Language and Thought: Investigating the Relationship Through Structure and Semantics

As mentioned earlier, there are two common approaches to studying the relationship between language and thought. One is to focus on language structure, collecting visible and inducible features (such as grammar and phonology) and analyzing the differences between different languages to draw conclusions about how structural features affect people's ways of thinking. The other approach is to observe the differences in language use among speakers of different languages in the same context in order to draw conclusions about how language influences thought. Although the conclusions reached by these two methods are very different, the scope and difficulty of the arguments involved are completely different. Structuralists believe that studying semantics is impractical, and that grammar is the only thing that can be captured and used for scientific research. Sapir-Worf's later research shifted from language structure to the semantic level, believing that only by combining the speaker's environment can the true meaning of their discourse be inferred.

Several scholars have attempted to empirically verify the Sapir-Worf hypothesis by using various standards. Wolf used the average standard European language to confirm whether the Hopi language has the concept of 'duration,' while Kaplan used the ability to use the language to infer text as a criterion to argue that Chinese speakers are not good at logical reasoning in hypothetical contexts. American linguist John A. Lucy compared the number markers of Yucatan Mayan and American English to determine whether the use of quantifiers in the plural marker of English is more focused on the shape of objects, while the Yucatan Mayan uses plural markers that are more related to the object's own characteristics. Based on this data comparison, John believes that English speakers need to pay more attention to the structure and characteristics of objects when recognizing things, while Yucatan Mayan speakers need to pay more attention to their external shape.

All three experiments above demonstrate that language influences ways of thinking from different perspectives, but the methods used still have shortcomings. In the thinking of Hopi language users, linear concepts such as tomorrow, yesterday, and the day after tomorrow do exist. If there are no such words in the language, they will use other words to express these concepts. Regarding the comparison of Chinese and English reasoning abilities, a Chinese linguist believes that the lack of authenticity in the Chinese translation affected the students' reading comprehension. When the Chinese questions were re-translated and given to Chinese students, the results showed that the reasoning abilities of Chinese speakers were basically the same as those of English speakers.

From the shortcomings of the above experiments, it can be concluded that in order to analyze and compare the different thinking modes influenced by two languages, existing language models cannot be used as standards, and a universal approach should be adopted. In addition, the texts used in the experiments should strive to replicate the expressions of the mother tongue.