The inherent flexibility of the digital format has favored the rise of editions that enable access to every witness of a particular textual work. These types of editions might have different goals and seek to answer different research questions, but they usually coincide in drawing attention to the importance of textual variants. To maximize the computational analysis that may be practiced with the variants in different witnesses, a complex taxonomy that reflects the diversity of cases is required.

Many scholars have taken into consideration the recommended TEI method to encode the types of variants – that is, through the attributes cause or type inside the element1 – and most agree in evaluating it as insufficient.2 These attributes are not able to enclose the hierarchy intrinsic to complicated taxonomies or the overlap of classes in an efficient way. However, the TEI Guidelines do offer a valid module for encoding this complex issue: feature structures3. This proposal does not advocate for a controlled vocabulary to categorize types of variants. What it offers instead is a pliable encoding method that allows the editor to include multiple layers of information in each apparatus tagset. As a way to examine the advantages of this method, I will present a practical case in which two "classical" taxonomies (textual: addition, deletion, transposition, and mutation; and substantive against non-substantive types) are combined with a goal-specific multi-layered categorization in a highly efficient way.