Contenu de l'article

Titre Le Factum Textus : fait de grammaire, fait de linguistique ou fait de cognition ?
Auteur L. Lundquist
Mir@bel Revue Langue française
Numéro no 121, février 1999 Phrase, texte, discours, sous la direction de Étienne Stéphane Karabétian
Page 56-75
Résumé anglais People are able to distinguish with a large degree of consensus between texts and non-texts. In the present article, I discuss the theoretical implications of this fact for text linguistics. I demonstrate, partly via a small experiment, that sentences in text can contain three categories of linguistic expressions which contribute to signal their position in the text : anaphors, argumentative signals and mental space builders. Taking a partly psycholinguistic point of view, I argue how text linguistics must integrate cognitive as well as linguistic aspects, but suggest that it can nevertheless obtain status as a linguistic science by satisfying the three criteria proposed by Milner (1995) of falsification, « mathematisation » and « technical application ».
Source : Éditeur (via Persée)
Article en ligne http://www.persee.fr/doc/lfr_0023-8368_1999_num_121_1_6279