Abstract

One of the essential tasks of several research fields, including philology, is to determine whether the matching of two or more elements within a data set can be regarded as evidence of a connection or as a random coincidence. This essay tackles the definition of ‘polygenesis’ (i.e., the accidental creation of identical scribal innovations), which is a crucial concept for both the classification of variant readings and the critical edition of ancient texts. Two main procedures for detecting possibly polygenetic variants are discussed and tested on a selection of Medieval Romance texts. The results of this survey show that several classes of variant readings, which are not commonly recognised as typically accidental, can be classified as polygenetic. In turn, this observation brings us to reconsider another key notion of philology, that is the distinction between ‘substantial’ (i.e., textual) and ‘formal’ (i.e., linguistic) variants.