COSME² - Complexities: 30 Years Of Research Of Medievalists DH Concerning A Thousand Years Of Medieval Sources

International audienceCosme² is the Consortium Sources Médiévales (linked to TGIR Huma-Num from CNRS : https://cosme.hypotheses.org/ ), it is dedicated to digital approach of the medieval sources, for french mediaevalists. It wants to bring together a large part of the French medievalist community (mainly historians) around the digital processing of medieval sources, mainly written (but not only). Its aims: to gather all mediaevalists around their sources, to build the community by proposing frameworks, standards, ideas around Digital Humanities, in collaboration with the group "Menestrel" http://www.menestrel.fr/ , and help to the creation and publication on line of digital corpora. The complexity of the medieval digital landscape is due to its ancientness: medievalists were among the first to be concerned about the digital processing of their sources. Databases and corpora digitised in various forms are therefore many and varied, many remain dormant or need to be upgraded, others lack metadata, others are no longer online or on outdated media, others lack interoperability, even if their content allows them to do so. The hundreds of thousands of digitised medieval charters are not yet effectively linked. Medievalists were among the first to design electronic publishing platforms (such as TELMA http://www.cn-telma.fr/ or Scripta http://www.unicaen.fr/scripta/ or CBMA http://www.cbma-project.eu/ ) but they are not yet interoperable. These corpora themselves are complex, associated with the sources which are just as complex, by their diversity, by the concepts implemented but also by the chronological span (1000 years of Middle Ages). Pioneering strategic choices have been made, such as the adoption of the TEI, but paradoxically many projects stand out from it, because the creation of interoperability no longer necessarily requires fine markings but automatic indexing tools based on complex referentials, fed by artificial intelligence training. The implementation of data exploitation tools, text mining procedures, instruments for extracting named entities and qualifying data within the corpuses of digitised medieval sources, but also the development of tools for analysing these corpuses (distant and close reading), within COSME², make it possible to consider the connection of these complex and still poorly connected corpuses. The development of good practice guides and support manuals, the organisation of summer schools to train researchers in these tools and techniques are essential complements to strengthen this connection. The poster aims to present the complexity of documentary files, corpora and data, the complexity of historical uses and the complexity of the communities concerned, but also to evoke the operations envisaged to solve the complexities of communication and structuring by implementing integrated solutions