Is the Brain Mechanism for Hierarchical Structure Building Universal Across Languages? An fMRI Study of Chinese and English

Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 7852 - 7861 December 7-11, 2022 ©2022 Association for Computational Linguistics Is the Brain Mechanism for Hierarchical Structure Building Universal Across Languages? An fMRI Study of Chinese and English Xiaohan Zhang1,2, Shaonan Wang1,2, Nan Lin3,4, Chengqing Zong1,2,5 1 National Laboratory of Pattern Recognition, Institute of Automation, CAS 2 School of Artificial Intelligence, University of Chinese Academy of Sciences 3 CAS Key Laboratory of Behavioural Sciences, Institute of Psychology 4 Department of Psychology, University of Chinese Academy of Sciences 5 CAS Center for Excellence in Brain Science and Intelligence Technology {xiaohan.zhang, shaonan.wang, cqzong}@nlpr.ia.ac.cn; linn@psych.ac.cn Abstract Evidence from psycholinguistic studies sug- gests that the human brain builds a hierarchi- cal syntactic structure during language compre- hension. However, it is still unknown whether the neural basis of such structures is univer- sal across languages. In this paper, we first analyze the differences in language structure between two diverse languages: Chinese and English. By computing the working memory requirements when applying parsing strategies to different language structures, we find that top-down parsing generates less memory load for the right-branching English and bottom- up parsing is less memory-demanding for Chi- nese. Then we use functional magnetic reso- nance imaging (fMRI) to investigate whether the brain has different syntactic adaptation strategies in processing Chinese and English. Specifically, for both Chinese and English, we extract predictors from the implementations of different parsing strategies, i.e., bottom-up and top-down. Then, these predictors are separately associated with fMRI signals. Results show that for Chinese and English, the brain utilizes bottom-up and top-down parsing strategies sep- arately. These results suggest that the brain adopts parsing strategies with less memory load according to different language structures. 1 Introduction A hallmark of human language ability is combin- ing linear sequential word inputs into a hierarchical structure using abstract syntactic rules. This abil- ity enables us to create infinite expressions from finite words. Previous studies have shown that several brain regions are involved in building the hierarchical syntactic structure (Hagoort and Inde- frey, 2014; Zaccarella et al., 2017). However, the parsing strategies that the brain uses to build such structures are less discussed. Moreover, most of these studies only have been conducted on one lan- guage. It is still unknown whether the neural basis of such structures is universal across languages. This paper investigates the neural basis of syn- tactic structure-building, particularly the parsing strategies that the brain uses, in two diverse lan- guages: Chinese and English. We focus on two parsing strategies: top-down and bottom-up. Since existing work has demonstrated that a parsing strat- egy requires different memory space when pro- cessing sentences with different branching direc- tions(Resnik, 1992), we first analyze the branching structures and the memory load generated by two parsing strategies for Chinese and English. Re- sults show that the dominant branching directions and the memory space required by parsing strate- gies are both different between the two languages. For the right-branching English, the top-down pars- ing …

Content truncated to 500 words