Center for Brain and Cognitive Sciences, Peking University, Beijing, China.
J Exp Psychol Learn Mem Cogn. 2013 May;39(3):915-30. doi: 10.1037/a0029182. Epub 2012 Jul 9.
Prosodic boundaries can be used to guide syntactic parsing in both spoken and written sentence comprehension, but it is unknown whether the processing of prosodic boundaries affects the processing of upcoming lexical information. In 3 eye-tracking experiments, participants read silently sentences that allow for 2 possible syntactic interpretations when there is no comma or other cue specifying which interpretation should be taken. In Experiments 1 and 2, participants heard a low-pass filtered auditory version of the sentence, which provided a prosodic boundary cue prior to each sentence. In Experiment 1, we found that the boundary cue helped syntactic disambiguation after the cue and led to longer fixation durations on regions right before the cue than on identical regions without prosodic boundary information. In Experiments 2 and 3, we used a gaze-contingent display-change paradigm to manipulate the parafoveal visibility of the first constituent character of the target word after the disambiguating position. Results of Experiment 2 showed that previewing the first character significantly reduced the reading time of the target word, but this preview benefit was greatly reduced when the prosodic boundary cue was introduced at this position. In Experiment 3, instead of the acoustic cues, a visually presented comma was inserted at the disambiguating position in each sentence. Results showed that the comma effect on lexical processing was essentially the same as the effect of prosodic boundary cue. These findings demonstrate that processing a prosodic boundary could impair the processing of parafoveal information during sentence reading.
韵律边界可以用于指导口语和书面语句子理解中的句法分析，但目前尚不清楚韵律边界的处理是否会影响即将到来的词汇信息的处理。在 3 项眼动实验中，参与者默读句子，在没有逗号或其他提示指定应采用哪种解释的情况下，这些句子允许有 2 种可能的句法解释。在实验 1 和实验 2 中，参与者听到句子的低通滤波听觉版本，该版本在每个句子之前提供了一个韵律边界提示。在实验 1 中，我们发现边界提示有助于提示后的句法消歧，并导致提示前的注视持续时间长于没有韵律边界信息的相同区域。在实验 2 和实验 3 中，我们使用注视相关的显示变化范式来操纵目标词在消歧位置后的第一个成分字符的前区的 parafoveal 可见性。实验 2 的结果表明，预览第一个字符会显著减少目标词的阅读时间，但当在该位置引入韵律边界提示时，这种预览优势会大大降低。在实验 3 中，在每个句子的消歧位置插入了一个视觉呈现的逗号，而不是声学提示。结果表明，逗号对词汇处理的影响与韵律边界提示的影响基本相同。这些发现表明，处理韵律边界可能会损害句子阅读过程中对 parafoveal 信息的处理。
