Neurocognitive Mechanisms for Language Learning through Input-Output Processing

本研究通过日语母语者和日语学习者对语言的理解（输入）和产生（输出）过程以及语言学习的输入/输出处理进行了调查。采用神经网络是通过经验形成的工作假设，我们进行了行为、眼球追踪和事件相关的潜在实验。我们发现，基于语言线索的语义推理和通过句法/语义信息的增量加工获得的前向预测在语言的加工和学习中起着至关重要的作用。

This study investigated into the process of comprehension (input) and production (output) of language as well as language learning through input/output processing by native Japanese speakers and Japanese language learners. Adopting the working hypothesis that neural networks are shaped through experience, we conducted behavioral, eye-tracking, and event-related potential experiments. We found that semantic inferences based on linguistic cues and forward predictions obtained by incremental processing of syntactic/semantic information play crucial roles in the processing and learning of language.