This paper addresses the problem of the difficulty in establishing stemmatic relationships in the repertoire of the Italian Trecento because of the strong tendency to innovation that characterises most medieval Romance traditions, on the one hand, and of the tendency to readily correcting evident copying mistakes that tends to characterise musical traditions, on the other.Various solutions to the problem are proposed from trying to recognise variants that, while not evident errors, can however give indications of directionality in the copying process, to devising methods for “triangulating” the distribution of nondirectional variants with other streams of evidence, in order to construct at least partial stemma, or nondirectional maps of the placing of the reciprocal position of the witnesses in the history of the transmission of the repertoire (the so-called phylogenetic trees).
Giacomo Ferraris (Mon,) studied this question.