Abstract
This article critically reflects on the overwhelming presence of literary texts in French diachronic linguistics, especially concerning medieval discourse markers and orality. After an attempt to linguistically define orality markers and to distinguish them from discourse markers, the question of why 'practical' non-literary documents (such as parliamentary records, trials or remission letters) have received little attention in linguistics will be discussed. We will then briefly present some of these documents, which could be studied together with literary texts, pointing out interesting meta-linguistic (status of swear-words, reception of these documents) and linguistic (pragmatic use of coordinators, 'oral' syntactic phenomena) aspects. This paper does not to seek to criticize prior studies on medieval orality in literature but rather to review the various sources available for the examination of medieval orality as well as to suggest new ways to deal with the problem of data.