Department of Spanish, Italian and Portuguese, Center for Language Science, The Pennsylvania State University.
Cogn Sci. 2023 Oct;47(10):e13366. doi: 10.1111/cogs.13366.
Previous work on individual differences has revealed limitations in the ability of existing measures (e.g., working memory) to predict language processing. Recent evidence suggests that an individual's sensitivity to detect the statistical regularities present in language (i.e., "chunk sensitivity") may significantly modulate online sentence processing. We investigated whether individual chunk sensitivity predicted the online processing of gender cues, a core linguistic feature of Spanish. In a self-paced reading task, we examined native speakers' processing of relative clauses in which gender cues were variably exploited to induce processing costs. Even after considering the effect of working memory and cognitive control, the results revealed a significant effect of chunking ability in modulating online sentence processing. Critically, higher-chunking ability speakers' reading times showed online sensitivity to core linguistic cues online; while low-chunking ability readers showed no sensitivity to manipulations, indicating shallow real-time processing of their native language.
先前关于个体差异的研究揭示了现有测量方法（例如工作记忆）在预测语言处理能力方面的局限性。最近的证据表明，个体对检测语言中存在的统计规律的敏感性（即“组块敏感性”）可能会显著调节在线句子处理。我们研究了个体的组块敏感性是否可以预测性别线索的在线处理，这是西班牙语的核心语言特征之一。在自定步调的阅读任务中，我们检查了母语为西班牙语的人对关系从句的处理，其中性别线索的变化会引起处理成本。即使考虑到工作记忆和认知控制的影响，结果仍显示出组块能力在调节在线句子处理方面的显著效果。关键是，具有较高组块能力的说话者的阅读时间表现出对核心语言线索的在线敏感性；而低组块能力的读者则对这些操作没有敏感性，表明他们对母语的实时处理较浅。
