Syntaxe, prosodie et structure informationnelle : une approche prédictive pour l'analyse de l'intonation dans le discours

Abstract

Résumé Cet article présente une approche pour l’analyse de la prosodie en français, basée sur une prédiction de l’intonation par défaut non marquée, qui reproduit les propriétés lexicales et syntaxiques de l’énoncé. La mise en regard des deux réalisations intonatives, l’intonation par défaut et celle réalisée par le locuteur, permet d’identifier de façon univoque les formes intonatives marquées utilisées par le locuteur, en vue d’examiner leur contribution au sens global du message. Le calcul de l’intonation par défaut tient compte de l’accent lexical, du nombre de syllabes, des relations de dépendance syntaxiques et des constructions syntaxiques. La procédure vise également la définition d’unités discursives à partir de critères syntaxiques et prosodiques.

Syntactic and Prosodic Forms of First Names in Institutional Interaction Involving Multiple Participants

“…with an IA starts a new prosodic group (Mertens, 2008;Wichmann, 2000). In this context, the IA emphasizes the term of address.…”

“…Thus, as can be seen in the graph, the rise starts a bit below the mean pitch level of the speaker and moves towards the upper end of the speaker's range. According to Mertens' (2008) "predictive theory" on French intonation, this type of major interval om down to up (LH) 7 invites the interlocutor to react. Morel and Danon-Boileau (1998) presented similar fi ndings.…”

“…However, the fact that there is a pause between the recipiency token "yes" and the fi rst name aff ects the form of the prosodic marking: the pause creates a prosodic boundary a er the word "yes". This leads to the production of an initial accent (IA) (Mertens, 2008) in the beginning of the fi rst name (Morel, 2010;Mertens, 2008). The occurrence of the pause, in turn, is related to the action of the participants, i.e., to the fact that Maryam does not continue talking a er the "yes", but she waits until she hears the teacher utter her name.…”

“…However, a movement of this size is still suffi cient to create the eff ect of a rising tone. A pitch contour of this type, starting with an IA, works to emphasize an entity and/or to start a new information unit (Mertens, 2008), since where a rise of the type LH links the word prosodically to the preceding group, a word starting 8. In addition to a raised pitch level, the IA of French also typically entails an increased level of loudness, a strengthened phonatory force and a longer duration of the syllable (Di Cristo, 1998;Mertens, 2008).…”

“…Thus, the rising pitch movement occurring here is similar to the one in excerpt [1], [Mou?ni::↑r]: in both cases, the size of the pitch movement is a major interval om low tone to high tone (LH), and the size of the rise is more than 9 semitones. As already mentioned, according to theories on French intonation, this type of contour (LH) typically invites the interlocutor to react (Mertens, 2008;Morel & Danon-Boileau, 1998). Raising the pitch in a quasi-interrogative fashion may accompany summons items, which is not necessary when the summons stands alone as in "Jim" (Schegloff , 1968(Schegloff , : 1081.…”