LINGUIST List 18.853
Wed Mar 21 2007
Diss: Discourse Analysis/Ling & Lit/Text&Corpus Ling: Dargnat: 'L'o...'
Editor for this issue: Hunter Lockwood
<hunterlinguistlist.org>
Directory
1. Mathilde
Dargnat,
L'oral comme fiction / Orality as fiction
Message 1: L'oral comme fiction / Orality as fiction
Date: 18-Mar-2007
From: Mathilde Dargnat <mathilde.dargnatfree.fr>
Subject: L'oral comme fiction / Orality as fiction
Institution: University of Provence (F) and University of Montreal (CA)
Program: Linguistics and Literary Studies
Dissertation Status: Completed
Degree Date: 2006
Author: Mathilde Dargnat
Dissertation Title: L'oral comme fiction / Orality as fiction
Dissertation URL: http://mathilde.dargnat.free.fr/index_fichiers/pageaccueilthese.html
Linguistic Field(s):
Discourse Analysis
Ling & Literature
Text/Corpus Linguistics
Writing Systems
Dissertation Director:
Lise Gauvin
Marie-Christine Hazaël-Massieux
Dissertation Abstract:
How do people construct a representation of spoken language in linguistictranscriptions and dramatic art? Is that representation stable? These twoquestions underlie the present dissertation.
Specifically, I offer a comparison between the images of spoken languageconveyed by a particular dramatic work (five plays by the Quebecois writerMichel Tremblay) and by a linguistic transcription (the Sankoff-Cedergrenand Montreal 84 corpora). Two conclusions are drawn. First, systematicallycomparing the linguistic and literary corpora allows one to highlight thedifferent constraints that shape the coding of orality, e.g. in thegraphical, syntactic and discourse dimensions. Second, in the literarycorpus, orality is more than a socio-linguistic parameter. It plays a rôlein the organization of narrative fiction. This shows that orality is doublyfictional, as a piece of social imagination about language and as a part ofa given narrative and emotional universe.
Methodologically, I used the Weblex software(http://weblex.ens-lsh.fr/wlx/), which allows one to compare the varioustranscriptions of words and idioms typically found in spoken language andto uncover the technical or aesthetical choices of transcribers or writers.Moreover, the software and two non parametrical statistical tests (Welchand Fisher tests) help extracting the linguistic profiles of characters andthe stylistic changes in the intertwining of language and narrative fictionover thirty years (1968-1998).
Theoretically, the main question is what «filters» are used in representingspoken language. The present work argues that two requirements have to bemet. (a) Linguistic categories have to be defined in a precise way, inorder for researchers to build and exploit non-standard language corpora,be they transcriptions or literary works. (b) Linguistic description shouldbe connected to a study of cultural and emotional factors, for a betterunderstanding of the three components (linguistic, symbolic and aesthetic)that make up orality.
|