Project

Discipline

first language acquisition; pragmatics; second language acquisition; empirical methods; discourse; cross-linguistic studies

Lead
Dans un texte, les phrases s’enchaînent de manière cohérente. Ces liens de cohérence peuvent être explicités par l’usage de mots comme ‘mais’ et ‘parce que’, que nous appelons des ‘connecteurs pragmatiques’. Des études ont montré que les problèmes de cohérence dans un texte sont bien souvent liés à un mauvais usage de ces mots. A l’inverse, lorsque les connecteurs pragmatiques sont correctement utilisés, ils améliorent la rapidité de lecture, ainsi que la compréhension et la mémoire d’un texte.
Lay summary
Maîtriser l’usage des connecteurs pragmatiques est capital pour articuler clairement ses idées, mais également particulièrement complexe. Certains travaux montrent que jusqu’à dix ans, les enfants n’utilisent pas de manière appropriée des connecteurs simples marquant des relations temporelles ou de causalité. Dans une perspective multilingue, la maîtrise des connecteurs pragmatiques est encore plus problématique, car les langues varient beaucoup dans l’usage qu’elles en font et peu de correspondances exactes existent, même entre des langues proches. Le projet MULDIS a trois objectifs : (1) étudier de manière systématique les correspondances entre les connecteurs pragmatiques du français, de allemand et l’anglais afin de cartographier précisément les zones de divergence ; (2) étudier l’acquisition des connecteurs typiques de la langue écrite en allemand et en français chez des enfants de 8 à 16 ans ; (3) étudier l’apprentissage des connecteurs en anglais langue seconde et comparer les difficultés des apprenants francophones et germanophones. Ces trois domaines d’étude complémentaires permettront d’améliorer notre compréhension du fonctionnement des connecteurs dans les langues ainsi que leur représentation dans le lexique mental des locuteurs, avec des retombées significatives pour la pédagogie en langue première et seconde et la lexicographie. Ce projet s'inscrit dans le cadre du centre scientifique de compétences sur le plurilinguisme.

Direct link to Lay Summary

Last update: 10.01.2014

Active participation

Number	Title	Start	Funding scheme

The MULDIS project proposes a cross-linguistic investigation of the way discourse connectives such as 'however', 'while', or 'therefore' are used to structure texts in three widespread languages in Switzerland: French, German and English. The empirical analysis of language use will be paired with cognitive studies bearing on the acquisition of connectives in the minds of first and second language learners. MULDIS will contribute to important debates about the interface between language and cognition, namely: the impact of cross-linguistic differences of linguistic encoding on acquisition and processing; the acquisition of lexical items restricted to the written mode; and the integration of procedural meaning in the first and second language mental lexicons.Discourse connectives form a functional category intersecting several grammatical categories such as conjunctions and adverbs. Their function is to convey discourse relations like cause or contrast between units of text or discourse (e.g. Halliday & Hasan, 1976; Mann & Thomson, 1988; Sanders et al., 1992; Knott & Dale, 1994). Both language pedagogy and psycholinguistics confirm that their mastery is essential for effective verbal communication, but also particularly challenging. Indeed, children as old as 10 years still perform less well than adults on tasks designed to assess their comprehension and use of connectives (e.g. Cain & Nash, 2011). The causes of the acute difficulty of connectives are numerous: (1) connectives require an understanding of sometimes complex discourse relations; (2) they can convey different senses in different contexts; (3) their senses may be hard to define because they convey procedural rather than conceptual meaning; and (4) their use may vary considerably according to genre and register. It has been shown that when texts are difficult to understand, it is often because of problems at the discourse structure level (Beck et al, 1991, McNamara et al., 1997), and conversely the adequate use of connectives helps readers with text processing and comprehension (Sanders et al., 2007; Canestrelli et al., 2012).From a multilingual perspective, comparing discourse connectives is a challenge, because languages differ widely in how and when they mark discourse structures and what additional information the connectives convey. In most cases no one-to-one mapping exists, even between typologically related languages (e.g. Degand, 2004; Taboada & de los Ángeles Gómez-González, 2012; Zufferey & Cartoni, 2012). For this reason, connectives are also particularly challenging for second language learners (e.g. Crewe, 1990; Granger & Tyson, 1996). These difficulties are increased by the deficiencies of traditional language learning tools, such as dictionaries and grammars, which fail to adequately explain the meanings and functions of connectives. MULDIS will bring new insights on the meaning, use and acquisition of discourse connectives through a series of empirical investigations conducted across three interrelated dimensions of study.1. Cross-linguistic analyses of connectives: systematic corpus-based analyses of convergences and mismatches between connectives that convey relations of cause, condition, concession and contrast (“the four Cs”) in English, French and German. Results from these studies will also contribute to a larger European consortium aiming at providing equivalences across connectives in eleven languages using convergent data and methodology.2. Late first language acquisition of connectives: controlled experiments to measure the development of late-acquired connectives that are typical of the written mode, such as 'toutefois', 'néanmoins' in French or 'demzufolge' and 'gleichwohl' in German. These experiments will have important applications for L1 syllabus design. Indeed, using a connective leads to longer and potentially more complex sentences for children to process, but this negative effect can be compensated for as soon as young readers can use the clues provided by connectives for text processing. This transition has however not yet been firmly established in the literature.3. Second language acquisition of connectives: controlled experiments paralleling those for first language acquisition. The focus will be on measuring the development of connective usage and comprehension by French-speaking and German-speaking learners of English, with an emphasis on the causes of misuses. These analyses will have important practical implications for foreign language pedagogy and will shed new light on the development of procedural meaning in an L2 mental lexicon. Through the coherent integration of these dimensions of study, MULDIS will pull the strings of language use and its mental representation together, and open new avenues of enquiry for the study of discourse structure across languages and minds.