Wrong to Right: A Cognitive Linguistic Approach to the Analysis of Second Language Errors

Error Analysis (EA) responded to how contrastive analysis (CA) failed to predict second language (L2) errors from L1 and L2 contrasts (e.g. Richards 1971; 1974a; 1974b, Corder, 1973; 1981, Dulay and Burt, 1974, James, 1998). Interest in EA waned because it assessed learner language from a non-developmental perspective and found identifying the sources of errors problematic (e.g. Shachter and Celce-Murcia, 1977; Ellis and Barkhuizen, 2007). Cognitive Linguistics (CL) renews interest, however, by treating all form as meaningful and meaning as originating in the conceptualizations through which an enactive cognition structures experience (e.g. Talmy, 1978, Lakoff, 1987, Langacker 1990). CL has revitalised Sapir (1949) and Whorf¡¦s (1956) proposition that different languages give cultures different ways of structuring reality (e.g. Lucy and Gaskins, 1979; Lucy and Shweder 2001) thus emphasizing how language learners must cross a conceptual divide. Concomitantly, some L2 errors begin in the use of second language forms with first language meanings and the conceptualizations that these carry (Slobin, 1985; 1996, Valeva, 1996; Holme, 2004) requiring from the student what has come to be called conceptual fluency¡¦ (Valeva, 1996). This project has two phases. In Phase 1 we exploit a one million word corpus of academic writing from Chinese L1 users of academic English compiled in a previous internally financed project. We use this corpus to search for consistent errors and place them on the lexico-grammatical continuum categorizing them as grammatical and semantically schematic or lexical and semantically substantive. We search the error type to determine its relative frequency, disregard the infrequent and thus compile an inventory of significant errors and their exemplars for this learner group. In phase 2 we translate the errors into Chinese and search for their frequency on the ICLE (International Corpus of Learner English) looking at files from non-Chinese learners. Erro

Funding Source: RGC - General Research Fund (GRF)