因为中文文本的词和词之间没有空格分隔，读者需要依据自己的知识进行词切分。以往的研究表明在词切分的过程中，因为认知资源的限制及加工策略的影响，第一遍阅读时有时会出现切分错误现象。以往的研究表明，读者能够检测到错误并会引起眼睛回视，对错误进行纠正。然而，读者究竟如何检测和纠正词切分错误目前还不得而知。拟采用眼动跟踪、脑电记录及fMRI脑成像技术对这个问题进行深入研究。研究一将考察读者如何检测阅读过程中出现的词切分错误；研究二将研究读者发现切分错误后，采用怎样的认知加工过程进行纠错；研究三将研究中文读者为何不仅仅依靠工作记忆存储的信息进行纠错，而需要回视；研究四中我们研究词切分错误纠错的脑机制。研究结果能帮助我们理解中文读者在词切分过程中出现错误后的检错和纠错的认知神经机制。研究结果还将为提高计算机分词系统的词切分准确率提供基础，为促进阅读教学及评估阅读材料的可读性做出贡献。

Since there are no inter-word spaces to demarcate word boundaries in Chinese text, Chinese readers have to depend on their knowledge to segment words during Chinese reading. Previous studies showed that Chinese readers sometime make mistakes in word segmentation due to limited processing resources and different processing strategies. When a word segmentation error happens, readers usually make a regressive eye movement to fix the error. However, it is not known how Chinese readers do that. In the proposed project, we plan to address this question by using eye tracking, EEG recording, and fMRI techniques. In the first study, we will look into how Chinese readers detect word segmentation errors. In study 2, we will explore how Chinese readers correct the error when they re-fixate on the character string which caused the word segmentation error. In study 3, we will investigate why Chinese readers make time-consuming regressive eye movements to fix the errors instead of using information stored in working memory. In study 4, we will test which brain regions are involved to detect and fix word segmentation errors. The findings of the proposed project will help us to understand the mechanisms of word segmentation error detection. Moreover, the findings will facilitate word segmentation algorithms for artificial language processing system. The findings will also benefit reading education and readability evaluation.