Le texte dans tous ses états. Philosophie d’encodage du projet Khartês

When editing Medieval sources, one always encounters the same dilemma : how to make the text readable nowadays without spoiling its testimony ? « Cleaning » the text from all that makes it difficult to read for someone
erases marks that would look valuable from another point of view. Hence, there is an unavoidable gap between the
source and the edition. The Khartês project (which intends to edit and study original Old French charters written
in Liège before 1300) has encountered this problem. Traditional « solutions » were inadequate and computers
opened an interresting path. However, Khartês had to think about creating an encoding scheme that would suit its
aim and remain easy to use. This is the way KML (the Khartês Markup Language, an XML-conformant language)
was born, together with a small set of handy encoding rules. A wide range of applications are made possible. For
instance, thanks to the way the corpus has been created, it has been possible to search a particular document for
marked letters (which look like our uppercase letters) and to study them.