本项目旨在通过无意识阅读来研究中文阅读过程中注意的动态分配。无意识阅读是当眼睛还在文本上运动的时候注意已经不在其上的现象，即“走神”。研究表明人们在日常生活中30-50%的时间里都在“走神”。无意识阅读也可以看成注意分配的极端例子——注意在深度和广度上趋近于零。研究者普遍接受眼动控制受注意参与程度的深度和广度影响。本项目为研究中文阅读过程中的无意识现象的首次探索。我们将实施一系列基础的实验，通过这些实验研究“走神”发生的情况。根据现有的理论，“走神”反映了大脑“默认网络”活动的情况。科学地研究“走神”也给现有的研究方法带来挑战，人们往往没法意识到自己已经“走神”。令人庆幸的是，眼动和大脑活动的的变化可以被用来较早地检测“走神”。本项目将结合眼动仪和脑电在自然阅读中更自然地检测“走神”。我们还将结合上述结果整合注意分配和眼动控制的动态性建立中文阅读过程中眼动控制的模型，并用来模拟无意识阅读。

The proposed research will examine episodes of mindless reading, with special focus on the novel insights that such episodes may offer into dynamic attention allocation during the reading of Chinese. Mindless reading is a phenomena wherein readers continue moving their eyes across the text, but their attention has disengaged from text processing. People are adept at decoupling thought from action, allowing well-practiced behaviors to proceed automatically, while the conscious mind is elsewhere occupied. It is known that mind-wandering occupies anywhere from 30-50% of everyday cognition. Mindless reading can be considered as an extreme case of attention allocation—the depth and scope of attention engagement approach to zero. A prominent hypothesis is that eye-movement control during reading can be influenced by the depth and scope of attention engagement. The proposed research represents the first investigation of mind-wandering in Chinese reading. We will conduct foundational experiments on mindless reading in numerous methods, as an ecologically valid means to investigate the nature of dynamic attention allocation. According to recent theory, mind-wandering reflects activity of the brain's "default network," which cyclically trades off with more controlled activity in frontal-parietal networks. The scientific study of mind-wandering poses methodological challenges, as it often requires self-report by the participant, who is typically unaware that attention has lapsed. In mindless reading, however, there are characteristic changes in eye-movements and brain signals that allow early, covert detection. Our research will use eye-tracking and EEG to identify mindless reading in an unobtrusive manner during natural reading. Combined these results, we will build a computational model for eye-movement control during the reading of Chinese, which will implement the dynamics of attention allocation and eye-movement control. Finally, we will simulate mindless reading by using this model.