Is the processing of words during eye fixations in reading strictly serial?
Is the processing of words during eye fixations in reading strictly serial?
复制标题
DOI：
10.3758/bf03212147
复制
发表时间：
2000-10-01
期刊：
影响因子：
--
通讯作者：
Shindler, KL
中科院分区：
文献类型：
--
作者：
Inhoff, AW;Starr, M;Shindler, KL
现有的阅读过程中的眼球运动控制模型认为，注意力的分配将单词识别限制在一次一个(目标)单词上，并且在注意力转移到新的(目标后)单词之前计算对该新单词的眼动。为了检验这些假设，在注视前一个目标词的过程中，操纵了后目标词的预览特性。主要结果显示，当目标后预览在视觉上与众不同或在拼写上非法时，目标观看持续时间更长。目标语后文本的意义并不影响目标词的初始阅读，但会影响目标词的重读时间。为了解释这些发现，现有的注意力转移模型必须假设读者在注意到一个副可见的单词之前，至少从它获得了视觉空间和拼写信息。然而，这种观点存在缺陷，有几个考虑因素倾向于限制性较低的模型假设，根据该假设，注意力可以一次分配到一个以上的单词。
Extant models of oculomotor control during reading maintain that allocation of attention confines word recognition to one (target) word at a time, and that an eye movement to a new (posttarget) word is computed before attention is shifted to it. To test these assumptions, properties of the posttarget's preview were manipulated during the fixation of the preceding target word. The main results revealed longer target viewing durations when the posttarget preview was visually distinctive or when it was orthographically illegal. The meaning of posttarget text did not affect initial target word reading, although it affected the time spent rereading the target. To account for these findings, extant attention-shift models must assume that readers obtain at least visuospatial and orthographic information from a parafovelly visible word before it is attended. This view has shortcomings, however, and several considerations favor less restrictive model assumptions according to which attention can be allocated to more than one word at a time.