We know now how inadequately we have been served by the traditional metaphor for the novel: that it “holds a mirror up to nature.” The metaphor does not fail because there is no nature, no reality out there to mirror. It fails because the novel offers us words, not the direct visual images that a mirror reflects. The reality those words reveal is both there and not there. On the basis of widely shared cues and conventions each reader’s mind must to a large extent project and create that imaginary reality. Thus literature may be both the most abstract and the most personal of the arts.

When Flaubert heard of plans to issue an illustrated edition of one of his novels, a kind of reading aid, he exploded.

Never, as long as I live, shall I allow anyone to illustrate me, because: the most beautiful literary description is eaten up by the most wretched drawing. As soon as a figure [type] is fixed by the pencil, it loses that character of generality, that harmony with a thousand known objects which make the reader say: “I’ve seen that” or “That must be so.” A woman in a drawing looks like one woman, that’s all. The idea is closed, complete, and every sentence becomes useless, whereas a written woman makes one dream of a thousand women. Therefore, since this is a question of aesthetics, I absolutely refuse any kind of illustration.

(To Ernest Duplan, June 12, 1862)

In 1874 Flaubert did finally permit an illustrated edition of Madame Bovary, but to his niece he wrote deprecatingly that the illustrations had as much to do with the book as with the moon. He was not objecting to their poor quality; he felt in his bones the contaminating, paralyzing effect that any particularized image can have on the suggestiveness of the word. An image short-circuits the reader’s imagination and prevents him from conjuring up a character or a scene out of his own associations and fantasies. Flaubert’s outburst insists that the novel mirrors not nature but words—words that in turn evoke a special universe of virtual entities and events. He provides a needed gloss on Conrad’s words in the preface to The Nigger of the “Narcissus”: “My task…is before all to make you see.” A novel presents something deeper than visual images; it assembles laminations and overlappings and dissociations of thought that do not coincide with any simple sensation of sight or sound.

Eisenstein opened his essay, “Dickens, Griffith, and the Film Today” by quoting the first sentence of Dickens’s The Cricket on the Hearth: “The kettle began it!” And four pages later: “Certainly this kettle is a typical Griffith-esque close-up.” Flaubert knew better. A picture—even a movingpicture—of a kettle fixes our attention on its concrete existence and can convey nothing of the swift-running thought in the rest of that brief sentence. No reverse dolly shot or shift of focus or trick editing can translate the inexhaustible abstraction contained in the predicate “began it.”

We also know now that even the most faithful film adaptation of a literary work will not function simply as an illustration—adding images to words as we might set words to music. Any discussion of adaptation will be hampered by the fact that film theory has come in two somewhat tendentious bursts: first, from Soviet pioneer filmmakers who had in mind primarily the silent film (e.g., Dziga Vertov, Pudovkin, Kuleshov, and Eisenstein); and after World War II from non-filmmakers drawn primarily toward linguistics (e.g., Béla Bálazs, Christian Metz, Jurij Lotman, and Peter Wollen). In the most useful book on the subject, De la littérature au cinéma (1970; no English translation) Marie-Claire Ropars-Wuilleumier moves beyond the silent image and beyond linguistic theory to propose a robust symbiosis between film and literature, not a mutual withdrawal.

We would do well to keep three questions unobtrusively in the back of our minds when we think about turning Proust into film. The first is: Has sound film finally broken up the big estates handed out by Lessing to word, as the means of narrative, and to image, as the means of representation? If, as Flaubert implies, the novel is to a significant degree narrative without representation, is a sound film a true synthesis of the two? But why then did Eisenstein write in 1932 (apropos of “inner murmurings” in Dreiser): “The true material for the sound film is, of course, the monologue”? And why did André Bazin write in 1951 (of Bresson’s “pure cinema”): “The screen emptied of images and given back to literature marks the triumph of cinematographic realism”? Are they refuting Flaubert along with many of their own beliefs and proposing that in sound film word will finally subdue image?


Secondly, dealing directly with physical, visible reality, has film developed adequate conventions of acting, cinematography, sets, lighting, and editing to convey abstract ideas, inner states of mind, past and future tenses, and the first person? We would like to think so, but unlike language, film has no established vocabulary and syntax. What does a dissolve “mean”? Passage of time? Memory? Fantasy? Compared to the novelist, the director begins virtually exnihilo. Half a century ago the famous Kuleshov experiment—in which the identical shot of an actor’s neutral face was variously interpreted (according to the way it was edited into a sequence) as expressing hunger, grief, or joy—demonstrated that each film must reinvent the conventions. You cannot be sure whether the cowboy dressed in black is the hero or the villain. There is no dictionary. The film medium probably survives on a widely shared sense of narrative causality, and the sheer fascination of any image that moves. It’s still close to magic.

Third, who is active mentally, a novel reader or a film viewer? Presumably Flaubert would still vote for reading, a process in which we must ourselves transform words into images. Both Eisenstein and Bazin argue strongly for the film viewer and disagree completely about the reasons. Eisenstein insists “montage technique”—cutting and editing—“obliges spectators themselves to create” (“Word and Image”); Bazin argues that depth of focus in the extended shot (planséquence) encourages “a more active mental attitude on the part of the spectator and even his positive contribution to the mise en scène” compared to the way montage manipulates our thought processes (“The Evolution of Film Language”). In the era of television the debate remains wide open and increasingly pertinent.

The challenge of film today lies in part in the number of unanswered questions it will not let us overlook.

Almost all the major masters of the twentieth-century novel before World War II have been translated into film. We have had at least one movie, in some cases several, based on the work of Joyce, Mann, Faulkner, and Kafka. Two authors remain untried: Virginia Woolf and Céline. Marcel Proust has just joined the former group. The promotional machinery of Swann in Love has inflated the background of this production into an epic story about a procession of great directors and scriptwriters who approached Proust’s awesome 3,000-page novel over a period of twenty years. Except for René Clément, the French directors quickly shied away. The non-French (Luchino Visconti, Joseph Losey, Peter Brook) all allowed themselves a lengthy dalliance with Proust, produced some form of script, and then were either unable or unwilling to carry through. The full-length (and probably five-hour) screenplay Harold Pinter wrote for Losey was published in 1977 in a gesture of impatience and appeal. In 1982 when Peter Brook postponed filming his own scenario, Nicole Stéphane, proprietor of the Proust rights since 1962, was immensely relieved to find the German director Volker Schlöndorff ready to take over Brook’s version and begin production without delay.

It is a little hard to understand the feeling of haste and urgency described in these negotiations. In 1984 the original French of A la recherche du temps perdu has, for all intents and purposes, fallen into the public domain. Any director who finds backing can use whatever parts of the novel he chooses. The best way to approach this international version of “Swann in Love” is to regard it as the first in a series of film adaptations of Proust. I hope that financing will now be found for a production of Pinter’s probing version, and that other directors will decide to try their hand.* The handsomely made but somewhat overstated semibiography of Proust’s final years, Céleste (directed in German by Percy Adlon), contributes further evidence that the time for Proust on film has arrived.

Swann in Love has been set before us with fanfares and drum rolls to announce that finally, against fearsome odds, a section of this gigantic modern novel has been filmed. In the glossy brochure given out at the first Paris screenings, Schlöndorff explains that he rejected the idea of using unknown actors because only recognized stars would give the film the aura of “larger than life” that Proust’s book requires. At the same time Schlöndorff has been telling special audiences and interviewers that they should regard his film as an independent work, something not to be measured against Proust. Is this Proust or isn’t it Proust? Schlöndorff wants to have it both ways. The question will have to be answered better than that. Most of the audience won’t know Proust anyway, for, in spite of all the fuss, Remembrance of Things Past is probably the most unread of the great modern novels—even the first volume, Swann’s Way. Therefore I shall begin by describing and commenting on Swann in Love from the point of view of an occasional moviegoer unfamiliar with Proust. Afterward I shall introduce what might be the thoughts of a reasonably literate film critic who has read at least the first volume of the novel. Only then shall I allow the professor and Proust scholar to say his piece.


Moviegoer: The costumes did the most for me. The movie follows Charles Swann, a Paris dandy with very English manners. He gets invited everywhere even though his parents were converted Jews. At a very elegant reception he gives the cold shoulder to a beautiful young duchess who has been panting for him for years, it seems. But he has other things on his mind. For months, maybe for years—it’s hard to tell—he’s been chasing a slippery little kitten called Odette. She’s been around with everyone and knows how to excite Swann. She also makes him jealous of every other male or female who so much as sniffs at her. Swann is always smelling the flower she wears between her breasts. She has her own group of rich friends and they don’t much like Swann. Well, she finally lets him get into bed with her. He likes it from behind. The next day he tells his friend, Charlus, that it’s all over after one night of love. Charlus knows better and asks him when he’s going to marry her. In the last scene, which takes place about ten years later, Swann is dying and pretty well shut out by the snobbish crowd that used to appreciate his cool wit. He and Charlus walk around the park trying to figure out what they have done with their lives.

The film has beautiful shots of Paris and furniture and costumes. But the action seems hard to follow. The film is all cut up and put together fast like a modern film, but it really wants to be oldfashioned. The brothel scene seems out of place. Swann, with most of his clothes on and smoking a cigarette in a holder, takes one of the girls from behind, while he asks her about Odette. I suppose that’s taken from the novel.

Critic: The film is about collecting, collecting things and memories. Schlöndorff and Brook have laid all the clues out nicely, at least at the beginning and the end. The first shot behind the credits—it’s a visual riddle—turns out to be a close-up of the inside of the big secretary in which Swann keeps together in a kind of shrine his most precious and fetishistic possessions: some small Flemish paintings, the leather case that contains his cash reserve, and Odette’s love letters. Very neat. Later he adds to the treasure a flower she has given him. Schlöndorff makes much of Swann being a rich man of exquisite taste who moves around in interiors laid out like a museum. Swann wants to collect Odette too, vulgar as she is. Odette drives him wild, but he does finally collect her by marrying her. Meanwhile he moves in and out of the various social circles from the book.

At the end as old men Swann and Charlus talk about death and the meaning of life. During the scene Odette drives across the park in her carriage looking like the Queen of Paris. Just before the final freeze frame Swann says, “All those old feelings are very precious to me now. It’s like a collection. I can look back over my old loves and I say to myself it would be too bad to leave all that.” Trivial sentiments, but they strike the right note. The film belongs to Swann (though Odette almost pushes him aside), and he comes off looking vapid. Just a collector, a dilettante in everything, and that’s close to what Swann represents in Proust’s novel.

Still, all kinds of things don’t work or are obviously borrowed. Brook apparently had the idea of collapsing the action into one day of Swann’s life, dawn to dawn. Not a bad idea, but the script makes it very difficult to follow the flashbacks unless you already know something about the story. For instance, in the book all the byplay around the cattleya orchid occurs on the night Swann first makes love to Odette. It’s barely fifty pages into “Swann in Love.” Most of the rest of the story—four times as much—is about his jealous pangs over the other men and women who may enjoy the same privilege, past, present, and future. Since the film doesn’t make clear that he’s been her lover for a long time, their one big scene in bed near the end appears to be his long-delayed reward. That’s standing the plot on its head. Of course the movie doesn’t have to follow Proust exactly, but it is confusing.

In Danton (also with a script by Jean-Claude Carrière) and Céleste there are wonderful sequences when the coiffeur comes in to give a ritualized shave. Schlöndorff uses the device here not once but twice, and not so effectively. But he found true masters to do the sets and the costumes. All those chairs and corsets and veils and bric-à-brac everywhere—even a lot of the characters who are played not by actors but by “real French aristocrats”—they all fit the collecting motif. Schlöndorff is something of an ambitious hack. His taste for the perverse worked better in Günter Grass’s The Tin Drum (1979) than it does in Swann in Love.

Professor: I have no quarrel with detaching “Swann in Love” from the rest of Proust in order to make a film of it. These 250 pages form a partially extruded, self-sufficient unit that relates a unified story with beginning, middle, and end. Every critic says all of Proust lies here in embryo and this is so. Since it’s told in the third person, scriptwriter and director don’t face the initial problem of first-person point of view.

The trouble with this version of “Swann” is that someone wanted to salvage the rest of Proust, as if all that wonderful literature just couldn’t go to waste—or wait for another film. The very first sequence shows a man writing in bed. When we hear his words voice-over using the first person to analyze intimate feelings, we have been given a misleading allusion to the writer-narrator Marcel, or even to Proust. In the novel Swann keeps no journal, doesn’t work in bed. Someone has to call a foul on Schlöndorff here. He needs the journal as a device to introduce some exposition of the action, which is already almost over in this truncated script. But the device is a red herring and evokes elements of the novel that are not pertinent to “Swann in Love.”

Too many other items have seeped back into the film from the rest of the book: the boy-poet to whom Charlus makes homosexual advances; the Guermantes clan, whose elevated social status almost displaces the Verdurin “noyau” more central to Swann’s story; the inappropriate brothel scene siphoned in from later stories about Albertine and about Charlus (I heard one earnest discussion at the critics’ screening about whether Schlöndorff intended Swann to engage in sodomy or coitus a tergo); and the last ten minutes of the film that show Swann in failing health a decade later. Brook and Schlöndorff imply classical unity with their twenty-four-hour structure and undermine it with these diversions.

If I had to write a screenplay for “Swann in Love,” I’d try hard to be faithful to three aspects of the novel and allow the rest to fall into place according to the needs of the film medium. First, everyone in the story is afflicted with one or several strains of a hereditary and infectious disease: snobbery. It governs how they are placed in and move among distinct social strata: the demi-monde of a fashionable cocotte; rich bourgeois with chic artistic tastes; and the landed aristocracy guarding their perimeter. Swann, having reached the top layers, succumbs to reverse snobbery and falls victim to a woman of the demi-monde. The bourgeois Verdurin clan suppresses or at least camouflages its desire to rise socially. Proust’s insights into human character and his constant comic perspective spring from his unrelenting scrutiny of snobbery at all levels. The film can’t help catching some of this feeling; Jeremy Irons portrays Swann as a kind of yawning fashion plate. I find him tense when he should be languid, and vice versa.

I would trace some of the trouble to Irons’s heroic effort to learn French for the part. His tutor, who remained with him during the shooting, did wonders. For prints of the film distributed in France, another actor dubbed Irons’s voice in native French. For the version with English subtitles, Irons redubbed his own good but identifiably foreign French. (He apparently thought his fans would expect to recognize his voice.) Most non-French will not notice, but some viewers who know the language will find his longer speeches irritating, even comic. Obviously for some spectators and subtly for others who are unaware of the discrepancy, Irons’s schooled French works against his role and the film. Voice is one of the principal instruments of snobbery, and in the redubbed version Swann is deprived of it. The actor whose carriage and voice best express the snobbery appropriate to his role is Jacques Boudet, who plays the Duc de Guermantes. A big bluff man, the duke remains detached from everything he does and says, and carries his hand like a flabby fin at shoulder height for his male guests to touch. Boudet gives him a powerfully distasteful presence.

The second essential element is the story of Swann’s “love”—not sexual thralldom, quickly surpassed, not romantic love of an unobtainable ideal, but the obsessive stages of an illusion. The ill-fated timing, the perverse afterlife of attachment in the form of jealousy, and the ceaselessly shifting temporal sequences of “successive loves” almost disappear in the twenty-four-hour telescoping of the action. “Not a chronicle but a crisis” is the comment of Jean-Claude Carrière. Precisely: in this capsule version one loses what I would call the Bolero effect of the novel. In Proust, Swann’s feelings go on and on, over and over, with gradual variations and changes in decor and tempo until he finally wakes from his dream. Schlöndorff comes closest to this obsessive quality of the action when he has Swann return alone to his house after watching Odette dress to go to the opera with the Verdurins. He wanders restlessly among the tastefully chosen paraphernalia of his life and recalls in convincing flashbacks Odette’s first, tentative, almost girlish visit to his place. When he shows her the Giotto painting he thinks she resembles, her reply is perfect. “I’m not a museum piece.” The drifting compulsiveness of Swann’s mind comes out far better in this unhurried meditation than in the sequence where unexpectedly hearing Vinteuil’s melody he gasps and leaves in the middle of an elegant concert.

Schlöndorff and his associates have done well by the third essential element: the idolatry of beautiful things. Swann epitomizes a culture of good taste, fashion, decorum, and art. He is a slave to the elevated social milieu to which he has been admitted. Every shot, exterior and interior, contains the outward and visible signs of an aesthetic point of view. For Swann, life must be certified by art; Odette is vulgar and “not his type” until she begins to look like a Giotto.

Yes, this may be a film about its costumes. It also shows that Schlöndorff and Brook have read the rest of Proust, and they try to inject too much of it into this wonderfully self-contained story.

What these three commentaries fail to bring out is the generally plodding quality of the film once you look beyond sets and costumes. Jeremy Irons acts primarily by abstaining from acting except when he gasps to express heightened emotion. A certain degree of stiffness is appropriate to Swann’s character, but Irons creates neither an underlying humanity nor the powerful aura of a dandy who has mastered life by distancing it. Alain Delon does better in infusing a hammedup intensity into the role of Charlus. I had the feeling that the only actor whom Schlöndorff really tried to direct was Ornella Muti. Having lived for several decades with my own half-particularized image of Odette (deeply influenced by a real woman who seemed to incarnate Odette socially and temperamentally), I needed a little time to fit this sultry, dark-complexioned lynx into my mental stereoscope. By evasion and posing and moments of apparent candor Muti finally occupies the psychoerotic space necessary to bewitch Swann. The fact that her beauty and body are far from stunning gives a certain reverberation to the magnificent sentence that closes the story but, unfortunately, not the film. “To think that I wasted years of my life, that I wanted to die, that I had my greatest love for a woman who didn’t really attract me, who wasn’t my type.”

A number of details in the film deserve mention. In spite of dubious statements by Schlöndorff that music in films can express inner states communicated by words in the novel, the best thing about the score is its rarity. The sound track is used effectively and includes chimes to keep us informed about Swann’s tight schedule. But in a production committed to historical accuracy in everything visual, it is jarring to hear music that sounds more like Schoenberg than like César Franck or Fauré. So far as I could discern after seeing and hearing the film twice, Vinteuil’s air that feeds Swann’s love consists of an insistent descending seventh and not much more.

The fifth paragraph of the section opens with a sentence about the protocol of the Verdurins’ dinners. “Evening dress was forbidden because they were among ‘old friends’ and in order to avoid ‘stuffiness.”‘ The one big Verdurin scene in the film presents them having a supper party on the Rue de Rivoli after the opera, and therefore in evening dress. Otherwise, though cruelly truncated, the supper party as filmed catches the raucous, forced wittiness and merciless gossip of the group. To symbolize Madame Verdurin’s authority someone had the idea of giving her a little bell like the one Madame Aubernon employed to regulate the conversation at her literary dinners. The painter Biche (Elstir in his youthful days) tells his risqué story about the latest art exhibit of works painted with God knows what—even caca. But the timing and dynamics with which he animates the story fall short of Proust’s lively narrative.

Quite appropriately Schlöndorff blows up a parenthetical clause in Proust into a scene of fine grotesquerie. Madame Verdurin (well played by Marie-Christine Barrault) dislocates her jaw by laughing too hard; the young Dr. Cottard melodramatically resets it on the spot with a commanding manner and a firm slap; everyone admires her extravagance and his professional skill. The party ends with Swann’s public humiliation when Odette takes Madame Verdurin’s carriage home from the party rather than his. The sequence skillfully fuses action, setting, and character with remarkable fidelity to the novel. Then, walking ahead of his carriage, Swann stages an outburst of anger against Odette and the Verdurins and himself. This episode misses the mark because the script gives him lines that add up to an overly neat résumé of the story—needed somewhere perhaps but inappropriate here.

Unlike many of the poets and painters of his time, Proust was not drawn to the new art of film. He knew enough about the medium to refer to it late in his novel and to reject “cinematographic narrative” as a linear representation of reality devoid of imagination. He did not know that film is equally reliant on a nonlinear, highly plastic montage principle. Many critics (including Paul Goodman in “The Proustian Camera Eye” in 1935) have pointed out that Proust carried on Balzac’s and Dickens’s and Flaubert’s work of discovering novelistic techniques that we now see as cinematic. When I ask students to choose a passage in Proust that lends itself well to film treatment, they complain about the embarras de choix. My own preference favors a matched pair of scenes about Odette (not from “Swann in Love”), the second of which might have been included in this film with appropriate effect.

In the first scene, the still adolescent Marcel watches Odette from the sidelines during her ritual promenade with attendant men through the Bois. Marcel summons up the courage to doff his hat and address to her an exaggeratedly admiring salutation. “People laughed.” Three hundred pages later in the second scene, Marcel has gained entry among her attendants and watches from inside the magic circle a passing grandee wheel his horse and make it rear up in greeting and homage to the notorious courtesan. Swann, now her husband, murmurs in her ear, “Le Prince de Sagan”—the most elegant and widely publicized aristocrat in Paris at the time and also a historical person, not fiction. It’s pure music hall—and superb film.

As Pinter’s screenplay demonstrates, the psychological, temporal, and aesthetic probings in Proust’s novel can inspire superbly inventive recombinations of the film medium. At the risk of incoherence and possibly of boredom, Pinter explored thematic structure, fragmented narrative, the blank screen, and visual rhymes. Proust deserves this boldness. Schlöndorff’s version almost never reaches beyond straightforward editing and following a story line—no dissolves or double exposures, no witty transitions as in Tony Richardson’s Tom Jones, no reaching for some equivalent of Proust’s undulating style. Near the start Swann spins the shaving mirror in which the camera is watching him, and the disorienting effect shatters the visible universe for a few seconds. But the stunning shot signifies nothing, connects with nothing, comes to nothing. After what could have been the last scene of the film, when Swann speaks of wasting years of his life over a woman who was not even his type, we watch two nearly identical, out-of-focus shots of the ladies in the Guermantes salon rising to their feet as if in surprise or outrage. I assume it represents Swann’s imagining what might happen if he married Odette and tried to introduce her into such a snobbish circle. The only other times we enter Swann’s mind take the form of conventional and therefore recognizable flashbacks. This repeated subjective passage seems cryptic and out of place.

Schlöndorff’s film remains plodding in part because it does not lead us back to the key questions—the relations between word and image, the potential of film to depict inner states of mind, and the comparable mental activity of viewer and reader. In Guide Michelin parlance Swann in Love merits only a short detour not a special visit. Moving pictures at their best offer us simultaneously the breathtaking freedom of dream and the convincing particularity of documentary. Proust brings his story to a close with a subtle, semi-comic dream that recapitulates Swann’s delusions and reverse snobbery. Brook and Schlöndorff generally keep their distance from the dream side of Proust. They go heavy on documentary.

Proust does not need illustration any more than Flaubert. Still, now that we have an ambitious and inadequate version of one detachable section of Remembrance of Things Past, the exploration should not cease. Who’s next?

This Issue

August 16, 1984