Early sensitivity of left perisylvian cortex to relationality in nouns and verbs

Neuropsychologia(2017)

引用 60|浏览20
暂无评分
摘要
The ability to track the relationality of concepts, i.e., their capacity to encode a relationship between entities, is one of the core semantic abilities humans possess. In language processing, we systematically leverage this ability when computing verbal argument structure, in order to link participants to the events they participate in. Previous work has converged on a large region of left posterior perisylvian cortex as a locus for such processing, but the wide range of experimental stimuli and manipulations has yielded an unclear picture of the region's exact role(s). Importantly, there is a tendency for effects of relationality in single-word studies to localize to posterior temporo-parietal cortex, while argument structure effects in sentences appear in left superior temporal cortex. To characterize these sensitivities, we designed two MEG experiments that cross the factors relationality and eventivity. The first used minimal noun phrases and tested for an effect of semantic composition, while the second employed full sentences and a manipulation of grammatical category. The former identified a region of the left inferior parietal lobe sensitive to relationality, but not eventivity or combination, beginning at 170ms. The latter revealed a similarly-timed effect of relationality in left mid-superior temporal cortex, independent of eventivity and category. The results suggest that i) multiple sub-regions of perisylvian cortex are sensitive to the relationality carried by concepts even in the absence of arguments, ii) linguistic context modulates the locus of this sensitivity, consistent with prior studies, and iii) relationality information is accessed early – before 200ms – regardless of the concept's event status or syntactic category.
更多
查看译文
关键词
Relationality,Magnetoencephalography,Inferior parietal lobe,Superior temporal cortex,Argument structure,Grammatical category
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要