In this paper, we evaluate the effectiveness of Granger's Integrated Contrastive Model for describing real language use and predicting correct and incorrect L2 productions with a detailed corpus‐based study of the structural and semantic similarities and divergences between the French and Dutch demonstrative determiner systems in L1 and their precise impact on written L2 productions. This study allows us to formulate six objective recommendations for developing pedagogical grammars and thus illustrates to what extent the combination of L1 and L2 corpora analysis should become an obligatory practical step rather than a theoretical one between primary forged linguistic analyses and the elaboration of well‐balanced and representative didactic material.
Dans cet article, nous évaluons l'efficacité de l'Integrated Contrastive Model de Granger pour décrire la langue réelle et pour prédire les productions L2 correctes et erronées par le biais d'une étude de corpus détaillée des ressemblances et divergences structurelles et sémantiques entre les systèmes du déterminant démonstratif en français et en néerlandais en L1 et leur impact précis sur les productions écrites en L2. Cette étude permet de formuler six recommandations objectives pour le développement de grammaires pédagogiques et illustre, par conséquent, dans quelle mesure la combinaison d'analyses de corpus L1 et L2 devrait devenir une étape pratique obligatoire plutôt qu'une étape théorique entre les premières analyses linguistiques et l'élaboration de matériaux didactiques représentatifs.
RÉSUMÉCet article étudie les différences distributionnelles entre les emplois référentiels du SNdém en français et en néerlandais. En soumettant les données issues de deux corpus comparables (Dutch Parallel CorpusetCorpus de Namur) à notre modèle triangulaire systématique des emplois du SNdém, nous constatons que deux explications majeures permettent de comprendre la plus grande partie des divergences attestées au niveau des emplois référentiels du SNdém en français et en néerlandais, à savoir des normes stylistiques différentes en relation avec des différences linguistiques et des divergences au niveau du degré de déicticité des SNdém en français et en néerlandais.
This paper focuses on translational shifts with respect to the demonstrative determiner in French and Dutch in parallel corpora. The paper aims to identify the types of translation shifts that occur systematically, and to explore the underlying mechanisms and semantic effects of this process. For this purpose, a well-balanced sub-corpus of the Dutch Parallel Corpus is used, making it possible to analyze both directions (French – Dutch and Dutch – French). In this corpus, 50% of the demonstrative determiners are translated by a demonstrative in the target text (in both directions). In 20% of the cases, the demonstrative is translated by a definite article, or vice versa, while 30% are translated by another grammatical element (e.g., indefinite determiner, adverb, personal pronoun) or vice versa. The parallel corpus study reveals that translational shifts with respect to French and Dutch demonstratives can be attributed to three different mechanisms: (1) translator preference related to translation universals at the level of the noun phrase (omissions, additions and reformulations of the noun phrase), (2) specific manifestations of translation universals within the noun phrase (syntagmatic and paradigmatic explicitation and implicitation involving demonstrative shifting) and (3) structural divergences between the French and Dutch demonstrative determiner systems (fixed expressions and semantic differences). This analysis demonstrates the usefulness of a detailed parallel corpus study, which clearly distinguishes between changes occurring at different levels, in accounting for divergent translations of the demonstrative determiner in different languages. To this end, several types of explanation drawn from various fields (such as translation studies and contrastive linguistics), must be considered.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.