Before Fiction. Nicholas D. Paige

plausibility.⁵³ A third novel, Méroüée, fils de France (by a certain “H.F.M.”), took as its backdrop Merovingian France. Here, because of the illustriousness of the title character, documentation did exist and was cited in detail in the work’s preface; historical pretensions are more evident than in the previous works, but the mode of invention was not qualitatively different. The same “H.F.M.” also brought out Tachmas, prince de Perse; like a good number of other novels of the period, not to mention Racine’s tragedy Bajazet (1672), Tachmas takes place in the Turkish court of the mid-seventeenth century, and moreover is claimed to be derived from an oral source. Rousseau de La Valette’s Le Comte d’Ulfeld, grand maître de Danemarc, is similarly set in the recent past; the author points out that some people still living in France knew the hero, and claims to derive his story from unpublished memoirs from Denmark. The final work, and the one that looks most like what we would tend to expect from the genre, is Dom Juan d’Autriche; this novel, attributed to Courtin and unfolding during the heated rivalry of Charles V and François I, contains a gallery of distinguished protagonists taken from the usual Renaissance memoirs. The nouvelle historique appellation could apply, then, to quite different works—covering periods from the recent to the remote past, incorporating protagonists of varying degrees of familiarity to readers and documentation of disparate sorts and quantity. The sense of the genre as a continuum, one that made room for historically “hard” and “soft” works, is not changed by widening the corpus to include two additional titles of 1678 that were not explicitly labeled nouvelles historiques, Cotolendi’s Mademoiselle de Tournon and Préchac’s Yolande de Sicile, both brought out by Lafayette’s publisher, Claude Barbin: the former advertises its careful work with Renaissance sources, while the latter’s late-medieval plot remains dreamily vague.

      Given that La Princesse de Clèves was the talk of the town, it is no accident that Valincour chose it and not another of these works to critique. But a look at the competing works from 1678 makes clear that had he wanted to he could not have accused them of the kind of violence to accepted rules of poetic invention he found in Lafayette: all followed Aristotelian precepts. I have alluded to the fact that Villedieu’s Désordres de l’amour gives a prominent place to the invented Maugiron: the invention of “plausible” supporting characters was generally accepted and practiced. Yet nouvelles historiques on the well-documented Renaissance usually did not stretch history even this far. Boursault’s Le Prince de Condé (1675), whose historical frame overlaps partially with the one subsequently chosen by Lafayette, is more typical of novellas focusing on this period: all its characters, primary and secondary, are attested in the historical record; the sexual liaisons that are its main subject mix well-known affairs (say, between the title character and the Maréchale de Saint-André), more covert “dirt” of the sort we have seen in Lafayette’s Montpensier (Boursault furnishes the date Mademoiselle de Limeuil gave birth to Condé’s illegitimate son in the Louvre), and purely hypothetical trysts (between François II and Mademoiselle de Saint-André). Not one of these works on Renaissance France, or the many others like them, invents a heroine, never mind a heroine whose existence is proved impossible by historical sources.⁵⁴

      If other examples from before or after the publication of La Princesse de Clèves would only repeat what we have already seen, one small subcategory of nouvelles historiques merits special consideration. For, like La Princesse de Clèves, a few novels give over their titles and narratives to female characters whose family names ring familiar but whose precise identity no doubt hovered below the threshold of a contemporary reader’s historical awareness. Here too, however, we search in vain for proof of Charnes’s assertion that the historical novella commonly traded in Lafayette’s admixture of fact and visible falsity. The liberties taken in Bremond’s La Princesse de Monferrat (1676) are unsurprising because it takes place in medieval Italy; just as Lafayette herself did in Zayde, Bremond uses familiar family names and the license provided by sketchy sources to create characters whose historical counterparts may or may not have existed. The precise identity of Mademoiselle de Tournon, from Cotolendi’s 1678 eponymous novel, would probably have been a mystery to most readers, even if members of her clan occupied positions of power in the Renaissance. But this particular Mademoiselle de Tournon did exist, and the author tells us in his preface that he retrieved her story from Marguerite de Valois’s memoirs, where, sure enough, she really figures. Boyer’s early La Comtesse de Candale (1672), situated toward the end of the reign of Louis XI (1461–83), is a more promising possible antecedent for Lafayette’s novel: once again, the name of the titular character would have been familiar (the house of Candale still existed), but it is a fair guess that the comtesse de Candale herself was unknown to Boyer’s readership. Any analogy with the princesse de Clèves quickly falls apart, however: Candale had existed, even if little was known about her; and she is used to explain the antipathy—duly noted by professional historians—between Anne de France and the duc d’Orléans. The heroine thus performs the familiar function of plausibly motivating an event in the historical record. Monferrat, Tournon, and Candale are not, after all, the figurative sisters of Clèves.⁵⁵

      The omission of the generic tag nouvelle historique from the title of La Princesse de Clèves in itself means nothing: more than half of seventeenth-century historical novellas contain no such designation. Yet in Lafayette’s case, the omission seems almost willful, so much does her variation on the genre’s admittedly elastic characteristics seem a transgression of its very reason for being—telling us about heroes of the past. French literary history has long consecrated La Princesse de Clèves as the best example of the nouvelle historique. The irony is that it may be about the worst.

      Chapter 2

      Quixote Circa 1670 (Subligny)

      Subligny’s La Fausse Clélie, a modest success in the years Lafayette was at work on La Princesse de Clèves and for some years after, is long forgotten.¹ On first inspection, it might not seem worth resurrecting. Its protagonist, Juliette d’Arviane, is subject to moments of madness during which she thinks she is the heroine of Scudéry’s famed historical romance Clélie, histoire romaine (1654–60). What could be clearer? Like Cervantes taking on chivalric romance, like Cervantes’s emulator Sorel attacking pastoral in Le Berger extravagant (1627), Subligny ridicules—now with a gendered twist—the heroic romances that had refused to leave the literary stage. But his confrontation of airy feminine illusion with the manly prose of everyday life seems doubly redundant. First, because the joke was old: numerous novels and plays besides Le Berger extravagant had ridiculed readers whose tastes were behind the times.² Second, because by 1670, the year La Fausse Clélie was published, even sympathetic observers seemed to know there would be no more Clélies. Hence, the critic Chapelain recognized the symbolism of the death, in 1663, of the author of the 10-volume Cassandre (1642–1650), writing, “romances [romans] … have fallen along with La Calprenède”; and Scudéry herself had already abandoned her old habits with Clélie’s last volume and started to experiment with shorter forms.³ So Subligny merely rehearsed a lesson everyone, even Scudéry, had absorbed: romance was dead. Finally.

      And it would soon have its replacement. “Little histories [petites histoires] have completely destroyed big romances [grands romans],” proclaimed critic and novelist Du Plaisir in 1682; by the following century his observation had hardened into banality.⁴ Such unanimity makes sense in France, whose literary production was particularly suited to proclamations of a revolutionary upheaval: the episodic, multivolume grands romans that ruled the roost from L’Astrée’s first book in 1607 to Clélie’s last in 1660 quite simply ceased to appear, never to return, and the vogue for the historical novella took up the slack. The English data are less dramatic: though England avidly consumed in translation both the French historical romance and the French historical novella, it did not produce enough of either to leave the same telling pattern. Nevertheless, historians of the English novel have located a few Du Plaisirs of their own—in Behn, who distances her “history” Oroonoko