An understanding of the definitive characteristics of the fabliau genre began with the great manuscript collections. Some of these combined copies of the Fables of Marie de France (e.g., MS Paris B.N. fr. 1593), and of Le Chastoiement d'un Père à son Fils (e.g., MS Paris B.N. fr. 19152) with a collection of short, humorous, generally narrative pieces in French octosyllabic couplets circulating independently of any synoptic framework. These independent pieces were also assembled in extensive miscellanies which might contain forty (MS Berne 354) or as many as fifty-eight (MS Paris B.N. fr. 837) fabliau texts. Some perception of fabliaux as a distinct genre with its own characteristic features clearly influenced this process of selection and assemblage, but the theoretical principles underlying this procedure are not articulated. In practice, as subsequent scholarship has demonstrated, it led to the tentative establishment of a corpus which has a group of identifiably related texts at its nexus, but extends through a cortex of heterogeneous materials with diminishingly recognisable association to the core group.

The earliest printed editions, those of Barbazan, Barbazan and Méon and Legrand d'Aussy, preserved the mix of fable anthology and fabliau miscellany established through the manuscript tradition. This policy was subjected to radical revision in the Recueil Général et Complet des Fabliaux des XIIIe et XIVe Siècles, a landmark edition which initiated modern fabliau scholarship. While aiming to provide critical editions of all known fabliau texts, the editors omitted fable anthologies, presumably because the majority of the stories which they contained were beast fables with animal protagonists, and thus clearly distinguishable from fabliaux.