Abstract: The reading task has different effects on the eye movements during reading. There are few studies to compare the perceptual spans of the normal reading and proofreading. The present study is to investigate the perceptual spans for the normal reading and proofreading in Chinese. The perceptual span is the effective vision field from which a reader can obtain useful information during a fixation in reading. There are studies which investigated the perceptual span of different languages, but few have studied the perceptual span of proofreading in Chinese language. In the present study, gaze-contingent moving-window technique was used to explore the perceptual span in normal Chinese reading and proofreading. 37 college students participated in the experiment. The experiment design was 2 (task: normal reading, proofreading) × 6 (window size: R0, L1R1, L2R2, L3R3, L4R4 and whole line) mixed design. The reading task was the between-participant factor, and the window size was the within-participant factor. The eye movements were recorded with a SR Research EyeLink 2K eye tracker. The sampling rate was 1000Hz. The refresh rate of the subject monitor is 150Hz. The distance between the subject and the screen was 69.6cm. Each Chinese character was 28×28 pixels and subtended 0.9°. There were 60 experimental sentences and 60 filler sentences. The stimuli were rotated in Latin square. There were one or two errors in the filler sentences in proofreading condition. The results were as followings: (1) There were slower reading rate and shorter mean saccade length in proofreading task than that in normal reading. (2)For the normal reading the perceptual was two to three characters to the right of the fixed word, for the proofreading the perceptual span was one character space. Reading rate implies temporal aspect of eye movements, and mean saccade length implies spatial aspects of eye movements. Reading task had influences on both temporal and spatial aspects of readers’ eye movement measures. These results indicated that the perceptual span was modulated by reading tasks. The results also suggested that readers adopted different reading strategies in the proofreading and normal reading. Compared to normal reading task, proofreading reading task required readers not only to comprehend the sentences, but also to look for errors. To do the task well, readers must scrutinize the sentences character by character. Furthermore, according to Kahneman (1973), who claimed cognitive resources were capacity-limited, the more complex the task was, the more capacity it would occupied. The cognitive load of proofreading task is heavier than that of normal reading, so in proofreading task, reading rate is slower, and mean saccade length is shorter, therefore, the perceptual span is smaller.

Key words: Key words:　perceptual span in reading, moving window technique, proof reading.

摘要： 采用经典的移动窗口技术，以Eyelink2000型眼动仪为工具，探讨了自然阅读和校对阅读两种不同的任务对中文阅读知觉广度的影响，结果表明：（1）与校对阅读条件相比，自然阅读条件下的阅读速度更快、眼跳幅度更大。（2）自然阅读条件下被试的知觉广度为注视点右侧2-3个字；校对阅读条件下被试的知觉广度仅为注视点1个字的空间。结果说明，读者在校对阅读和自然阅读条件下采用了不同的阅读策略，阅读知觉广度的大小受任务要求的影响。

关键词: 关键词：阅读知觉广度, 移动窗口技术, 校对阅读