神经科学视角下的会话含意理解  

A Neuroscience Perspective on the Recovery of Conversational Implicatures

作　　者：冯望舒 周晓林 FENG Wangshu;ZHOU Xiaolin(Artificial Intelligence and Human Languages Lab,Beijing Foreign Studies University;不详)

机构地区：[1]北京外国语大学人工智能与人类语言重点实验室 [2]上海外国语大学语言研究院 [3]华东师范大学心理与认知科学学院

出　　处：《当代语言学》2024年第2期259-273,共15页Contemporary Linguistics

基　　金：中央高校基本科研业务费专项资金“语言智能理论与技术应用创新”(2023TD001)的资助。

摘　　要：格赖斯(Grice 1957)提出了会话含意这一概念,并将其分为一般会话含意和特殊会话含意。当前三种主流的语用理论对两种含意之间的关系持不同观点,对两种含意理解的神经基础有不同的预期。基于神经科学的最新研究成果,本文发现,相比于默认理论和关联理论,语义最小论的观点更加贴合含意理解的实证研究结果。具体来说,特殊含意与一般含意理解所涉及的语用系统共享语言加工的核心网络,负责含意理解所需的额外语义激活与整合加工;但特殊含意理解还涉及负责心理理论的脑网络,以支持个体对说话人在当前特定语境下的心理状态进行推测;一般含意理解所涉及的更加胶囊化的语用系统还包括背内侧前额皮层,其计算过程需要根据以往会话交际中有关说话人意图的统计事实进行。这些发现将对认识含意理解的认知过程有重要的参考价值。In daily conversation,the listener needs not only to decode the contextinvariant“"sentence meaning",but also to infer the implicated meaning(conversational implicature)beyond the literal expression.Grice(1957)proposes a concept of conversational implicature and divides it into generalized conversational implicature(GCI)and particularized conversational implicature(PCI).Three mainstream pragmatic theories hold different views concerning the relationship between the two types of implicatures and have different predictions about their neurocognitive substrates.According to default theory,GCI is computed by an automatic and effortless system that supports default inferences,whereas PCI is computed by a separate one that supports context-sensitive inferences.In contrast,relevance theory holds that both PCI and GCI are derived from the same cognitive processes,which take contextual considerations into account from the beginning.Semantic minimalism offers a more balanced view that PCI is derived from fully context-based inferences,while GCI is derived from constraint-based inferences.Comparing the neurocognitive mechanisms underlying PCI and GCI comprehension would allow us to choose between these theoretical approaches.

关 键 词：含意 一般会话含意 特殊会话含意 语言网络 心理理论 

分 类 号：H0[语言文字—语言学]