【摘要】 长期以来，语言相对论关于语言深刻影响思维的命题一直得不到实证和证伪，主要是缺少一种“语言人”和“非语言人”的比较。AI大语言模型让AI获得了自然语言理解和生成能力，AI的能力也产生飞跃，甚至出现涌现能力，这证明自然语言符号系统对AI的认识活动有强烈的建构作用。这引导我们去观察人类大脑在具备某种特殊符号系统的情况下认识能力有什么变化。通过比较证明，数学符号系统对数学认识活动有建构作用，音乐符号系统对音乐认识活动有建构作用。总之，无论是人工智能AI还是人脑，都受到了语言符号系统的深刻建构。语言不仅仅是记录思维和交际的工具，更重要的是认识活动得以建立的建筑材料。由于语言符号系统的差异，这种建构也存在差异，在一定程度上支持语言相对论，即不同的语言符号系统对认识活动有不同的建构作用。至于这种建构是否深刻影响思维，有待进一步研究。更多还原

【Abstract】 For a long time, the proposition of linguistic relativity — that language profoundly influences thinking — has remained difficult to empirically verify or falsify, largely due to the lack of a comparison between “linguistic humans” and “non-linguistic humans.” AI large language models have enabled AI to acquire natural language understanding and generation capabilities, leading to a leap in AI’s abilities, and even in emergent. This demonstrates that natural language symbol systems play a strong constructive role in AI’s cognitive activities. This insight guides us to observe how human brain’s cognitive capacities change when equipped with specific symbol systems. Through comparative verification, it has been shown that mathematical symbol systems play a constructive role in mathematical cognitive activities, while musical notation systems play a constructive role in musical cognitive activities. In short, no matter whether it is artificial intelligence or human brain, both are profoundly constructed by language symbol systems. Language is a tool not only for recording thought but also for communication. More importantly, it serves as the building material through which cognitive activities are established. Due to the differences in language symbol systems, such constructions are also different, providing support for linguistic relativity to a certain extent, that is, different language symbol systems produce different constructive influences on cognitive activities. As to whether such constructions may profoundly affect thinking, further research is needed.更多还原