Abstract: Oral reading means reading aloud. It is very important for language learning. However, compared to the abundant research in the silent reading, much fewer research concerns the cognitive control in oral reading. A unique index in eye movement recording for oral reading is the eye-voice span (EVS), which directly links the eye movement and the speech voice. Thus, unveiling the psychological meaning of the EVS through related studies is a proper way to understand the processing mechanism in oral reading The EVS has been put forward for a long time. However, it has not received much attention until the last ten years. The EVS can be defined either in terms of space (the distance between the currently articulated item and the currently fixated one, spatial EVS), or in terms of time (how long it takes to articulate the item after having fixated it, temporal EVS). When EVS is defined in terms of time, a distinction is made between the time from the onset of word fixation to the onset of word naming (onset EVS), and the time from the offset of word fixation to the onset of word naming (offset EVS). According to the definition of the temporal EVS, the key roles of its calculation are (1) to keep the recoding of eye movement and the voice into the same reference in time; (2) to separate different items of a fluent speech in time. The first question can be resolved by a proper procedure in the Experiment Builder. A Praat script could be used to locate the beginning and end of spoken parts by crossings of an intensity threshold. The onset EVS for each item is equal to its naming latency, thus contains all the cognitive processing before the voice onset. Previous studies showed that the temporal onset EVS was around 500ms, while the offset EVS about 250ms, much longer that the time consuming for the motor programming of single word. Thus, the offset EVS was considered as an index for parallel processing for the current item and the next one. Both the onset and offset EVS can be used as predictors for reading deficits, and influence the eye movement control during oral reading. To be specific, readers can adjust the fixation duration to keep the EVS not too long; if this adjustment did not sufficiently reduce the EVS, a regression rather than a refixation followed with high probability. The onset EVS has been wildly explored in the rapid automatized naming (RAN) task in the previous researches. In the future, the offset EVS in this task needs to get more attention, especially considering its implications for parafoveal preview processing during oral reading. The EVS can also help understand the mechanism of language processing deficits in other special groups, such as stutters. However, a limitation for the application of the EVS is that, locating the beginning and end of spoken parts by the Praat script needs great manual work and lacks standardized criteria. The development of an automatic and standardized procedure to record the EVS is in great need for the promotion of the EVS application and comparison between scientific researches.

Key words: Eye-voice span (EVS), Onset-EVS, Offset-EVS, Rapid automatized naming (RAN)

摘要： 朗读是一种出声的阅读方式，而测量朗读过程的一个重要的眼动指标就是眼音距。眼音距是指朗读过程中注视点和发音点之间的时间或空间距离，虽然这个概念早在1920年就提出来了，但直到近十年才受到较多关注，而对于其适用范围及心理学意义，认识尚不充分。本文从眼音距的概念提出和发展入手，分别分析了起始眼音距和结尾眼音距的计算方法和心理学意义，以及它们在快速命名和自然文本朗读任务中对阅读能力和认知控制的反映。未来研究可借助眼音距的特点和优势，进一步探讨朗读过程中副中央凹预视加工的特点和认知控制，以及揭示特殊群体在语言加工缺陷上的发生机制。

关键词: 眼音距, 起始眼音距, 结尾眼音距, 快速命名范式