Dahan Delphine, Tanenhaus Michael K
Max Planck Institute for Psycholinguistics.
J Exp Psychol Learn Mem Cogn. 2004 Mar;30(2):498-513. doi: 10.1037/0278-7393.30.2.498.
The authors used 2 "visual-world" eye-tracking experiments to examine lexical access using Dutch constructions in which the verb did or did not place semantic constraints on its subsequent subject noun phrase. In Experiment 1, fixations to the picture of a cohort competitor (overlapping with the onset of the referent's name, the subject) did not differ from fixations to a distractor in the constraining-verb condition. In Experiment 2, cross-splicing introduced phonetic information that temporarily biased the input toward the cohort competitor. Fixations to the cohort competitor temporarily increased in both the neutral and constraining conditions. These results favor models in which mapping from the input onto meaning is continuous over models in which contextual effects follow access of an initial form-based competitor set.
作者进行了两项“视觉世界”眼动追踪实验，以研究荷兰语结构中的词汇通达情况，其中动词对其后的主语名词短语施加或未施加语义限制。在实验1中，在约束动词条件下，对一组竞争对象图片（与所指对象的名字即主语的起始部分重叠）的注视与对干扰项的注视没有差异。在实验2中，交叉拼接引入了语音信息，该信息暂时使输入偏向于一组竞争对象。在中性条件和约束条件下，对一组竞争对象的注视都暂时增加了。这些结果支持这样的模型，即从输入到意义的映射是连续的，而不是那种语境效应在基于初始形式的竞争对象集通达之后才出现的模型。
