It is no wonder, then, that eighteenth-century women made wide use of epistolary strategies to achieve what privileged-class men could write in overtly authorial forms. For the letter is defined not by its content but by its rhetorical frame; it can embrace virtually any topic and encompass virtually any discursive mode. Epistolarity thus becomes a ready refuge, what Bakhtin calls a “compositional surrogate of the author’s discourse.”⁶ In the novel, this compositional surrogacy, by which the reader might assume an ideological equivalence between the author and the fictional letter-writer, seems best accomplished through an epistolary structure that is dominated by a single voice. When women like Haywood, Grafigny, and Riccoboni chose such a univocal structure rather than the polyphonic form of a Richardson or Laclos that incorporates multiple letter-writers with divergent points of view, they may have been motivated not simply by representational purposes but by narrational one. For example, single-voiced epistolary novels by women were often designed, as both Susan Carrell and Linda Kauffman have shown, to produce and prolong passion, but erotic desire does not account for all women’s uses of the single letter-writing voice. At the same time, the epistolary surrogate remains a fictional figure without the historical status accorded the extradiegetic, authorial voice. An impatience with the surrogacy of epistolarity may explain why Aphra Behn’s first novel, Love Letters between a Nobleman and His Sister (1684—87) makes a progressively smaller use of letters, from fifty-seven in the first volume to nine in the third, so that the novel virtually loses its claim to epistolarity.

On the other hand, I have already suggested that practices of overt authoriality opened women novelists to the same charges of transgression leveled at women writing in other public-voiced genres. As several scholars have shown, such charges often confounded the economic with the sexual and textual: according to Katharine Rogers, “women who aggressively competed with men” in the literary marketplace “were attacked as if they supported themselves by prostitution” (which, she notes wryly, would have been an easier means of sustenance).⁷ As I suggested in Chapter 2, the explicitly female authoriality for which a writer such as Aphra Behn was paid (and paid) in the 1680s had become more elusive by the 1740s and 1750s as notions of propriety, femininity, and the new (gendered) opposition between private and public spheres were emerging and commodification threatened the old literary aristocracies. As John Sitter describes it, in England “practical success and ambition” came to be seen as “sordid, venal, or mean,” cynicism about the marketplace led to a dissociation of quality from profitability, and “ambition and wit” were censured on moral grounds.⁸ In France, where it was not the gentry but the nobility in whose hands literary judgment resided, success for unconnected bourgeois men was linked to a salon culture in which the favors of aristocratic women were said to control literary careers.⁹

In such a climate, which more or less prevailed in the half-century from 1730 to 1780, women writers would be unlikely to adopt the overtly authorial voice of a Tom Jones or a Jacques le fataliste—the kind of narrator Anna Barbauld described in 1804 as one “supposed to know every thing,” who “indulge[s], as Fielding has done, in digressions, and thus deliver[s] sentiments and display[s] knowledge which would not properly belong to any of the characters.”¹⁰ I have not found a novel written in the first half of the century that explicitly represents an authorial narrator as female, let alone as the text’s dominant ideological voice, as Aphra Behn represented her narrators in the 1680s, and I am speculating that this absence of authoriality, like the presence of epistolarity, maintains the illusion of the novel as private discourse and helps to preserve the gendered public/private dichotomy that Dale Spender describes. French women, writing within the class and gender system of the ancien régime, seem to have avoided heterodiegetic narration almost entirely, while those British women who used heterodiegesis generally confined ideological comments to the voices of characters and avoided the extrafictional and extrare-presentational structures that I have identified as constituents of authoriality: substantive prefaces, generalizations in the narrator’s voice, explicit allusions by the narrator to literature or history, direct addresses to a public narratee, and explicit references to the narrating subject or the narrative act. Barbauld herself makes an implicit gender differentiation when she contrasts Fanny Burney’s “dramatic” practices to Henry Fielding’s “display.”¹¹

In those few instances where I have found mid-century women’s novels venturing overt authoriality—if only, as Margaret Doody says, “in short breaths at a time,”¹² what gets inscribed is not only the impulse toward authoriality but the difficulty of sustaining it. For women’s authorial narrators in this period are likely to reauthorize masculinity even in their attempts at assertiveness. Lennox’s The Female Quixote (1752) promises an authorial stance, for example, in a chapter titled “Definition of love and beauty—The necessary qualities of a hero and heroine” (4.3), but the “definition” gets worked out through a polyphony of characters’ voices. And the wise voice that sermonizes through the novel’s penultimate chapter is a male “doctor” so learned that Samuel Johnson has been thought not only his model but his creator (possible irony, given the fact that Johnson used without attribution Lennox’s 1753 Shakespear Illustrated for his own Preface to Shakespeare).¹³ Eliza Haywood’s History of Miss Betsy Thoughtless (1751), written late in her career, does constitute an overtly authorial voice, but from the first sentence of the novel the narrative voice dissociates itself from the class “woman” by deploring “the sex” for vanity, heartlessness, cattiness and lack of self-scrutiny.¹⁴ Such a narrator reinscribes her identity status as masculine in spirit if not in name.

The tight position of female authorial voice in the middle of the eighteenth century is sharply crystallized in the alternatives taken by Sarah Fielding’s The Countess of Dellwyn (1759), which visibly appropriates authoriality, and Marie-Jeanne Riccoboni’s L’Abeille (1761–65), which visibly refuses it. In using Fielding’s novel as a prelude to my exploration of Riccoboni’s anomalous text, I want to open the category of “women novelists” to distinctions beyond gender by associating the different narrative strategies in these works with differences of class—of education, affiliation, relation to audience, and especially material necessity—that in turn result from both national and personal differences. While Fielding’s certainly seems the more authoritative text, I will argue that The Countess of Dellwyn’s overt appropriation of “masculine” practices cannot escape the gender-class system that produces it, while L’Abeille’s overt self-silencing—in which the impossibility of escape is acknowledged—stands as an equivocal rejection of that system without yet challenging its terms. Riccoboni’s narrator’s “failure” of voice, in other words, makes clear what constitutes a “successful” female authoriality.

Riccoboni’s dramatic display of self-silencing punctuates a longer and less evolutionary career. Of her eight original novels, five are epistolary and one is a private memoir addressed to a confidante; the two heterodiegetic novels, the Histoire du Marquis de Cressy (1758) and the brief Hi...