3 Towards the construction of a stemma

The elaboration of a stemma codicum, representing the filiation between the witnesses that transmit a text whose original is lost, is the core of the genealogical method: on the one hand, only once these relationships have been determined can text restoration be tackled; on the other hand, the stemma may be the goal of the work of synthesising a certain textual tradition. In order to construct a stemma, some preliminary steps are needed; these steps are specifically treated in the sections of the present chapter. The first step of the stemmatic workflow – namely, the identification of both direct and indirect witnesses (technically: heuristics) – is the subject of Gabriel Viehhauser’s contribution (3.1). After sketching a brief history of the concept, he addresses the issue of how the heuristic process is carried out after the material turn in the twentieth century, providing useful information about both the traditional and the more recent tools that researchers have at their disposal. Particularly relevant is the advent of digital catalogues and digital facsimiles, which can offer easier and faster access to primary sources. This development has profound consequences for framing the history of transmission of a text, as shown in the critical review of various Parzival editorial projects based on different heuristic approaches. Caroline Macé (3.2) deals with a frequently neglected aspect of editorial practice: the use of the indirect tradition of a given text (e.g. translations and rewritings, quotations, interpolations, glosses, and marginal notes) for stemmatological purposes. The conclusion reached, namely that “the main point of using indirect witnesses is that their text has been preserved ‘outside’ of the main tradition; they can therefore be used as an ‘outgroup’ [...] to orientate the stemma”, is central from a methodological point of view. The indirect tradition can also be used to document the early history of textual traditions – especially when indirect witnesses are older than the oldest extant direct ones of a given work – as well as the appearance of (hyp)archetypes. Despite their relevance for stemmatic analysis, she warns us to use indirect witnesses with great caution due to the methodological difficulties inherent to them. In her section (3.3), Tara Andrews addresses the problems of transcribing and then comparing (technically: collating) the different instances of a text preserved in several witnesses. In so doing, she presents both non-digital and digital ways of transcribing and collating witnesses, providing also some insights into the current theoretical debate on what these processes and the results they produce mean to different scholars and scholarly communities. She offers a definition of the central notion of a “variant location”, which arises when different witnesses show different readings at a point that can be considered “the same place” in the text. The discovery of these places is key to the establishment of a stemma, as the set of variant locations is the information with which a stemmatic analysis is performed. In a