10.5281/zenodo.4317407
https://zenodo.org/records/4317407
oai:zenodo.org:4317407
Stokes, Peter
Peter
Stokes
0000-0002-9060-9340
École Pratique des Hautes Études – Université PSL
Digital and Computational Palaeography: Some Promises and Problems
Zenodo
2020
Palaeography
Manuscript Studies
Artificial Intelligence
Machine Learning
Digital Humanities
2020-12-11
eng
10.5281/zenodo.4317406
https://zenodo.org/communities/dmwr
Creative Commons Attribution 4.0 International
Video, abstract and draft transcript of presentation given at the 13th Annual (Virtual) Schoenberg Symposium on Manuscript Studies in the Digital Age.
As the Schoenberg Symposium and the Schoenberg Institute for Manuscript Studies have long been demonstrating, access to digital images of manuscripts has grown enormously in recent years, particularly with IIIF and the movement towards increasingly open and rights-free content. At the same time, we also have increasing access to relatively powerful computers, fast internet connections, and now access even to hardware and software for deep learning such as free software libraries, increasingly affordable GPUs and, in some cases, high-performance computing clusters. At the same time, however, librarians are often reporting less time for cataloguing and metadata, and at least in some cases access to original items is becoming increasingly difficult, not to mention the increasing complexity of computational methods and the widely-recognised fact that the ‘reasoning’ of modern machine learning is inaccessible even to specialists. The current health crisis seems to be accelerating these trends, which leaves the question: what of palaeography? Looking at the past and present of digital and computational methods, in this talk I attempt to identify some of the promises and problems of where we are now and where we seem to be going.
eScriptorium has received funding from the PSL IRIS programme, and from the European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement No. 871127