Lemmatiser un corpus de textes hagiographiques : enjeux et modalités pratiques

This paper deals with the lemmatisation process of a corpus of Latin hagiographic
texts translated from Greek from the early Middle Ages onwards. The discussion
concerns the well-known LASLA (Laboratoire d’Analyse Statistique des Langues
Anciennes, University of Liège) method, initially developed for Classical Greek and
Latin texts, and the conflicts that arise when it is applied to medieval texts. Some of
the most complicated cases are produced regarding uses differing from Classical
Latin (especially changes of declension) with a suggestion of how to solve them.