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.