Before Fiction. Nicholas D. Paige

slip in standards among authors of his day. If Lafayette’s book is the only one he incriminates, it is also the only one he can incriminate: Lafayette’s contemporaries all adhere at least in spirit to the Aristotelian rigor set forth by Castelvetro one hundred years earlier, as a quick look at the three works commonly seen as predecessors of La Princesse de Clèves confirms: Lafayette’s very own La Princesse de Montpensier, routinely credited with hastening the demise of the voluminous French heroic romances; Saint-Réal’s Dom Carlos (1672), which coined the term nouvelle historique in its subtitle; and Villedieu’s triptych Les Désordres de l’amour (1675). All these books moved carefully within the blanks of a historical record that was easily available through works of vulgarized history—Davila’s Histoire des guerres civiles (translated from the Italian in 1657), the 1668 abridgement of Mézeray’s massive Histoire de France (orig. 1643–51), the publication of the memoirs of Castelnau (1659) and Brantôme (1665–66), and even something like Le Grand Alcandre (1651), a transparently keyed and much-reprinted account of Henri IV’s dalliances with his mistresses.⁹ The blanks involved love, whose entanglements—completely invented or merely amplified with respect to what was known—served to explain or motivate noteworthy past events, thus producing the sensation of reading what Valincour called “secret history” (V 70).¹⁰ It is true that Lafayette and Villedieu, against the stipulations of Castelvetro, permit themselves an invented character, but two factors absolve them of poetic wrongdoing. First, both characters, the comte de Chabannes of Montpensier and Madame de Maugiron of Les Désordres’ third novella, are spurned lovers whose presence adds tension and pathos to the main love plots, which take place between bona-fide historic personages. In this respect, they are close to Sabine or Dircis, otherwise unknown figures who appear for dramatic reasons in, respectively, Corneille’s 1640 retelling of the Horatii legend and his 1659 version of Oedipus Rex. Second, nothing in the historical record permits one to conclude that the characters could not have existed: their family names were real enough; it was just that available sources could neither affirm nor deny the existence of the individuals themselves. In a famous remark in his 1614 History of the World, Raleigh had asserted that “filling up the blanks of old histories” in this manner need not be cause for concern: “For it is not to be feared that time should run backward, and by restoring the things themselves to knowledge, make our conjectures appear ridiculous.”¹¹

      Even if we can no longer quite feel the discomfort Valincour felt on seeing an invented woman pushed to the forefront of La Princesse de Clèves, we understand the reasons for his objection. Certainly, invented characters were not unheard-of in novels, but, as in the above examples, the invention almost always remained historically plausible and was typically confined to secondary characters.¹² Moreover, how would readers know that the future princesse de Clèves, Mademoiselle de Chartres, and her mother to boot, never existed? What made them “visibly false,” in contrast to characters like Chabannes and Maugiron? We need not suppose perfect mastery of Renaissance genealogy by Lafayette’s contemporaries: all they had to do was to reach for the volumes of the historian Mézeray and the memorialist Brantôme, which were not the dusty tomes they are now but best-sellers devoured by an aristocratic public fascinated by their glorious ancestors. If Valincour’s confrontation of the novel with the historical record strikes us as odd—who could imagine indicting Balzac for having invented Goriot?—in these years such a confrontation made much more sense: in a literary culture in which literary heroes were real almost by definition, the first thing readers of La Princesse de Clèves would do would be to ask for more information on the heroine.¹³

      And it must be said that Lafayette fairly solicits the reader’s curiosity: Valincour’s recourse to histories of the French Renaissance is incited by the book itself. By its detail, of course, familiar to any reader of the novel’s dense opening pages. But also by what, in the midst of such exacting descriptions of the court, of the legendary affairs, of the tournaments, was so suspiciously out of focus—the origins of Mademoiselle de Chartres. Her dead father is evoked but not named; her family name would have been odd, since the house of Chartres did not exist; and, a final goad, this fatherless young woman of whom no one had heard was supposed to be “one of the greatest heiresses of France.”¹⁴ Mézeray and Brantôme were silent on the subject, but since silence was the accomplice of possibility, what permitted Valincour to declare with such certainty that the novel featured “a Mademoiselle de Chartres who never was” (qui n’a jamais existé au monde; V 62)? The best proof that Chartres did not exist was her marriage to someone who did figure in the histories—to one of the three sons of Jacques de Clèves, duc de Nevers, none of whom ever married anyone of that (odd) name. What should we possibly think, Valincour says triumphantly, of “a duc de Clèves, who marries [Mademoiselle de Chartres] even though he was never married” (V 62–63)? It turns out that Valincour is wrong in the details (he confuses the de Clèves brothers), but the contradiction is indeed there, and it gives the critic all he needs to declare the Princess factually implausible, false.¹⁵

      Smelling blood, perhaps, Valincour manages to locate one more contradiction. Reading in Lafayette of the tournament in which Nemours competes under the colors yellow and black in homage to the Princess, he turns to the pages where Brantôme describes the same event. The detail, astonishingly, was true: Brantôme notes that the real Nemours, Jacques de Savoie, wore the colors in homage to an unnamed woman. This would seem be the perfect example of a historical blank, since the Princess could plausibly be just that woman about whom the historian is silent. Unfortunately, history isn’t silent enough, for Brantôme adds that Nemours chose the colors of a woman whose sexual favors he was then enjoying.¹⁶ Lafayette’s maintaining of the reference to yellow and black, Valincour reasons, is tantamount to identifying her virtuous heroine with a courtesan. The author would have done better to give alternative colors so as not to raise such a specter in the minds of readers—some contradictions, we can assume, are better than others.¹⁷

      Charnes’s response to this detective work, and to the entire Aristotelian thrust of Valincour’s argument, is that contradictions mean nothing: since the Princess is “an invented person,” Charnes remarks with respect to the colors of Brantôme’s courtesan, “I cannot fathom how one can apply what was said about the historical lady to what the novel’s author tells us about the imaginary lady” (C 84). It is on account of this that François Weil, Charnes’s modern editor, detects in the Conversations the seed of a post-Aristotelian line of thought that will eventually lead to modern fictionality and the novel as we know it. While it is indeed tempting to construe Charnes’s remark in this way, his thought on the issue is a good deal less “coherent” (C xxiii) than Weil alleges, and deserves some pressure. Weil’s main proof rests on a passage in which Charnes says that novels of his day, which he calls “gallant histories,” represent a “third species” of “fiction,” a word he uses in the general sense of poetry. Neither “those pure fictions in which the imagination allows itself complete freedom, without regard for truth,” nor “those in which the author takes a historical subject in order to embellish and improve it with his inventions,” La Princesse de Clèves belongs to a different category entirely: here, “one invents a subject, or takes a subject that is not universally known, and one decorates it [l’orne] with historical details that increase its plausibility [vraisemblance] and heighten the reader’s curiosity and attention” (C 129–30).

      Charnes’s argument, however, is problematic on at least three counts. First, the difference between his third species—the one that holds a tantalizing resemblance to modern fictionality—and fiction’s second species seems close to a quantitative argument: the novel, Charnes suggests, is simply less historical than history.¹⁸ Second, and more serious, both species resemble the poetic composition Valincour himself prescribes: contemporary authors, Charnes writes, try to produce such “faithful copies of true history” that readers take them “for history itself”; they concentrate on the actions of “private persons” recounted in a way in which the reader often understands those actions “as the secret motivation [ressorts] behind the memorable events History has taught us” (C 135–36). The difference between species two and three is that the former treats the great actors of history, whereas the latter treats unknown people whose lives