engleski

On Methodological and Acquisitional Issues in Analysing Coherence in Non-native and Native Written Discourse

Communicatively competent non-native language users should be able to use their knowledge of language elements to form accurate and articulate written texts. However, this ability seems to be among those difficult to acquire, since it is largely language- and culture- specific. Most communicative competence models (cf. Canale 1983, Bachman & Palmer 2010) incorporate this ability into the notion of discourse competence which includes knowledge about cohesion and coherence. Previous relevant studies of non-native written discourse predominantly analysed aspects of cohesion – which can probably be attributed to the fact that cohesion is seemingly easier to operationalise – but coherence remains understudied, mostly because it has been inconsistently defined and conceptualised by text linguists and, consequently, applied linguists. Motivated by insufficient knowledge about ways of achieving coherence and factors that influence (in)coherence in non-native written discourse, this study aims at describing and comparing coherence dimensions in texts created by non-native users of English and compare them with coherence dimensions in texts they wrote in their first language as well as in those written by native speakers of English. Despite a number of existent methods of coherence analysis, our research aim warranted development of an appropriate method that considers distinctive features of non-native discourse (e.g. less coherent parts and coherence breaks). Subsequently, the Non-Native Text Coherence Analysis (NN-TCA) was designed building on the methods by Daneš (1970) and Lutamatti 1978). NN-TCA generates quantitative data on the number of t-units, topics, types of progressions, and coherence breaks thus allowing for statistical analysis and cross- sample comparisons. We analysed 30 argumentative essays written by non-native users of English, their essays in their first language and 30 essays written by English native speakers. The results point to differences among participants' texts in terms of types and numbers of progressions that might emanate from various conceptualisations of coherence in the two languages and cultures. Problems in achieving coherence prominent in non-native discourse can be explained by non-native writers' lack of awareness of what constitutes a coherent text in the target language. Also, non-native writers rely heavily on cohesive devices in establishing coherence. Methodological as well as acquisitional implications are discussed.

coherence, non-native written discourse, coherence analysis

nije evidentirano