Explicit vs. implicit L2 grammar knowledge in written error correction

Abstract

Error correction is undoubtedly an important part of the process of drafting and producing written texts. The aim of the paper is to analyse the learners’ ability to correct grammatical errors in relation to the type of knowledge they employ in this task. Green & Hecht (1992), in an often quoted study, found a low correlation between L2 learners’ knowledge of explicit grammar rules and their ability to correct errors. They interpret this as suggesting that in error correction, learners rely primarily on their implicit knowledge. However, certain design features of their study might have caused the subjects to simply guess the correct forms, which, in turn, as DeKeyser (2003) suggests, may have led to the overestimation of implicit knowledge. This paper reports the results of an experiment where 150 Polish learners of English were administered a corpus-based error correction task, the design of which, however, differed from that of Green & Hecht (1992). These alterations resulted in finding a much closer link between the subjects’ knowledge of rules and their ability to correct grammatical errors.