But what is it, exactly? Certainly not, as we might imagine or hope, some ur-collection of pensées and reflections precisely illuminating the sources, pulse, and articulation of his novels. The Carnets de travail are haphazard and chaotic, accessible only with the help of considerable critical apparatus; they are incomplete, and those parts which survive are numbered in a batty, non-sequential manner (by a “berserk librarian,” Biasi suggests). The most obviously readable sections of Caroline's legacy to the Musée Carnavalet, the travel journals, have long since been hived off and separately published. What we have left is the rump, the brute beginning thoughts, the stutterings and jottings, the false early certainties of writing. Biasi, without trying to force categorization on to the eighteen notebooks he here transcribes, divides them into carnet and calepin, into those containing research for particular novels and the more general grand carnet d'idées or grand carnet de projets; but subjects criss-cross from one notebook to another, and working problems raised on one page may suddenly be treated as if long solved on the next. Writers don't keep notebooks with a view to making things easy for their subsequent editors; they jot on the run, use shorthand, know what they mean when at their most cryptic, cross things out, have second, third, fourth thoughts. Flaubert, moreover, was a writer who never proceeded with bland orderliness from one project to another; his books were long pondered, and ideas relating to any one of several novels may occur side by side. For instance, the main body of notes for Bouvard et Pécuchet that have survived are to be found in Carnets i%bis, 18, 11, and 6 (covering the period 1874–9: it's a sign of how chaotic the numbering system is that the lower numbers refer to later years). However, his preparatory reading-list for the novel (at least 1,500 books, and probably many more, according to Biasi) occurs back in Carnet 15 (1869–74), while the first references to Bouvard et Pécuchet—originally Les deux cloportes—are found in Carnet 19 (1862–3). The Carnets de travail are both a palimpsest and a cat's cradle.
Writing to George Sand in 1873, Flaubert describes how he is “ruining himself” with his book-purchases for the background reading to Bouvard et Pécuchet, and humorously complains about a journey he made in search of a particular piece of countryside for the novel. He goes from Paris to Rambouillet by chemin de fer, Rambouillet to Houdan by calèche, Houdan to Mantes by cabriolet, then “re-chemin de fer” to Rouen. Total expenditure 83 francs— “such is the cost of conscientious literature!” This whole volume is evidence of the cost—in the main not monetary, but measured in time, energy, travel, and lection—of making conscientious literature. It could be submitted to the tax authorities as an example of the hidden but undeductible disbursements of a writer's life.
Flaubert is celebrated as a writer dedicated to research. But what we mean by “research” varies greatly from novelist to novelist. At the simple level of the popular novel there are diligent writers who find out everything they can about a subject—banking, say, or airports or the motor industry—and put this into their work, often in barely assimilable chunks, as proof that they know what they're talking about. More sophisticated novelists pile up research like a compost heap, but then leave it alone, let it sink down, acquire heat, and degrade usefully into fertilizing elements. Thus at the moment of note-taking a novelist may often have no idea how useful his scribble might prove. A good example of this is found in Carnet 19, where Flaubert makes three successive notes concerning women of flexible virtue. The first is about Mlle X., former chambermaid of a young lorette (a lorette being a woman half-way between a grisette and a femme entretenue on the closely calibrated French sexual scale), who set up next door to her former mistress. Gentlemen leaving the lorette would routinely call on the ex-femme-de-chambre who, after performing her exercices de fellation, would enquire humbly, “Was that as good as Madame?” The second story is about an upwardly mobile bar girl in Algeria who celebrates her risen status by equipping herself with some “ancestors”; surrounded by secondhand pictures of fake relatives, she now plays salon hostess and receives officers she formerly pleasured for 10 francs a go. The third, and briefest note, is about a sixteen-year-old girl waiting in a boudoir to lose her virginity. She is served dinner, only eats the confitures, and then falls asleep on top of a pile of erotic engravings. Of the three, the first is a jolly story, complete in itself, but with something falsely neat about it— often the case in laddish anecdote; the second is a colourful social vignette; the third an unfinished moment (where is the man? did she lose her virginity? what happened to her afterwards?) which could be funny or sad or anything in between depending how it is told. The first two might be called “closed” stories, the third “open”; and it is a version of the third story (initially communicated to him by Suzanne Lagier) that Flaubert used in L'Education sentimentale.
“Research,” therefore, isn't something finite and clear-cut which a writer does before “getting down to” the book. It's something wider and vaguer than that, a state of mind, a sort of dreaming which the writer goes in for even though to the outsider the research looks very little like dreaming and probably more like “real work” than anything the writer is subsequently seen to do. It comes in various forms. At its simplest, we see Flaubert posing himself straightforward questions for L'Education sentimentale: where in the environs of Paris is Wetnurse country? And what are the requirements for a funeral among the wealthy classes? The first question is answered quickly, a couple of pages later (Taverny, Saint-Leu, Pontoise, vallée de Montmorency); the second at more delighted length.
Research means jotting down possibly useful names (he likes the sound of Cahours as a place; Tardival and Vaudichon as surnames; relishes a prostitute who calls herself “Crucifix”). It means tramping the countryside until you find the right cliff for Bouvard and Pécuchet to be terrified by (he enlists Maupassant's help on this quest), and the correct plateau stupide on which to site their village. It means reading enormous quantities of books, and even—if we interpret correctly a couple of brief notes—relying on the loathed craft of photography (“which is never what one has actually seen”). It means doing whatever is necessary: when preparing himself to describe the beheading of Iaokanann in Hérodias, he writes to his niece, “I need to have a good look at a head that's recently been cut off.” There isn't any evidence that he did; but he might have recalled those childhood walks with Oncle Parain and the bloodied cobbles.
Research isn't just “finding something good you can use.” Flaubert is grandly dismissive of this approach: “Goncourt,” he writes to George Sand in 1875, “is very happy when he picks up in the street a word he can then shove into a book.” At the time Flaubert makes this complaint about light-fingered literalism he is himself researching Saint Julien L'Hospitalier and ploughing through medieval cynegetic treatises, finding out what animals you hunt with, how you hunt with them, where such beasts come from, and so on. In the course of listing the birds with which Julien might hunt, Flaubert notes the Tartaret or Barbary Falcon: “ Le tartaret, taller and plumper than the peregrine falcon, comes from Barbary.” This he knows, this he has established. But when he comes to write his story, he tries out various other provenances for the tartaret. The drafts show that he had it coming from Norway, from Iceland, from Scandinavia, before he decided on un grand tartaret de Scythie. In the end, euphony and association are allowed to win over documentary exactitude.
Realism versus Beauty? Realism or Beauty? Beauty attained through Realism? Flaubertians chase these formulae around their skulls. Flaubert, in his letter scorning Goncourt's skip-hunting type of research, states very clearly that “I regard technical and local details—the precise, historical side of things—as very much of secondary importance.” If Goncourt found satisfaction in picking up a tasty word, Flaubert found it, by contrast, “when I have written a page which avoids assonance and repetition.” This ought to be clear enough, and indeed Flaubert's denials of realistic intent are frequent: he sought only beauty. But writers' declarations of intent and writers' practices don't al
ways match up. The novelist takes delivery of the researcher's tartaret de Barbarie, turns it around, tries out some Northern icy habitats for the bird, and comes down for tartaret de Scythie. But not tartaret de Pont-l'Evêque, after all. The researcher has given the novelist a body of information from which he can take off and glide and display his feathers; but there are limits to the range of manoeuvres possible. And when Flaubert was attacked for getting things wrong in Salammbô, first by Sainte-Beuve, and second, more famously, by Guillaume Froehner (assistant curator in the Department of Antiquities at the Louvre, who reviewed the novel in the Revue Contemporaine), he did not retreat behind the screen of Beauty. He didn't argue that he had changed things to avoid assonance or increase euphony. Quite the contrary: he defended himself stubbornly, systematically, violently, referring his assailants to all the sources and authorities he had consulted in the course of his researches.
Research for Flaubert is not preliminary but central; not a matter of “checking” but part of the writing process. On location for a bit of countryside he needs in Bouvard et Pécuchet, the novelist sometimes refers automatically to ils: his two protagonists are already wandering around in his research notes, as if preparing for their fictional life to come. And what of the Flaubert who can be found wandering around in these Carnets de travail? What do we learn about him? Biasi makes an audacious but finally convincing claim about the person we encounter here. In the novels, he suggests, we find the writer; in the letters we find the man; here in the Carnets we finally discover that famous hybrid, rare as a Scythian tartaret, l'homme-plume (the pen-man). It is a strange and reclusive bird, whose tail-feathers are ready-trimmed quills, and whose capacious gizzard grinds everything—life, books, whole countrysides—into literature.
Biasi admits that his edition is a risk (though it sounds more respectable and Pascalian in French: “Cette édition est un pari”). The Carnets aren't a novel, nor can they be read like an intimate journal; what they amount to is a catalogue of the writer's thought in its roughest, most nascent form. There is no doubt that it makes difficult reading, as we dodge in and out of text and commentary and try to remember the various editorial signs for such categories as “words added but subsequently crossed out.” It does present us with a “third” Flaubert alongside the writer of the novels and the man of the letters, but this Number Three is only approachable after a thorough acquaintance with Numbers One and Two. Thus there are many points at which Flaubert is roughing out ideas and phrases not just for his novels but also for his Correspondance. For instance, his epigram “Honours dishonour, titles degrade, office-holding ossifies” (which he was clearly proud of, sending it to three different correspondents in the winter of 1878–9) is found here in its earliest form:
[Le grade dégrade]
La Fonction [abbêtit] bêtifie-
Le Titre déshonore.
Those familar with Flauberts One and Two will not find the Carnets full of new ideas, though there are some unfamiliar formulations. Flaubert is always savage and sound on the fallacy of Progress, but I hadn't before come across this theory from Carnet 15: “The dogma of Progress is a reaction to the dogma of the Fall.” Also new was his epigram on Paris, which he tries out in two formulations: “No longer being able to love Paris is a sign of decadence; no longer being able to do without it is a sign of stupidity.” Was Flaubert reworking Dr. Johnson?
But despite the occasional in-character surprises—appreciation for Joseph Wright of Derby, a paragraph on gloves, an account of a busy forty-first birthday, some treasured lubricities—the Carnets de travail are almost entirely what they sound like: working notes, often of the smallest and most circumstantial detail. (It comes, for instance, as a severe shock when in the middle of notes on the marital state in L'Education sentimentale, Flaubert suddenly jots down a Whole Theme: “Show how, since 1830, Sentimentalism has followed Politics and reproduced all its phases.”) If this is the case, is the book of the narrowest interest, best left to close-text scholars needing to demonstrate the genesis of some particular phrase or incident in the novels; or is there any wider appeal? What of Biasi's wager?
It is a very long shot, but it does, surprisingly, come off. One of the many examples of ridiculous, patronizing, and pretentious observations that Flaubert gleefully collected for the “Copie” of Bouvard et Pécuchet comes from “Marcellus,” or Louis-Marie-Auguste de Martin du Tyrac, a deputy under the Restoration, who in 1825 published Conseils d'un ami à un jeune homme studieux. This work includes the following advice: “As for the arts, it's a good idea to be familiar with them, to like them, even to cultivate them up to a certain point.” How ironically that quote reads in among these Carnets de travail, whose page-by-page credo is that you do not cultivate art up to a certain point but way above and beyond that point; you live it, breathe it, inhabit it—the man and the pen become one, superglued together by a hyphen. And the process of research and writing, that back-to-back activity, comes to its closest fraternity, its most intense pitch, in Bouvard et Pécuchet, the ultimate work of l'homme-plume, his “indescribable posthumous novel,” as Henry James termed it.
In Carnet 2 (1859–78) Flaubert makes a note about Jean Magnon, a friend of Molière, who proposed to write a work in ten volumes—each consisting of 20,000 lines of verse—which would sum up human knowledge, and be so well devised and wrought that libraries would thereafter become mere useless ornaments. Biasi gives no suggested date for this entry, no indication of its possible use, and no annotation; but it seems plausible to suggest that Flaubert's interest in Magnon's project (which came to nothing, the writer being murdered on the Pont-Neuf in 1662) lay not far away in cerebral circuitry from the idea that turned into Bouvard et Pécuchet.
He describes this final project of his life in a letter to Mme Brainne in 1872, when he is deep in the study of medical and educational theory, and reckons he has two or three more years of research to go: “All this for the sole purpose of spitting out on my contemporaries the disgust they inspire in me. I shall finally proclaim my way of thinking, exhale my resentment, vomit my hatred, expectorate my bile, ejaculate my anger, sluice out my indignation …”
As a motive for a book, this doesn't sound good enough (though simply from the way Flaubert gets carried away with “exhale … vomit … expectorate …” we can imagine him laying it on for effect—just as in his letters about Madame Bovary he is contemptuous of Emma to a degree which, if let into the novel, would have been severely destructive). But it does indicate the curious fusion of rage and research, of driven encyclopaedianism, which fuels the novel. The two copyists would seek enlightenment and understanding in everything from gardening to history, and would everywhere be fooled and disappointed. Then they would return to their old trade of copying, but even here would be deceived, for their “Copie” (the never-completed Part Two of the book) would be a transcription of what they took to be the wisdom of the world, but which would appear to the reader as further proof of its folly.
It is a great idea for a novel: but could it ever have been a great novel? It sounds like something by Borges—and indeed, Turgenev told Flaubert to keep it swift and funny (“If you make it too heavy, if you make it too scholarly …”). His advice was disregarded; the monstrous project was already under way. There are 4,000 pages of notes and drafts for the book; the background reading-list is immense and immensely various—Comte 's Principes de la philoso-phie positive on the same three-monthly roster as Visca's Du vagi-nisme and Histoire d'un atome de carbone. Notes originally taken for Salammbô, L'Education sentimentale, and La Tentation de Saint-Antoine are annexed and redeployed. Flaubert scours the countryside; he visits a model farm; he asks Maurice Sand for souvenirs agronomiques, especially of mistakes Sand might have made, and the reasons for those mistakes. He is out to prove something— again and again, chapter by chapter.
Nor, with Bouvard et Pécuchet, does he stop researching when he starts writing; the book becomes a rhythmic va-et-vie
nt as each digested subject is processed into a chapter. What becomes clear in a quite unexpected and poignant way as you read the Carnets de travail is how Flaubert in his final novel created an enormous, weighty, complicated machine, which could be operated solely by pedal power. The cockpit only has room for one, the seat is moulded to his rotund form, and there he sits pedalling furiously, sweating profusely, knowing that only he can get the damn thing off the ground. When we read a writer's letters complaining about the tyranny of his work we may sometimes be sceptical (Flaubert often pleads commitment to Emma Bovary as a means of keeping Louise Colet away); but there is something transparently undeceptive and oddly moving as we follow self-inflicted agonies of composition: page after page of seemingly arid notes, lists of books read, idiocies identified and work still to do. By the end of the Carnets the daily reality of l'homme-plume has become thunderously present, and M. Biasi has won his wager.
Flaubert told Mme Roger des Genettes that he must be mad (“fou et triplement phrénétique”) to take on Bouvard et Pécuchet. He also said that his secret plan with Part Two of the novel, the “Copie,” was to stun the reader into madness: “ahurir tellement le lecteur qu'il en devienne fou.” The crazy doggedness of the novelist invites, and demands, an answering crazy doggedness in the reader (sometimes the dupes Bouvard and Pécuchet seem the only sane people around). In 1876, writing his Trois contes at the rate of fifteen or sixteen hours a day, he told Caroline that he feared one day he would simply explode like a shell and the bits of him would be found scattered over his desk. Four years later, strapped into the fearsome machine of Bouvard et Pécuchet, pedalling frantically away, he exploded. Conscientious literature costs money; and sometimes the price is higher still. It cost Flaubert his life.