Poortman Eva B, Pylkkänen Liina
Utrecht Institute of Linguistics OTS, Utrecht University, Trans 10, 3512 JK Utrecht, The Netherlands.
Department of Linguistics, New York University, 10 Washington Place, New York, NY 10003, USA; Department of Psychology, New York University, 6 Washington Place, New York, NY 10003, USA; NYUAD Institute, New York University Abu Dhabi, PO Box 129188, Abu Dhabi, United Arab Emirates.
Brain Lang. 2016 Sep;160:50-60. doi: 10.1016/j.bandl.2016.07.006. Epub 2016 Jul 28.
Though a large literature implicates the left anterior temporal lobe (LATL) for combinatory operations, recent MEG studies have suggested that it is specifically involved in the composition of complex concepts, rather than syntactic or semantic composition in a more general sense. To further specify the computational contribution of the LATL, we tested whether LATL effects as observed in MEG require a situation in which features combine to form a single coherent entity representation or whether the relevant computation simply requires the attribution of features to a set but not necessarily to the same members of the set. Under the former hypothesis, the LATL would be sensitive to the number of features added to the representation of a single entity whereas under the latter account, LATL activity would reflect the total number of features integrated across different members of a set. To test this, we employed conjunctions of two adjectives whose lexical semantics were varied such that they either allowed or disallowed the attribution of their denoted properties to the same members of a set, i.e., the properties were either compatible or incompatible. The compatible properties resulted in so-called intersective and the incompatible in so-called collective readings. Our results show that the LATL tracks the number of features attributed to an individual as opposed to the number of features attributed to a set. Interestingly, the reverse pattern was found in the right ATL, demonstrating that although this region often shows parallel effects to the LATL, its functional contribution is clearly distinct.
尽管大量文献表明左前颞叶（LATL）参与组合操作，但最近的脑磁图（MEG）研究表明，它具体涉及复杂概念的构成，而非更广义的句法或语义构成。为了进一步明确LATL的计算贡献，我们测试了MEG中观察到的LATL效应是否需要一种特征组合形成单个连贯实体表征的情况，或者相关计算是否仅仅需要将特征归因于一个集合，而不一定是该集合的相同成员。在前一种假设下，LATL会对添加到单个实体表征中的特征数量敏感，而在后一种情况下，LATL活动将反映跨集合不同成员整合的特征总数。为了测试这一点，我们使用了两个形容词的连词，其词汇语义有所变化，使得它们要么允许要么不允许将其表示的属性归因于集合的相同成员，即属性要么兼容要么不兼容。兼容属性导致所谓的交集解读，不兼容属性导致所谓的集合解读。我们的结果表明，LATL追踪归因于个体的特征数量，而不是归因于集合的特征数量。有趣的是，在右前颞叶（ATL）中发现了相反的模式，这表明尽管该区域通常显示出与LATL平行的效应，但其功能贡献明显不同。
