Anticipating at a Distance: Classifier–Noun Agreement Processing in Japanese by Native Speakers and Chinese JFL Learners

Abstract This study examined how native Japanese speakers and Chinese learners of Japanese as a foreign language (JFL) process classifier–noun agreement involving short‐ and long‐distance dependencies. Using a maze task, participants read sentences in which classifiers agree locally either with the subject noun or with a distant object noun. Native speakers showed faster reaction times for short‐distance dependencies and a significant slowdown at the verb region under long‐distance conditions, but no difference at the head noun, indicating anticipatory processing using verb cues. In contrast, Chinese JFL learners showed slower processing at both the classifier and following noun regions, but faster responses at the head noun in the long‐distance condition. This suggests a bottom‐up, non‐predictive strategy, with resolution occurring only when the head noun is encountered. These findings highlight a key difference in real‐time sentence processing: Native speakers engage in predictive integration, while JFL learners rely more on reactive parsing.