This chapter approaches the roman-nouvelle distinction by treating these artifacts as Wittgensteinian “families”: what makes a roman a roman (or a nouvelle a nouvelle) is not one particular shared feature but a pattern of overlapping similarities. Using the criteria of length, linearity, and truth posture to establish a series of different scenarios for determining what might be considered a roman (or a nouvelle), the transition between the two forms can be clearly mapped. Against the commonplaces of traditional French literary history, it reveals that the roman was in decline before the arrival of the nouvelle. It also shows that the roman was in no way a traditional or archaic form, but one with its own relatively short history.