Codex as Corpus : Using TEI to unlock a 14th-century collection of Old French short texts

Paper given at the TEI2022 conference.

Abstract:

Medieval manuscript collections of short texts can act as ready-made corpora, from which data may be drawn to help scholarship understand the circumstances of their production and early readership. As physical items they are of course materially discrete, yet this can obscure their internal diversity, containing stints by various scribes which often overlap asynchronously with texts from various exemplars, themselves from multiple geographical and temporal points of origin. The utility of traditional palaeographical and codicological techniques in examining such codices is well established, but there is also value to be found in the use of quantitative, digital methods that take the whole manuscript as the source of their data.

This paper will discuss the role played by TEI in an ongoing mixed-method study into one such codex-corpus: Bibliothèque nationale de France, fonds français, 24432, a fourteenth-century manuscript containing some 90 short texts written in Old French. The aim of the project has been to display how fruitful it can be to combine traditional and data-driven approaches in the holistic study of individual manuscripts, and its first phase has focussed on aspects of the manuscript’s production, particularly through an examination of scribal hands.

TEI has been critical to this analysis, enabling discoveries about the manuscript which have eluded less technologically enabled generations of scholarship. For example, quantitative analysis of the rates and methods of scribal abbreviation, made possible through the manuscript’s encoding, has supported and refined hypotheses reached through the qualitative analysis of palaeographic features regarding the number and division of hands within the manuscript. In addition to this confirmatory role, the quantitative analysis itself has led to further hypotheses which are less easily reached through qualitative methods, in particular in determining which variation is evidence of differences in scribal preferences, and which may be a continuation of variation found within the exemplar manuscripts.

As with any project of this nature, the process of encoding BnF fr. 24432 in TEI has not been without difficulty, and so this paper will also discuss the ways in which attempts have been made to streamline the process through automation and UI tools. These include the use of XForms in creating an input tool for the bare-bones encoding, and the use of XSLT workflows, both to perform tasks that would be overly time consuming if completed by hand, and also to create bespoke renders of the text encoded in TEI that are useful at a particular moment of the transcription or editing process. We will consider the benefits and drawbacks of creating one’s own tools for such tasks, and the ways in which the tools created for this project may be usefully replicated or adapted by those conducting similar projects.