对“中文屋”思想实验四个主要版本的考察

A Study on Four Major Versions of "Chinese Room" Thought Experiment

"中文屋"思想实验提出30年来,学界对该思想实验在批判强人工智能的有效性问题上一直争论不休。特别是强人工智能的支持者往往通过转换视角,添加问题或构建新理论模型等方式试图躲避"中文屋"的责难。为应对各种回应,塞尔在保持"中文屋"论证的基本逻辑结构和要点不变的前提下,通过修改思想实验的描述形式进行一一答复,从而保证了该思想实验对强人工智能批判上的效力。借助对"中文屋"几个主要版本及其相关答复的考察,表明"中文屋"在批判强人工智能上的有效性,同时指出这种有效性并不依赖于思想实验外在描述形式的变化,而依赖于蕴含其中的论证结构和论证形式。

John Searle has proposed his ＂Chinese Room＂ Thought Experiment over thirty years.During this time,there were heated debates over the validity of its criticism of strong artificial intelligence.In particular,supporters of strong artificial intelligence often try to avoid the censure from Chinese Room in several ways：changing perspective,adding questions or building new theoretical models.In response to a variety of responses,Searle has modified the descriptive form of Chinese Room without changing the basic argumentative structure and gist of this experiment.Thus he ensured the validity of Chinese Room on criticizing strong artificial intelligence.This paper attempts to show the validity of Searle＇s criticism of strong artificial intelligence by exploring several major versions of Chinese Room and its related responses.It also tries to point out that the validity of Chinese Room is not dependent on the changing of its external form,but on its argumentative structure.