由辛顿的获奖感言反思语言学研究的路向  

The Routes and Directions of Linguistic Studies: Reflections on Geoffrey E. Hinton's Criticism of Chomskyan School

作　　者：袁毓林 YUAN Yulin

机构地区：[1]澳门大学人文学院中国语言文学系,中国澳门999078

出　　处：《语言文字应用》2025年第1期86-99,共14页Applied Linguistics

基　　金：澳门大学讲座教授研究与发展基金(CPG2025-00008-FAH)和启动研究基金(SRG2022-00011-FAH)支持。

摘　　要：本文由辛顿在获诺贝尔奖感言中对乔姆斯基语言学的批评,反思语言学研究的路线与方向。首先,扼要引述辛顿在多个场合对乔姆斯基的批评,就此提出两个问题:(1)语言大模型的成功能不能颠覆语言天生理论?(2)生成语法理论有没有把语言学的方向带偏了?然后通过简介Piantadosi(2023)和进化神经生物学等有关观点和实验,说明生成语法的研究范式不符合人类朴素的心智。接着,简单比较近70年来人工智能和生成语言学的发展历程,说明学派竞争和客观评测是推动学术进步的有效路径。最后,通过分析乔姆斯基对其学术思想及其历程的自述,说明语言天生理论和普遍语法虽然不易证伪,但是更难证明其成立。This paper reflects on Geoffrey E. Hinton's criticism of the Chomskyan linguistic theory following his acceptance of the Nobel Prize, highlighting the routes and directions of linguistic studies. First, a brief overview is provided of Hinton's critiques made on several occasions of the Chomskyan linguistic theory and the following two key questions are posed:(1) Can the success of large language models(LLMs) overthrow the principles of linguistic innatism?(2) Has the theory of generative grammar misled the pursuit of linguistic research?Next, we introduce relevant perspectives from Piantadosi(2023) and findings from evolutionary neurobiology to demonstrate that the generative grammar paradigm does not align with the mentality simplicity of human beings. We then compare the developments in artificial intelligence and generative grammar over the past 70 years, illustrating how competition among different schools and objective evaluating can drive academic advancement. Finally, we analyze Chomsky's personal reflections on his academic thoughts and their evolution, highlighting the challenges of falsifying linguistic innatism and universal grammar, which are often even more difficult to verify.

关 键 词：辛顿 乔姆斯基 语言大模型 语言天生理论 生成语法 普遍语法 

分 类 号：H0-0[语言文字—语言学]