1 Introduction

This article introduces the concept of meta discourse and its role in academic discourse, pointing out that meta discourse helps authors express their views and positions, and promotes readers' understanding. At the same time, existing research has focused more on the differences in the use of metadiscourse in different contexts, lacking theoretical explanations of its underlying motivations. The theory of conformity provides pragmatic evidence for analyzing the use of meta discourse in academic discourse. This study will analyze the usage characteristics of meta discourse in interdisciplinary academic discourse at home and abroad from the perspective of adaptation theory, based on Hyland's meta discourse classification framework, and reveal the underlying rationale and motivation behind it.

2 Literature Review and Theoretical Framework

Hyland (2005) proposed an interpersonal model of meta discourse, which divides meta discourse into two categories: guided and interactive. The former constructs text coherence, while the latter expresses the author's viewpoint. Verschueren's (1999) adaptation theory suggests that language users adjust their language expression to conform to their environment, and the variability and negotiability of meta discourse provide possibilities for their adaptability. The four dimensions of adaptation theory constitute its basic framework, influencing language choice and expression. Previous studies have analyzed metadiscourse from the perspective of adaptation theory, but research on academic discourse is limited. This article will compare and analyze the use of meta discourse in English interdisciplinary scientific terminology texts by Chinese and foreign scholars from the perspective of conformity theory, summarize its characteristics and forming rationale, and provide reference for domestic scholars to write English academic discourse.

3 Research Methods

This chapter introduces research methods, including data collection and organization, and data statistical methods. The study collected SCI/SSCI high impact factor journal articles from the past decade, and classified them into English native speaker and Chinese native speaker papers based on the author's name and mailing address to determine their language background, covering both hard and soft disciplines. A total of 428 papers were included, with irrelevant content such as table of contents and references removed. Antconc software was used to calculate the frequency of 10 types of meta discourse markers, excluding words that do not achieve meta discourse function and words with extremely low frequency. A total of 221 markers were finally included in the statistics. The original frequency was standardized and subjected to chi square test.

4 Results and Discussion

This chapter compares and analyzes the differences in the use of meta discourse between ENS (native English speakers) and CNS (non-native English speakers) in academic discourse, revealing the characteristics and reasons for the use of meta discourse in different corpus categories. Research has found that the total use of meta discourse in ENS is significantly higher than that in CNS, especially in the use of transitional language and code annotation, which helps to enhance the logical and clear expression of the article, and conforms to the principles of communication in the social world. In contrast, CNS has a higher frequency of using evidence markers, especially the use of "conforming to", reflecting CNS's compliance with communicative principles in the psychological world. In terms of interactive meta discourse, ENS and CNS have significant differences in the use of fuzzy constraints and attitude markers. ENS tends to use multiple structures to express fuzzy constraints, while CNS tends to use verbs with a single speculative/suggestive meaning. In addition, there are significant differences in the use of meta discourse between hard and soft disciplines, with hard disciplines tending towards objective description and soft disciplines tending towards subjective expression. Soft subject papers promote communication and interaction with readers through the use of transitional language, code annotations, and fuzzy restrictive language, enhancing readers' resonance and understanding of the content of the paper. The physical world, social world, and psychological world in the theory of conformity affect the selection and use of meta discourse in academic discourse in different ways, and these three worlds are intertwined, jointly influencing the characteristics of meta discourse use.

5 Conclusion

We explored the pragmatic adaptability of meta discourse use in papers from different language backgrounds and disciplinary types, and found that there are differences in the use of guiding and interactive meta discourse among different authors. Hard disciplinary papers tend to use self proclaimed language to enhance logic and persuasiveness, while soft disciplinary papers establish emotional connections through transitional language, etc. These differences reveal the underlying rationale behind the use of meta discourse, providing a reference for the study of academic discourse writing.