no one knows where she has gone. She has, it seems, no more agency than Balthazar, who has never been able to do anything about his buffeted life, his fate as the imprisoned witness of human vice.

I don’t think the film escapes interpretation, only symbolism of the obvious, too easily evoked sort. We are given shot after shot of Balthazar’s eyes, and it is impossible not to make something of these images. I see patience and a mild surprise, an implied question rather like the thought that Elizabeth Costello, in J.M. Coetzee’s novel of that name, attributes to Kafka’s ape: “Why do men behave like this?” The most intimately troubling sequence in the film, perhaps, occurs when Balthazar meets at the circus, one after another, in isolated takes, a tiger, a polar bear, a chimpanzee, and an elephant. Each time the camera cuts from the donkey to the other animal and back, repeating the sequence for the tiger and the bear, but not for the chimp. By the time we get to the elephant there is just a cold scary eye followed by Balthazar’s kindly but worried gaze. The human creature watching the film feels excluded from the visual dialogue, too far away, although he or she can hardly fail to register the sense of a threat to Balthazar, as if the mere difference of these species were a form of enmity.

When Balthazar dies among the sheep, biblical allegories assault the mind from all sides. And then fade away. Thoughts of the bewildered crowd and the lonely self remain, but it’s not clear how far we can take them. The sheep obviously don’t know what it’s like to be a donkey. Balthazar would probably rather have been alone in his dying, but since when did he have a choice about such a matter, or any matter? At least he wasn’t surrounded by the three or four thousand sheep Bresson wanted.