School of Psychological Sciences, La Trobe University, Bundoora 3086, Victoria, Australia.
J Child Lang. 2011 Jan;38(1):222-34. doi: 10.1017/S0305000909990316. Epub 2010 Mar 3.
In this paper we report on a visual world eye-tracking experiment that investigated the differing abilities of adults and children to use referential scene information during reanalysis to overcome lexical biases during sentence processing. The results showed that adults incorporated aspects of the referential scene into their parse as soon as it became apparent that a test sentence was syntactically ambiguous, suggesting they considered the two alternative analyses in parallel. In contrast, the children appeared not to re-analyze their initial analysis, even over shorter distances than have been investigated in prior research. We argue that this reflects the children's over-reliance on bottom-up, lexical cues to interpretation. The implications for the development of parsing routines are discussed.
在本文中，我们报告了一项视觉世界眼动实验的结果，该实验调查了成人和儿童在重新分析过程中使用参照场景信息的不同能力，以克服句子处理过程中的词汇偏向。结果表明，一旦测试句子在句法上变得模棱两可，成年人就会将参照场景的某些方面纳入到他们的解析中，这表明他们同时考虑了两种可供选择的分析。相比之下，孩子们似乎并没有重新分析他们的初始分析，即使在比之前研究中调查的更短的距离内也是如此。我们认为，这反映了孩子们过度依赖于自下而上、词汇线索来进行解释。本文还讨论了这对解析程序发展的影响。
