The summer school in Digital Palaeography at the Göttingen University Institute for Digital Humanities is an intensive training course for graduate students from all around the world. It has been held every year since 2022 in the summer at the institute’s premises in Göttingen. In a two-week-long program, participating students learn about medieval Latin palaeography and digital approaches to manuscript studies, in both theoretical and practical sessions. With generous support from the Volkswagen Stiftung and in collaboration with the Göttingen State and University Library as well as the Herzog August Library Wolfenbüttel, the summer school covers all essential areas of digital palaeography at a state-of-the-art level. Its unique program means to support new researchers on their way to becoming future experts in palaeography and digital manuscript studies alike. Key areas that are covered in the summer school include palaeography and codicology, hyper- and multispectral imaging, text encoding with TEI XML, handwritten text recognition (HTR), digital editing, cataloguing, IIIF and digitisation standards, bookbinding, and more. Applications to the summer school are accepted from around February to May each year.
Multispectral imaging is a non-invasive, non-destrucive method for recovering damaged, erased, or otherwise illegible text from cultural heritage objects such as manuscripts and documents. The setup is simple: two banks of LED lights , a camera, a filter wheel. At Göttingen, we use system designed and built by Phase One and operated by their Rainbow MSI software. This has the advantage of being portable, so that we can take the system with us during (for example) our Digital Paleography Spring School in Vercelli and Verona, allowing us to study precious manuscripts wherever they reside.
Digital images are taken in a series of discrete wavelengths of light, from the ultraviolet backwards through the rainbow to the infrared. The filter wheel enables us to squeeze more data of these wavelengths, or bands: we can, for example, illuminate a manuscript with ultraviolet, then use a filter to block out all the reflected light, capturing only the fluorescent light that the page itself is emitting. Or we can do the reverse, and capture only the reflected light. The goal is to obtain as much spectral data — light data — from each folio as possible.
Cod MS Syr 19c, a palimpsest from the collections of the University of Göttingen, as it appears to the naked eye.
This, though, is just the beginning of the process. It is very unlikely that any one image will reveal everything one would like to read on the page; none of them will show much more than standard UV imaging. To that end, we use a number of different programs including Hoku, ImageJ, and Harris Geospatial’s ENVI to process the data. In this stage we sort and sift, transform and analyze, using techniques like principal components analysis to enhance and isolate targeted areas of the text. When successful, this allows us to (for example) digitally fade out the overtext of a palimpsest while darkening and revealing the erased text below.
Cod MS Syr 19c after processing.
Multispectral imaging is a new and growing field. An increasing number of libraries and institutions are purchasing MSI systems, but there remain few programs for training users to capture and process the data these systems can obtain. Thanks to a grant from the Lower Saxony Ministry for Science and Culture (MWK), the University of Göttingen has been able not only to purchase an operate an MSI system of its own, but also to offer courses and training to students interested in learning about this burgeoning area of the digital humanities.