Untangling locality and orientation constraints in the L2 acquisition of anaphoric binding: a feature-based approach

This study offers a Minimalist analysis of the L2 acquisition of binding properties whereby cross-linguistic differences arise from the interaction of anaphoric feature specifications and operations of the computational system (Reuland 2001, 2011; Hicks 2009). This analysis attributes difficulties in the L2 acquisition of locality and orientation properties in binding to problems reanalysing the features responsible for reflexivisation in the target language. Such an approach is shown to predict, in contrast to previous accounts, that if the locality and orientation behaviour of English reflexives arise due to syntactic operations on their features (Agree), acquisition of locality cannot be achieved unless orientation is also acquired; a picture-verification task completed by 70 Korean L2 speakers of English fully bears out this prediction. We show that for independent reasons, Korean speakers could still behave apparently nativelike for locality (by means of L1 transfer), but not for orientation. Crucially, this analysis can explain how two properties traditionally subsumed under the same UG principle can appear to pose different learning difficulties to L2 speakers