The TEI msDesc module was deliberately created as a flexible scheme, which would "accommodate the needs of many different classes of encoders" . The potential price of this flexibility is the development of discrete communities of practice among different users of the module resulting in small but significant differences in encoding which complicate interoperability and reuse of metadata. Many such differences can indeed be identified. Not only has the module been relatively widely adopted by digital catalogues of Western medieval manuscripts (the manuscripts for which the module was originally most obviously designed). It has also been adopted by catalogues of non-Western manuscripts and other text-bearing objects, such as Fihrist for Arabic manuscripts and EpiDoc for ancient inscriptions. But there have been few assessments of the overall strengths and weaknesses of the module, areas in which it typically succeeds or fails.

This paper examines the use of the module using as case study a current project at the Bodleian Library, Oxford and the University Library, Cambridge, which is assessing how various cataloguing projects within both institutions have implemented the msDesc module. The aim is to simplify the technical infrastructure for storing and processing the TEI, to provide clearer guidance and improved workflow for future cataloguers, to improve interoperability within and between the respective institutions, and potentially to provide technical frameworks and metadata standards for other institutions looking to create msDesc records.

Both institutions have developed a wide range of TEI manuscript catalogues: In addition to Fihrist, both have contributed to the catalogue of Hebrew Genizah texts, and there are catalogues of Tibetan, Sanskrit, and other manuscripts and documents. There are considerable challenges in developing a TEI framework that is flexible enough to accommodate the needs of these different sources, while strict enough to meet the demands outlined above.