Artificial Grammar Learning Ability Predicts L2 Processing of English Number Agreement

Abstract

This study investigates why L2 learners differ in their success in learning their second language, by examining the correlation between learning ability in artificial grammar involving an adjacent dependency and ability in processing English number agreement. Thirty-four Korean L2 learners of English completed two tasks: an artificial grammar learning task and a self-paced reading task. The results showed that (a) the L2 learners are able to track a frequent pattern (e.g., adjacent dependency) occurring in the artificial language; (b) the advanced L2 learners were not sensitive to violations of English number agreement in determiner phrases; and (c) the ability to learn a rule in the artificial grammar learning task was positively correlated with the L2 learners insensitivity to disagreement in the self-paced reading task. These findings shed light on the effect of the ability to track a rule in an artificial language on L2 sentence processing. The results thus suggest one explanation for why L2 learners variations appear in language learning, at least in the case of the comprehension of English adjacent number agreement.