A commonplace of French literary history holds that around 1660 an archaic novelistic form called the roman was suddenly replaced by the nouvelle, and that this replacement amounts to the birth of the novel in a modern sense. In this quantitative analysis, I tag of a sample of novels appearing between 1601 and 1730 for a variety of characteristics long said to distinguish romans from nouvelles (length, use of inset narratives, historical setting); I add the further variables of protagonist type (drawn from history or not) and truth posture (assertions of veracity and admissions of invention). Such analysis reveals that although romans do predominate in the first half of the century while nouvelles flourish in the second, 1660 cannot be confirmed as a threshold. In fact, far from being diametrically opposed, romans and nouvelles are in many respects merely different moments in the evolution of the same basic artifact, one to be eventually replaced by the first-person forms familiar from the eighteenth century. More broadly, a quantitative approach suggests that the novel’s history should be thought of less as a story of stability and rupture than as continual — but patterned — flux.

The artifacts examined in this and the following chapter are “document” novels: starting in the late seventeenth century, novelists pretend their works were written by the protagonists themselves, chiefly in the form of memoirs or letters. Examining memoirs and other non-epistolary document novels, this chapter shows a clear process of isomorphism, as first-person documents come to resemble each other as they become more popular. The chapter also demonstrates that while the use of the first person was much more abundant in the seventeenth-century novel than typically acknowledged, that first person was not written but voiced (characters told their own stories to others). As such, there is no continuity between the dominant first-person technologies of the seventeenth and eighteenth centuries.

This chapter is the first of three that attempt to empirically measure the commonly attested rupture between the roman and the nouvelle around the year 1660. (This rupture is a version of the frequent opposition in English literary history between romance and novel.) According to Du Plaisir’s 1683 Sentiments sur les lettres, the use of inset narratives would appear to be a defining formal characteristic of the roman. After exploring the haphazard spread of this device from the French translation of Heliodorus’s Aethiopica in the 1540s to the opening of the seventeenth century, the chapter details the rise and fall of various sorts of insetting up to 1750.