【作者】 仝文； 刘志方； 张骏； 李利平； 任梦雪；

【Author】 TONG Wen;LIU Zhi-Fang;ZHANG Jun;LI Li-Ping;REN Men-Xue;Department of Psychology, Shanxi Normal University;College of Education, Hangzhou Normal University;

【机构】 山西师范大学心理系； 杭州师范大学教育学院；

【摘要】 阅读知觉广度反映的是读者阅读时的视觉注意分布范围,其大小受注意资源限制,且在一些主客观因素的影响下发生变化。历史上,视觉注意被认为是基于空间而不是基于对象分布的,但是后来的研究发现对象的属性也会影响注意的分布,即注意也可以基于"对象"分布,但是如何定义"对象"仍不清楚。为检验中文阅读中注意分布的对象究竟是"字"还是"词",本研究通过观察词长(对象属性)对阅读知觉广度(注意分布)的影响,考察中文阅读中注意分布的依据单元(对象大小)问题。若词长大的句子阅读知觉广度更大,则注意分布的基本单元为词;若阅读知觉广度不随词长的增大而增大,则注意分布的基本单元为字。研究由三个实验组成,每个实验有42名母语为汉语的本科生参加,他们的平均年龄为20岁(年龄范围:18-22)。研究采用移动窗口范式,通过三项实验分别两两比较单字词句(完全由单字词组成的句子)、双字词句(完全由双字词组成的句子)和混合句(所有词由1-4个汉字组成,且平均词长为1.5个字)的阅读知觉广度。实验1发现,单字词句子的阅读知觉广度为L1R3(注视点左侧一个字,注视点右侧三个字,共五个字),双字词句的阅读知觉广度为L1R2(共四个字);实验2发现,单字词句子的阅读知觉广度为L1R3(共五个字)或者L2R2(共五个字),混合词长句子的阅读知觉广度为L1R3(共五个字);实验3发现,双字词句子阅读知觉广度为L1R3(共五个字),混合词长句子的阅读知觉广度为L2R3(共六个字)。综合三项实验结果可以确定,中文阅读知觉广度不随句子词长的增大而增大,说明阅读中的注意资源以字为单元分布在阅读知觉广度内。更多还原

【Abstract】 The perceptual span reflects the deployment of visual attention during reading, which is variable and constrained by attention resources(Henderson & Ferreira, 1990). Historically, visual attention was thought to be space based rather than object-based, later, studies revised this view that the attributes of objects also affect the deployment of attention, so the attention can also be based on "object", but how to define "object" is still not clear(Li & Logan, 2008). In order to investigate whether the objects of visual attention deployment in Chinese reading was "word" or "character", this study explored the effects of word length(one of the word’s attributes) on reading perceptual span(attention deployment region), inferring the basic unit of attention deployment in Chinese reading. If the perceptual span would be larger when the length of words that consisted the sentences was lengthier, the basic unit was word. On the contrary, if the perceptual span did not grow with the word length, the basic unit of attention deployment was character. The study consisted of three experiments, all of which used the moving window paradigm, and employed a 2(Sentence Type) × 6(Window: L1 R1, L1 R2, L1 R3, L2 R2, L2 R3, Normal) within-subjects design. 126 undergraduates and native Chinese speakers were randomly divided into three equal groups and respectively assigned to one of the three experiments. Their ages ranged from 18 to 22 years(M = 20). Eye movements were monitored via an EyeLink II tracker with eye position sampled at 500 Hz. In each experiment, one group of participants read the two of three sentences types(each 60): 1-character-word sentences which consisted of single character words entirely, 2-character-word sentences which consisted all by 2-character-words and mixed sentences which consisted of words that contain 1-4 characters and the average word length was 1.5 characters. The three experiments respectively compared the relative perceptual span of the two of three sentence types, inferring the basic unit of Chinese reading attention deployment. Experiment 1 found that the perceptual span of 1-character-word sentences ranges from 1 character to the left to 3 characters to the right of the fixation, and from 1 character to the left to 2 characters to the right for 2-character-word sentences. The latter had a narrower perceptual span than the former. Experiment 2 found that the perceptual span of 1-character-word sentences ranges from 1 character to the left to 3 characters to the right or from 2 characters to the left to 2 characters to the right of the fixation. The perceptual span of the mixed sentences ranges from 1 character to the left to 3 characters to the right. It was equal in perceptual span for the two sentence types. Experiment 3 found that the 2-character-word sentences had a perceptual span from 1 character to the left to 3 characters to the right of the fixation, and the perceptual span of mixed sentences ranges from 2 characters to the left to 3 characters to the right of the fixation. 2-character-word sentences had a narrower perceptual span than mixed sentences. Based on the results of the three experiments, it could be concluded that the perceptual span did not increase with the length of the words that constituent the sentences in Chinese reading, which indicates that the attention resources during reading are deployed on the characters rather than words within the perceptual span. This conclusion provides a basis for the localization of the models of eye-movement control in alphabetic language reading and the establishment of further revision of the word segmentation and recognition model in Chinese reading, making it suitable for a wider range of situations.更多还原