Towards a classification of text features highly indicative of context-appropriate L2 writing competence

Abstract

The quality of written L2 performance is mirrored in the interplay between different text features potentially revelatory of the underlying competence. Such features may include vocabulary, grammar, cohesion, content, length, task realization, and more. The features appearing in such a form/use whereby they may be indicative of incomplete L2 acquisition, relative to the operating expectations, are commonly termed errors. Coding texts for errors, as context-inappropriate text features, has been a staple of applied linguistics for decades. The features indicating expected L2 acquisition having taken place (i.e., diametrically opposite to errors in terms of what they represent) do not have such a well-established conceptual or terminological equivalent nor a tradition of being marked in performances. In fact, there are no readily available classifications for systematically coding such context-appropriate text features when performed. To address this gap, the paper puts forward an intuitive approach to detecting and classifying L2 feedback-triggering text features exemplary of context-appropriate writing performance, termed here assessment-positives (mirroring errors as assessment-negatives). The paper then goes on to discuss the practical usefulness of the proposed coding method.