中国手语模态对句子加工的影响
发布时间：2019-01-23 14:33
【摘要】：该文系统地研究了中国手语模态对句子加工的影响,使用事件相关电位(ERP)技术来探索句子加工过程中手势模态的语义和语音信息对词汇通达的影响和时间过程。设计实验材料时,对句尾变化条件下目标手势的语义和音韵形式进行了系统的人为操纵。通过18个被试的实验,观察到将初级视觉特征映射到手势音韵特征的手势感知证据(时长100~200ms),紧接着是将目标手势音韵特征与语境信息匹配的负成分(时长200~400ms),以及表征语篇语境对语义制约的动态建构的负成分(时长400~650ms),它促进了手势的音义信息与句子信息的整合。音近和义近手势在时间进程和效应方向上存在着相似性,表明手语模态的词汇通达与有声语言类似,并消耗类似的在线处理成本。同时发现了手语模态独有的表征手语音韵特征加工已经完成的N300效应,它代表了手势音韵元素识别和手势识别之间的直接联系。实验结果表明:在进行句子加工时,被试首要考虑的是语境和目标手势匹配,并不是与手势的音韵特征匹配。这种结果可能是词汇选择时音近和义近相关信息相互作用的标志。
[Abstract]:In this paper, the effects of Chinese sign language modes on sentence processing are systematically studied. The Event-related potential (ERP) technique is used to explore the semantic and phonological effects of gesture modes and phonological information on lexical access in sentence processing. When designing experimental materials, the semantic and phonological forms of target gestures are systematically manipulated. Through the experiments of 18 subjects, we observed the gestural perception evidence (100~200ms) which mapped primary visual features to gesture phonological features, followed by negative components (time-length 200~400ms), which matched the phonological features of target gestures with contextual information. And the negative component (400~650ms), which characterizes the dynamic construction of semantic constraints in discourse context, promotes the integration of phonological and sentence information of gestures. There are similarities in time process and effect direction between near and semantic gestures, which indicates that the lexical accessibility of sign language mode is similar to that of sound language, and the online processing cost is similar. At the same time, the N300 effect, which is unique to sign language modal, is found to represent the N300 effect, which represents the direct relationship between gesture rhyme element recognition and gesture recognition. The experimental results show that the context and target gesture matching is the primary consideration in sentence processing, but not with the phonological feature of the gesture. This result may be a sign of the interaction between phonological and semantic information in lexical selection.
【作者单位】： 清华大学人文学院计算语言学实验室;清华大学心理学与认知科学研究中心;北京市信息服务工程重点实验室(北京联合大学);
【基金】：国家自然科学基金重大项目(91420202);国家自然科学基金重点项目(61433015) 国家社会科学基金重大项目(14ZDB154) 教育部人文社会科学研究规划基金青年项目(14YJC740104) 清华大学自主科研项目两岸清华大学专项(20161080056)
【分类号】：H126.3
，
本文编号：2413895
[Abstract]:In this paper, the effects of Chinese sign language modes on sentence processing are systematically studied. The Event-related potential (ERP) technique is used to explore the semantic and phonological effects of gesture modes and phonological information on lexical access in sentence processing. When designing experimental materials, the semantic and phonological forms of target gestures are systematically manipulated. Through the experiments of 18 subjects, we observed the gestural perception evidence (100~200ms) which mapped primary visual features to gesture phonological features, followed by negative components (time-length 200~400ms), which matched the phonological features of target gestures with contextual information. And the negative component (400~650ms), which characterizes the dynamic construction of semantic constraints in discourse context, promotes the integration of phonological and sentence information of gestures. There are similarities in time process and effect direction between near and semantic gestures, which indicates that the lexical accessibility of sign language mode is similar to that of sound language, and the online processing cost is similar. At the same time, the N300 effect, which is unique to sign language modal, is found to represent the N300 effect, which represents the direct relationship between gesture rhyme element recognition and gesture recognition. The experimental results show that the context and target gesture matching is the primary consideration in sentence processing, but not with the phonological feature of the gesture. This result may be a sign of the interaction between phonological and semantic information in lexical selection.
【作者单位】： 清华大学人文学院计算语言学实验室;清华大学心理学与认知科学研究中心;北京市信息服务工程重点实验室(北京联合大学);
【基金】：国家自然科学基金重大项目(91420202);国家自然科学基金重点项目(61433015) 国家社会科学基金重大项目(14ZDB154) 教育部人文社会科学研究规划基金青年项目(14YJC740104) 清华大学自主科研项目两岸清华大学专项(20161080056)
【分类号】：H126.3
，
本文编号：2413895
本文链接：https://www.wllwen.com/wenyilunwen/yuyanyishu/2413895.html
