ERROR ANALYSIS IN A LEARNER CORPUS. （学习者语料库的误差分析。）.pdf

ERROR ANALYSIS IN A LEARNER CORPUS. WHAT ARE THE LEARNERS’ STRATEGIES? Willelmira Castillejos López1 Universidad Autónoma Chapingo (México) Abstract This corpus-based work is intended to find transfer and intralingual errors made by native Spanish speakers with a high proficiency level of the English language, particularly when the production is written. Recent trends in language teaching are in tune with second language acquisition (SLA) research results, so studying learner outputs is important to help define parameters on the way a second language is learned. At the same time, this means a contribution to teaching. By analyzing errors of learners’ production it is possible to distinguish the strategies they trigger in their process of SLA. The learner corpus used for this research is named ICLE (International Corpus of Learner English) and belongs to the Center for English Corpus Linguistics from the Université Catholique de Louvain (UCL). This corpus contains over three million words, a collection of essays written by learners of English with diverse mother tongues. The Spanish sub-corpus was error-tagged by researchers at the UCL, this was the material used for the study. Key words: learner corpus, transfer error, intralingual error, error-tagging. I. INTRODUCTION I.1. Corpus linguistics and error analysis For many years linguistics was criticized for lacking methods that demanded scientific rigour. Linguistic studies were considered part of other disciplines: logics, philosophy, sociology, psychology, etc (Corder, 1992: