1 Introduction

Abstract plays a crucial role in academic writing, as it not only summarizes the research content, but also promotes the dissemination and understanding of the research. The social function of an abstract is reflected in its use of rhetorical strategies to guide readers and establish relationships with them. The meta discourse theory focuses on these rhetorical and linguistic features in the abstract, especially the use of stance language, which includes vague restrictive language, dominant language, attitude and emotion markers, and self referential language. These elements are crucial for constructing the author's stance and expressing their attitude. Science popularization abstract, which is an abstract aimed at a wider readership, using more understandable language, aiming to expand the influence of research. Although popular science abstracts have become a new trend in academic publishing, there is relatively little research on their language and genre characteristics. This study adopts the meta discourse theory framework to explore the differences in the use of stance language between popular science abstracts and traditional abstracts, in order to fill the gap in existing research.

2 Research Design

This chapter explores the role of stance language in international journal abstracts, particularly in the use of stance language in popular science abstracts. Based on Hyland's classification of stance language, this study analyzed the differences in the use of stance language between traditional abstracts and popular science abstracts. The study constructed an abstract corpus containing two important journals in the field of earth sciences, and used AntConc software and SPSS for word frequency statistics and significance testing, aiming to reveal how authors adjust their identity construction strategies according to different audiences, reflect new trends in international journal writing, and provide reference for paper writing.

3 Results and Discussion

In international journal articles, the frequency of using stance language is slightly higher in popular science abstracts compared to traditional abstracts, with 34.5 words and 28.6 words per thousand words, respectively. Journal authors tend to use meta discourse resources in their abstracts to express the degree of certainty or rigor in the proposition or argumentation process. The frequency of using self referential language in popular science abstracts is significantly higher than that in traditional abstracts, which may be related to targeting a wider readership. The author uses language strategies to reflect their cognition and attitude towards information or discourse, in order to highlight their identity and construct their stance. Both types of abstracts tend to use verbs, such as "may," "suggest," and "indicate," in the use of fuzzy restrictive language to avoid excessive absolutization and leave room for negotiation. The frequency of using dominant language is slightly lower, but popular science abstracts and traditional abstracts show consistency in word expression, mainly relying on verbs such as "show", "find", "demonstrate", and "know" to enhance a positive and confident tone. The use of self referential language is more frequent in popular science abstracts, usually using the first person plural pronoun "we" to enhance persuasiveness and establish authorship, reflecting the team collaboration nature of scientific research and the development trend of journal paper writing.

4 Conclusion

Both popular science abstracts and traditional abstracts frequently use position language in international journals in the field of earth science to interact and express the author's position. Science popularization abstracts tend to use self referential language to enhance credibility and reader interaction. The limitations of the research lie in the small amount of data and limited disciplinary scope. Future research needs to expand the corpus and disciplinary scope, and explore more factors that affect the writing of popular science abstracts.